Datasets:

TaylorAI

/

pubmed_noncommercial

like 5

PMC005xxxxxx/PMC5371711.txt

==== Front Contemp Oncol (Pozn)Contemp Oncol (Pozn)WOContemporary Oncology1428-25261897-4309Termedia Publishing House 2890910.5114/wo.2016.64608Case ReportCastleman’s disease mimicking lymph node metastases in a young woman with laryngeal cancer Serkies Krystyna Łazar-Poniatowska Małgorzata Seredyńska Joanna Biernat Wojciech Jassem Jacek Medical University of Gdansk, PolandAddress for correspondence: Krystyna Serkies, Medical University of Gdansk, Dębinki 7, 80-211 Gdańsk, Poland. e-mail: kserkies@gumed.edu.pl20 12 2016 2016 20 5 421 424 27 10 2014 14 4 2015 Copyright: © 2016 Termedia Sp. z o. o.2016This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) License, allowing third parties to copy and redistribute the material in any medium or format and to remix, transform, and build upon the material, provided the original work is properly cited and states its license.Laryngeal cancer occurs rarely in adolescents and young people. Castleman’s disease is a rare lymphoproliferative disorder of uncertain etiopathogenesis and heterogeneous clinicopathological forms. Involved lymph nodes and extranodal lesions in the course of Castleman’s disease may mimic malignant involvement. We report a case of an 18-year-old woman with T2N0M0 laryngeal glottis cancer treated with definitive radiotherapy. During the irradiation, the patient underwent an excision of incidentally discovered left-sided enlarged cervical lymph nodes located outside the irradiated area. Coincidental hyaline vascular type of Castleman’s disease was diagnosed. During six-year follow-up she has been free of cancer relapse and Castleman’s disease symptoms. laryngeal carcinomaCastleman’s diseaseadolescentradiotherapysurvival ==== Body Introduction Carcinoma of the larynx typically occurs in adult males. Primary risk factors for this tumour include active and passive smoking, alcohol abuse, and poor oral hygiene [1]. Human papilloma virus (HPV) and human immunodeficiency virus (HIV) infections, impaired immunity, gastro-oesophageal reflux disease, as well as several molecular abnormalities are additional risk factors for this malignancy [2]. The course of laryngeal cancer is strictly related to its location: cancers located in the glottis have relatively slow growth and low risk of nodal metastases. Among children and young adults, the disease, most frequently located in the glottis, is extremely rare. Laryngeal carcinoma in adolescents is believed to have particularly aggressive behaviour. Castleman’s disease (CD) is a noncancerous lymphoproliferative disorder of unclear aetiology [3–6]. This entity occurs in heterogeneous clinicopathological forms and comprises at least two distinct diseases: a unicentric CD (UCD; 90%) subtype and a multicentric (MCD) subtype, which develops in both HIV-negative or HIV-positive patients. There are three classical histological subtypes of CD: a hyaline-vascular (HV, most common, 60%), plasma cell (PC, 10%), and mixed variants. The recently described plasmablastic subtype is associated with MCD and mostly with human herpes virus 8 (HHV-8) and HIV co-infection. Epstein-Barr virus infection may also result in nodal CD. Finally, immune disorders and autoimmunisation have been postulated as pathogenic factors of CD. The unicentric HV variant of CD occurs mainly in adolescents as a slowly growing tumour located in the mediastinum or neck region [5, 6]. After surgical excision of the lesion recurrence is extremely rare. MCD, typically a PC subtype, occurs in adults and affects mainly peripheral lymph nodes. MCD represents a systemic disease with more aggressive behaviour and uncertain prognosis. MCD frequently manifests with systemic symptoms, such as fever, weight loss, weakness, and night sweats, considered to be a result of overproduction of interleukin-6 (IL-6) by lymph nodes (primary under influence of HHV-8) [5]. The coincidence of CD with solid tumours is extremely rare. We present a case of an 18-year-old woman who developed CD in cervical lymph nodes during radical irradiation for laryngeal cancer. Case report An 18-year-old woman with T2N0M0 laryngeal cancer was admitted in October 2008 to receive radical radiotherapy. The patient reported an increasing hoarseness lasting for five months. She negated alcohol or tobacco use, laryngeal papillomatosis, and prior neck radiotherapy. Laryngoscopic examination revealed a lesion involving the left true vocal cord, left laryngeal pocket, and posterior part of the ventricle false cord. Biopsy of the lesion revealed moderately differentiated squamous cell carcinoma of the larynx (Fig. 1A, B). CT demonstrated thickening of 8 mm of the left false vocal cord, shallowing the left laryngeal ventricle, and infiltration extending to the upper part of the vocal cord and small (up to 5 mm in the greatest diameter) unspecific lymph nodes of the neck. Staging work-up, including chest CT and ultrasound examination of the neck and abdomen, was negative for metastases. The patient was referred to radical radiotherapy with four lateral 6 MEV photon beams (5.5 cm × 7.5 cm) to the total dose of 66 Gy in 33 fractions. She started treatment two weeks after the diagnosis. Another laryngeal examination, performed on the first day of radiotherapy, showed previously absent infiltration of the anterior commissure and subglottis. Owing to these findings, after delivering 6 Gy, the lower radiation fields were extended by 0.5 cm. Additional CT, performed for radiotherapy modification revealed previously absent, enlarged, left-sided cervical lymph nodes of the second group, outside the irradiated region. An ultrasound of this region showed enlarged lymph nodes with a central hyperechoic area. The biggest lymph node was 2.5 cm in diameter and was localised at the left mandibular angle. Radiotherapy was resumed after a three-day pause. The patient refused fine-needle aspiration biopsy of the neck lymph node. PET-CT performed without interrupting radiotherapy revealed metabolically active lesions in the larynx and in the left cervical lymph nodes at the level of the third cervical vertebra (SUV of 3.8 and 2.4, respectively). After 19 fractions the enlarged cervical lymph nodes were removed. The surgery caused an additional four-day pause in radiotherapy. The tissue material contained two lymph nodes (17 and 12 mm in diameter), with macroscopic features of typical reactive inflammatory response. Histopathology examination showed HV type of CD (Fig. 1C, D). During six-year follow-up, she has been free of cancer recurrence and CD disease symptoms. Fig. 1 Laryngeal squamous carcinoma was diagnosed from a small biopsy (A) that revealed well differentiated epithelial tumour with slight atypia and invasive growth into the fibrous stroma (B). Lymph node biopsy showed small, atrophic germinal centres surrounded by expansion of mantle zone and prominent interfollicular region (C). Typical “onionskin” pattern is visible at the germinal centre (D) Discussion Laryngeal carcinoma in patients aged below 40 years accounts for less than 3% of all laryngeal cancer cases [7–9]. The rarity of the disease among adolescents hampers the diagnosis and may lead to delay in treatment, as was the case in our patient. It has been estimated that in glottis cancer patients a one-month delay from the onset of symptoms to start of radiotherapy was equivalent to a 4.5% decrease in recurrence-free survival [10]. The management of laryngeal cancer in young people is similar to that in adults. Radiotherapy continues to be the preferred method, allowing in most cases preservation of laryngeal function and voice. On the other hand, this modality is associated with the risk of post-irradiation effects including second malignancies. The patient presented in this report received conventional radiotherapy at a dose of 66 Gy to a region confined to the larynx. As recently reported, a 13-year-old boy with T2N0MX moderately differentiated HPV-positive squamous cell carcinoma of the glottis was administered a hyperfractionated dose of 81.6 Gy [11]. Diagnosis of CD is based on histological evaluation of lymph nodes, but fine-needle biopsy may be sufficient to confirm CD in the extranodal sites, although it is challenging for the pathologist. On clinical examination and radiological images, CD can mimic various lesions, including malignant disorders [12–14]. If accompanied by cancer, it may be misdiagnosed as progression or recurrence. Specific calcifications may occur in the HV type; however, as in our case, they are usually absent from the involved lymph nodes. Intensively enhancing CD-involved nodal masses present in CT scans can be difficult to distinguish from lymphomas or cancer metastases. PET is helpful in resolving ambiguous CT findings [6]. As in our patient, SUV values of CD-involved lymph nodes are lower than those typical for lymphoma or cancer metastases. The association between CD and cancer remains unclear. Several authors suggested that CD is associated with an increased risk of lymphomas and lymphoma-associated diseases [15, 16]. HHV-8 infection, typical for MCD, is often present in AIDS patients with or without Kaposi sarcoma, and those with other lymphomas. It was suggested that IL-6 and VEGF are the paracrine factors related to CD involved in carcinogenesis of malignant tumours that coexisted with CD [5, 17]. Anecdotal coincidences of CD with lymphomas, B-cell non-Hodgkin, and Hodgkin lymphomas and with multiple myeloma were reported [18, 19]. A few reports described co-existence of CD (frequently unicentric and HV type) and solid neoplasms including ovarian, colorectal, bladder, renal, lung and tongue cancer, hepatocellular carcinoma, and retroperitoneal leiomyosarcoma, diagnosed in both early or advanced stages of disease [17, 20–25]. To our knowledge this is the first published report of the co-existence of CD and laryngeal carcinoma. The possible relationship between CD and this rare case of laryngeal cancer in an adolescent remains unclear. Castleman’s disease should be kept in mind as a differential diagnosis for nodal enlargement. The authors declare no conflict of interest. ==== Refs References 1 Maasland DH van den Brandt PA Kremer B Goldbohm RA Schouten LJ Alcohol consumption, cigarette smoking and the risk of subtypes of head-neck cancer: results from the Netherlands Cohort Study BMC Cancer 2014 14 187 10.1186/1471-2407-14-187 24629046 2 Niu J Huang YJ Wang LE Sturgis EM Wei Q Genetic polymorphisms in the PTPN13 gene and risk of squamous cell carcinoma of head and neck Carcinogenesis 2009 30 2053 8 19892796 3 Bowne WB Lewis JJ Filippa DA Niesvizky R Brooks AD Burt ME Brennan MF The management of unicentric and multicentric Castleman’s disease Cancer 1999 85 706 17 10091744 4 Casper C The aetiology and management of Castleman disease at 50 years; translating pathophysiology to patient care Br J Haematol 2005 129 3 17 15801951 5 Cronin DM Warnke RA Castleman disease: an update on classification and the spectrum of associated lesions Adv Anat Pathol 2009 16 236 46 19546611 6 Madan R Chen J-H Trotman-Dickenson B Jacobson F Hunsaker A The spectrum of Castleman’s disease: Mimics, radiologic pathologic correlation and role of imaging in patient management Eur J Radiol 2012 81 123 31 20643523 7 Prasad KC Abraham P Peter R Malignancy of the larynx in a child Ear Nose Throat J 2001 8 508 11 8 Rutt AL Hawkshaw MJ Sataloff RT Laryngeal cancer in patients younger than 30 years: a review of 99 cases Ear Nose Throat J 2010 89 189 92 20397149 9 Shvero J Hadar T Segal K Abraham A Sidi J Laryngeal carcinoma in patients 40 years of age and younger Cancer 1987 60 3092 5 3677032 10 Hansen O Larsen S Bastholt L Godballe CH Jorgensen KE Duration of symptoms: impact on outcome of radiotherapy in glottis cancer patients Int J Radiat Biol Phys 2005 61 789 94 11 Joos B Joos N Bumpous J Burns C French ChA Farghaly H Laryngeal squamous cell carcinoma in a 13 year-old child associated with human papillomaviruses 16 and 18: a case report and review of the literature Head and Neck Pathol 2009 3 37 41 20596987 12 Hakimi AA Faiena I Kaleya RN Ghavamian R Retroperitoneal Castleman’s disease Urology 2010 76 1379 20138654 13 MacDonald SR Lurain JR Hoff F Variakojis D Fishman DA Castleman disease presenting as a pelvic mass Obstet Gynecol 1996 87 875 7 8677123 14 Petrina A Eugeni E Badolato M Boselli C Covarelli P Rondelli F Noya G Unicentric Castleman’s disease approached as a pancreatic neoplasm: case report and review of literature Cases J 2009 2 9090 20062727 15 Abdel-Reheim FA Koss W Rappaport ES Arber DA Coexistence of Hodgkin’s disease and giant lymph node hyperplasia of the plasma-cell type (Castleman’s disease) Arch Pathol Lab Med 1996 120 91 6 8554455 16 Nicoli P Familiari U Bosa M HHV8-positive, HIV-negative multicentric Castleman’s disease: early and sustained complete remission with rituximab therapy without reactivation of Kaposi sarcoma Int J Hematol 2009 90 392 6 19756920 17 Chun YS Calderaro J Zucman-Rossi J Synchronous hepatocellular carcinoma and Castleman’s disease: the role of the interleukin-6-signaling pathway Hepatology 2012 56 392 3 22611056 18 Unsal Tuna EE Ozbek C Arda N IIdogan E Dere H Ozdem C Development of a Hodgkin disease tumor in the neck of a patient who previously had undergone complete excision of a hyaline-vascular Castleman disease neck mass Ear Nose Throat J 2010 89 20 3 19 Yuan Z Dun X Li Y Hou J Treatment of multicentric Castleman’s disease accompanying multiple myeloma with bortezomid: a case report J Hematol Oncol 2009 2 19 22 19400935 20 Bertero D Buzio M Albertino B Giaccone M Ricci E A case report of Castleman’s disease in a patient with endometrioid adenocarcinoma of the ovary Minerva Chir 1991 46 989 93 1754099 21 Chan J Loh A Sim H Tan MH Toh ChK Coexistence of unicentric Castleman’s disease and locally advanced papillary renal cell carcinoma: more than a coincidental association? Ann Acad Med 2010 39 584 5 22 Deshmukh M Ball M Deshpande P Jambhekar NA Synchronous squamous cell carcinoma of tongue and unicentric cervical Castleman’s disease clinically mimicking a stage IV disease: a rare association or coincidence? Head Neck Pathol 2011 5 180 3 21240573 23 Gomez-Raposo C Nistal M De Castro Carpieno J Rotunda GS Belda-Iniesta C Casado E Baron MG Retroperitoneal Castleman’s disease with colon cancer. A rare association Clin Transl Oncol 2008 10 238 40 18411199 24 Horio H Hjima T Sakaguchi K Kuwabara K Mediastinal Castleman disease associated with pulmonary carcinoma, mimicking N2 stage lung cancer Jpn J Thorac Cardiovasc Surg 2005 53 286 9 15952326 25 Ikari J Kojima M Tomita K A case of IgG4-related lung disease associated with multicentric Castleman’s disease and lung cancer Inter Med 2010 49 1287 91

PMC005xxxxxx/PMC5371718.txt

==== Front ACG Case Rep JACG Case Rep JcrjACG Case Reports Journal2326-3253American College of Gastroenterology crj.2017.5010.14309/crj.2017.50ImageEsophagusHeterotrophic Ossification of Intercostal Muscle Flap Causing Refractory Esophageal Stricture Alali et alHeterotrophic Ossification of IMFAlali Ali MBBCh, BAO, FRCPCWaschke Kevin MD, CM, FRCPC, FASGEDivision of Gastroenterology, McGill University Health Center, McGill University, Montreal, Quebec, CanadaCorrespondence: Kevin Waschke, Division of Gastroenterology, McGill University Health Center, McGill University, 1650 Cedar Ave, Montreal, Quebec, Canada H3A 0G4 (kevin.waschke@mcgill.ca).2017 29 3 2017 4 e5023 12 2016 17 2 2017 Copyright © Alali et al.2017This is an open-access article. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ ==== Body Case Report A 78-year-old man presented with a history of progressive solid food dysphagia for a few weeks. His history was notable for a recent diagnosis of gastroesophageal junction adenocarcinoma that was treated with Ivor-Lewis esophagectomy, with esophagogastric anastomosis and intercostal muscle flap (IMF) placement for leak and fistula management. Upper endoscopy showed stenosis of the esophageogastric anastamosis precluding the passage of a diagnostic gastroscope. There was no evidence of tumor recurrence. Balloon dilation was performed using a through-the-scope dilating balloon. The patient underwent multiple dilatations over the next 2 months (every 2 weeks), up to a diameter of 16.5 mm. However, the patient’s symptoms recurred rapidly, with evidence of restenosis seen on endoscopy (Figure 1). Computed tomography (CT) showed extrinsic compression and narrowing of the esophageogastric anastomosis caused by heterotrophic ossification of the pedicled IMF (Figure 2). A fully covered esophageal stent was inserted to manage his refractory esophageal stricture. The stent was removed after 6 weeks, followed by rapid recurrence of his symptoms. The patient opted for nonoperative management with serial dilations. Figure 1 Recurrence of the stricture with evidence of ulceration after dilation. Figure 2 Chest CT showing (A) heterotrophic ossification of the intercostal muscle flap with resultant esophageal luminal narrowing, and (B) heterotrophic ossification of the intercostal muscle flap. IMFs were initially described by Shenstone for the use in thoracic surgery in 1936.1 IMFs are used during bronchial or esophageal surgeries to reinforce the anastomosis.2 They have been used in cases of esophageal perforation related to benign or malignant etiologies.3 Heterotrophic ossification of IMF is a process of abnormal bony formation of these flaps. This entity has been reported as an incidental finding on imaging. However, bronchial obstruction with resultant pain and infection is well described.4 The characteristic radiological features seen on CT imaging include discontinuous linear calcification that appears either as a single stripe or two parallel stripes with an average thickness of 4 mm (range, 1–8 mm). The radiological changes can be seen as soon as 1 week postoperatively.5 Disclosures Author contributions: A. Alali wrote the manuscript. K. Waschke edited the manuscript and is the article guarantor. Financial disclosure: None to report. Informed consent was obtained for this case report. ==== Refs References 1. Shenstone NS The use of intercostal muscle in the closure of bronchial fistulae . Ann Surgery . 1936 ;104 (4 ):560 –71 . 2. Rendina EA , Venuta F , De Giacomo T , Ricci C Intercostal pedicle flap in tracheobronchial surgery . Ann Thorac Surg . 1996 ;62 (2 ):630 –1 . 3. Whyte RI , Iannettoni MD , Orringer MB Intrathoracic esophageal perforation: The merit of primary repair . J Thorac Cardiovasc Surg . 1995 ;109 (1 ):140 –6 .7815790 4. Prommegger R , Salzer GM Heterotopic ossification in pedicled intercostal muscle flaps causing clinical problems . J Thorac Cardiovasc Surg . 1998 ;115 (2 ):466 –7 .9475544 5. Kwek BH , Wain JC , Aquino SL The radiologic appearance of intercostal muscle flap . Ann Thorac Surg . 2004 ;78 (2 ):432 –5 .15276491

PMC005xxxxxx/PMC5371719.txt

==== Front ACG Case Rep JACG Case Rep JcrjACG Case Reports Journal2326-3253American College of Gastroenterology crj.2017.4610.14309/crj.2017.46Case ReportInflammatory Bowel DiseaseChronic Granulomatous Disease Mimicking Colonic Crohn’s Disease Successfully Treated with Infliximab Peixoto et alCGD Mimicking Crohn’s DiseasePeixoto Armando MD12Coelho Rosa MD12Maia Tiago MD3Sarmento António PhD4Magro Fernando PhD1256Macedo Guilherme PhD121 Department of Gastroenterology, Centro Hospitalar São João, Porto, Portugal2 WGO Porto Training Center, Porto, Portugal3 Department of Pathology, Centro Hospitalar de São João, Porto, Portugal4 Department of Infectious Diseases, Centro Hospitalar São João, Porto, Portugal5 Department of Pharmacology and Therapeutics, Faculty of Medicine, University of Porto, Porto, Portugal6 MedInUP, Center for Drug Discovery and Innovative Medicines, University of Porto, Porto, PortugalCorrespondence: Armando Peixoto, Alameda Prof. Hernâni Monteiro, 4200-319, Porto, Portugal (armandoafp5@gmail.com).2017 29 3 2017 4 e4613 7 2016 2 2 2017 Copyright © Peixoto et al.2017This is an open-access article. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/4.0/Chronic granulomatous disease (CGD) is a genetically induced disease caused by mutations in one of the components of the NADPH-oxidase in phagocytes, characterized by life-threatening bacterial and fungal infections and granuloma formation. Treatment includes prevention of infectious complications and immunomodulation. However, a standard strategy is not yet defined. The authors report an X-linked CGD female carrier who presented during adulthood with diarrhea and colorectal ulcers, with high impairment of quality of life. Induction with infliximab 5 mg/kg (weeks 0, 2, and 6) with infectious prophylaxis was initiated. She continued infliximab 5 mg/kg every 8 weeks with complete symptomatic response at 15 months. ==== Body Introduction Chronic granulomatous disease (CGD) is a rare primary immunodeficiency disease caused by a defect in a subunit of NADPH oxidase; 70% of cases are X-linked and associated with CYBB mutation, with defective production of gp91phox.1 This mutation leads to decreased macrophage capacity to eliminate bacteria and fungi. Female carriers can present with gastrointestinal (GI) symptoms similar those seen in inflammatory bowel disease.1 Treatment of symptomatic carriers includes prevention of infectious complications and immunomodulation, although a standard strategy is not defined.2 Case Report A 41-year-old woman presented with chronic diarrhea for 3 years associated with significant weight loss. An ileocolonoscopy showed multiple well-defined rectal ulcers (Figure 1). Biopsies revealed reactive hyperplasia of the superficial and cryptic epithelium with extensive ulceration translated by the presence of exudate, as well as epithelioid granulomas with nucleated giant cells (Figure 2). Histiocytes containing brown pigment in the cytoplasm were observed in the lamina propria. Biopsies from the colon showed mild chronic inflammatory changes with an epithelioid granuloma. Immunohistochemistry for cytomegalovirus, periodic acid–Schiff, and Ziehl-Neelsen stains were negative. Abdominopelvic computed tomography enterography showed only circumferential thickening of the rectum. Figure 1 Initial ileocolonoscopy showing multiple, well-defined rectal ulcers resembling Crohn’s disease. Figure 2 Rectal biopsies revealing reactive hyperplasia of the superficial and cryptic epithelium with extensive ulceration translated by the presence of exudate, and epithelioid granulomas with nucleated giant cells (arrow). After reviewing the patient’s medical history, we found that one of her children had died at the age of 9 years with a septic complication due to CGD. A cellular immunity and oxidative burst of granulocytes study showed a shortfall in the production of oxidative metabolites in 35% of granulocytes, therefore establishing the diagnosis of CGD carrier linked to chromosome X. A repeat colonoscopy was performed due to her high risk for recurrent infections. A rectal ulcer biopsy was positive for Nocardia through polymerase chain reaction (PCR). After excluding cerebral disease, she was started on cotrimoxazole for 3 months, followed by continuous weekly prophylaxis. Forty-two days after colonic biopsies were performed, a positive cultural result for mycobacterium tuberculosis (TB) was depicted. Standard anti-TB therapy was started, consisting of isoniazid, rifampicin, ethambutol, and pyrazinamide; symptoms persisted after 12 months of treatment. Three months after the end of antibiotics, a new colonoscopy revealed that the ulceration had extended to the proximal colonic segments. Histological evaluation showed similar histological findings to what was described initially, including pigmented histiocytes and epithelioid granulomas (Figure 3). Biopsies for Mycobacteria and Nocardia were negative; however biopsies were now positive for herpes simplex 1 PCR and Aspergillus PCR. There was no evidence of herpes viremia. Blood cultures for fungi were also negative. The patient was treated with valacyclovir and subsequently maintained prophylaxis with itraconazole. After controlling these infections, she continued to complain of diarrhea (>10 bowel movements per day) with further weight loss. The patient was reluctant to start systemic steroids, so a multidisciplinary meeting led to the initiation of infliximab 5 mg/kg with cotrimoxazole and itraconazole prophylaxis. Figure 3 Histological re-evaluation after anti-TB therapy showing similar histological findings, including pigmented histiocytes and epithelioid granulomas (arrow). After 15 months of infliximab 5 mg/kg every 8 weeks, she had decreased her bowel movements to 3 per day and gained 5 kg. Subsequent laboratory tests showed consecutively normal complete blood counts, increased serum albumin, and normal C-reactive protein. During this period, the patient did not have any infectious complications. As such, anti-tumor necrosis factor (anti-TNF) therapy will be maintained indefinitely under close surveillance for the appearance of complications, especially infectious complications. Discussion CGD is a rare, genetically heterogeneous disease occurring in 1:250,000 births and caused by mutations in any one of the five components of the NADPH oxidase in phagocytes.3 It is characterized by recurrent life-threatening bacterial and fungal infections and granuloma formation, and, contrary to our case, most patients are diagnosed before age 5 years. Although female carriers of X-linked CGD have been previously considered to be unaffected, it is increasingly recognized that they may suffer from problems similar to those experienced by CGD patients.1 It is known that female carriers generally do not have an increased rate of infections. Nonetheless, as with our patient, they may be more predisposed to have inflammatory manifestations associated with CGD.2 GI manifestations of CGD include abdominal pain, diarrhea, colitis, proctitis, strictures, fistulae, and obstruction. In a series of 140 CGD patients, 43% of X-linked and 11% of autosomal recessive CGD patients had GI manifestations. Because Crohn’s disease is much more frequent than CGD, it is important to identify distinctive histopathological patterns. However, knowledge is still scarce.4,5 Pigmented histiocytes within the lamina propria present in colonic biopsies from patients with CGD have been suggested as the major distinctive feature.4,5 Corticosteroids are the most commonly used therapy for inflammatory manifestations of CGD.6 However, immunosuppressors are usually are required for long-term maintenance. Several options are described in the literature, including messalazine, thalidomide, gamma-interferon, azathioprine, and anti-TNF agents, but none are formally recommended.7,8 In extreme cases, surgery and hematopoietic cell transplantation can also be considered, with the latter being potentially curative, although associated with a significant mortality.9 In addition to effective maintenance of clinical remission induced by corticosteroids, anti-TNF alpha agents can be effective in inducing clinical remission. However, an increased risk of serious infectious complications has been described compared with other conditions. So, if their use is to be considered, the intensification of antimicrobial prophylaxis is critical, as well as an aggressive treatment of infectious complications.8 There are several reports of invasive aspergillosis in adults with Crohn’s disease treated with infliximab. Because Crohn’s-like colitis can be the initial or sole manifestation of CGD in some cases, it is possible that these cases may actually represent undiagnosed CGD. In a cases series including 5 patients, infliximab was highly effective in the treatment of refractory CGD-associated colitis but was associated with serious bacterial and fungal infections, including 2 deaths.8 However, these were patients with the full spectrum of the disease, so in these cases careful attention should be given before starting these therapies because of the high risk of serious infections. There is not enough evidence of the use in symptomatic carriers to draw a conclusion on the risks and benefits. However, due to the unwillingness of our patient to start corticosteroids, and the need for rapid and effective induction of clinical response, we chose to start infliximab therapy under close surveillance for infectious complications. Disclosures Author contributions: A. Peixoto, R. Coelho, and F. Magro wrote and critically revised the manuscript. T. Maia, A. Sarmento, and G. Macedo revised the manuscript for important intellectual content. F. Magro is the article guarantor. Financial disclosure: None to report. Informed consent was obtained for this case report. ==== Refs References 1. Battersby AC , Cale AM , Goldblatt D , Gennery AR Clinical manifestations of disease in X-linked carriers of chronic granulomatous disease . J Clin Immunol . 2013 ;33 (8 ):1276 –84 .24078260 2. Wolach B , Scharf Y , Gavrieli R , de Boer M , Roos D Unusual late presentation of X-linked chronic granulomatous disease in an adult female with a somatic mosaic for a novel mutation in CYBB . Blood . 2005 ;105 (1 ):61 –6 .15308575 3. Roos D , de Boer M Molecular diagnosis of chronic granulomatous disease . Clin Exp Immunol . 2014 ;175 :139 –49 .24016250 4. Liu S , Russo PA , Baldassano RN , Sullivan KE CD-68 expression is markedly different in Crohn’s disease and the colitis associated with chronic granulomatous . Disease. Inflamm Bowel Dis . 2009 ;15 :1213 –7 .19229980 5. Mitomi H , Mikami T , Takahashi H , et al Colitis in chronic granulomatous disease resembling Crohn’s disease: Comparative analysis of CD-68 positive cells between two disease entities . Dig Dis Sci . 1999 ;44 :452 –6 .10063937 6. Quie PG , Belani KK Corticosteroids for chronic granulomatous disease . J Pediatr . 1987 ;111 :393 –4 .3625407 7. Noel N , Mahlaoui N , Blanche S , et al Efficacy and safety of thalidomide in patients with inflammatory manifestations of chronic granulomatous disease: A retrospective case series . J Allergy Clin Immunol . 2013 ;132 :997 –1000 .23791514 8. Uzel G , Orange JS , Poliak N , Marciano BE , Heller T , Hollan SM Complications of tumor necrosis factor-α blockade in chronic granulomatous disease-related colitis . Clin Infect Dis . 2010 ;51 (12 ):1429 –34 .21058909 9. Seger RA Hematopoietic stem cell transplantation for chronic granulomatous disease . Immunol Allergy Clin North Am . 2010 ;30 :195 –208 .20493396

PMC005xxxxxx/PMC5371720.txt

==== Front ACG Case Rep JACG Case Rep JcrjACG Case Reports Journal2326-3253American College of Gastroenterology crj.2017.4910.14309/crj.2017.49ImageEndoscopyPumpkin Seed Bezoar Causing Lower Gastrointestinal Bleeding Nehme et alPumpkin Seed Bezoar in Distal RectumNehme Fredy MDRowe Kyle MDNassif Imad MDDepartment of Internal Medicine, Kansas University School of Medicine, Wichita, KSCorrespondence: Fredy Nehme, 1010 N Kansas St, Wichita, KS 67214 (nehme.fredy@gmail.com).2017 29 3 2017 4 e4916 11 2016 17 2 2017 Copyright © Nehme et al.2017This is an open-access article. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ ==== Body Case Report A 38-year-old woman with no significant past medical history presented to the emergency department with a 1-day history of diffuse abdominal pain. She reported normal bowel movements up to 1 day prior and was unable to pass anything rectally since. There was no report of fever, nausea, vomiting, hematemesis, or melena. Physical exam demonstrated diffuse abdominal tenderness and normal bowel sounds. Laboratory evaluation was unremarkable, and a computed tomography scan of the abdomen and pelvis showed fecal impaction without evidence of obstruction (Figure 1). She was initially treated with tap water enemas without improvement. In the following 24 hours, the patient reported several episodes of hematochezia with a drop of 4.2 g/dL hemoglobin. Figure 1 Sagittal computed tomography showing fecal impaction in the rectosigmoid colon (arrow). Colonoscopy revealed a pumpkin seed bezoar impacted at the distal rectum (Figure 2). Removal of the phytobezoar was accomplished with multiple washings and digital removal under general anesthesia. Underneath the bezoar, a circumferential area of ulcerated mucosa was found, and a small bleeding mucosal tear was noted in the proximal ascending colon (Figure 3). The patient tolerated the procedure, and her diet was progressively advanced with significant symptomatic improvement and no further episodes of bleeding. Figure 2 Pumpkin seed bezoar at the distal rectum. Figure 3 (A) Large circumferential area of exudative ulcerated bleeding mucosa at the distal rectum. (B) Small area of non-exudative bleeding ulcerated mucosa in the proximal ascending colon. A phytobezoar, defined as a solid mass of indigestible food materials, is usually reported in the stomach. Rectal seed bezoars are a rare cause of fecal impaction, particularly in Western countries, and are likely related to dietary habits.1 Various types of seeds have been reported to cause fecal impaction, although only 2 cases to date were related to pumpkin seeds.2,3 This is the first reported case of a pumpkin seed bezoar complicated by GI bleeding. Stercoral ulceration is an ulcer secondary to pressure necrosis usually caused by fecal impaction, and can result in gastrointestinal (GI) bleeding or perforation.4 Some reports suggest development of stercoral ulceration by various foreign bodies.5 In our patient, the bezoar is likely to have formed in the upper GI tract and to have caused mucosal damage during its passage through the proximal colon. Currently there are no treatment guidelines for the management of rectal bezoars. Therapeutic options range from conservative modalities (laxatives, stool softeners) to endoscopic therapy (water jet flush, snare, retrieval basket, forceps) and surgical interventions in refractory cases. Disclosures Author contributions: F. Nehme and K. Rowe wrote the manuscript. I. Nassif revised the manuscript for intellectual content. F. Nehme is the article guarantor. Financial disclosure: None to report. Informed consent was obtained for this case report. ==== Refs References 1. Eitan A , Bickel A , Katz IM Fecal impaction in adults: Report of 30 cases of seed bezoars in the rectum . Dis Colon Rectum . 2006 ;49 (11 ):1768 –71 .17036204 2. Manne JR , Rangu VM , Motapothula UMR , Hall MC A crunching colon: Rectal bezoar caused by pumpkin seed consumption . Clin Med Res . 2012 ;10 (2 ):75 –7 .22031478 3. Mirza MS , Al-Wahibi K , Baloch S , Al-Qadhi H Rectal bezoars due to pumpkin seeds . Trop Doct . 2009 ;39 (1 ):54 –5 .19211433 4. Huang C-C , Wang I-F , Chiu H-H Lower gastrointestinal bleeding caused by stercoral ulcer . CMAJ . 2011 ;183 (2 ):E134. 21173065 5. Park S , Keum B , Jeen YT Stercoral ulcer due to chicken bones in rectum: Overtube used as a conduit for endoscopic removal . Clin Gastroenterol Hepatol . 2010 ;8 (3 ):A32.

PMC005xxxxxx/PMC5371721.txt

==== Front ACG Case Rep JACG Case Rep JcrjACG Case Reports Journal2326-3253American College of Gastroenterology crj.2017.4810.14309/crj.2017.48Case ReportLiverNovel Application of Extracorporeal Photopheresis as Treatment of Graft-versus-Host Disease Following Liver Transplantation Brown et alECP in Post-Transplant GVHDBrown Timothy J. MD1Gentry Cathy BS2Hammer Suntrea T. G. MD3Hwang Christine S. MD4Vusirikala Madhuri MD5Patel Prapti A. MD5Matevosyan Karén MD3Tujios Shannan R. MD6Mufti Arjmand R. MD6Collins Robert H. Jr., MD51 Department of Medicine, University of Texas Southwestern Medical Center, Dallas, TX2 University of Texas Southwestern Medical School, Dallas, TX3 Department of Pathology, University of Texas Southwestern Medical Center, Dallas, TX4 Department of Surgery, University of Texas Southwestern Medical Center, Dallas, TX5 Division of Hematology and Oncology, University of Texas Southwestern Medical Center, Dallas, TX6 Division of Digestive and Liver Diseases, University of Texas Southwestern Medical Center, Dallas, TXCorrespondence: Timothy J Brown, University of Texas Southwestern Medical Center, Department of Medicine, 5323 Harry Hines Blvd, Dallas, TX 75390 (Timothy.brown2@phhs.org).2017 29 3 2017 4 e4829 11 2016 17 2 2017 Copyright © Brown et al.2017This is an open-access article. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/4.0/A 48-year-old man with hepatitis C virus (HCV) cirrhosis complicated by hepatocellular carcinoma underwent liver transplantation. His course was complicated by fever, diarrhea, abdominal pain, and pancytopenia. He developed a diffuse erythematous rash, which progressed to erythroderma. Biopsies of the colon and skin were consistent with acute graft-versus-host disease. Donor-derived lymphocytes were present in the peripheral blood. The patient was treated with corticosteroids and cyclosporine; however, he had minimal response to intensive immunosuppressive therapy. Extracorporeal photopheresis was initiated as a salvage therapy. He had a dramatic response, and his rash, diarrhea, and pancytopenia resolved. He is maintained on minimal immunosuppression 24 months later. ==== Body Introduction Graft-versus-host disease (GVHD) after liver transplantation (LT) has been well described in the literature.1 Donor-derived T-lymphocytes attack the recipient’s skin, bone marrow, and gut, leading to rash, cytopenia, and diarrhea. The management of GVHD after LT is mostly derived from treatment of GVHD following hematopoietic stem cell transplantation (HSCT).2,3 Extracorporeal photopheresis (ECP) is effective in post-HSCT GVHD but has only rarely been used for solid organ transplant-associated GVHD, with only one reported survivor thus far.4-6 It has never been reported with GVHD post-LT. Case Report A 48-year-old man with hepatitis C virus (HCV) cirrhosis and hepatocellular carcinoma underwent LT from a 62-year-old woman. The donor human leukocyte antigen (HLA) type was A2,30; B18,44; Bw6,4; C5,–; DR1,17; DR52; DQ2,5. The recipient’s HLA type was A26,29; B44,49; Bw4,–; C7,16; DR1,7; DR53; DQ2,5. Induction immunosuppression was basiliximab followed by tacrolimus, mycophenolate mofetil, and prednisone. The early postoperative period was complicated by biopsy-confirmed mild acute cellular rejection, which resolved with steroids. Seventeen days after transplantation, the patient was admitted with fevers, diarrhea, and an erythematous, desquamating rash. He was pancytopenic with a white cell count 400/uL, hemoglobin 8.3 g/dL, and 65,000 platelets/uL. The rash progressed to involve the entire integument and was pronounced in the palms and soles. An infectious workup was negative. Liver function tests were normal, although ferritin was elevated to 27,232 ng/mL (reference range 30–450 ng/mL). A colon biopsy revealed scattered epithelial apoptosis with crypt destruction (Figure 1). Skin biopsy showed interface dermatitis with dyskeratotic keratinocytes, epidermal necrosis, and superficial perivascular lymphocytic infiltration. The bone-marrow biopsy had decreased cellularity with rare hemophagocytic histiocytes, but normal morphology. Bone-marrow chimerism analysis utilizing short tandem repeats revealed 4% donor DNA, 96% host DNA, and no third-party DNA. Fluorescence in situ hybridization of the marrow showed 1.5% donor-derived cells; 7% of the peripheral blood T-lymphocytes was of female origin. Figure 1 Photomicrograph depicting findings consistent with acute GVHD, including prominent crypt epithelial apoptosis present in the left half of the field and prominent crypt drop-out on the right half of the field (200x magnification). There was initial concern for secondary hemophagocytic lymphohistiocytosis, and given critical illness, the patient underwent plasma exchange. On the basis of the clinical picture, histology, and bone-marrow findings of donor-derived DNA, it became clear that acute GVHD was the predominant syndrome. The patient was treated with dexamethasone and intravenous immunoglobulins. Tacrolimus was changed to cyclosporine due to thrombocytopenia. He continued to deteriorate over 5 days, and ECP was initiated after obtaining informed consent for off-label use. There was clear improvement of his skin rash, and his fever resolved. Diarrhea, mucositis, and cytopenia resolved, and he underwent 32 sessions of ECP. Repeated peripheral blood chimerism studies 10 days after ECP showed elimination of peripheral female T-lymphocytes. ECP was continued for another 9 months, with a gradual increase in the interval between treatments with improved clinical status. The patient is currently alive and well on tacrolimus monotherapy with undetectable HCV 24 months after transplant. Discussion GVHD in LT is rare, but has a mortality rate of 85%.1,7,8 In LT, immunocompetent donor-derived lymphocytes undergo activation following exposure to recipient-derived antigens. Activated donor T-lymphocytes mediate an immune response against recipient tissue.7 The target tissues in GVHD following LT are bone marrow, skin, and the gut, with notable sparing of the liver.7 The pathogenesis of GVHD in LT occurs in three continuous phases. In the first phase, surgery induces a pro-inflammatory state where host macrophages release tumor necrosis factor α (TNFα) and IL-1, resulting in increased host antigen-presenting cell activity. In the second phase, donor-derived T-lymphocytes residing in the donor liver activate, stimulated by HLA/peptide complex interactions, resulting in IL-2 receptor expression and clonal expansion, ultimately leading to the release of pro-inflammatory cytokines IL-2 and IFN-γ.3,7 In the third phase, anti-host T-cells release granzyme and perforin, leading to further inflammation and promotion of GVHD.7,9 In LT, GVHD presents 1–8 weeks after transplant.1,7 In 15% of cases, the patient presents with findings only of a rash involving the palms and soles, with eventual bullous transformation and desquamation.6,7 GVHD resulting in multi-system organ failure has an 85% mortality rate. The natural history of post-LT GVHD is a relapsing-remitting pattern of diarrhea, rash, fever, and neutropenia, culminating in sepsis and death.7 A diagnosis of GVHD after LT is confirmed by biopsy of the affected tissues, which will demonstrate donor-derived lymphocytic infiltration in the appropriate clinicopathological setting.7,10 GVHD should be suspected if chimerism is present by polymerase chain reaction or HLA typing of lymphocytes in the peripheral blood with symptoms concerning for GVHD.1,6,8,10 Early recognition requires thoughtful synthesis of pathological, laboratory, and clinical data, and prompt treatment is essential.10 There are no guidelines for treatment of post-LT GVHD. Current therapy is borrowed from GVHD treatment following HSCT and consists of immunosuppression with high doses of corticosteroids or calcineurin inhibitors.7-10 However, this is frequently complicated by toxicity and infections.6-11 Salvage therapies in steroid-refractory GVHD after LT are mostly ineffective. Administration of OKT3 or anti-thymocyte globulin (ATG) produces a profound immunosuppression, but is not associated with remission and frequently results in fatal infections.7,10 Targeting T-lymphocytes with daclizumab and basiliximab is associated with remission of skin GVHD, but are usually inadequate in suppressing gut GVHD.7,8,10 Anti-TNFα agents are also not effective in treating GVHD and are associated with fungal infections.8,10 Decreasing immunosuppression following LT allows host lymphocytes to regain activity against donor lymphocytes with the risk of graft rejection.6,8,10 Lastly, T-lymphocyte elimination with ATG or irradiation has not proven to be worthwhile due to the unknown risk of impairing engraftment.1,7,12 ECP has been effective in GVHD following HSCT and is regarded as a second-line treatment, resulting in response rates as high as 80% and up to 50% long-term survival.11,13-15 Additionally, ECP has also been somewhat effective in mitigating rejection post-LT.16 In ECP, patients undergo leukapheresis and incubation with 8-methoxypsoralen, exposure to ultraviolet-A light, and reinfusion.9,14,17 Currently, investigations are underway into cryopreservation of apheresed white blood cells for patients unable to tolerate "classic" ECP; however, this method has not yet been utilized in LT.18 ECP-treated lymphocytes undergo apoptosis and absorption by dendritic cells and macrophages in the reticuloendothelial system, resulting in the secretion of tissue growth factor-β and IL-10, and promoting immunotolerance.9,17,19 Dendritic cells differentiate into type-2 dendritic cells, enhancing the anti-inflammatory response.17,20 Additionally, ECP induces a significant increase in circulating CD4+ CD25+ Treg cells, associated with a tolerogenic phenotype that is present for up to one year following treatment.11,16,17,21 Lastly, ECP preserves host responses to foreign antigens, avoiding the increased risk of infection due to immunosuppression.4,6,11,15,17,22,23 We present a case of refractory post-LT GVHD that responded to the novel application of ECP. The patient currently has no evidence of GVHD at 2 years post-transplant. ECP can potentially be used as a salvage therapy in patients with GVHD refractory to treatment following LT, although the optimal use of ECP requires continued study. Disclosures Author contributions: All authors wrote and edited the manuscript. STG Hammer provided the pathology images. RH Collins, Jr., is the article guarantor. Financial disclosure: None to report. Informed consent was obtained for this case report. ==== Refs References 1. Smith DM , Agura E , Netto G , et al Liver transplant-associated graft-versus-host disease . Transplantation . 2003 ;75 (1 ):118 –26 .12544883 2. Collins RH Jr. , Cooper B , Nikaein A , Klintmalm G , Fay JW Graft-versus-host disease in a liver transplant recipient . Ann Intern Med . 1992 ;116 (5 ):391 –2 .1736772 3. Collins RH Jr. , Anastasi J , Terstappen LW , et al Brief report: Donor-derived long-term multilineage hematopoiesis in a liver-transplant recipient . N Engl J Med . 1993 ;328 (11 ):762 –5 .8094887 4. Kaloyannidis P , Mallouri D The role of the extracorporeal photopheresis in the management of the graft-versus-host disease . Transfus Apher Sci . 2012 ;46 (2 ):211 –9 .22123355 5. Kitko CL , Braun T , Couriel DR , et al Combination therapy for graft-versus-host disease prophylaxis with etanercept an extracorporeal photopheresis: Results of a Phase II clinical trial . Biol Blood Marrow Transplant . 2016 ;22 (5 ):862 –8 .26551636 6. Houston BL , Yan M , Tinckam K , et al Extracorporeal photopheresis in solid organ transplant-associated acute graft-versus-host disease . Transfusion . 2016 ;56 (4 ):962 –9 .26892365 7. Taylor AL , Gibbs P , Bradley JA Acute graft versus host disease following liver transplantation: The enemy within . Am J Transplant . 2004 ;4 (4 ):466 –74 .15023138 8. Rogulj IM , Deeg J , Lee SJ Acute graft versus host disease after orthotopic liver transplantation . J Hematol Oncol . 2012 ;5 :50. 22889203 9. Kitko CL , Levine JE Extracorporeal photopheresis in prevention and treatment of acute GVHD . Transfus Apher Sci . 2015 ;52 (2 ):151 –6 .25748231 10. Perri R , Assi M , Talwalkar J , et al Graft vs. host disease after liver transplantation: A new approach is needed . Liver Transpl . 2007 ;13 (8 ):1092 –9 .17663410 11. Perotti C , Sniecinski I A concise review on extracorporeal photochemotherapy: Where we began and where we are now and where are we going! Transfus Apher Sci . 2015 ;52 (3 ):360 –8 .25910538 12. Korngold R , Sprent J Lethal graft-versus-host disease after bone marrow transplantation across minor histocompatibility barriers in mice. Prevention by removing mature T cells from marrow . J Exp Med . 1978 ;148 (6 ):1687 –98 .363972 13. Richter HI , Stege H , Ruzicka T , Soehngen D , Heyll A , Krutmann J Extracorporeal photopheresis in the treatment of acute graft-versus-host disease . J Am Acad Dermatol . 1997 ;36 (5 Pt 1 ):787 –9 .9146547 14. Edelson R , Berger C , Gasparro F , et al Treatment of cutaneous T-cell lymphoma by extracorporeal photochemotherapy. Preliminary results . N Engl J Med . 1987 ;316 (6 ):297 –303 .3543674 15. Malagola M , Cancelli V , Skert C , et al Extracorporeal photopheresis for treatment of acute and chronic graft versus host disease: An Italian multicentric retrospective analysis on 94 patients on behalf of the Gruppo Italiano Trapianto di Midollo Osseo . Transplantation . 2016 ;100 (12 ):e147 –55 .27861297 16. Urbani L , Mazzoni A , Catalano G , et al The use of extracorporeal photopheresis for allograft rejection in liver transplant recipients . Transplant Proc . 2004 ;36 (10 ):3068 –70 .15686696 17. Lamioni A , Parisi F , Isacchi G , et al The immunological effects of extracorporeal photopheresis unraveled: Induction of tolerogenic dendritic cells in vitro and regulatory T cells in vivo . Transplantation . 2005 ;79 (7 ):846 –50 .15818329 18. Pochon C , Reppel L , Halle P , et al Cryopreservation as a way to maintain extracorporeal photopheresis regimen for GvHD treatment while circumventing patient temporary inability to undergo apheresis . Bone Marrow Transplant . 2017 ;52 (1 ):167 –70 .27643864 19. Di Renzo M , Rubegni P , Sbano P , et al ECP-treated lymphocytes of chronic graft-versus-host disease patients undergo apoptosis which involves both the Fas/FasL system and the Bcl-2 protein family . Arch Dermatol Res . 2003 ;295 (5 ):175 –82 .12883827 20. Edelson RL Mechanistic insights into extracorporeal photochemotherapy: Efficient induction of monocyte-to-dendritic cell maturation . Transfus Apher Sci . 2014 ;50 (3 ):322 –9 .23978554 21. Bladon J , Taylor P Extracorporeal photopheresis normalizes some lymphocyte subsets (including T regulatory cells) in chronic graft-versus-host-disease . Ther Apher Dial . 2008 ;12 (4 ):311 –8 .18789119 22. Perfetti P , Carlier P , Strada P , et al Extracorporeal photopheresis for the treatment of steroid refractory acute GVHD . Bone Marrow Transplant . 2008 ;42 (9 ):609 –17 .18660840 23. Adamski J , Kinard T , Ipe T , Cooling L Extracorporeal photopheresis for the treatment of autoimmune diseases . Transfus Apher Sci . 2015 ;52 (2 ):171 –82 .25886694

PMC005xxxxxx/PMC5371722.txt

==== Front ACG Case Rep JACG Case Rep JcrjACG Case Reports Journal2326-3253American College of Gastroenterology crj.2017.4710.14309/crj.2017.47Case ReportStomachGastrointestinal Bleeding from Metastatic Prostate Adenocarcinoma to the Stomach Koop et alGI Bleeding from Metastatic Prostate CancerKoop Andree MD1Brauhmbhatt Bhaumik MBBS2Lewis Jason MD3Lewis Michele D. MD121 Department of Internal Medicine, Mayo Clinic, Jacksonville, FL2 Department of Gastroenterology and Hepatology, Mayo Clinic, Jacksonville, FL3 Department of Laboratory Medicine and Pathology, Mayo Clinic, Jacksonville, FLCorrespondence: Michele D. Lewis, 4500 San Pablo Rd S, Jacksonville, FL 32224 (Lewis.Michele@mayo.edu).2017 29 3 2017 4 e476 12 2016 17 2 2017 Copyright © Koop et al.2017This is an open-access article. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/4.0/We present a rare case of gastrointestinal (GI) bleeding associated with metastatic prostate adenocarcinoma to the stomach. Prostate cancer, which is the most common noncutaneous malignancy among men, rarely spreads to the stomach, with only 7 cases reported in the English literature. Symptoms may include abdominal pain, nausea, vomiting, and GI bleeding. Our patient was treated with epinephrine injection and bipolar cautery, but GI bleeding recurred 7 months later when he had worsening of his thrombocytopenia while using ibuprofen. ==== Body Introduction Hemorrhage is common in advanced cancer, occurring in up to 10% of patients with varying causes, including chemotherapy, radiotherapy, anticoagulants, surgery, disseminated intravascular coagulation, and abnormalities in platelet number and functioning.1,2 Malignant lesions, although frequently associated with oozing and chronic gastrointestinal (GI) blood loss, are an uncommon cause of acute upper GI bleeding, accounting for less than 5% of acute bleeding episodes.3,4 Case Report A 51-year-old man presented to the emergency department for coffee-ground emesis and melena. He had a history of castration-resistant prostate adenocarcinoma diagnosed 7 years earlier, with metastases to his thoracic spine, ribs, and left femur. He had received radiation to his thoracic spine and was on medical therapy with leuprolide and denosumab, a human monoclonal antibody against RANKL that inhibits bone destruction in patients with castration-resistant prostate cancer and bone metastases.5 A bone-marrow biopsy performed 2 months earlier for pancytopenia following treatment with cabazitaxel showed diffuse bone-marrow replacement with prostate adenocarcinoma. His medications included 20 mg omeprazole daily. He had a sleeve gastrectomy for obesity 10 years earlier. At presentation, his vital signs were stable, and he had pale conjunctiva. His white blood cell count was 1.0 x 109 cells/L, hemoglobin 7.6 g/dL, and platelet count 54 x 109/L. His prostate-specific antigen measured 2 weeks earlier was >4,500 ng/mL. Upper endoscopy revealed a mucosal nodule in the gastric body 1 cm in diameter with active bleeding (Figure 1). The nodule was biopsied for histology with a cold forceps and felt firm. The nodule was injected with epinephrine and cauterized with a bipolar probe, and hemostasis was obtained (Figure 2). Following upper endoscopy, the patient’s hemoglobin level remained stable, and he was discharged the next day with a scheduled follow-up with oncology. The biopsy revealed high-grade prostate adenocarcinoma (Figure 3). Figure 1 Endoscopy of a 1-cm nodule in the gastric body with active bleeding. Figure 2 The gastric nodule with resolution of bleeding after injection with epinephrine and cauterization with a bipolar probe. Figure 3 (A) Hematoxylin and eosin stain of the stomach biopsy demonstrating hyperchromatic, individual malignant cells infiltrating between benign-appearing gastric glands. (B) Immunohistochemical stain for prostate-specific antigen under 100x magnification highlighting the malignant cells (brown), while the benign gastric glands are unstained. The patient presented 7 months later with a second episode of coffee-ground emesis and melena while taking 800 mg ibuprofen daily for pain. His hemoglobin was 8.9 g/dL, and platelet count was 23 x 109/L. Disseminated intravascular coagulation was ruled out with a normal prothrombin time, activated partial thromboplastin time, and fibrinogen level with no evidence of fibrin monomers. His clinical course was complicated by the spread of prostate adenocarcinoma to the lungs with recurrent malignant pleural effusions and oxygen dependence. Given his recurrent bleeding and widespread metastases, he was treated supportively with platelet transfusion, intravenous proton pump inhibitor, and desmopressin with control of bleeding. Discussion Prostate cancer is the most common noncutaneous malignancy among men and generally has an indolent course. The lymph nodes and bone are the most common sites of metastatic disease, and the lungs and liver are the most common sites of visceral metastases.6 Metastatic prostate cancer to the stomach, as found in our patient, is rare. The most common malignancies to metastasize to the stomach include lung, breast, and esophageal cancer.7 In a post-mortem study of solid malignant tumors, the incidence of gastric metastases in patients with prostate cancer was estimated at 2.5%.7 Nausea, vomiting, abdominal pain, and GI bleeding are the most common presenting symptoms.8 A recent case report cited only 7 cases of metastatic prostate cancer to the stomach in the English literature, including itself.8 Six of the patients were known to have prostate cancer prior to the diagnosis of gastric metastases, and prostate cancer was diagnosed in one patient after upper endoscopy with biopsy and confirmation on immunohistochemical staining.8-13 Although rare, metastases to other sites in the GI tract, such as the small bowel and esophagus, have been reported.14,15 The exact mechanism of GI metastasis is unclear, but it may include lymphatic, hematogenous, and direct spread.8 Our patient was known to have widespread metastases of prostate cancer to his axial and appendicular skeleton, with no known visceral metastases at the time of his initial episode of GI bleeding. The median survival of patients with bone and visceral metastases is 14 months, compared to patients with isolated lymph node (43 months) or bone (24 months) involvement. 6 The patient’s initial presentation was concerning for peptic ulcer disease induced by nonsteroidal anti-inflammatory drug use. In patients with metastatic prostate cancer who present with symptoms of GI bleeding, gastric metastases should be considered as an etiology. Despite extensive literature on the etiology and management of upper GI bleeding in the general population, an understanding of the disease course and management in patients with cancer remains limited.16 Endoscopic therapy is generally effective for initial hemostasis of malignant lesions, but rebleeding is common and occurs in up to 80% of patients with poor survival.3,16 Endoscopic therapies include argon-plasma coagulation, laser, heater probe, and injection, although no gold-standard treatment has been established.16 The role of endoscopic ultrasound (EUS) in the treatment of GI bleeding is rapidly evolving and may be a therapeutic option in patients with severe nonvariceal GI bleeding including malignancies, who are unsuitable candidates for other interventions. In a study of 17 patients with severe refractory GI bleeding, 7 out of 8 patients with a malignant etiology had successful EUS-guided therapeutic intervention with no rebleeding at a median follow-up of 10 months. The patient with rebleeding had rectally invasive prostate cancer with continued daily bleeding despite a marked decrease in vessel flow on EUS after alcohol injection.17 Desmopressin, an analogue of vasopressin, has been successfully used in the prevention and treatment of bleeding associated with uremia, cirrhosis, and medication-induced platelet dysfunction, although there are no clear guidelines for use in these settings.18,19 Ibuprofen is associated with platelet dysfunction.20 Although not formally indicated, desmopressin was administered to our patient in the setting of severe GI bleeding and daily ibuprofen use. Disclosures Author contributions: All authors critically revised the manuscript for intellectual content and approved the final version. A. Koop and B. Brauhmbhatt wrote the manuscript. A. Koop and J. Lewis provided the images. A. Koop is the article guarantor. Financial disclosure: None to report. Informed consent was obtained for this case report. ==== Refs References 1. Pereira J , Phan T Management of bleeding in patients with advanced cancer . Oncologist . 2004 ;9 (5 ):561 –70 .15477642 2. Dutcher JP Hematologic abnormalities in patients with nonhematologic malignancies . Hematol Oncol Clin North Am . 1987 ;1 (2 ):281 –99 .3308824 3. Loftus EV , Alexander GL , Ahlquist DA , Balm RK Endoscopic treatment of major bleeding from advanced gastroduodenal malignant lesions . Mayo Clin Proc . 1994 ;69 (8 ):736 –40 .8035627 4. Savides TJ , Jensen DM , Cohen J , et al Severe upper gastrointestinal tumor bleeding: Endoscopic findings, treatment, and outcome . Endoscopy . 1996 ;28 (2 ):244 –8 .8739741 5. Fizazi K , Carducci M , Smith M , et al Denosumab versus zoledronic acid for treatment of bone metastases in men with castration-resistant prostate cancer: A randomised, double-blind study . Lancet . 2011 ;377 (9768 ):813 –22 .21353695 6. Gandaglia G , Karakiewicz PI , Briganti A , et al Impact of the site of metastases on survival in patients with metastatic prostate cancer . Eur Urol . 2015 ;68 (2 ):325 –34 .25108577 7. Oda I , Kondo H , Yamao T , et al Metastatic tumors to the stomach: Analysis of 54 patients diagnosed at endoscopy and 347 autopsy cases . Endoscopy . 2001 ;33 (6 ):507 –10 .11437044 8. Bhandari V , Pant S Carcinoma prostate with gastric metastasis: A rare case report . J Cancer Res Ther . 2015 ;11 (3 ):659. 9. Bilici A , Dikilitas M , Eryilmaz OT , Bagli BS , Selcukbiricik F Stomach metastasis in a patient with prostate cancer 4 years after the initial diagnosis: A case report and a literature review . Case Rep Oncol Med . 2012 ;2012 :292140. 23243531 10. Christoph F , Grünbaum M , Wolkers F , Müller M , Miller K Prostate cancer metastatic to the stomach . Urology . 2004 ;63 (4 ):778 –9 . 11. Holderman WH , Jacques JM , Blackstone MO , Brasitus TA Prostate cancer metastatic to the stomach. Clinical aspects and endoscopic diagnosis . J Clin Gastroenterol . 1992 ;14 (3 ):251 –4 .1564301 12. Hong KP , Lee SJ , Hong GS , Yoon H , Shim BS Prostate cancer metastasis to the stomach . Korean J Urol . 2010 ;51 (6 ):431 –3 .20577612 13. Onitilo AA , Engel JM , Resnick JM Prostate carcinoma metastatic to the stomach: Report of two cases and review of the literature . Clin Med Res . 2010 ;8 (1 ):18 –21 .20305145 14. Malhi-Chowla N , Wolfsen HC , Menke D , Woodward TA Prostate cancer metastasizing to the small bowel . J Clin Gastroenterol . 2001 ;32 (5 ):439 –40 .11319319 15. Nakamura T , Mohri H , Shimazaki M , et al Esophageal metastasis from prostate cancer: Diagnostic use of reverse transcriptase-polymerase chain reaction for prostate-specific antigen . J Gastroenterol . 1997 ;32 (2 ):236 –40 .9085174 16. Maluf-Filho F , Martins BC , de Lima MS , et al Etiology, endoscopic management and mortality of upper gastrointestinal bleeding in patients with cancer . United European Gastroenterol J . 2013 ;1 (1 ):60 –7 . 17. Law R , Fujii-Lau L , Wong Kee Song LM , et al Efficacy of endoscopic ultrasound-guided hemostatic interventions for resistant nonvariceal bleeding . Clin Gastroenterol Hepatol . 2015 ;13 (4 ):808 –12 .25245627 18. Mannucci PM Desmopressin (DDAVP) in the treatment of bleeding disorders: The first twenty years . Haemophilia . 2000 ;6 Suppl 1 :60 –7 .10982270 19. Ozgönenel B , Rajpurkar M , Lusher JM How do you treat bleeding disorders with desmopressin? Postgrad Med J . 2007 ;83 (977 ):159 –63 .17344569 20. Goldenberg NA , Jacobson L , Manco-Johnson MJ Brief communication: Duration of platelet dysfunction after a 7-day course of ibuprofen . Ann Intern Med . 2005 ;142 (7 ):506 –9 .15809462

PMC005xxxxxx/PMC5371723.txt

==== Front ACG Case Rep JACG Case Rep JcrjACG Case Reports Journal2326-3253American College of Gastroenterology crj.2017.5110.14309/crj.2017.51Case ReportEndoscopyEndoscopic Treatment of Bleeding Diversion Pouchitis with High-Concentration Dextrose Spray Nyabanga et alDiversion Pouchitis TreatmentNyabanga Custon T. MD1Shen Bo MD, FACG21 Cleveland Clinic Lerner College of Medicine, Case Western Reserve University, Cleveland, OH2 Center for Inflammatory Bowel Disease, Digestive Disease and Surgery Institute, The Cleveland Clinic Foundation, Cleveland, OHCorrespondence: Bo Shen, Center for Inflammatory Bowel Diseases, Digestive Disease and Surgery Institute-A31, Cleveland Clinic, 9500 Euclid Ave, Cleveland, OH 44195 (shenb@ccf.org).2017 29 3 2017 4 e5127 1 2016 9 2 2017 Copyright © Nyabanga et al.2017This is an open-access article. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/4.0/Surgical closure of stoma with the reestablishment of gut continuity is the only curative intervention available for inflammatory bowel disease patients with diversion pouchitis, proctitis, or colitis. For patients who are not candidates for surgical reestablishment of bowel continuity, the alternative nonsurgical approaches, such as topical therapy with mesalamine, corticosteroids, or short-chain fatty acids, have only shown modest efficacy. The management of massive bleeding from diversion pouchitis has not been described. We present a patient with ulcerative colitis with severe hematochezia and diffuse mucosal bleeding in a diverted ileal pouch, which was successfully treated with endoscopic spray of hypertonic glucose. ==== Body Introduction Despite advances in medical therapy in patients with inflammatory bowel disease (IBD), a significant number of patients with Crohn’s disease or ulcerative colitis (UC) eventually require bowel resection surgery. Restorative proctocolectomy with an ileal pouch-anal anastomosis (IPAA) has become the surgical treatment of choice for UC patients who require colectomy. During the construction of the ileal pouch, a protective diverting ileostomy is often required to promote the maturation and anastomosis of the newly built ileal pouch. In addition, temporary or permanent ileostomy may be created in patients with pouch failure, leaving the pouch body in situ. The diverted pouch can develop pouchitis with bleeding. Similar to diversion colitis, diversion pouchitis is an inflammatory disorder occurring in the ileal pouch resulting from the exclusion of the fecal stream and a subsequent lack of nutrients from luminal bacteria. Patients generally present with varying symptoms such as tenesmus, urgency, bloody or mucus discharge, and abdominal pain. Chronic inflammation, even when asymptomatic, can lead to the formation of severe strictures and accumulation of mucus, which may cause formation of symptomatic bezoars. The diagnosis of diversion pouchitis is based on clinical and surgical history and endoscopic and histologic inflammation. Though rare, some patients may present with profuse mucosal bleeding, and the management of such a presentation can be challenging.1 Common therapeutic approaches to gastrointestinal mucosal bleeding include irrigation with iced normal saline and saline with adrenaline enemas, thermocoagulation, endoclipping, and transanal suturing.1,2 These approaches can prove ineffective in cases of diffuse bleeding with no discernable source. Hypertonic glucose spray, which is used in the treatment of variceal and peptic ulcer-associated bleeding, has been shown to initiate hemostasis in radiation enteritis-associated mucosal bleeding.3,4 Case Presentation The patient was a 44-year-old man who presented to our Interventional IBD Unit with profuse bleeding from the anus. His past medical history was significant for UC complicated by colitis-associated low-grade dysplasia, status postlaparoscopic total proctocolectomy with 2-stage IPAA in 2008. His J-pouch surgery was complicated by pouchitis, cuffitis, and the leaks at the tip of the "J" and at the anastomosis. In 2012, a presacral sinus developed from the chronic anastomotic leak, which was successfully treated with needle-knife sinusotomy. However, the tip of the "J" leak persisted despite multiple sessions of endoscopic treatment with endoclips and an over-the-scope clipping system, necessitating diverting loop ileostomy. He later presented in June 2015 with profuse hematochezia resulting in an emergency room visit. An emergent pouchoscopy under conscious sedation revealed diffuse active bleeding from the diverted afferent limb and pouch body. The mucosa was edematous and coated with old and fresh blood, with active oozing. No single bleeding source was found. We believe that the bleeding stemmed from active oozing from severe diversion pouchitis (Figure 1). The mucosa was endoscopically sprayed with 150 mL 50% dextrose via a catheter, which resulted in immediate hemostasis (Video 1; Figure 2). After the dextrose administration, the patient reported complete resolution of hematochezia during the hospitalization. Follow-up pouchoscopy 2 weeks after the dextrose spray showed normal pouch mucosa with no evidence of bleeding or mucosal friability (Figure 3). The patient was closely followed up by clinical assessment for 6 months, when he underwent a pouch-redo surgery after the initial endoscopy therapy. The patient did not experience further episodes of recurrent bleeding during the 6-month follow-up. The patient was not on any anticoagulation medicines before or after the bleeding episodes. No prescribed medicines were given after the endoscopic therapy. Figure 1 Pouchoscopy showing diffuse active bleeding of the diverted pouch. Video 1. Endoscopic spraying of the pouch mucosa with 150 mL 50% dextrose via catheter, resulting in immediate hemostasis. http://s3.gi.org/media/links/shen_video.mp4Figure 2 (A) D50 spray in action, and (B) the appearance of the pouch mucosa immediately after treatment. Figure 3 Appearance of the pouch mucosa 2 weeks later. Discussion Diversion colitis and diversion proctitis are common in patients with fecal diversion. The common clinical presentations of diversion colitis/proctitis include pelvic pain, urgency, and rectal discharge of mucus or blood. A large quantity of spontaneous blood discharge is uncommon. The best treatment option is the reestablishment of gut continuity. Various agents have been used to treat diversion colitis/proctitis, including topical mesalamines and corticosteroids, and short-chain fatty acid (SCFA) enema, with various efficacies for mucosal inflammation. There are no published studies on the efficacy of those agents in controlling brisk bleeding. The incidence of diversion pouchitis is unknown. It appears more commonly in patients with underlying IBD. Nonsurgical approaches for the treatment of diversion pouchitis include the use of SCFA enemas, topical 5-aminosalicylic acids, and topical glucocorticoids. Outcomes from studies of the efficacy of these medical approaches have been conflicting, and only surgical reanastomosis with the reestablishment of gut continuity remains the curative approach. There are some unique aspects of diversion pouchitis. For example, SCFAs have been considered as the essential nutrients for colonic epithelial cells. The efficacy for the inflammation of pouch mucosa is not clear. We felt that medical therapy with such agents might not be effective in controlling active bleeding. The main presentation of our case was acute and profuse bleeding. The patient was reluctant to have the surgery due to multiple previous complications. Since there was no discernible spot source of bleeding on endoscopic exam, routine endoscopic treatment techniques, such as intralesional injection, heat probe, and argon plasma coagulation, were not feasible. We opted for endoscopic spray with 50% dextrose, and long-lasting hemostasis was achieved. It is thought that hypertonic glucose works thorough osmotic dehydration and sclerosant effects, inducing long-term mural necrosis and fibrotic obliteration of mucosal vessels.3,4 We present the first case of hypertonic glucose spray for the management of diffuse diversion ileal pouch bleeding. Glucose spray is safe and inexpensive, and it has a very low chance of causing transient hyperglycemia because there is no direct injection of the hypertonic solution into blood vessels. The approach has the potential to reduce bleeding recurrence and need for surgical interventions. Its superiority over other approaches will have to be explored through case series or randomized controlled trials. Disclosures Author contributions: CT Nyabanga drafted the initial version of the manuscript. B. Shen revised the manuscript for intellectual content and is the article guarantor. Financial disclosures: Dr. Bo Shen is supported by the Ed and Joey Story Endowed Chair. Informed consent was obtained for this case report. ==== Refs References 1. Gorgun E , Remzi FH Complications of ileoanal pouches . Clin Colon Rectal Surg . 2004 ;17 :43 –55 .20011284 2. Fazio VW , Ziv Y , Church JM , et al Ileal pouch-anal anastomoses complications and function in 1005 patients . Ann Surg . 1995 ;222 :120 –7 .7639579 3. Chang KY , Wu CS , Chen PC Prospective, randomized trial of hypertonic glucose water and sodium tetradecyl sulfate for gastric variceal bleeding in patients with advanced liver cirrhosis . Endoscopy . 1996 ;28 :481 –6 .8886633 4. Tian C , Mehta P , Shen B Endoscopic therapy of bleeding from radiation enteritis with hypertonic glucose spray . ACG Case Rep J . 2014 ;1 :181 –3 .26157869

PMC005xxxxxx/PMC5371724.txt

==== Front ACG Case Rep JACG Case Rep JcrjACG Case Reports Journal2326-3253American College of Gastroenterology crj.2017.4510.14309/crj.2017.45Letter from the EditorPerspectives in Colonoscopy Perforation During Gastroenterology Fellowship MachicadoColonoscopy Perforation During GI Fellowship Machicado Jorge D. MDUniversity of Pittsburgh Pittsburgh, PA Associate Editor ACG Case Reports JournalCorrespondence: Jorge D. Machicado, UPMC Presbyterian, M2, C-Wing, 200 Lothrop St, Pittsburgh, PA 15213 (machicadoj@upmc.edu).2017 29 3 2017 4 e45Copyright © Machicado.2017This is an open-access article. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ ==== Body Diagnostic and therapeutic colonoscopy represents one of the mandatory skillsets of gastroenterology (GI) training. This is typically done in a step-up fashion of learning basic (e.g., colonic navigation, prevention of looping), intermediate (e.g., forceps biopsy, basic polypectomy), and advanced techniques (e.g., large polyp resection).1 At any point during the training process, complications can develop, and colonic perforation still stands as the most disruptive complication for patients, families, and healthcare professionals, with reported mortality rates of up to 5%.2 While feared by trainees, colonoscopy perforations are rare and may not even occur during a 3-year GI fellowship. Several studies have shown that the risk of perforation is 1:1,400 for diagnostic and 1:1,000 for therapeutic colonoscopy.2 Therefore, a trainee who performs 400–500 colonoscopies has approximately 1:2 to 1:3 odds of encountering a colonic perforation during training. These estimates may be higher in trainees exposed to endoscopic mucosal resection (EMR) or submusosal dissection (ESD), which carry perforation risk of 1% and 5%, respectively.2 Limited endoscopic experience is a risk factor for colonoscopy perforation. Compared with GI physicians, colonoscopy performed by GI trainees or surgeons carries a higher risk of perforation.3,4 In early stages of training, this may be associated with mechanical damage (e.g., retroflexion in a small rectum, fixed angulation, excessive looping) or air insufflation. Later during fellowship, perforations are more likely to be caused by therapeutic interventions. Removal of difficult polyps (e.g., large polyps, cecal location) is associated with higher risk of perforation. During the first years in practice, gastroenterologists should be cautious and ask the opinion of a more experienced endoscopist before starting a possibly complex polypectomy. Selecting which polyps are removable and recognizing personal technique limitations are skills that every competent gastroenterologist should develop to reduce colonoscopy complications and assure complete polyp resection. Colonic perforations caused by mechanical injury are typically large and can be detected during the procedure. In contrast, perforations caused by endoscopic interventions are smaller in size and more frequently have a delayed presentation. When a perforation is found during the procedure, the first rule is not to panic. The gastroenterologist in charge should decide whether conservative, endoscopic, or surgical management is preferred. Calling a surgeon, an advanced endoscopist, and a senior gastroenterologist into the endoscopy suite as quickly as possible helps to make a multidisciplinary decision based on local expertise. In the last 20 years, there has been increasing experience with the use of different devices for endoscopic closure of colonic perforations. This ranges from the simple use of through-the-scope clips to more complicated methods, such as over-the-scope clips, endoscopic suturing, colonic stents, or glue.5 While these techniques are usually learned during advanced endoscopy training, any GI fellow should be familiar with the use of at least through-the-scope clips for perforation closure. Surgery should be avoided if possible, as this is associated with increased mortality, prolonged hospital stay, and postoperative morbidities.6 Fellows may encounter consults or patient calls for abdominal pain following a colonoscopy. This requires immediate clinical evaluation to rule out delayed perforation. In the presence of diffuse peritoneal signs or hemodynamic instability, surgery is typically indicated. If the pain is localized and the patient is stable, an upright X-ray or cross-sectional imaging must be obtained. If free air is not the cause, the pain could be explained by postpolypectomy coagulation syndrome or other reasons, and it likely can be managed conservatively. The presence of free air on imaging typically leads to surgery, but conservative management should be considered in select cases. Sometimes free air can be detected on imaging following EMR or ESD in the absence of true perforation.7 In these situations, pain may be transient or completely absent, and conservative management is preferred. Operator stress and burnout can follow a colonoscopy perforation.8 GI fellows are less used to morbid complications than trainees in surgical specialties, despite the mandatory training in internal medicine. For any fellow who encounters this complication, sharing the experience with other fellows, experienced faculties, mentors, or the program director can be helpful to reduce stress and to learn from others’ experiences with colonoscopy perforation. Fellows should present this complication in a morbidity and mortality conference to enrich the education of other fellows who may not encounter this during training. A discussion should be focused on the mechanism of the perforation, precipitating factors, treatment options, and future preventive strategies. ==== Refs References 1. Sedlack RE , Shami VM , Adler DG , et al Colonoscopy core curriculum . Gastrointest Endosc . 2012 ;76 :482 –90 .22898404 2. Panteris V , Haringsma J , Kuipers EJ Colonoscopy perforation rate, mechanisms and outcome: From diagnostic to therapeutic colonoscopy . Endoscopy . 2009 ;41 :941 –51 .19866393 3. Anderson ML , Pasha TM , Leighton JA Endoscopic perforation of the colon: Lessons from a 10-year study . Am J Gastroenterol . 2000 ;95 :3418 –22 .11151871 4. Bielawska B , Day AG , Lieberman DA , Hookey LC Risk factors for early colonoscopic perforation include non-gastroenterologist endoscopists: A multivariable analysis . Clin Gastroenterol Hepatol . 2014 ;12 :85 –92 .23891916 5. Thirumurthi S , Raju GS Management of polypectomy complications . Gastrointest Endosc Clin N Am . 2015 ;25 :335 –57 .25839689 6. Iqbal CW , Cullinane DC , Schiller HJ , Sawyer MD , Zietlow SP , Farley DR Surgical management and outcomes of 165 colonoscopic perforations from a single institution . Arch Surg . 2008 ;143 :701 –6 .18645114 7. Heerasing N , Dowling D , Alexander S Abdominal pain post endoscopic mucosal resection: Treat the patient not the CT scan . World J Gastrointest Endosc . 2013 ;5 :455 –6 .24044046 8. Keswani RN , Keefer L , Surawicz CM Burnout in gastroenterologists and how to prevent it . Gastroenterology . 2014 ;147 :11 –4 .24861640

PMC005xxxxxx/PMC5372015.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01357810.1136/bmjopen-2016-013578Diabetes and EndocrinologyResearch1506184317181725Influencing factors on compliance of timely visits among patients with proliferative diabetic retinopathy in southern China: a qualitative study Duan Fang 1Liu Yuhong 1Chen Xiang 1Congdon Nathan 1Zhang Jian 1Chen Qianyun 1Chen Lingling 1Chen Xi 2Zhang Xiulan 1Yu Chengpu 2Liu Yizhi 1 1 Zhongshan Ophthalmic Centre, State Key Laboratory of Ophthalmology, SunYat-sen University, Guangzhou, China 2 Centre of Medical Anthropology and Behavioural Health, School of Sociology and Anthropology, Sun Yat-sen University, Guangzhou, ChinaCorrespondence to Dr Yizhi Liu; yzliu62@yahoo.comYL and CY contributed equally to this work. 2017 27 3 2017 7 3 e01357828 7 2016 28 1 2017 6 3 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective To identify the reasons for low adherence among patients with diabetic retinopathy (DR) in southern China using a qualitative method. Methods Exploratory indepth interviews were conducted in 27 diabetic patients with proliferative diabetic retinopathy who required vitrectomy surgery at Zhongshan Ophthalmic Centre, Sun Yat-sen University, from March to August 2015. Qualitative data analysis and research software (ATLAS.ti7) was used for data processing and analysis. Results Factors influencing the occurrence of timely visits included lack of DR related knowledge, fear and worries about insulin, interactions between patients and society combined with the complexity of emotions and social culture, and the economic burden of treatment. Conclusions Although the reasons for low adherence involved social, emotional, cultural and economic factors, the key issue was the lack of awareness and knowledge of DR. Our findings have several practical implications for health policymakers and programme planners in China. compliancein-depth InterviewsQUALITATIVE RESEARCH ==== Body Strengths and limitations of this study This is the first study to use qualitative method to study adherence of diabetic retinopathy patients in China. Indepth interview of the patients and their family members simultaneously allowed us to understand the interaction between them. The study was conducted by anthropologists and ophthalmologists working together, which allowed us to investigate this complex issue from multiple angles. Our study was limited by using a hospital based selected sample. All selected participants developed proliferative diabetic retinopathy prior to the interview. Introduction The prevalence of diabetes mellitus (DM) in China has increased nearly 10-fold to 11.6% from 1980 to 2010.1 China has the largest number of people with DM in the world (113.9 million). However, less than a third have been diagnosed. Furthermore, 50.1% of the Chinese adult population may have pre-diabetes,1 an important risk factor for the development of overt DM. With the aging of the Chinese population, the number of patients with DM is expected to double from 2000 to 2030.2 Therefore, DM and its complications is presenting a major public health problem in China. Diabetic retinopathy (DR) is a leading cause of blindness worldwide,3 4 and a priority disease in the ‘VISION 2020’ initiative for the global elimination of preventable blindness. Population based research in the USA showed that the incidence of DR among patients with DM aged >40 years was 28.5%. Vision threatening DR accounts for 4.4% of cases and, of those, proliferative diabetic retinopathy (PDR) accounts for 1.5%.5 Nearly half of people with diabetes (43.1%) in a population based rural Chinese study in Handan had DR, while 6.3% had vision threatening DR.6 Two population based studies in Beijing identified the population prevalence of DR as 6.5% in 20017 and 24.7% in 2009.8 These two studies showed that the prevalence of DR in China is increasing. As the average lifespan of the Chinese population extends and the prevalence of DM increases, the prevalence of DR will also increase. Therefore, the urgent problem is how can we implement early DR interventions and reduce the corresponding loss of labour and incurred medical expenditures. It has been suggested in Chinese guidelines for the prevention and treatment of DR that patients with DM should receive a dilated fundus examination once a year.9 10 However, two-thirds of patients with DM in China have not undergone an examination for more than 1 year, and nearly half (43.2%) have never been examined.11 Many adults living with DM remain unaware of their ocular condition until their DR has progressed to a stage at which treatment is difficult.12 This phenomenon is more severe in rural areas. Only 54.5% of patients with DR who need panretinal photocoagulation have completed treatment.13 Currently, there is no comprehensive national diabetes prevention or screening system available in China, although the government is hoping to establish a chronic disease prevention service programme gradually; it will take some time to see its effects. It is important to identify the reasons for this very low adherence with the recommended treatment, to seek solutions to improve the situation and to reduce the burden of blindness due to DR. In previous questionnaire based studies of adherence with DR care in China, several factors were identified, such as lack of knowledge about the condition, older age, low income, inadequate education, lack of health insurance and transportation.11 14 15 However, quantitative studies have their limitations. Qualitative research is complementary to quantitative research methods and is used in small groups of patients to explore complex issues in greater depth, allowing findings and themes to emerge from the data rather than testing predetermined hypotheses.16 In our study, we conducted indepth interviews with patients with PDR who did not adhere to timely treatment and developed the need for vitrectomy. Our aim was to explore possible intervention methods to prevent the occurrence of DR, and to lay the foundations for future intervention studies. Research design and methods Topic guides were preliminarily formed based on previous literature11 17–19 and discussion with experienced ophthalmologists and anthropology experts; the final versions were developed after the first five patients. All indepth interviews were conducted with diabetic patients with PDR who required vitrectomy surgery at Zhongshan Ophthalmic Centre, Sun Yat-sen University. Inclusion criteria included PDR patients who were older than 18 years and had not received a dilated fundus examination for over 1 year before developing PDR. Ethics approval was obtained from the ethics committee of Zhongshan Ophthalmic Centre (2015MEKY067). After being informed about the objectives of the study, informed consent was obtained from all participants who were assured of confidentiality, de-identification of all data and adequate protection against release of confidential information. Recorded interviews of approximately half an hour in length were conducted by a professor (CY) of anthropology and an undergraduate student (XC) from the anthropology department, from May to August 2015. The interview methods have been described elsewhere.20 In brief, consecutive patients were interviewed with or without accompanying family members, according to patients' willingness. The audio recorded interview data were transcribed verbatim to capture the cultural concepts and nuances embedded in the language. The transcripts were reviewed to ensure accuracy (by LC and XC, with qualitative and ethnographic training). In the process of analysis, the code lists were compared to resolve any discrepancies (by FD and XC). Once agreement had been achieved, a coding framework was developed to capture key themes, and each coded theme was subjected to further analysis to identify subthemes and illustrative quotes. All of the above steps were performed using software for qualitative data processing and analysis (ATLAS.ti7). To protect family members' anonymity, unique identifiers were used. Results We screened 30 patients with PDR, and 27 agreed to participate in the interviews, including 26 patients with type 2 DM and 1 patient with type 1 DM. Of these, 20 patients were accompanied by family members during the interviews. The demographic characteristics of the 27 patients are shown in table 1. The frequency of the key words that emerged from the interviews are listed in table 2. Table 1 Demographic characteristic of the 27 diabetes mellitus patients with proliferative diabetic retinopathy Characteristic Value Male sex (n (%)) 13 (48.1) Age (years) (mean±SD, range) 53.1±11.1, 32–71 Married (n (%)) 25 (92.6) Duration of DM (years) (mean±SD, range) 10.9±7.6, 0.2–28 Occupation (n (%))  Professional (health, education, etc) 2 (7.4)  Semiskilled (office worker) 5 (18.5)  Unskilled 3 (11.1)  Shopkeeper/business 5 (18.5)  Agriculture 5 (18.5)  Retired 4 (14.8)  Unemployed 3 (11.1) DM, diabetes mellitus. Table 2 Frequency of the key words that emerged from the interviews Theme Frequency Lack of DR knowledge 26 Fear, worry and misunderstanding about the use of insulin 6 Interactions between patients and society  Family 18  Fellow patients 4  Working 3 Economic burden of DM and DR treatment 24 DM, diabetes mellitus; DR diabetic retinopathy. Knowledge of DR and need for information on DR From quantitative studies it has been found that patients with DR generally lack DR related knowledge,11 21 which is consistent with what we found in our interviews. Patient No 16 was a 26-year-old unmarried nurse at a township hospital and visited the doctor with her 60-year-old father. There was no history of DM in her family. She was found to have type 1 DM when she was 7 years old. When asked about how she viewed her DM complications, she said:“I did not know DR existed, and no one reminded me to undergo fundus examinations. I have looked up some knowledge on the complications of DM. However, I never thought that the complications would occur to my eyes; in my opinion, they generally occur to the heart and kidney, which affects one's life and must be attended to first. People may pay less attention to the eyes.” This patient was a medical worker, had some DM related knowledge and the possibility of some occurrence of complications, but knew nothing about eye complications. This suggests that DR is not familiar to patients and is not taken as seriously as other life threatening complications. Additionally, this patient, as a medical worker, had no knowledge of DR, which indicates that local medical workers might also lack sufficient knowledge on DR. We also found in our interviews that most patients knew they should regularly have their blood sugar monitored but did not know the potential eye complications and the importance of regular fundus examinations. Of the 27 patients in our study, only one patient (patient No 9) explicitly mentioned that her doctor told her about DR and suggested that she should receive an eye examination annually when she was diagnosed with DM. This suggests the need for patients with DM to improve their DR related knowledge and for the awareness and importance of DR to be strengthened. Because the patients lacked knowledge of DR and were unaware of the importance of periodic fundus examinations, they did not see their doctor when ocular symptoms occurred. They thought they may have been due to other common non-serious eye diseases, such as myopia or cataracts. For example, patient No 26, a university teacher in Guangdong, often needed to search for documents on the computer. He did not pay attention when his vision became blurry he as thought it was because he was too tired. The eyesight in one of his eyes worsened dramatically from 1.0 to 0.1 within 2 months. A 60-year-old patient (patient No 10) used a few eyedrops when his vision became cloudy. Later, his family thought it was a cataract and sent him to the hospital until his right eye deteriorated due to secondary glaucoma and he experienced headaches. A 32-year-old patient (patient No 11) thought that it was the progression of myopia that caused her blurred vision. Furthermore, three patients (patient Nos 2, 11 and 19) were first diagnosed with DM only after identifying the occurrence of serious PDR affecting their eyesight. These patients manifested successive symptoms of DR, but their lack of knowledge and awareness of DR resulted in PDR diagnoses when they visited the hospital. This not only leads to treatment difficulties but also affects the prognosis of the disease. Therefore, patients were recommended to seek medical help when noticing blurred vision. Fear, worry and misunderstanding about the use of insulin Insulin is the most effective hypoglycaemic agent and is widely used in the treatment of patients with DM to reduce their long term complications. We found in our study that some patients had major concerns, or even misunderstood the use of insulin. Patient No 27, a 60-year-old woman, demonstrated apparent fear when she gave her opinions on using insulin as a neighbour had died because of the frequent use of insulin.“There may be other reasons for this, but we still resist and dare not to use insulin”. The patient connected the death of a neighbour with the use of insulin, and had a great misunderstanding of the use of insulin, which affected her adherence to insulin use, and led to PDR. Patient No 26 refused to use insulin because of concerns about insulin dependence and only agreed to use insulin when his oral medicine had become completely ineffective. Now he thinks that his DR was mainly caused by the poor control of his blood sugar caused by the delay in the use of insulin.“If I had used insulin earlier, my DR may have been better”. Patient No 2, a 71-year-old, farmer, was diagnosed with DM when she went to the hospital because her vision was blurred. She said,“the doctor suggested that I use insulin when my oral medicine could not control my blood sugar. I worry about insulin dependence too, but there is no other way”. In addition, the patient worried about the inconvenience of insulin injections because she could not perform the injections on her own due to her poor eyesight. The main concern regarding the use of insulin was fear of insulin dependence and its side effects. These cases indicate that lack of insulin related knowledge in some patients resulted in insulin use disorders or poor adherence. These concerns are likely to affect patients' use of insulin, resulting in poor control of blood sugar.22 Intensive therapy with insulin in patients with type 1 diabetes was associated with a substantial reduction in the long term risk of ocular surgery.23 Therefore, it is necessary for medical staff to help patients objectively understand this so that patients' unnecessary concerns can be reduced and adherence can be improved, achieving appropriate control of blood sugar, and thereby reducing the occurrence and development of PDR. Interactions between patients and society involve the complexities of emotion and social culture, presenting a double-edged sword As part of social cultures, social relationships can directly affect people's lifestyles and psychological states. Social relationships consist of many types. Family relationships, working relationships and fellow patients’ relationships were mentioned most often by patients during the interviews. The influence of these relationships on the development of disease had two sides. Patient No 7, a 57-year-old man, enjoyed eating and drinking with friends in teahouses. His family began to supervise his life and diet after he was diagnosed with PDR. His wife (57 years old) told us“He used to eat sweets and fats. Now I do not allow him to eat these foods. If a man is not good, the whole family worries about him”. She always reminded her husband to control his diet. This case is an example of the positive influence of family support in the treatment of disease. The mother of patient No 16, the only patient with type 1 DM, stopped working and took care of patient No 16. The mother reduced the consumption of starchy foods, provided a reasonable mix of meat and vegetables, and increased the quantity of vegetables when taking care of her daughter's diet. The mother looked at many types of folk prescriptions to cure her daughter, including many Chinese herbal medicines which sometimes caused vomiting. The patient said that her mother's actions were useless, even foolish. However, she continued to try what her mother suggested as she understood that her mother was trying to cure her. This case provides an example of the complexity of family relationships and interactions. Such excessive family care may cause a psychological burden. Patient No 2, a 71-year-old woman, lived in the countryside with her husband. Every time the children returned home, she was busy taking care of them and enjoyed the rare family gatherings. She reluctantly talked about her DM, even when asked by her children. She said,“I can endure it as long as it is not too severe. I do not want to affect my children's work and life”. With the extension of the average lifespan and low birth rate, China has become an ageing society. Patient No 2's family represents an example of a lonely elderly family, which is very common. With this type of family structure, the children work outside throughout the year, cannot take good care of their sick parents and often cannot take their parents to hospital in time for examinations and treatment when an illness develops. Simultaneously, the elderly parents are unwilling to trouble their children and become too much of a burden. Therefore, this type of family interaction may delay the treatment of DM/DR, leading to worsening illnesses. In addition, some patients often have to engage in social activities because of work, especially if they are men. Some patients are unwilling to tell others about their illness and even“fear that others know they suffer from the disease” (patient No 25). Patient No 4, a 52-year-old salesman, often engaged in social activities.“Once, I was drunk and taken to the hospital. At that time, no one knew that I had DM, and the doctor transfused glucose. I felt more and more uncomfortable and almost lost my life”. After the ‘death’ experience, patient No 4 no longer concealed his illness, and discovered that his friends understood and considered his lifestyle and diet. This was a great comfort and source of encouragement for him. As a factor, the complexity of social relationships can affect the progress of the disease. When a person becomes sick, his social relations will expand to his group of fellow patients with the same diseases. Sometimes, the influence of fellow patients is not inferior to that of the family, but it is often overlooked. Patient No 8 initially needed to receive four laser treatment sessions according to his doctor's suggestion, but only one treatment was completed. He decided not to accept the remaining three laser treatment sessions after hearing his fellow patients mention that his eyesight would become worsen. Patient No 21, a businessman, felt thirsty and hungry and lost a substantial amount of weight over 10 years. His friend with DM said that he might have DM and gave him some medicine. Since then, he had been taking the medicine that his friend gave him and had not gone to the hospital for many years. These patients listened to the suggestions of their fellow patients, and missed or delayed their treatment. This suggests a lack of trust between patients and doctors. Interactions between patients and society involve the influences of emotion and social culture, including all types of social relationships. These influences can be positive or negative. Although people often overlook them, these factors greatly influence the attitudes of patients with DR towards treatment and are sometimes key factor influencing the patient's treatment decision. Economic burden of DM and DR treatment As a chronic disease, people with DM need lifelong medication. In our interviews, most patients had medical insurance but the reimbursement ratios were different by region, and therefore patients reported financial burdens of treating DM, in particular the costs of surgery for treating PDR. Patient No 10, a 60-year-old woman, was a primary schoolteacher in the town of Guangxi before retirement. She could receive a higher reimbursement ratio if she accepted treatment locally. However, she had to bear all of the treatment costs if she accepted PDR treatment in Guangzhou. Patient No 10's son said,“a lot of people in our hometown have DM, some have swollen feet and blind eyes, but they are not treated because they cannot afford it. Eventually, they end up not seeing a doctor”. Patient No 5, a 51-year-old woman , had DM for 16 years and was living in the countryside. She said that she could not work after she was diagnosed with DM, and her source of income was her husband's small salary.“For economic reasons, I take cheaper drugs, buy some when we have money, and do not take them when we do not have money”. She also said that farmers in her hometown had rural medical insurance but had to apply for special outpatient reimbursement. Their insurance could reimburse about RMB 100 a month. However, most drug costs and surgery costs could not be reimbursed. Patient No 14, a 52-year-old driver, was unemployed due to his poor eyesight. In the beginning, when his right eye first became blind, he thought that“one eye can see, and it doesn't matter if I provide self-care”. He thought that there were parents and children at home who needed to be taken care of, and he was the main source of income for his family. He did not go to the hospital for treatment until both of his eyes became blind.“It is a burden for my family. We don't have money and borrowed it from others”. These cases reflect the influence of economic status on patients' behaviour in terms of seeking medical advice. Their illnesses become severe as a result of the delay in treatment for economic reasons. The efficacy and cost effectiveness of early detection and treatment of DR are well established.24 It is necessary to emphasise the importance of early detection and treatment. Discussion In our interviews, we found that lack of knowledge of DR was common in patients with PDR and was present in over 90% (26/27) of patients. A previous DR questionnaire study found that a lack of DR knowledge was associated with poor patient adherence,11 14 15 21 but it did not explicitly illustrate why patients lacked this knowledge or the missing link that led to the lack of knowledge. We found that the lack of knowledge was due to various reasons. Firstly, DM patients knew that they needed to control their blood glucose but did not know that DM could affect their eyes. This suggests that patients did not obtain information on DR during the course of their diagnosis and treatment of DM. Secondly, patients did not take the initiative to acquire DR knowledge, suggesting that they lacked effective channels for obtaining DM related knowledge. Furthermore, patients who did not know about DR tended to associate their eye problems with other eye diseases that they were more familiar with, such as cataracts or myopia. This suggests that DR remains unfamiliar to the public, although the prevalence of DM is as high as 11.6%.1 In addition, we found that some DM patients had concerns about the use of insulin. Tong et al22 reported that psychosocial and emotional barriers were associated with poor glycaemic control in people with type 2 DM using insulin. Such psychological concerns leading to poor glycaemic control could promote the progression of DR. This also suggests that patients' lack of knowledge is not solely the lack of DR knowledge but may also include lack of knowledge about DM. Campaigns and education that improve awareness of DR may need to be strengthened. We found that there was a discrepancy between patients' recognition of symptoms and the characteristics of DR. The patients ‘see a doctor when they feel ill’. However, the gradual occurrence of DR can easily lead to a gap that results in patients treating DR lightly and ignoring the importance of regular examinations. For patients, the criterion used to judge whether they are healthy is whether their eyesight is affected. There are no obvious symptoms in the early stages of DR. Therefore, patients can easily form the idea that ‘DR is not that serious’ and thus miss their best treatment opportunity. Previous studies have suggested that patients with less advanced disease and those who do not believe that the disease poses a threat to their vision are less adherent to scheduled follow-up appointments.21 25 As an integral component of their life, social relationships affect patients' psychological and behavioural habits. The influence of relationships between patients and society inevitably interacts with the complexity of emotions and social culture and has both positive and negative roles in the patient’s disease course. Some families attach great importance to patients after they are diagnosed with PDR, repeatedly reminding them to receive treatment and taking good care of them. Patients may reverse bad habits and actively cooperate with treatment due to a sense of responsibility and ‘pressure’. However, excessive care can sometimes increase the psychological burden of patients. In other situations, elderly people living alone may delay their treatment because they do not want to affect their children's work and life, and some even conceal the severity of their illness. This demonstrates the emotional and cultural influences in treatment, indicating it is no longer an issue of simply lack of care. There is a contradiction between patients' consideration of their families and the concern for themselves. This also suggests that existing health insurance should provide medical assistance to lonely elderly people when they need it, instead of requiring them to rely on their children. In addition, we found that patients believed their fellow patients more than the doctors. This indicates that fellow patient relationships are an important factor that cannot be ignored. In conclusion, we believe that the treatment and care of patients with DR should be extended to their social relationships to develop more effective treatment measures. Economic conditions influenced patients' behaviour in receiving medical treatment. More than three-quarters of the interviewed patients with PDR reported experiencing economic difficulties and concerns. Some patients concealed their illness, or delayed or did not receive treatment because of economic problems. China is a developing country with a population of 1.3 billion. Per capita gross domestic product is way behind most developed countries, and at present there is no national level medical insurance service available. There is an established socioeconomic gradient in the development of type 2 DM and related complications, with people in lower socioeconomic groups at a higher risk of type 2 DM and its complications.26–28 Social and economic problems also manifest in accessibility of medical services. For example, some patients in remote areas chose the nearest hospital for treatment in the early stages but usually encountered limited medical resources and inadequate treatment. This suggests that these patients had difficulty accessing high quality medical care. All of these problems may result in patients with DR failing to be detected at an early stage and failing to be effectively treated. Early detection of DR is critical, as prompt treatment increases the likelihood of preserving vision.29 At present, China lacks an effective system for preventing DR, in contrast with studies in western developed countries. To provide effective care for DR with limited resources, interventions to promote adherence among patients are needed. Adherence is the product of the interaction between various factors. Our study demonstrated the factors influencing the occurrence of timely visits among patients with PDR: lack of DR related knowledge, fear and worries about insulin, the interaction between patients and society combined with the complexity of emotions and social culture, and the economic burden of treatment. Although it involves social, emotional, cultural and economic reasons, the key is lack of knowledge and awareness of DR. Our findings have several practical implications for health policymakers and programme planners in China. Our study had some limitations. Firstly, this study was conducted in a small sample of 27 patients, and so the potential generalisability of the findings may be limited. Secondly, 20 patients were accompanied by family members which may have caused insufficient expression of the patients' opinions. Hence further research, including a large scale quantitative study, should be conducted to better understand the problems identified in the current study. Contributors: FD was involved in question development, data collection and analysis, and manuscript preparation. YL participated in data analysis and manuscript preparation. XC was involved in question development and manuscript preparation. NC was involved in literature review and manuscript preparation. JZ was involved in question development and manuscript preparation. QC was involved in data collection and analysis. LC was involved in data collection and analysis. XC was involved in the interviews of the patients, data collection and analysis. XZ was involved in question development and manuscript preparation. CY was involved in interview of the patients, data analysis and manuscript preparation. YL was involved in question development and manuscript preparation. All authors read and approved the final manuscript. Funding: Supported by National Natural Science Foundation of China (81400381), Medical Scientific Research foundation of Guangdong Province, China (B2013140), Science and Technology Programme of Guangdong Province, China (2013B020400003) and Science and Technology Programme of Guangzhou, China (15570001). Competing interests: None declared. Ethics approval: The study was approved by the ethics committee of Zhongshan Ophthalmic Centre. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Xu Y , Wang L , He J Prevalence and control of diabetes in Chinese adults . JAMA 2013 ;310 :948 –59 . 10.1001/jama.2013.168118 24002281 2 Wild S , Roglic G , Green A Global prevalence of diabetes: estimates for the year 2000 and projections for 2030 . Diabetes Care 2004 ;27 :1047 –53 . 10.2337/diacare.27.5.1047 15111519 3 Congdon NG , Friedman DS , Lietman T Important causes of visual impairment in the world today . JAMA 2003 ;290 :2057 –60 . 10.1001/jama.290.15.2057 14559961 4 Cheung N , Mitchell P , Wong TY Diabetic retinopathy . Lancet 2010 ;376 :124 –36 . 10.1016/S0140-6736(09)62124-3 20580421 5 Zhang X , Saaddine JB , Chou CF Prevalence of diabetic retinopathy in the United States, 2005–2008 . JAMA 2010 ;304 :649 –56 . 10.1001/jama.2010.1111 20699456 6 Wang FH , Liang YB , Zhang F Prevalence of diabetic retinopathy in rural China: the Handan Eye Study . Ophthalmology 2009 ;116 :461 –7 . 10.1016/j.ophtha.2008.10.003 19168222 7 Xie X , Xu L , Yang H Frequency of diabetic retinopathy in the adult population in China: the Beijing Eye Study 2001 . Int Ophthalmol 2009 ;29 :485 –93 . 10.1007/s10792-008-9272-9 19221695 8 Xu J , Wei WB , Yuan MX Prevalence and risk factors for diabetic retinopathy: the Beijing Communities Diabetes Study 6 . Retina (Philadelphia, Pa) 2012 ;32 :322 –9 . 10.1097/IAE.0b013e31821c4252 9 Klein R , Klein BE , Moss SE The Wisconsin epidemiologic study of diabetic retinopathy. II. Prevalence and risk of diabetic retinopathy when age at diagnosis is less than 30 years . Arch Ophthalmol 1984 ;102 :520 –6 .6367724 10 Klein R , Klein BE , Moss SE The Wisconsin epidemiologic study of diabetic retinopathy. III. Prevalence and risk of diabetic retinopathy when age at diagnosis is 30 or more years . Arch Ophthalmol 1984 ;102 :527 –32 .6367725 11 Wang D , Ding X , He M Use of eye care services among diabetic patients in urban and rural China . Ophthalmology 2010 ;117 :1755 –62 . 10.1016/j.ophtha.2010.01.019 20471689 12 Bragge P , Gruen RL , Chau M Screening for presence or absence of diabetic retinopathy: a meta-analysis . Arch Ophthalmol 2011 ;129 :435 –44 . 10.1001/archophthalmol.2010.319 21149748 13 Hua W , Cao S , Cui J Analysis of reasons for noncompliance with laser treatment in patients of diabetic retinopathy . Can J Ophthalmol 2013 ;48 :88 –92 . 10.1016/j.jcjo.2012.09.012 23561600 14 Paksin-Hall A , Dent ML , Dong F Factors contributing to diabetes patients not receiving annual dilated eye examinations . Ophthalmic Epidemiol 2013 ;20 :281 –7 . 10.3109/09286586.2013.789531 23662945 15 Lee DJ , Kumar N , Feuer WJ Dilated eye examination screening guideline compliance among patients with diabetes without a diabetic retinopathy diagnosis: the role of geographic access . BMJ Open Diabetes Res Care 2014 ;2 :e000031 10.1136/bmjdrc-2014-000031 16 Britten N Qualitative interviews in medical research . BMJ 1995 ;311 :251 –3 . 10.1136/bmj.311.6999.251 7627048 17 Chun KM , Chesla CA Cultural issues in disease management for Chinese Americans with type 2 diabetes . Psychol Health 2004 ;19 :767 –85 . 10.1080/08870440410001722958 18 Whitmarsh I Medical schismogenics: compliance and “culture” in Caribbean biomedicine . Anthropol Q 2009 ;82 :447 –75 . 10.1353/anq.0.0060 19 Lewis K , Patel D , Yorston D A qualitative study in the United Kingdom of factors influencing attendance by patients with diabetes at ophthalmic outpatient clinics . Ophthalmic Epidemiol 2007 ;14 :375 –80 . 10.1080/09286580701375195 18161611 20 Yu C , Holroyd E , Cheng Y Institutional incentives for altruism: gifting blood in China . BMC Public Health 2013 ;13 :524 10.1186/1471-2458-13-524 23721212 21 Thompson AC , Thompson MO , Young DL Barriers to follow-up and strategies to improve adherence to appointments for care of chronic eye diseases . Invest Ophthalmol Vis Sci 2015 ;56 :4324 –31 . 10.1167/iovs.15-16444 26176869 22 Tong WT , Vethakkan SR , Ng CJ Why do some people with type 2 diabetes who are using insulin have poor glycaemic control? A qualitative study . BMJ Open 2015 ;5 :e006407 10.1136/bmjopen-2014-006407 23 DCCT/EDIC Research Group Aiello LP , Sun W , Das A Intensive diabetes therapy and ocular surgery in type 1 diabetes . N Engl J Med 2015 ;372 :1722 –33 . 10.1056/NEJMoa1409463 25923552 24 Vijan S , Hofer TP , Hayward RA Cost-utility analysis of screening intervals for diabetic retinopathy in patients with type 2 diabetes mellitus . JAMA 2000 ;283 :889 –96 . 10.1001/jama.283.7.889 10685713 25 Lee PP , Feldman ZW , Ostermann J Longitudinal rates of annual eye examinations of persons with diabetes and chronic eye diseases . Ophthalmology 2003 ;110 :1952 –9 . 10.1016/S0161-6420(03)00817-0 14522771 26 Robbins JM , Vaccarino V , Zhang H Socioeconomic status and diagnosed diabetes incidence . Diabetes Res Clin Pract 2005 ;68 :230 –6 . 10.1016/j.diabres.2004.09.007 15936465 27 Williams ED , Tapp RJ , Magliano DJ Health behaviours, socioeconomic status and diabetes incidence: The Australian Diabetes Obesity and Lifestyle Study (AusDiab) . Diabetologia 2010 ;53 :2538 –45 . 10.1007/s00125-010-1888-4 20740271 28 Chaturvedi N , Stephenson JM , Fuller JH The relationship between socioeconomic status and diabetes control and complications in the EURODIAB IDDM Complications Study . Diabetes Care 1996 ;19 :423 –30 . 10.2337/diacare.19.5.423 8732703 29 Antonetti DA , Klein R , Gardner TW Diabetic retinopathy . N Engl J Med 2012 ;366 :1227 –39 . 10.1056/NEJMra1005073 22455417

PMC005xxxxxx/PMC5372016.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01452810.1136/bmjopen-2016-014528Research MethodsProtocol150617301842171917231704Understanding the acceptability and adherence to paediatric antiretroviral treatment in the new formulation of pellets (LPV/r): the protocol of a realist evaluation http://orcid.org/0000-0002-5318-8815Nebot Giralt Ariadna 1Nöstlinger Christiana 1Lee Janice 2Salami Olawale 3Lallemant Marc 2Ouma Onyango 4Nyamongo Isaac 4Marchal Bruno 1 1 Department of Public Health, Institute of Tropical Medicine, Antwerp, Belgium 2 Drugs for Neglected Diseases Initiative, Geneva, Switzerland 3 Drugs for Neglected Diseases Initiative, Nairobi, Kenya 4 Institute of Anthropology, Gender and African Studies, University of Nairobi, Nairobi, KenyaCorrespondence to Ariadna Nebot; anebot@itg.be2017 29 3 2017 7 3 e01452829 9 2016 27 1 2017 2 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Background Improving access to paediatric HIV treatment requires both large-scale treatment programmes and medication that is adapted to infants and children's needs. The WHO recommends lopinavir/ritonavir as first-line antiretroviral therapy for all HIV-infected children younger than 3 years. There is currently little evidence on the acceptability of, and adherence to, a formulation of this combination treatment if given in the form of pellets. This protocol presents how we will carry a realist evaluation to assess the factors that contribute to the acceptability and adherence to the new pellets formulation in 3 hospitals in Kenya. Methods We structured the protocol along the realist evaluation cycle following 4 steps: (1) the initial programme theory, (2) the study design, (3) the data collection methods and (4) the data analysis plan. Theories of behavioural sciences were reviewed for frames that could provide insights into how using such new formulations may contribute to better acceptability and adherence. Ethics and dissemination This study was approved by the Institutional Review Board of the Institute of Tropical Medicine, the Ethical Committee of the University Hospital Antwerp and the Kenyatta National Hospital/University of Nairobi Ethics and Research Committee. We aim to disseminate the findings through international conferences and peer-reviewed journals and to share them with Drugs for Neglected Diseases initiative's (DNDi) programme managers and with the Kenyan healthcare providers. Discussion In developing this study, we encountered some challenges. First, methods to measure the acceptability of any formulation and adherence to it are not standardised. The second challenge is common in realist evaluation and relates to how to choose from different potentially interesting theoretical frameworks. We identified relevant and empirically tested theories from behavioural science that may be helpful in our study. We will test them in 3 settings by exploring the multilevel factors that influence acceptability and adherence of this new paediatric Antiretroviral (ARV) formulation. Paediatric antiretroviral treatmentrealist evaluationadherenceacceptabilityKenya ==== Body Strengths and limitations of this study Considering that there are no standardised methods to measure the acceptability and the adherence of any formulation, the proposal to adopt the realist evaluation approach for unpacking the black box between prescription of the drug and the actual long-term adherence is a major strength of this study. The lack of standardised methods to measure acceptability and adherence may be seen as a limitation of our study. Given the time constraints of a 1-year study, a complete literature review and full testing of partially contradicting hypotheses was not possible. Observations at home may suffer from selection bias. Background The introduction of antiretroviral therapy (ART) in children has substantially reduced the risk of mortality and morbidity and has transformed the prognosis of HIV infection from a fatal disease into a manageable chronic illness.1–4 However, children living with HIV continue to experience persistent diagnostic challenges and treatment gaps. Globally, there are 2.6 million children under 15 years of age living with HIV and only one in three are on treatment. During 2014 alone, 220 000 children were newly infected with HIV (almost 90% of them in sub-Saharan Africa). Another 150 000 children died worldwide because of HIV, the vast majority before the age of 2, and many without having been diagnosed.5 Most of these deaths could have been avoided if the diagnosis had been made in the first weeks of infection and immediate access to treatment was ensured.6 The 2010 WHO guidelines recommended immediate initiation of ART in all children under 2 years of age regardless of immune or clinical status.7 In 2013, the WHO guidelines re-emphasised the immediate initiation of ART and recommended it for all children below 5 years.8 The WHO recognised the challenge related to the limited availability of drugs in appropriate formulations for infants and toddlers, the long-term toxicities of ARV drugs, and the difficulty with adherence. It is estimated that about half of the 1.2 million children who are exposed to HIV were tested before the age of 2 months, a proportion that is stable since 2014. More progress was made with treatment. Almost half of the HIV-infected children under 15 years of age are now on treatment, more than double the number compared with 2009.9 In Kenya, according to the UNAIDS database, 41% of children (aged 0–14 years) living with HIV had access to ART.10 Improving access to paediatric HIV treatment requires both large-scale treatment programmes and medication that is adapted to the needs of infants and children. Existing paediatric ART typically combines nevirapine (NVP), a non-nucleoside reverse transcriptase inhibitor (NNRTI), with two other nucleoside reverse transcriptase inhibitors (NRTIs): lamivudine (3TC) plus either abacavir (ABC) or zidovudine or stavudine (d4T). Young children with HIV who are exposed to NNRTIs used for prevention of mother to child transmission (PMTCT) may have demonstrable viral resistance in more than half of the cases,11 which compromises the response to NVP-containing first-line ART.12 13 For this reason, the revised 2013 WHO guidelines recommend the use of lopinavir/ritonavir (LPV/r) as first-line ART for all children infected with HIV younger than 3 years (36 months) of age, regardless of NNRTI exposure.14 While accumulated evidence demonstrates its superior clinical efficacy over NVP, regardless of the child's prior exposure to ART for PMTCT,15 16 LPV/r is still not widely used because of its liquid pharmaceutical formulation. The LPV/r syrup has a high alcohol and solvent content, and a bitter taste; it is therefore difficult to administer. Furthermore, it requires cold chain storage and is prohibitively expensive for many countries. In order to improve the administration of LPV/r as well as its acceptability, Cipla Pharmaceuticals developed mini ‘melt’ tablets in the form of oral pellets. This formulation has clear advantages to syrup in logistical terms (easy storage and transport) and acceptability (it can be administered to children who cannot swallow tablets and has a better taste).13 The LPV/r pellets received tentative approval by the US Food and Drug Administration (FDA) in 2015. In 2011, the Drugs for Neglected Diseases initiative (DNDi) created a paediatric HIV programme with the commitment to make the LPV/r pellets immediately available for treatment of children infected with HIV in resource-poor settings, while working with Cipla to improve the formulation, assembling in a single-unit LPV/r taste-masked granules and dual NRTI powder. A randomised controlled trial, the LIVING study,i has been set up in three hospitals in Kenya (Nairobi and Kisumu) with the primary objective of evaluating the effectiveness of LPV/r pellets in addition to the ABC/3TC or AZT/3TC paediatric fixed dose combination tablet under routine treatment conditions in HIV-infected infants and young children who cannot swallow tablets. The study's secondary objectives are to document the safety of LPV/r pellets and associated dual NRTIs, to assess the population pharmacokinetics of LPV/r and dual NRTIs in pellet form, to measure the adherence and to evaluate the acceptability of the pellets and their ease of use by the caregiver and healthcare workers. Initial observations by LIVING researchers indicated a differential uptake of the new formulation by the caregivers and differences in adherence to the treatment.17 Adherence to ART is an important determinant of treatment outcome. It could be argued that effective antiretroviral treatment made HIV a chronic disease, and therefore issues affecting children living with HIV are in essence similar to those affecting children with chronic diseases in general, in terms of ‘living with the disease for always' and requiring a chronic treatment. However, HIV is quite different compared with other chronic diseases in that it is multigenerational, affects primarily vulnerable populations, and has a unique social stigma attached to its transmission modes.18 HIV also requires almost perfect levels of adherence to achieve long-lasting viral suppression, and suboptimal adherence to ART is the most common cause of virological failure.19 Non-adherence to ART, which can take many different forms, can have important public health implications, such as treatment failure, promoting viral mutations and drug resistance.20 Acceptability of drug formulations by caregivers and children is an important determinant of long-term adherence. Acceptability may be viewed as a precondition for adherence, but since many intertwined contextual factors may interfere, it does not lead automatically to adherence. These unique HIV-related characteristics make adherence behaviour in infants and children quite complex. To unpack the black box between prescription of the drug and the actual long-term adherence, we developed a protocol for a study that adopts the realist evaluation (RE) approach. The study is a substudy to the LIVING study, and conducted in collaboration between DNDi, the Institute of Tropical Medicine (ITM), Antwerp and the Institute of Anthropology, Gender and African Studies in Nairobi (IAGAS), University of Nairobi. In this paper, we present the protocol of the RE study, labelled ‘RE-LIVING’, which aims at better understanding the factors that contribute to the acceptability and adherence to the new paediatric ART formulation (LPV/r 40/10 mg, oral pellets) in the three Kenyan settings where the LIVING study is currently enrolling study participants. As mentioned before, acceptability may be perceived as a precondition for adherence but it does not lead automatically to adherence. In addition, adherence to ART is an important determinant of treatment outcome. Therefore, the RE-LIVING main findings may contribute, in the short term, to better understand the factors influencing the adherence to oral pellets introduced by the LIVING study and in the Kenyan-specific context. In the long term, these findings can contribute to better manage some of the adherence-related issues and therefore to have better acceptability to the paediatric treatment as well as to get better treatment outcomes. Methods The objective of the RE-LIVING study is to assess the factors that contribute to the acceptability of and adherence to the new paediatric ART formulation LPV/r 40/10 mg pellets in real-life settings. The study sets out to develop a good understanding of the acceptability of the new formulation, to assess the factors that contribute to correctly administering the new formulation when starting the treatment (ie, initiation to adherence) and to assess the factors and the conditions that enable caregivers to maintain adherence to the treatment over time. For this substudy, we adopted the RE approach, which was developed by Pawson and Tilley21 and based on scientific realism.22 23 RE belongs to the school of theory-driven inquiry. Pawson and Tilley argue that evaluation and research need to answer the questions ‘What works in which conditions for whom?’, rather than merely ‘Does it work?’ in order to gather useful data for decision-makers. RE considers that interventions work (or do not) because actors take up what is provided by the intervention (or not). The interaction between ‘intervention’ and ‘actors’ in specific ‘contexts’ triggers ‘mechanisms’ that cause ‘outcomes’ to occur. RE is theory-guided, starting and ending with middle range theory, often called ‘programme theory’.24 The latter should be considered as a detailed hypothesis that describes how the intervention is expected to lead to its effect and in which conditions. This hypothesis is then tested through empirical research. An RE ends by refining the initial programme theory, setting the scene for a new round of research.21–23 25 Figure 1 presents the realist research cycle, along which we structured our protocol. Figure 1 The realist cycle. Initial observations by LIVING researchers indicated a differential uptake of the new formulation by the caregivers and differences in adherence to the treatment among the three Kenyan settings where the study was implemented. RE will allow us to identify explanations in the form of context-mechanism-outcome (CMO) configurations that explain these differences. Step 1: formulating the initial programme theory An RE starts with eliciting the initial programme theory, which will guide further choices of design, and data collection and data analysis methods. A number of sources can be used to elicit the programme theory (figure 2). Figure 2 Eliciting the programme theory. The first source of information includes the programme documents, plans and logical frameworks of the intervention under study. In our case, we have reviewed the protocol of the ongoing LIVING study in order to identify the assumptions of the designers on how the new formulation would lead to better adherence. A second source of information consists of interviews and discussions with designers, funders and/or implementers of the programme and on-site observations can be used to obtain additional information. In our case, we engaged with DNDi staff at the Geneva headquarters and researchers involved in the LIVING study. DNDi staff identified a number of hypothetical factors influencing adherence to the new formulation. These included easier transport and storage and reduced conspicuousness of pellets compared with syrup formulation, less problems with swallowing because of the less bitter taste, etc. Reviewing the literature is the third way to develop the initial programme theory, aiming at assessing findings of previous evaluations or research studies, and the state of the art in general. First, we examined papers presenting clinical trials that looked at the effectiveness and acceptability of the new formulation (for instance, the Cher trial in South Africa,26 the Promote paediatric trial in Uganda27 or the Chapas-2 study).13 Second, we explored theories of behavioural sciences for concepts and frameworks that could provide insights into how using such new formulations may contribute to better acceptability and adherence. While we did not carry out any exploratory empirical research to elicit the initial programme theory, reviewing the available evidence shows that a complex set of psychosocial, sociocultural and economic factors, which are often intertwined and mutually influencing each other, determine whether a new medication or treatment will be accepted and adhered to. This has been shown, for instance, with areas in early infant testing for HIV and uptake of HIV care.28 This resulted in a categorisation of factors that influence adherence, including caregiver-related factors, health services and system factors, and sociocultural and community-related factors (box 1). Box 1 Multilevel factors related to adherence to treatment Caregiver-related factors We identified a number of studies investigating the role and impact of the caregiver on children's adherence. There is evidence that the quality of the caregiver–child relationships, including family functioning,29 influences adherence. This relationship has been shown to be moderated through emotional bonding and the ability to fully care for a child.30 There is no clear evidence if and how a parent's biological relationship affects adherence. Having one person who feels responsible to oversee the child's treatment is likely to result in better adherence than having multiple caregivers.31 Whoever the caregiver, it has been shown that caregivers' personal beliefs about HIV and the treatment, caregivers' education and antiretroviral therapy (ART) comprehension as well as their psychosocial functioning29 are crucial for motivation to support the child. Caregivers who accepted the child's infection were more likely to internalise the information given and developed a stronger motivation for adherence support.32 In the case of vertical transmission, how mothers deal with their own HIV infection may have an influence, especially if they have feelings of guilt or emotional and mental health problems. Caregivers who are HIV positive and on ART may draw on their own experience when supporting the adherence of their children. A systematic review of 38 quantitative studies found substantial empirical evidence on the relationship between perceived HIV-related stigma and HIV medication adherence. Proposed mechanisms included enhanced vulnerability to mental health conditions, reduction in levels of self-efficacy due to exposure to stigma and concerns about unintended HIV disclosure.33 We found that one of the most widely used behavioural theories to predict adherence outcomes is the information-motivation-behavioural skills (IMB) model.34 It has been applied to various health promotion behaviours, including HIV prevention,35–38 antiretroviral medication adherence39–41 and self-breast examinations.42 It has been validated in resource-rich and resource-limited settings.43–45 While originally developed to predict HIV prevention behaviour, this model can serve as an explanatory framework to explain the caregivers' individual motivation to adhere to the treatment guidelines. In a nutshell, the IMB model stipulates that individuals who have correct information, sufficient motivation and behavioural skills will be more likely to achieve good adherence. It has been shown that information and motivation influence both directly and indirectly the behavioural outcome and that behavioural skills are an important mediator towards this goal. One central psychological construct in this theory is self-efficacy, that is, the belief to be able to perform the specific behaviour. Self-efficacy has been shown to have great predictive value for behaviour change. Introducing skills as a behavioural construct is important because it constitutes, contrary to personality factors or societal macrofactors, an immediately modifiable factor in interventions that can be addressed through, for instance, patient education and counselling. Health services and system factors Easy geographic and financial access to health services may lead to a better ART adherence outcome. The lack of coordination across services has been identified in previous studies as a barrier to timely paediatric ART initiation in low-income and middle-income countries.46 47 Failure to diagnose HIV in pregnancy, to provide prevention of mother to child transmission (PMTCT) services or to follow-up the HIV-exposed infant represent missed chances for prevention of HIV transmission, infant diagnosis and linkage to care. Previous research has shown that integrating antenatal services, PMTCT, early infant diagnosis and paediatric HIV care greatly improve outcomes for HIV-infected infants in resource-limited settings.48–50 Finally, good performance of health services may have a positive impact on the user's confidence in the services and trust in their personnel. Sociocultural factors Individual-level theories intrinsically focus on microlevel factors, with no specific linkage to macrolevel factors on the structural level.51 However, resources, cultural and social factors like HIV-related stigma, social control and community health beliefs, as well as structural factors such as food shortage and poverty,29 may affect adherence behaviour directly or indirectly. Food insecurity can contribute to enhanced HIV transmission, as well as to antiretroviral therapy non-adherence, treatment interruptions and missed clinic visits, all of which are strong determinants of worse HIV health outcomes.52 The exacerbation of hunger or ART side effects in the absence of adequate food and competing resource demands are potential mechanisms that explain the relationship between food shortage and adherence.53 Individual-level factors are shaped by social and community norms. Research has shown that relevant community perceptions explain why women may be unwilling to take ART including HIV-related stigma, guilt, lack of knowledge, denial, stress and despair or futility. Maternal depression and internalisation of HIV-related stigma, while situated at the individual level, have been shown to further decrease uptake of treatment.54 A literature review18 highlights that while much has been written about how child-specific factors (eg, psychosocial function, neurological development, development stage) and regimen characteristics (eg, drug formulation, changes to treatment plans) may affect children's ART adherence, little is known about social factors impacting children's ART adherence.55 Social support has been shown to enhance adherence to HIV medication, with practical support bearing the highest correlation. Certain factors such as family functioning act as a mediator between social support and adherence.55 Research in Zimbabwe, for instance, has highlighted that despite poverty, social disruption and limited health infrastructure, social support networks creating a social atmosphere of empowerment can be facilitating factors for improving children's adherence.56 We summarised the findings of the review of the programme documents, the discussions with the designers and the literature review in the form of an initial programme theory, which shows how the intervention (ART in the form of pellets, combined with adherence support given by health providers) influences the choice of the caregiver to initiate and maintain treatment (or not). It shows how the context in terms of caregiver–provider relation, health service and system, and sociocultural factors influence the caregiver–infant dyad in the different stages of the adherence pathway (initiation, implementation and/or discontinuation). This programme theory will serve as a hypothesis that will be tested subsequently through qualitative research (figure 3). Figure 3 The initial programme theory. ART, antiretroviral therapy. Step 2: study design RE is design neutral: the design should allow adequate testing of the key elements of the initial programme theory. It is thus the research question and not any methodological or design preference that steers the design of the study. We will adopt the multiple embedded case study design.57 This design is often used in realist research in health.58 Case definition We define the case in this study as the caregiver–infant couple or the extended family setting in which paediatric HIV care takes place. We will study them as they are embedded in the three facilities in which the LIVING study is being implemented and in their sociocultural context. Having three sites provides us with contrasting cases, as they present different service provision and management modalities, as well as cultural contexts. Case selection The cases will be purposively selected. We will include caregiver–infant couples who are enrolled in the LIVING study and who have transitioned from syrup to the new pellet formulation. This group may include dyads who (1) just recently initiated the new treatment, (2) are maintaining the treatment, and (3) initiated and maintain treatment but, because of encountering different problems, have not achieve a long-term adherence to the treatment. In addition, purposive selection criteria will include age, sociodemographic background and other user characteristics (according to the sociodemographic profiles of the LIVING study participants) aiming to achieve a maximum variation sampling. Step 3: data collection methods, tools and procedures In terms of data collection methods, RE is method-neutral. In health policy and systems research, often both quantitative and qualitative data are gathered in order to allow rigorous ‘testing’ of the initial programme theory.59 In this study, we will use complementary data collection techniques. First, we will describe the actual implementation of the ART intervention in the three sites. To achieve that, we will collect and analyse intervention guidelines and implementation reports. This document review will allow us to assess the LIVING study outcomes (treatment refusal, treatment initiation, treatment maintenance or implementation and discontinuation). Second, we will carry out in-depth interviews with caregivers of HIV+ infants and with healthcare providers and, if needed, with health service managers. Here, we expect to explore the personal experiences with the new formula and issues like the perception of adherence support, the role of provider–patient relationships and of communication skills. Candidates will be selected purposively (see Case selection section above) to capture a wide range of opinions, meanings and interpretations concerning the new formulation. We estimate to do at least 15 interviews in each site with caregivers and 5 with providers and managers until data saturation is reached. Interviews will be held in suitable locations and we will use a topic guide that uses open-ended questions. Third, in each site, one to two focus group discussions (FGDs) will be held with healthcare providers. FGD is an appropriate data collection tool, given its potential for identifying barriers and facilitators to adherence, adherence maintenance and perceived support strategies. We will aim at achieving homogeneous group composition, balancing power issues that may arise (for instance, between nurses and physicians due to their different professional functions and roles, linked to hierarchical positions). The FGDs will be moderated by a social scientist experienced in qualitative research and observed by another social science researcher who will function as a note taker. FGDs will be conducted at convenient locations (eg, meeting room of the clinic). They will last about 1½–2 hours. The FGDs will be recorded, using two recorders and notes will be taken as a backup. After each FGD, the moderator, researchers and the note taker will hold a debriefing meeting. This serves to identify most salient and emerging issues, which will inform the following FGDs to account for an iterative data collection approach. Finally, to further explore the barriers and facilitators to adherence of the new formula at the level of family/kinship and the service provision, we will conduct observations at the clinic and in study participants' home environments, focusing on but not restricted to the infant's response to the drugs as well as the caregivers' and other family members' reactions. This means that home visits will be planned to coincide with treatment administration. This ethnographic method allows for an in-depth insight into relevant contexts, relationships and behaviours that cannot be assessed through personal inquiry. In this study, we use observations as an overt and complementary method to validate participants' subjective reporting of what they believe and do. In this sense, observation is used for data triangulation and to increase the validity of the overall study findings. Data collection through observation has two objectives. Observations in the clinic setting aim at observing healthcare provider and study participants' interaction and communication patterns. Items include healthcare provider relationship (active–passive role), responsiveness to patient needs, timing and time constraints, ease of communication and addressing specific (sensitive) topics such as adherence, adherence support strategies and handling of confidentiality. Observations in the home setting aim at observing how study participants practically handle the administration of the pellets, and to get insights into the family-related and contextual factors that may influence this action. They will include ease versus difficulties of administering the pellets, timing and integration in daily routine (meals, etc), support for adherence received through family members, HIV-related stigma and other social predictors of child adherence (for instance, partner conflict, blame for infant's HIV infection, etc). Kenyan anthropologists will observe the patient–provider interaction (ie, provider attitudes, communication style, supportive motivational strategies), as well as interaction in the home environment. In total, we aim at carrying out five observations in each setting in each of the three study sites. To ensure neutrality, the researcher conducting the participant observation will not contribute to any of the practical aspects of the study (eg, signing ICs, interviewing, etc). Given the study settings, we apply observation in a moderate way, meaning that the researchers will maintain balance between ‘insider’ and ‘outsider’ roles, allowing for a sufficient degree of involvement while observing in a non-intrusive manner, and keeping the necessary detachment. Observations will be conducted in a systematic way using checklists with the main topics presented above. The researchers will assemble detailed field notes, interview summaries and memos, which form the raw data to be analysed (see below). Step 4: data analysis Since the RE is method neutral, we will use appropriate data analysis methods following the best practices of the disciplines from which it borrows the methods.21 In general, the analysis is guided by the CMO configuration, to which we add intervention and actors. Qualitative data analysis All recorded interviews and FGDs will be transcribed verbatim and translated where necessary. Interview transcripts and relevant documents will be entered in an N-Vivo database. In line with realist principles, a thematic coding approach will be used based on the core elements of the initial programme theory. We adopt ‘Framework’ analysis,60 which suits the realist approach. For each case belonging to the different clinical settings, we will identify key themes. To do so, the data will be categorised in a first round of analysis using the themes ‘intervention’, ‘actor’, ‘context’, ‘mechanism’ and ‘outcome’. In essence, realists use retroduction.61 In retroduction, the analysis starts from the observed outcomes and with exploring the context conditions and the mechanisms through which the outcomes can be explained. They use the initial programme theory to guide them in the analysis, but are open to new explanations that emerge from the data. Questions at this stage include: Identifying outcomes: What are the observed outcomes? Identifying current intervention: What is the actually implemented intervention? Identifying the process to reach this outcome: How was it carried out (duration, intensity)? Identifying actors: Who delivered the intervention? Who are the actors involved in this programme? How did the intervention reach them and to what degree (coverage)? Identifying links (mechanisms) between intervention and outcome: Can the observed results be linked to the actual intervention? Once we have intervention-actors-outcomes identified, new interpretations will emerge in subsequent rounds of coding, leading to a refined analysis. At this stage, we will use the following guiding questions: Identifying context features: Which context conditions facilitated the uptake of the intervention by the actors? Which conditions constrained the actors in taking up the intervention? Are there alternative explanations (ie, other interventions that took place and which may have contributed to the observed outcomes)? Identifying links (mechanisms) between intervention and outcome: What explains the link between the actual intervention and the actual outcomes? Typically, the following questions guide this phase: What are the observed outcomes? What is the actually implemented intervention? How was it carried out (duration, intensity)? Who delivered the intervention? Who are the actors involved in this programme? How did the intervention reach them and to what degree (coverage)? Can the observed results be linked to the actual intervention? Interpretations of how these themes are linked will emerge in subsequent rounds of coding. At this stage, guiding questions include: What explains the link between the actual intervention and the actual outcomes? Which context conditions facilitated the uptake of the intervention by the actors? Which conditions constrained the actors in taking up the intervention? Are there alternative explanations (ie, other interventions that took place and which may have contributed to the observed outcomes)? We will use methodological triangulation by combining the different qualitative data collection techniques, drawing on data from different source types (interviews, FDGs, observations and document reviews). This will increase the validity of the study's findings. Findings will first be summarised in case research reports and in a second step, a comparison of the findings of the three sites will be carried out. Documentation of the data collection Each day, the field research team members will write field notes and a research diary will be kept. This will be kept as a qualitative data log file, identifying the different sources of data collection. Contact summary sheets will be written after each interview to allow recording of impressions and ideas emerging form the interviews. Analytical memos will be written to allow for an iterative approach in data analysis. Case analysis meetings will be held regularly in the form of feedback and discussion meetings of all study collaborators. These will allow for a critical review of observations and preliminary findings and conclusions, as well as an internal peer review. Monitoring and quality control The timelines and study milestones will be monitored by the principal investigators at ITM. All digital study materials will be stored in password-protected folders and continuously backed up. All data collectors will receive individual training on conducting qualitative data using the study-specific data collection tools. All transcripts and the translations will be checked for quality and accuracy, and will be adapted if necessary. All digital study material will be stored in password-protected folders and continuously backed up. Discussion This paper describes a protocol of an RE that uses a multiple embedded case study design to understand the acceptability and adherence to a new paediatric ART formulation and to identify the factors that contribute to adherence. While developing this study, we encountered a number of challenges. First, methods to measure the acceptability of any formulation and adherence to it are not standardised.52 Acceptability of the formulation may be assessed on the basis of a caretaker’s report of his/her child’s behaviour when given the medication. Direct observation may also provide insights. For this reason, we decided to both interview the caregivers and carry out observations of the administration of the new formulation in the clinic or at home (provided that the caregiver agrees, see section on ethical issues). RE is a relatively new methodology. Guidance on how to carry out RE studies is limited and researchers with a broad knowledge of the approach are scarce in the health policy and research community, even more so in low-income and middle-income countries.58 Furthermore, RE requires good insights into the different dimensions of the context that are relevant to the issue under study. In our case, cultural and social context factors are likely to play an important role and, for these reasons, we set up a consortium with experienced social scientists and clinical researchers from Kenya. A third challenge is common in RE and relates to how to adjudicate between different potentially interesting theoretical frameworks.24 62 Especially in case of short project duration, as in our 1-year project, a thorough literature review and full testing of partially contradicting hypotheses will not be possible. We relied on expert opinion to identify a few relevant theories from behavioural science that may be helpful in explaining adherence. In conclusion, this study will allow for an initial testing of the programme theory as outlined above, and will deliver an in-depth exploration of the multilevel factors that influence acceptability and adherence of this new paediatric ARV formulation. At the same time, this study's findings will add new insights to adherence research in general and we hope that further studies will take off from our results. The authors wish to acknowledge the Drugs for Neglected Diseases initiative (DNDi) for its commitment to and funding of this realist evaluation study. Contributors: ANG, CN, BM, JL, ML, OO and IN conceived the study proposal. ANG, CN, BM, JL, OS, ML, OO and IN developed the study proposal. ANG and BM led the writing of this paper. CN, JL, OS, ML, OO and IN contributed to the manuscript. All authors read and approved the final version of the manuscript. Funding: The study protocol reported in this paper has been funded by the Drugs for Neglected Diseases initiative in line with the research collaboration agreement signed between DNDi and ITM. Competing interests: None declared. Patient consent: Obtained. Ethics approval: The study protocol has been reviewed and approved by the Institutional Review Board of the ITM (ref. 1088/16), and by the Ethical Committee of the University Hospital Antwerp (Belgian registration number B300201628563). In addition, the protocol has been approved by the Kenyatta National Hospital/University of Nairobi Ethics and Research Committee (reference KNH-ERC/A/293). The study will be carried out according to the principles stated in the Declaration of Helsinki, all related applicable regulations and established international scientific standards. The authors will follow the Code of Ethics of the American Anthropological Association (AAA) and distribute to each participant (before the data collection) an information sheet and an informed consent form. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: The authors aim to disseminate the findings through international conferences and international peer-reviewed journals. The authors aim to share these findings with DNDi's programme managers in the field and in the headquarters as well as with the Kenyan healthcare providers. i Implementation study protocol—protocol number: DNDIPedHIV002. More details on http://www.dndi.org/2016/clinical-trials/clinical-trials-hiv/ ==== Refs References 1 Newell ML , Coovadia H , Cortina-Borja M Mortality among infected and uninfected infants born to HIV-infected mothers in Africa: a pooled analysis . Lancet 2004 ;364 :1236 –43 . 10.1016/S0140-6736(04)17140-7 15464184 2 Brady MT , Oleske JM , Williams PL Declines in mortality rates and changes in causes of death in HIV-1-infected children during the HAART Era . J Acquir Immune Defic Syndr 2010 ;53 :86 –94 . 10.1097/QAI.0b013e3181b9869f 20035164 3 Puthanakit T , Aurpibul L , Oberdorfer P Hospitalization and mortality among HIV-infected children after receiving highly active antiretroviral therapy . Clin Infect Dis 2007 ;44 :599 –604 . 10.1086/510489 17243067 4 Sutcliff C , Van Dijk J , Bolton C Effectiveness of antiretroviral therapy among HIV-infected children in sub-Saharan Africa . Lancet Infect Dis 2008 ;8 :477 –89 .18652994 5 UNICEF . UNAIDS 2014 HIV and AIDS estimates . New York : United Nations , 2015 . 6 WHO . Global report: UNAIDS report on the global AIDS epidemic 2013 Geneva : World Health Organization , 2013 . 7 UNAIDS . On the fast-track to an AIDS free generation . Geneva : United Nations , 2016 . 8 WHO . The use of antiretroviral drugs for treating and preventing HIV infection . WHO consolidated guidelines , 2013 . 9 UNAIDS . On the fast-track to an AIDS free generation . Geneva : UNAIDS , 2016 . 10 UNAIDS . Progress report on the global plan . Report on the Global Plan United Nations Programme on HIV/AIDS (UNAIDS) , 2015 . 11 Arrivé E , Newell M , Ekouevi D Prevalence of resistance to nevirapine in mothers and children after single-dose exposure to prevent vertical transmission of HIV-1: a meta-analysis . Int J Epidemiol 2007 ;36 :1009 –21 .17533166 12 Lockman S , Saphirom RL , Smeaton LM Response to antiretroviral therapy after a single, peripartum dose of nevirapine . N Engl J Med 2007 ;356 :135 –47 . 10.1056/NEJMoa062876 17215531 13 Musiime V , Fillekes Q , Kekitiinwa A The pharmacokinetics and acceptability of lopinavir/ritonavir minitab sprinkles, tablets, and syrups in African HIV-infected children . J Acquir Immune Defic Syndr 2014 ;66 :148 –54 .24828266 14 WHO . Antiretroviral therapy for HIV infection in infants and children: towards universal access. Recommendations for a public health approach: 2010 revision . Geneva : World Health Organization , 2010 . 15 Violari A , Paed F , Lindsey J Nevirapine versus ritonavir-boosted lopinavir for HIV-infected children . N Engl J Med 2012 ;366 :2380 –9 .22716976 16 Palumbo P , Lindsey JC , Hughes MD Antiretroviral treatment for children with peripartum nevirapine exposure . N Engl J Med 2010 ;363 :1510 –20 . 10.1056/NEJMoa1000931 20942667 17 Kekitiinwa A , Musiime V , Thomason MJ Acceptability of lopinavir/r pellets (minitabs), tablets and syrups in HIV-infected children . Antivir Ther (Lond) 2016 ;21 :579 –85 . 10.3851/IMP3054 27128199 18 Haberer J , Mellins C Pediatric adherence to HIV antiretroviral therapy . Current HIV AIDS Rep 2009 ;6 :194 –200 . 19 Stone VE Strategies for optimizing adherence to highly active antiretroviral therapy: lessons from research and clinical practice . Clin Infect Dis 2001 ;33 :865 –72 . 10.1086/322698 11512092 20 WHO . Adherence to long-term therapies . Geneva : World Health Organization , 2003 . 21 Pawson R , Tilley N Realistic evaluation . London : Sage , 1997 . 22 Pawson R Evidence based policy. A realist perspective . London : SAGE publications , 2006 . 23 Pawson R , Manzano-Santanella A A realist diagnostic workshop . Evaluation 2012 ;18 :176 –91 . 24 Pawson R The science of evaluation: a realist manifesto . London : SAGE Publications , 2013 . 25 Marchal B Why do some hospitals perform better than others? A realist evaluation of the role of health workforce management in well-performing health care organisations. A study of 4 hospitals in Ghana . Brussels : Vrije Universiteit Brussel , 2011 . 26 Violari A , Cotton M , Otwombe K Does early initiation of ART in infants affect virological and resistance outcomes? Data from the CHER trial after 6 years of follow-up . J Int AIDS Soc 2012 ;15:18085 (Suppl 4 ). 27 Ruel T , Kakuru A , Ikilezi G Virologic and immunologic outcomes of HIV-infected Ugandan children randomized to lopinavir-ritonavir or nonnucleoside- reverse-transcriptase-inhibitor therapy . J Acquir Immune Defic Syndr 2014 ;65 :535 –41 .24326597 28 Donahue M , Dube Q , Dow A “They have already thrown away their chicken”: barriers affecting participation by HIV-infected women in care and treatment programs for their infants in Blantyre, Malawi . AIDS Care 2012 ;24 :1233 –9 .22348314 29 Haberer J , Mellins C Paediatric adherence to HIV antiretroviral therapy . Current HIV AIDS Rep 2009 ;6 :194 –200 . 30 Marhefka S , Koenig L , Allison S Family experiences with pediatric antiretroviral therapy: responsibilities, barriers, and strategies for remembering medications . AIDS Patient Care STDs 2008 ;22 :637 –47 .18627275 31 Fassinou P , Elenga N , Rouet F Highly active antiretroviral therapies among HIV-1-infected children in Abidjan, Cote d'Ivoire . AIDS 2004 ;18 :1905 –13 .15353976 32 Hammami N , Nöestlinger C , Hoerée T Integrating adherence to highly active antiretroviral therapy into children's daily lives: a qualitative study . Pediatrics 2004 ;114 :591 –7 . 33 Sweeney S , Vanable P The association of HIV-related stigma to HIV medication adherence: a systematic review and synthesis of the literature . AIDS Behav 2016 ;20 :29 –50 .26303196 34 Fisher J , Fisher W Changing AIDS-risk behavior . Psychol Bull 1992 ;111 :455 –74 .1594721 35 Fisher J , Fisher W , Amico K An information-motivation-behavioral skills model of adherence to antiretroviral therapy . Health Psychol 2006 ;25 :462 –73 .16846321 36 Anderson E , Wagstaff D , Heckman T Information-Motivation-Behavioral Skills (IMB) model: testing direct and mediated treatment effects on condom use among women in low-income housing . Ann Behav Med 2006 ;31 :70 –9 .16472041 37 Nöstlinger C , Nideröst S , Grediog D Condom use with steady partners among heterosexual people living with HIV in Europe: testing the information-motivation-behavioral skills model . AIDS Patient Care STDs 2010 ;24 :771 –80 .21091237 38 Nöstlinger C , Nideröst S , Platteau T Sexual protection behavior in HIV-positive gay men: testing a modified information-motivation-behavioral skills model . Arch Sex Behav 2011 ;40 :817 –27 .20878223 39 Starace F , Massa A , Amico K Adherence to antiretroviral therapy: an empirical test of the information-motivation-behavioral skills model . Health Psychol 2006 ;25 :135 –62 .16569104 40 Amico KR , Toro-Alfonso J , Fisher JD An empirical test of the information, motivation and behavioral skills model of antiretroviral therapy adherence . AIDS Care 2005 ;17 :661 – 73 . 10.1080/09540120500038058 16036253 41 Kalichman SC , Amaral CM , Cherryn C Monitoring medication adherence by unannounced pill counts conducted by telephone: reliability and criterion-related validity . HIV Clin Trials 2008 ;9 :298 –308 . 10.1310/hct0905-298 18977718 42 Misovich S , Martinez T , Fisher J Predicting breast self-examination: a test of the Information, Motivation, Behavioral Skills model . J Appl Soc Psychol 2003 ;33 :775 –90 . 43 Smith LR , Amico KR , Shuper PA Information, motivation, and behavioral skills for early pre-ART engagement in HIV care among patients entering clinical care in KwaZulu-Natal, South Africa . AIDS Care 2013 ;25 :1485 –90 . 10.1080/09540121.2013.775398 23477286 44 Peltzer K , Friend-du Preez N , Ramlagan S Antiretroviral treatment adherence among HIV patients in KwaZulu-Natal, South Africa . BMC Public Health 2010 ;10 :111 10.1186/1471-2458-10-111 20205721 45 Ware N , Wyatt M , Bangsberg D Examining theoretic models of adherence for validity in resource-limited settings: a heuristic approach . J Acquir Immune Defic Syndr 2006 ;1 :S18 –22 . 46 Posse M , Meheusn F , Van Asten H Barriers to access to antiretroviral treatment in developing countries: a review . Trop Med Int Health 2008 ;13 :904 –13 . 10.1111/j.1365-3156.2008.02091.x 18466183 47 Braun M , Kabue M , McCollum E Inadequate coordination of maternal and infant HIV services detrimentally affects early infant diagnosis outcomes in Lilongwe, Malawi . J Acquir Immune Defic Syndr 2011 ;56 :122 –8 . 48 Spensley A , Sripipatana T , Turner A Preventing mother-to-child transmission of HIV in resource-limited settings: the Elizabeth Glaser Pediatric AIDS Foundation experience . Am J Public Health 2009 ;4 :631 –7 . 49 Davies MA , Keiser O , Technau K Outcomes of the South African national antiretroviral treatment programme for children: the IeDEA Southern Africa collaboration . S Afr Med J 2009 ;99 : 730 –7 .20128272 50 Cherutich P , Inwani I , Nduati R Optimizing paediatric HIV care in Kenya: challenges in early infant diagnosis . Bull World Health Organ 2008 ;86 :155 –60 .18297171 51 Kaufman M , Cornish F , Zimmerman R Health behavior change models for HIV prevention and AIDS care: practical recommendations for a multi-level approach . J Acquir Immune Defic Syndr 2014 ;66 :250 . 52 Weiser SD , Young SL , Cohen CR Conceptual framework for understanding the bidirectional links between food insecurity and HIV/AIDS . Am J Clin Nutr 2011 ;94 :1729S –39S . 10.3945/ajcn.111.012070 22089434 53 Young S , Wheeler A , McCoy S A review of the role of food insecurity in adherence to care and treatment among adult and pediatric populations living with HIV and AIDS . AIDS Behav 2014 ;18 (Suppl 5) :505 –15 .24469223 54 Kohler PK , Ondenge K , Mills LA Shame, guilt, and stress: community perceptions of barriers to engaging in prevention of mother to child transmission (PMTCT) programs in western Kenya . AIDS Patient Care STDs 2014 ;28 :643 –51 . 10.1089/apc.2014.0171 25361205 55 DiMatteo MR Social support and patient adherence to medical treatment: a meta analysis . Health Psychol 2004 ;23 :207 10.1037/0278-6133.23.2.207 15008666 56 Campbell C , Skovdal M , Mupambireyi Z Building adherence-competent communities: factors promoting children's adherence to anti-retroviral HIV/AIDS treatment in rural Zimbabwe . Health Place 2012 ;18 :123 –31 . 10.1016/j.healthplace.2011.07.008 21975285 57 Yin R Case study research. Design and methods . 3rd edn London : SAGE Publications , 2003 . 58 Marchal B , Van Belle S , van Olmen J Is realist evaluation keeping its promise? A literature review of methodological practice in health systems research . Evaluation 2012 ;18 :192 –212 . 59 Ritchie J , Spencer L Qualitative data analysis for applied policy research . In: Ritchie J , Lewis J , eds. Qualitative research practice . London : Sage , 2004 :173 –94 . 60 Rapoff MA Adherence to paediatric medical regimens . New York : Kluwer Academic/Plenum Publishers , 1999 . 61 McEvoy P , Richards D Critical realism: a way forward for evaluation research in nursing? . J Adv Nurs 2003 ;43 :411 –20 .12887360 62 Pawson R Middle range theory and programme theory evaluation: from provenance to practice . In: Vaessen J , Leeuw F , eds. Mind the gap: perspectives on policy evaluation and the social sciences . Transaction Press , 2009 :171 –203 .

PMC005xxxxxx/PMC5372017.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01506710.1136/bmjopen-2016-015067Health EconomicsResearch150617011728Economic costs of automated and continuous ambulatory peritoneal dialysis in Taiwan: a combined survey and retrospective cohort analysis Tang Chao-Hsiun 1Wu Yu-Ting 12Huang Siao-Yuan 1Chen Hsi-Hsien 34Wu Ming-Ju 5Hsu Bang-Gee 6Tsai Jer-Chia 7Chen Tso-Hsiao 38Sue Yuh-Mou 38 1 School of Health Care Administration, College of Management, Taipei Medical University, Taipei, Taiwan 2 Institute of Health Policy and Management, College of Public Health, National Taiwan University, Taipei, Taiwan 3 Division of Nephrology, Department of Internal Medicine, School of Medicine, College of Medicine, Taipei Medical University, Taipei, Taiwan 4 Division of Nephrology, Department of Internal Medicine, Taipei Medical University Hospital, Taipei Medical University, Taipei, Taiwan 5 Division of Nephrology, Department of Internal Medicine, Taichung Veterans General Hospital and Rong Hsing Research Center for Translational Medicine, Institute of Biomedical Science, College of Life Science, National Chung Hsing University, Taichung, Taiwan 6 Division of Nephrology, Department of Internal Medicine, Buddhist Tzu Chi General Hospital and School of Medicine, Tzu Chi University, Hualien, Taiwan 7 Department of Internal Medicine, Faculty of Renal Care, Kaohsiung Medical University Hospital and, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan 8 Division of Nephrology, Department of Internal Medicine, Wan Fang Hospital, Taipei Medical University, Taipei, TaiwanCorrespondence to Dr Yuh-Mou Sue; sueym@tmu.edu.twC-HT, Y-TW contributed equally to this work. 2017 21 3 2017 7 3 e01506710 11 2016 5 1 2017 16 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives Taiwan succeeded in raising the proportion of peritoneal dialysis (PD) usage after the National Health Insurance (NHI) payment scheme introduced financial incentives in 2005. This study aims to compare the economic costs between automated PD (APD) and continuous ambulatory PD (CAPD) modalities from a societal perspective. Design and setting A retrospective cohort of patients receiving PD from the NHI Research Database was identified during 2004–2011. The 1:1 propensity score matched 1749 APD patients and 1749 CAPD patients who were analysed on their NHI-financed medical costs and utilisation. A multicentre study by face-to-face interviews on 117 APD and 129 CAPD patients from five hospitals located in four regions of Taiwan was further carried out to collect data on their out-of-pocket payments, productivity losses and quality of life with EuroQol-5D-5L. Outcome measures The NHI-financed medical costs, out-of-pocket payments and productivity losses of APD and CAPD patients. Results The total NHI-financed medical costs per patient-year after 5 years of follow-up were significantly higher with APD than CAPD (US$23 005 vs US$19 237; p<0.01). In terms of dialysis-related costs, APD had higher costs resulting from the use of APD machines (US$795) and APD sets (US$2913). Significantly lower productivity losses were found with APD (US$2619) than CAPD (US$6443), but the out-of-pocket payments were not significantly different. The differences in NHI-financed medical costs and productivity losses between APD and CAPD remained robust in the bootstrap analysis. The total economic costs of APD (US$30 401) were similar to those of CAPD (US$29 939), even after bootstrap analysis (APD, US$28 399; CAPD, US$27 960). No discernable differences were found in the results of mortality and quality of life between the APD and CAPD patients. Conclusions APD had higher annual dialysis-related costs and lower annual productivity losses than CAPD, which made the economic costs of APD very close to those of CAPD in Taiwan. peritoneal dialysismedical utilizationeconomic cost ==== Body Strengths and limitations of this study This is the first study to evaluate the overall economic costs of automated peritoneal dialysis (APD) and continuous ambulatory peritoneal dialysis (CAPD) modalities. This study comprises a retrospective cohort of patients receiving APD and CAPD from the National Health Insurance Research Database and a multicentre study by face-to-face interviews of APD and CAPD patients. The information about out-of-pocket payments and productivity losses collected from patient interviews, which were rarely assessed in previous studies, adds important economic data to the overall evaluation of the costs associated with patients undergoing APD and CAPD. The unidentified laboratory data from the database and other potential confounding factors such as patient preference, self-care ability and physician selection are drawbacks of this study. Productivity losses related to presentism (impaired productivity or reduced effectiveness at work associated with APD or CAPD) are not included in this study. Introduction Given that both the incidence and prevalence rates of end-stage renal disease (ESRD) in Taiwan are among the highest in the world, this particular disease has become an increasing burden on the Taiwan National Health Insurance (NHI) system's finances; indeed, by 2011, the cost of dialysis accounted for an astonishing 7.2% of the total annual NHI expenditure.1 2 In an attempt to contain the total costs of dialysis, in addition to applying a blanket budget cap on dialysis expenditure, a series of strategies were implemented by the NHI administrators to change the incentives relating to the choices of dialysis modalities. Since peritoneal dialysis (PD) has an all-cause mortality rate similar to that of haemodialysis (HD), but with lower medical costs,3–6 from 2005 onwards administrators within the Taiwan NHI have been actively promoting the use of PD, including automated PD (APD) and continuous ambulatory PD (CAPD), as a viable alternative treatment for dialysis. This is being achieved by a reduction in reimbursements for HD and a corresponding increase in those for PD, as well as the NHI payment scheme covering the APD machine costs. As a result, the use of PD in Taiwan has been gradually increasing, from 6.5% in 2003 to 8.5% in 2007 and 10.3% in 2009.7 In terms of the global trend, Taiwan has now succeeded in raising the proportion of PD usage to the average level within developed countries.8 With the increasing usage of PD, the costs of the APD and CAPD modalities have become an important issue. Although there appears to have been very limited analysis of the costs involved in APD and CAPD, two specifically focused studies have revealed that APD has higher medical costs than CAPD, with the greatest differences being the costs of the dialysis machines and disposables.9–11 The prior studies have not, however, examined the differences between APD and CAPD in terms of the non-dialysis-related medical utilisation, which may be attributable to complications and clinical outcomes brought about by APD or CAPD; neither did they report differences in the out-of-pocket payment, including expenses on caregivers, as well as productivity losses from patients and family. Therefore, we set out in this study, from a societal perspective, to compare economic costs between APD and CAPD patients in Taiwan using population-based NHI claims data and face-to-face interviews. Methods Data sources, study design, setting and population The joint institutional Review Board of Taipei Medical University approved this study (No. 201 503 057). The first part of this study was a nationwide retrospective cohort study covering the years 2004 to 2011 based on the claims data obtained from the National Health Insurance Research Database (NHIRD). Taiwan's mandatory-enrolment, single-payer NHI programme was launched on 1 March 1995, with the programme now covering more than 23 million enrollees, representing over 99% of the entire population. The NHIRD contains registration files and original claims data for the reimbursement of medical services by enrollees, potentially making it one of the largest and most comprehensive administrative healthcare databases worldwide. The NHIRD has been utilised for previous epidemiologic research, and the results have been validated for HD and PD patients.12 13 Comprehensive details on the files used from the NHIRD in this study have already been provided in our previous work.14 ESRD patients requiring long-term dialysis in Taiwan are identified as patients with a catastrophic illness, as confirmed by two nephrologists, who may then be exempted from co-payments within the NHI system. The cohort of incident patients receiving long-term PD in the present study (new ESRD patients) were identified from the registry for catastrophic illness (International Classification of Diseases, 9th Revision, Clinical Modification, (ICD-9-CM), 585) within the NHIRD. These PD patients were in receipt of long-term PD therapy for at least three consecutive months between 1 January 2006 and 31 December 2010. Clinical characteristics were collected on the sample patients, with these patients subsequently being followed-up by referring to the NHIRD database covering the period from 1 January 2004 to 31 December 2011. We further categorised the patients into two types, APD or CAPD, according to whether they were receiving PD via a cycler for at least 90 days after the initiation of PD. A total of 1801 patients receiving APD therapy, and 4205 patients receiving CAPD therapy, were included in this study. A schematic illustration of the inclusion and exclusion criteria of the study sample is provided in figure 1. The second part of this study was a multicentre study by cross-sectional interviews on patients over 18 years old, carried out at the nephrology outpatient clinics of five hospitals located in northern, central, southern and eastern Taiwan from April 2015 to March 2016. The patient interviews were performed face to face by well-trained nurses from the site or graduate students from Taipei Medical University. All the interviewers had attended interviewer training before they performed the interview. In total, there were 117 APD patients and 129 CAPD patients available for analysis. Figure 1 Study sample selection process. Automated peritoneal dialysis (APD) or continuous ambulatory peritoneal dialysis (CAPD) patients were further categorised according to whether they were receiving peritoneal dialysis (PD) via a cycler for at least 90 days after the initiation of PD. Clinical characteristics of the national cohort study were collected and followed-up covering the period from 2004 to 2011. Outcome measurement The NHI-financed medical costs and utilisation during the 5-year period after the index date are the main outcomes of interest in the national cohort study. The ‘index date’ is defined in this study as the first day on which the corresponding patient started receiving their APD or CAPD therapy for a period of at least 90 days. The patients were then followed-up for a period of up to 5 years until 31 December 2011, when the data were censored, or until the occurrence of: (1) the change from APD to CAPD modality for at least 60 days, or vice versa (to evaluate the pure therapeutic period of APD or CAPD); (2) the change from PD to HD for at least 90 days; (3) the receipt of a kidney transplant; or (4) the death of the patient. The medical costs that had been incurred by each patient were traced back, starting from the index date to the last day of the follow-up period, after which we then calculated the total medical costs classified by the ‘outpatient’, ‘emergency’ and ‘inpatient’ departments. The total medical costs were divided on the basis of whether they were ‘dialysis-related’ or ‘non-dialysis related’. The ‘dialysis-related’ costs were defined as: (1) costs of the APD machine and set; (2) costs of erythropoietin and PD solution; (3) administration, physician and nursing fee; and (4) intubation and extubation cost of the PD catheter. The total medical utilisation was subsequently calculated from the index date to the last day of the follow-up period, with all utilisation being classified by specific departments. The total medical costs and number of medical visits per patient-year were defined as the total medical costs and number of medical visits within the follow-up period divided by the number of days in the same period. There were no differences in terms of administration, physician and nursing fees between the APD and CAPD patients. There were also no regional differences in the amount reimbursed for medical services in Taiwan. The secondary outcomes examined the differences in out-of-pocket payments, productivity losses and health-related quality of life (QoL) between APD and CAPD patients. There were two sources of time loss evaluated: patients' and caregivers' time spent in seeking care, and time spent in operating dialysis apparatus at home. Productivity losses were valued using the ‘human capital-cost approach’,15 and were measured by multiplying the loss of time in hours or days with average hourly/daily wage rate reported by the Directorate-General of Budget, Accounting, and Statistics, Taiwan (see online supplementary table S1). Out-of-pocket payments included all expenses related to ESRD paid by the patient/family and not reimbursed by the NHI, such as expenses for medicines, medical materials and devices, herbal and complementary medicines, and nutritional supplements. QoL was elicited using the visual analogue scale, and the EuroQol-5D-5L (EQ-5D-5L) with five levels of severity. The EQ-5D-5L index score was calculated based on a scoring algorithm representing the preferences of Japanese residents because there is no population-based preference weight available in Taiwan.16 10.1136/bmjopen-2016-015067.supp1supplementary data Statistical analysis A logistic regression analysis was carried out on the APD versus CAPD patients using the variables listed in online supplementary table S2. We then used the propensity score analysis to match the APD patients with the CAPD patients.17 Further analyses began with a comparison of those patients who were in receipt of either the APD or CAPD therapy at the baseline. Second, an independent paired t-test was then carried out to analyse the normally-distributed continuous variables, as well as a Wilcoxon rank sum test to analyse the non-normally distributed continuous variables by their median level. In addition, χ2 tests were carried out on the categorical variables. Third, the log-rank test and the Kaplan-Meier method were used to examine the mortality and chart the survival curves. Then a Cox proportional hazard regression was performed to examine the differences in patient survival. Fourth, in analysing the patient interview survey data, the Mann-Whitney U test and Wilcoxon rank sum test were performed to assess the differences between the APD and CAPD patients. Finally, a bootstrap analysis was further performed on NHI-financed medical costs, as well as on out-of-pocket payments and productivity losses, by forming 1000 bootstrap samples of APD and CAPD patients of equal size (1749 vs 1749 the in national cohort study, and 117 vs 129 in the patient interview survey) with replacement. The difference between the groups was considered to be significant if the two-sided p value was <0.05. All of the analyses in this study were undertaken using the SAS 9.3 software (SAS Institute, Cary, North Carolina, USA). Additionally, the PD patients, who have a catastrophic illness, may receive lower wage rates than the general population, resulting in lower productivity losses. In order to assess the impact of productivity losses on the total economic costs, we adjusted the productivity losses for the mean Taiwan unemployment rate during the interview period18 and then set the productivity losses with a 20%, 30% or 40% decrement of wages as different scenarios to calculate the total economic costs. Results Demographic characteristics Figure 1 shows the study sample selection process. A total of 1749 APD patients and 1749 CAPD patients were enrolled in the national cohort study, and a total of 117 APD patients and 129 CAPD patients were interviewed in the multicentre cross-sectional study. Demographic characteristics are shown in online supplementary tables S2 and S3. There were 265 patients who were put on HD (15.2%) among the APD patients during the 45 192 patient-months of follow-up. The result was greater than the 206 patients (11.8%) among the CAPD patients during the 47 272 patient-months of follow-up (APD: 70.4 per 1000 patient-years, CAPD: 52.3 per 1000 patient-years; p<0.01). There were 324 deaths (18.5%) among the APD patients during the 45 192 patient-months of follow-up. This was greater than the 311 deaths (17.8%) among the CAPD patients during the 47 272 patient-months of follow-up (APD: 86.0 per 1000 patient-years, CAPD: 78.9 per 1000 patient-years; p=0.22). The survival rates at 60 months for the APD and CAPD patients were 61.5% and 62.3%, respectively (see online supplementary figure S1). The result of Cox proportional hazard regression showed the APD and CAPD patients had a similar risk of death after adjusting for age, Charlson comorbidity index score, diabetes and cardiovascular diseases (p=0.13). The NHI-financed medical costs and utilisation The NHI-financed medical costs per patient-year for both the APD and CAPD patients are reported in table 1. The median total medical costs were found to be US$3769 higher among the APD patients than the CAPD patients (US$23 005 vs US$19 237, p<0.01). Following further classification of the medical costs based on the various departments, the APD patients were found to have significantly higher median costs than the CAPD patients in both the outpatient care (US$20 158 vs US$16 883, p<0.01) and the inpatient care (US$1197 vs US$992, p=0.01). If classified based on dialysis-related or non-dialysis-related costs, the APD patients had higher median costs than the CAPD patients (US$19 235 vs US$16 050, p<0.01), with the largest difference between the two groups being in the outpatient care (US$18 579 vs US$15 594, p<0.01). Non-dialysis-related medical costs were also found to be higher among the APD patients than the CAPD patients (US$2975 vs US$2639, p<0.01), with the largest difference between the two groups being in the inpatient care (US$1007 vs US$863, p=0.03). In the bootstrap analysis, the mean total medical costs were found to be US$3589 higher among the APD patients than the CAPD patients (US$23 488 vs US$19 899, p<0.001). Table 1 NHI-financed medical costs per patient-year for PD patients APD (n=1749) CAPD (n=1749) Median difference p Value Variables Mean Median SD Mean Median SD Total medical costs (US$)* 25 498 23 005 10 646 21 879 19 237 16 635 3769 <0.01 Classified by departments  Outpatient department 20 010 20 158 3607 16 918 16 883 7419 3275 <0.01  Emergency department 263 100 524 237 85 470 15 0.14  Inpatient department 5225 1197 10 766 4724 992 12 423 206 0.01 Classified by non-dialysis or dialysis  Non-dialysis-related 6296 2975 9703 5964 2639 16 013 336 <0.01   Outpatient department 1574 1258 1505 1595 1207 6900 51 <0.01   Emergency department 256 99 510 232 84 463 14 0.15   Inpatient department 4466 1007 9338 4137 863 11 210 144 0.03  Dialysis-related (by department) 19 202 19 235 3340 15 914 16 050 2917 3185 <0.01   Outpatient department 18 436 18 579 3344 15 323 15 594 2885 2985 <0.01   Emergency department 7 0 43 4 0 26 0 0.04   Inpatient department 759 91 1720 587 41 1610 51 <0.01  Dialysis-related (by items) 19 202 19 235 3340 15 914 16 050 2917 3185 <0.01   APD set 2391 2913 1089 48 0 213 2913 <0.01   APD machine 813 795 208 16 0 61 795 <0.01   Cost of erythropoietin 1622 1763 654 1618 1736 620 27 0.80   PD-related costs† 3390 3466 416 3407 3470 409 –4 0.02   PD solution 9609 9762 2464 9887 10 107 2178 –346 <0.01   Others 1377 811 1873 938 503 1627 308 <0.01  Dialysis-related (by quantities)   Quantity of PD solution (L) 3653 3700 740 2651 2706 629 994 <0.01   PRBC transfusion (unit) 3.7 0 13.5 3.3 0 8.7 0 0.15   Oral iron supply (mg) 2956 0 10 357 2155 0 8787 0 0.40   Erythropoietin (DDD)‡ 145.2 5.1 86.2 145.1 5.2 85.5 0 0.96   Extubation of PD catheter 0.06 0 0.29 0.05 0 0.24 0 0.35 After bootstrap analysis  Total medical costs (US$)* 23 488 783 19 899 173 <0.001   Non-dialysis-related 4723 629 4340 168 <0.001   Dialysis-related 18 859 157 15 768 76 <0.001 *US$1=30 New Taiwan Dollars. †Costs include: (1) administration, physician and nursing fee; and (2) intubation and extubation cost of peritoneal dialysis catheter. ‡1 DDD=1000 IU epoetin alfa=1000 IU epoetin beta=5 μg darbepoetin alfa=4 μg methoxy polyethylene glycol-epoetin beta. APD, automated peritoneal dialysis; CAPD, continuous ambulatory peritoneal dialysis; DDD, defined daily dose; NHI, National Health Insurance; PD, peritoneal dialysis; PRBC, packed red blood cell. Expenditure on the PD solution was found to be the highest cost item among all of the categories in both patient groups. Although the PD solution costs of the APD patients were significantly lower than those of the CAPD patients (US$9762 vs US$10 107, p<0.01), the APD patients were found to require significantly higher quantities of the PD solution than the CAPD patients (3700 L vs 2706 L, p<0.01). The greatest differences in the dialysis-related costs were found to be the costs of the APD machines (US$795 vs US$0, p<0.01), and those of the APD sets (US$2913 vs US$0, p<0.01). No discernible differences were found in the costs and the defined daily dosage (DDD) of erythropoietin between the APD and CAPD patients. The annual NHI-financed medical utilisation rates for both the APD and CAPD patients are reported in table 2, whereas no differences were found in the all items between the two patient groups. Table 2 Annual NHI-financed medical utilisation per patient-year of peritoneal dialysis patients Variables APD (n=1749) CAPD (n=1749) Median difference p Value Mean Median SD Mean Median SD Total no. of visits/hospitalisations 44.58 40.63 18.86 44.51 40.50 20.09 0.13 0.53 Visits classified by department  Outpatient visits 41.56 38.13 17.77 41.59 37.89 18.90 0.24 0.68   Dialysis-related visits 14.35 12.43 4.39 13.49 12.40 3.48 0.03 0.24   Non-dialysis-related visits 27.22 23.73 17.54 28.11 24.38 18.67 −0.65 0.25  Emergency room visits 1.66 0.90 2.47 1.60 0.87 2.33 0.04 0.51  Hospitalisations 1.36 0.79 1.69 1.31 0.76 1.75 0.03 0.23 APD, automated peritoneal dialysis; CAPD, continuous ambulatory peritoneal dialysis; NHI, National Health Insurance. Multicentre cross-sectional study for out-of-pocket payments and productivity losses Results of the out-of-pocket payment, productivity losses and QoL are reported in table 3. There were no discernable differences between the APD and CAPD patients in the out-of-pocket payments (US$1075 vs US$855, p=0.62), although the APD patients had significantly higher co-payment for outpatient visits than the CAPD patients (US$60 vs US$48, p=0.03). Compared with the CAPD patients, the APD patients had lower annual productivity losses (US$2619 vs US$6443, p<0.001), resulting from less time spent seeking care (APD, 45.0 hours vs CAPD, 59.9 hours, p<0.001), and less time spent operating dialysis apparatus (APD, 330.4 hours vs CAPD, 821.3 hours, p<0.001). In terms of QoL, there were no significant differences between the APD and CAPD patients in each dimension of EQ-5D-5L. Details of EQ-5D-5L by five severity levels are shown in online supplementary table S4. The results of the mean out-of-pocket payments, productivity losses and EQ-5D-5L index remained unchanged in the bootstrap analysis. Table 3 Per patient-year out-of-pocket payments and productivity losses as well as quality of life by interview survey of peritoneal dialysis patients Variables APD (n=117) CAPD (n=129) Median difference p Value Mean Median SD Reporting moderate to extreme problems (%) Mean Median SD Reporting moderate to extreme problems (%) Out-of-pocket payments (US$) 2012 1075 2861 2170 855 3182 220 0.62  Co-payment to outpatient visits 165 60 283 126 48 274 12 0.03  Co-payment to hospitalisations 417 0 920 381 0 816 0 0.66  Medicine not covered by NHI 424 67 861 374 0 667 67 0.22  Medical equipment 188 120 265 165 67 312 53 0.19  Chinese medication 149 0 1247 4 0 25 0 0.12  Traditional medicine 36 0 311 6 0 70 0 0.02  Nutritional supplements 184 0 561 246 0 1214 0 0.52  Caregiver costs 381 0 1697 783 0 2438 0 0.35  Transportation costs 68 39 94 84 58 89 −19 0.12 Productivity losses (US$)* 3006 2619 2159 6125 6443 3234 −3824 <0.001  Seeking outpatient care   From patients 274 214 211 341 333 254 −118 0.04   From family caregivers 125 0 235 192 0 352 0 0.19  Seeking inpatient care   From patients 402 0 600 426 0 588 0 0.75   From family caregivers 254 0 489 334 0 539 0 0.16  Time spent operating dialysis apparatus 1952 2020 1626 4831 5059 2780 −3039 <0.001 Visual analogue scale 72.8 75.0 15.6 – 69.0 70.0 17.9 – 0.05 Utility from EuroQol-5D-5L index† 0.82 0.87 0.19 0.82 0.87 0.21 0.68  Mobility – – – 7.7 – – – 13.2 0.16  Self-care – – – 9.4 – – – 9.3 0.98  Usual activities – – – 7.7 – – – 13.2 0.16  Pain/discomfort – – – 12.8 – – – 16.3 0.44  Anxiety/depression – – – 11.1 – – – 14.7 0.40 After bootstrap analysis Out-of-pocket payments (US$) 2019 261 2171 272 <0.001 Productivity losses (US$) 3007 200 6125 285 <0.001 EuroQol-5D-5L index 0.82 0.02 0.82 0.02 0.52 *US$1=30 New Taiwan Dollars. †Using tariffs from Japanese version of EuroQol-5D-5L. APD, automated peritoneal dialysis; CAPD, continuous ambulatory peritoneal dialysis; NHI, National Health Insurance. Economic costs The total economic costs per patient-year, including the direct medical costs by NHI, out-of-pocket payments and productivity losses, are reported in table 4. NHI-financed medical costs comprised of a higher proportion of the total economic costs among the APD patients than those of the CAPD patients; in contrast, productivity losses contributed to a higher proportion of the total economic costs among the CAPD patients than among the APD patients. The total economic costs of APD (US$30 401) were similar to those of CAPD (US$29 939), even after bootstrap analysis (APD, US$28 399; CAPD, US$27 960). After considering the productivity losses under various scenarios, the differences in total economic costs between the APD and CAPD patients slightly increased in models 2–4. Table 4 Total economic costs per patient-year of APD and CAPD patients Variables APD CAPD Direct medical costs by NHI (US$)* 25 498 21 879 Out-of-pocket payments (US$) 2012 2170 Productivity losses (US$) 3006 6125  Adjusted for unemployment rate† 2891 5890  Adjusted for unemployment rate and a 20% decrement in wages 2312 4712  Adjusted for unemployment rate and a 30% decrement in wages 2023 4123  Adjusted for unemployment rate and a 40% decrement in wages 1734 3534 Total costs, model 1‡ 30 401 29 939 Total costs, model 2‡ 29 822 28 761 Total costs, model 3‡ 29 533 28 172 Total costs, model 4‡ 29 244 27 583 After bootstrap analysis Direct medical costs by NHI (US$) 23 488 19 899 Out-of-pocket payments (US$) 2019 2171 Productivity losses (US$) 3007 6125  Adjusted for unemployment rate† 2892 5890  Adjusted for unemployment rate and a 20% decrement in wages 2313 4712  Adjusted for unemployment rate and a 30% decrement in wages 2024 4123  Adjusted for unemployment rate and a 40% decrement in wages 1735 3534 Total costs, model 1‡ 28 399 27 960 Total costs, model 2‡ 27 820 26 782 Total costs, model 3‡ 27 531 26 193 Total costs, model 4‡ 27 242 25 604 *US$1=30 New Taiwan Dollars. †Adjusted for mean Taiwan unemployment rate (3.82%) between April 2015 and March 2016. ‡Models 1–4: Total costs include direct medical costs financed by NHI, out-of-pocket payments, and productivity losses adjusted for unemployment rate (model 1); adjusted for unemployment rate and a 20% decrement in wages (model 2); adjusted for unemployment rate and a 30% decrement in wages (model 3); adjusted for unemployment rate and a 40% decrement in wages (model 4). APD, automated peritoneal dialysis; CAPD, continuous ambulatory peritoneal dialysis; NHI, National Health Insurance. Discussion The main results of this study using Taiwan's population-based claims data and interview survey data demonstrate that total economic costs of APD (US$29 977) were close to the costs of CAPD (US$30 750). The annual medical costs for the APD patients were US$3769 higher than those of the CAPD patients, with the greatest difference (US$2985) being found in the costs of outpatient care for dialysis, accounting for 80% of the difference in the total costs. The main source of the differences between the APD patients and CAPD patients in terms of dialysis-related costs were those of the APD machines and APD sets. As regards the medical utilisation rates, no significant differences were discernible between the APD and CAPD patients. The results from the interview survey demonstrated lower productivity losses in the APD patients than the CAPD patients, but no differences were detected in the out-of-pocket payments among the two patient groups. The differences in direct medical costs paid by NHI and productivity losses between APD and CAPD remained robust in the bootstrap analysis. The NHI-financed medical utilisation between the APD and CAPD patients are quite similar; however, a significantly higher total of the NHI-financed medical costs were discernible among the APD patients than among the CAPD patients. This would seem to indicate that the costs per outpatient or emergency room visit are higher among the APD patients. Furthermore, while the APD patients were found to have similar numbers of hospitalisation to the CAPD patients (0.79 vs 0.76, p=0.23), the APD patients were found to have higher inpatient costs than the CAPD patients, particularly those relating to non-dialysis inpatient treatment. This would seem to indicate that the costs per hospitalisation are again higher or the hospital stays are longer among the APD patients than the CAPD patients, possibly owing to the higher incident rate of patients changed to HD among the APD patients (APD: 70.4 per 1000 patient-years, CAPD: 52.3 per 1000 patient-years; p<0.01). The NHI-financed dialysis-related costs of the APD patients are approximately 1.2 times higher than those of the CAPD patients, which is in line with the results of several related studies carried out in other countries.9 19–21 The major sources of the higher dialysis-related costs for the APD patients were found to be the APD machines (4.2%), APD sets and PD solution (62.5%), and erythropoietin (8.4%), while those for the CAPD patients were PD solution (62.4%) and erythropoietin (10.2%). The differences between the dialysis-related costs for the APD patients and the CAPD patients were mainly attributable to the costs of the APD machines and APD sets (table 1). These costs were US$3708 higher among the APD patients than the CAPD patients. When compared to CAPD, APD involves the use of a cycler and extra line sets, which explains the higher dialysis costs involved in the APD treatment. The results of a related UK study also indicated that the greatest differences in costs between APD and CAPD were attributable to the APD machines and disposables.9 In addition, APD patients had the lower costs and higher quantities of PD solutions. The most likely reasons were that the per-litre cost of the commonly used 5 L bag PD solution for APD is much cheaper than that of the commonly used 2 L bag PD solution for CAPD, and that the APD patients usually received shorter dwell time but more exchanges than the CAPD patients to reach adequate dialysis.22 The present study features an overall evaluation of the costs associated with patients undergoing APD and CAPD, with out-of-pocket payments and productivity losses incorporated into the analysis, which were rarely assessed in previous studies. From the payer's perspective, the NHI-financed medical costs of the CAPD seems to be a more cost-saving modality than those of the APD; however, from a societal perspective, the annual economic costs of APD were close to those of CAPD although differences exist in the proportion of key cost components. Based on the human capital-cost approach,15 the productivity losses were estimated as the reduced future gross income, including reduced paid or unpaid production, due to PD related mortality and/or morbidity and contributed to a lower proportion of the total economic costs for APD rather than that of CAPD. On the other hand, NHI-financed medical costs made up a larger portion of the total economic costs for APD than that of CAPD. This is reflected in the fact that the costs of the productivity losses, resulting from the time spent seeking care and operating dialysis apparatus, were significantly lower in the APD than CAPD patients. With the help of the APD machine, the APD modality spent less time in operating dialysis so as to decrease the productivity losses; however, it correspondingly increases the costs of the APD machines and APD sets. The mortality rates of the APD and CAPD patients were similar and the result did not differ from most of the previous findings of multicentre or nationwide cohort studies.13 19 23 24 In terms of QoL, the results of the visual analogue scale and EQ-5D-5L were quite similar in the APD and CAPD patients (table 3 and online supplementary table S4), which were consistent with the findings published from the literature.25–28 The results of the present study also have some limitations. First, the major drawbacks are those commonly found in administrative database research. Although comorbidities and medications were matched at cohort entry, the patient's weight, peritoneal equilibration test and dialysis clearance, and their level of residual renal function are not available from the NHIRD. These unidentified data are potentially confounding factors in the prescriptions of the peritoneal dialysate dose. Other potential confounding factors that are not available from the claims database when computing the propensity scores were patient preference, self-care ability and physician selection. Second, the productivity losses estimated in this study were the value of time lost by patients and their family caregivers when seeking care/operating dialysis apparatus. Productivity losses related to presentism (impaired productivity or reduced effectiveness at work associated with ESRD) were not included in this study and thus may have led to an underestimation of the economic costs. Finally, the ED-5D-5L index score was calculated based on a scoring algorithm representing the preferences of Japanese residents which may not perfectly represent those of Taiwanese people. Although NHI-financed annual medical costs of CAPD had a greater reduction in expenses than those of the APD, CAPD had higher annual productivity losses than APD. To extend the generalisability of our findings to other national health systems, our results highlight that the APD modality may appear to be more desirable in terms of its substantially lower productivity losses for countries with a higher value of time (hourly wage or daily wage), or with a younger dialysis patient population. In this study, we present a national cohort study to analyse NHI-financed medical utilisation and medical costs and a cross-sectional study to survey the out-of-pocket payments and productivity losses for APD and CAPD patients. From a societal perspective, although APD had higher annual dialysis-related costs financed by NHI, the overall economic costs of APD were very close to those of CAPD because CAPD had higher annual productivity losses than APD. Contributors: Obtained funding: SYM and TCH. SYM and TCH take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and questionnaire design: SYM, TCH, WYT, HSY, CHH, WMJ, HBG, TJC, and CTH. Acquisition of data: SYM, TCH, WYT, HSY, CHH, WMJ, HBG, TJC, and CTH. Analysis and interpretation of data: SYM, TCH, WYT, HSY, CHH, WMJ, HBG, TJC, and CTH. Statistical analysis: SYM, TCH, WYT, and HSY. All authors participated in writing the paper, reviewed it for important intellectual content and approved the final version. Final approval of the manuscript: SYM, TCH, WYT, HSY, CHH, WMJ, HBG, TJC, and CTH. Funding: This research received grants from National Science Council (NSC 102-2815-C-038-007-H) and from Wan Fang Hospital, Taipei Medical University, Taiwan (102TMU-WFH-08 and 104TMU-WFH-12). Competing interests: None declared. Patient consent: Obtained. Ethics approval: The Joint Institutional Review Board of Taipei Medical University. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Yang WC , Hwang SJ Incidence, prevalence and mortality trends of dialysis end-stage renal disease in Taiwan from 1990 to 2001: the impact of national health insurance . Nephrol Dial Transplant 2008 ;23 :3977 –82 . 10.1093/ndt/gfn406 18628366 2 National Health Insurance Administration, Ministry of Health and Welfare. http://www.nhi.gov.tw/webdata/webdata.aspx?menu=17&menu_id=662&webdata_id=805&WD_ID=698 (accessed 15 Jun 2015 ). 3 Mehrotra R , Chiu YW , Kalantar-Zadeh K Similar outcomes with hemodialysis and peritoneal dialysis in patients with end-stage renal disease . Arch Intern Med 2011 ;171 :110 –18 . 10.1001/archinternmed.2010.352 20876398 4 Chang YK , Hsu CC , Hwang SJ A comparative assessment of survival between propensity score-matched patients with peritoneal dialysis and hemodialysis in Taiwan . Medicine (Baltimore) 2012 ;91 :144 –51 . 10.1097/MD.0b013e318256538e 22525667 5 Karopadi AN , Mason G , Rettore E The role of economies of scale in the cost of dialysis across the world: a macroeconomic perspective . Nephrol Dial Transplant 2014 ;29 :885 –92 . 10.1093/ndt/gft528 24516226 6 Kao TW , Chang YY , Chen PC Lifetime costs for peritoneal dialysis and hemodialysis in patients in Taiwan . Perit Dial Int 2013 ;33 :671 –8 . 10.3747/pdi.2012.00081 23636434 7 Wu MS , Wu IW , Shih CP Establishing a platform for battling end-stage renal disease and continuing quality improvement in dialysis therapy in Taiwan--Taiwan Renal Registry Data System (TWRDS) . Acta Nephrologica 2011 ;25 :148 –53 . 8 Jain AK , Blake P , Cordy P Global trends in rates of peritoneal dialysis . J Am Soc Nephrol 2012 ;23 :533 –44 . 10.1681/ASN.2011060607 22302194 9 Baboolal K , McEwan P , Sondhi S The cost of renal dialysis in a UK setting--a multicentre study . Nephrol Dial Transplant 2008 ;23 :1982 –9 . 10.1093/ndt/gfm870 18174268 10 Cortes-Sanabria L , Rodriguez-Arreola BE , Ortiz-Juarez VR Comparison of direct medical costs between automated and continuous ambulatory peritoneal dialysis . Perit Dial Int 2013 ;33 :679 –86 . 10.3747/pdi.2011.00274 23547280 11 Cortes-Sanabria L , Paredes-Cesena CA , Herrera-Llamas RM Comparison of cost-utility between automated peritoneal dialysis and continuous ambulatory peritoneal dialysis . Arch Med Res 2013 ;44 :655 –61 . 10.1016/j.arcmed.2013.10.017 24211750 12 Tang CH , Wang CC , Chen TH Prognostic benefits of carvedilol, bisoprolol, and metoprolol controlled release/extended release in hemodialysis patients with heart failure: a 10-year cohort . J Am Heart Assoc 2016 ;5 :pii: e002584 . 13 Tang CH , Chen TH , Fang TC Do automated peritoneal dialysis and continuous ambulatory peritoneal dialysis have the same clinical outcomes? A ten-year cohort study in Taiwan . Sci Rep 2016 ;6 :29276 10.1038/srep29276 27388055 14 Tang CH , Chen TH , Wang CC Renin-angiotensin system blockade in heart failure patients on long-term haemodialysis in Taiwan . Eur J Heart Fail 2013 ;15 :1194 –202 . 10.1093/eurjhf/hft082 23671265 15 Liljas B How to calculate indirect costs in economic evaluations . Pharmacoeconomics 1998 ;13 (1 Pt 1) :1 –7 .10175982 16 Shiroiwa T , Ikeda S , Noto S Comparison of value set based on DCE and/or TTO Data: scoring for EQ-5D-5L health states in Japan . Value Health 2016 ;19 :648 –54 . 10.1016/j.jval.2016.03.1834 27565282 17 D'Agostino RB Jr Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group . Stat Med 1998 ;17 :2265 –81 .9802183 18 Unemployment rate. National Statistics, R.O.C. (Taiwan) 2016 https://www.stat.gov.tw/point.asp?index=3 (accessed 31 Dec 2016 ). 19 Badve SV , Hawley CM , McDonald SP Automated and continuous ambulatory peritoneal dialysis have similar outcomes . Kidney Int 2008 ;73 :480 –8 . 10.1038/sj.ki.5002705 18046315 20 Peeters P , Rublee D , Just PM Analysis and interpretation of cost data in dialysis: review of Western European literature . Health Policy 2000 ;54 :209 –27 .11154790 21 De Vecchi AF , Dratwa M , Wiedemann ME Healthcare systems and end-stage renal disease (ESRD) therapies--an international review: costs and reimbursement/funding of ESRD therapies . Nephrol Dial Transplant 1999 ;14 (Suppl 6) :31 –41 . 22 Akonur A , Firanek CA , Gellens ME Volume-based peritoneal dialysis prescription guide to achieve adequacy targets . Perit Dial Int 2016 ;36 :188 –95 . 10.3747/pdi.2014.00255 26293841 23 Mehrotra R , Chiu YW , Kalantar-Zadeh K The outcomes of continuous ambulatory and automated peritoneal dialysis are similar . Kidney Int 2009 ;76 :97 –107 . 10.1038/ki.2009.94 19340090 24 Michels WM , Verduijn M , Boeschoten EW Similar survival on automated peritoneal dialysis and continuous ambulatory peritoneal dialysis in a large prospective cohort . Clin J Am Soc Nephrol 2009 ;4 :943 –9 . 10.2215/CJN.04440908 19357244 25 Guney I , Solak Y , Atalay H Comparison of effects of automated peritoneal dialysis and continuous ambulatory peritoneal dialysis on health-related quality of life, sleep quality, and depression . Hemodial Int 2010 ;14 :515 –22 . 10.1111/j.1542-4758.2010.00465.x 20955286 26 Michels WM , van Dijk S , Verduijn M Quality of life in automated and continuous ambulatory peritoneal dialysis . Perit Dial Int 2011 ;31 :138 –47 . 10.3747/pdi.2010.00063 21357936 27 Balasubramanian G , McKitty K , Fan SL Comparing automated peritoneal dialysis with continuous ambulatory peritoneal dialysis: survival and quality of life differences? Nephrol Dial Transplant 2011 ;26 :1702 –8 . 10.1093/ndt/gfq607 20921296 28 Bilgic A , Akman B , Sezer S Daytime sleepiness and quality of life in peritoneal dialysis patients . Ther Apher Dial 2011 ;15 :565 –71 . 10.1111/j.1744-9987.2011.00987.x 22107693

PMC005xxxxxx/PMC5372018.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01342510.1136/bmjopen-2016-013425Diabetes and EndocrinologyResearch150618431722Psychological and behavioural patterns of stigma among patients with type 2 diabetes: a cross-sectional study Kato Asuka 12Fujimaki Yuko 3Fujimori Shin 3Isogawa Akihiro 4Onishi Yukiko 5Suzuki Ryo 6Yamauchi Toshimasa 6Ueki Kohjiro 6Kadowaki Takashi 6Hashimoto Hideki 1 1 Department of Health and Social Behavior, School of Public Health, The University of Tokyo, Tokyo, Japan 2 The Health Care Science Institute, Tokyo, Japan 3 Department of Internal Medicine, School of Medicine, Teikyo University, Tokyo, Japan 4 Mitsui Memorial Hospital, Tokyo, Japan 5 The Institute for Adult Diseases Asahi Life Foundation, Tokyo, Japan 6 Department of Diabetes and Metabolic Diseases, Graduate School of Medicine, The University of Tokyo, Tokyo, JapanCorrespondence to Dr Asuka Kato; asukakato-tky@umin.ac.jp2017 29 3 2017 7 3 e01342511 7 2016 31 1 2017 8 3 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives The aim of this study was to test the psychological and behavioural patterns of stigma (self-esteem and social participation) and their relationship to self-stigma, patient activation for engaging in self-care and glycaemic control among patients with type 2 diabetes mellitus (T2DM). Design A cross-sectional study. Setting 2 tertiary-level hospitals and 2 secondary-level hospitals in Japan. Participants A consecutive sample of 209 outpatients with T2DM. Inclusion criteria were as follows: presence of T2DM, age 20–74 years, no diagnosis of dementia and/or psychosis, and no need for urgent medical procedures. Outcome measures Study measures included a self-administered questionnaire to assess the Rosenberg Self-Esteem Scale (SES), the 3 subscales of 36-question Short Form Health Survey (SF-36; Social Function, Role Physical, Role Emotional), Self-Stigma Scale and Patient Activation Measure (PAM-13). Glycated haemoglobin was obtained from same day blood work. In our previous qualitative study, we found that psychological and behavioural patterns of stigma varied according to patients' levels of illness-related self-esteem as well as attitudes towards social participation. For quantitative consistency, we used the SES scale to measure self-esteem and the SF-36 subscales to measure social participation. We then divided participants into 4 groups by exhibited psychological and behavioural patterns: group A (high SES/high SF-36), group B (high SES/low SF-36), group C (low SES/high SF-36) and group D (low SES/low SF-36). Results Using analysis of covariance after controlling for age and sex, there was a significant difference in self-stigma levels between the four groups (F[3203]=15.70, p<0.001). We observed the highest mean self-stigma levels in group D. Group D also had significantly lower PAM-13 scores than those of groups A (p<0.001) and B (p=0.02). Conclusions The psychological and behavioural pattern of group D was found to be associated with higher levels of self-stigma and poorer patient activation for self-care. Psychosocial, behavioral medicineDiabetes educationStigma ==== Body Strengths and limitations of this study This is the first study to test the psychological and behavioural patterns of stigma (self-esteem and social participation) and their relationship to self-stigma, patient activation for engaging in self-care and glycaemic control among patients with T2DM. This study quantitatively demonstrated that there were differences between the distinct psychological and behavioural patterns of stigma by the two dimensions (self-esteem and social participation), which were derived from our previous qualitative study, suggesting that the group with low self-esteem/low social participation is associated with higher levels of self-stigma and poorer patient activation for self-care. Owing to the cross-sectional design of this study, causal relationships between self-stigma, self-esteem and attitudes towards social participation cannot be determined. Introduction Many people with type 2 diabetes mellitus (T2DM) experience diabetes-related stigma due to the fact that the condition is often perceived as a lifestyle-related disease by the general population.1–10 As a result, people with T2DM may feel ashamed of their illness and encounter discrimination, limited opportunities and negative stereotyping against them.6 Public stigma and self-stigma are two distinct but inter-related constructs. Public stigma refers to negative reactions of the general public towards a group based on stereotypical attributes distinguishing that group.11 12 Self-stigma, in contrast, refers to the internalisation of society's negative perceptions towards an illness by someone who has that particular illness.11 12 In order to internalise public stigma, an individual must first be aware that a public stigma for their illness exists, accept the stereotypes associated with their illness, and then develop a negative attitude towards the illness and thus themselves.11 12 This means that merely perceiving public stigma does not necessarily lead to the development of self-stigma.11 12 In this study, we decided to use the Self-Stigma Scale that consists of cognitive, affective and behavioural subscales.13 Since the Self-Stigma Scale was specifically designed to evaluate the self-stigma of groups that are not immediately identifiable,14 it is viewed as an appropriate tool for assessing the particularly concealable condition of T2DM. Using this scale, it was demonstrated that self-stigma is a significant predictor of decreased activation levels for engaging in self-care among patients with T2DM.15 It is a theoretically and empirically well-known fact that the two dimensions self-esteem and social participation are related to the ability to adjust to chronic illnesses in general.16 17 Previous studies show that self-stigma is associated with self-esteem, which is defined as the degree to which an individual has a favourable or unfavourable opinion of himself/herself and finds himself/herself worthy or unworthy as a whole including their condition.13 18–20 Depending on the degree of social participation, adjustment styles to illness differ.16 17 For T2DM specifically, Kato et al21 have found that patients' responses to self-stigma could be associated with social participation, which was defined as the degree to which patients are willing to participate in social life regardless of the severity of their condition. In our previous qualitative study,21 we showed that, in regard to responses to diabetes-related stigma, psychological and behavioural patterns of patients with T2DM could be explained by the following two dimensions: (1) a sense of self-esteem in relation to their illness and (2) attitudes towards social participation (for more details, see our previous paper21). These psychological and behavioural patterns can be categorised as follows: group A (high self-esteem/high social participation), group B (high self-esteem/low social participation), group C (low self-esteem/high social participation) and group D (low self-esteem/low social participation). We qualitatively found that, compared with other groups, groups C and D seemed to develop lower self-esteem as a result of their own negative experiences and perceptions of their illness (stigma). The internalisation of this stigma (self-stigma) was seen as a possible contributing factor to changes in social participation and treatment adherence.21 However, the associations between these distinct psychological and behavioural patterns of stigma (self-esteem and social participation) with self-stigma, patient activation for engaging in self-care and glycaemic control remain unproven quantitatively. Therefore, we need to investigate further in order to detail the characteristics of the psychological and behavioural patterns of stigma in T2DM. Accordingly, the aim of this study was to test the psychological and behavioural patterns of stigma (self-esteem and social participation) and their relationship to self-stigma, patient activation for engaging in self-care and glycaemic control among patients with T2DM. Methods Study participants A questionnaire-based cross-sectional study was conducted between November 2013 and March 2014. Consecutive sampling was used to recruit all outpatients with T2DM who had visited an endocrinologist on a specific date at four locations in Japan (ie, two university hospitals, one non-university-affiliated hospital and one non-university-affiliated clinic). Patients were recruited through their physicians. After the physicians obtained permission, the patients received an explanation of the study's purpose by the research staff, after which written informed consent was obtained. Inclusion criteria were as follows: presence of T2DM, age 20–74 years, ability to read and speak Japanese, no diagnosis of dementia and/or psychosis, and no need for urgent medical procedures or examinations. Patients completed a questionnaire that took ∼15–20 min. This study was approved in advance by the Research Ethics Committee of the University of Tokyo Graduate School of Medicine and Faculty of Medicine, and by each participating facility. Study measures Participants' sociodemographic factors included age, sex, education (in years) and marital status. We also collected participants' clinical information such as body mass index (BMI), time (in months) since diagnosis of T2DM, injection therapy, diabetes-related complications and glycated haemoglobin (HbA1c) level. We calculated the number of diabetes-related complications as the simple sum of six complications in reference to the Diabetes Complications Index.22 Scores ranged from 0 to 6. Self-esteem We used the Rosenberg Self-Esteem Scale (SES) to assess self-esteem levels.18 23 The SES is a widely accepted scale due to its high reliability and validity. It contains 10 items scored on a four-point Likert scale, from 1 (strongly disagree) to 4 (strongly agree). Five negative items were reverse-scored to compute the total scores of individual participants. In this study, internal consistency was 0.79. All participants were assigned to one of two groups (high self-esteem/low self-esteem) depending on whether their score was above or below the total mean score. Since the self-esteem levels were normally distributed, the mean value of the score was used as the threshold. Social participation We assessed patients' levels of generic functional states using the following three subscales in the 36-question Short Form Health Survey (SF-36), Role Physical (RP), Role Emotional (RE) and Social Function (SF),24 as a suitable approximation for attitudes towards social participation. In this study, we use the term social participation as it was defined in our previous study; essentially, social participation is defined as the degree to which patients are willing to participate in social life, which is influenced more by public stigma than by physical complications stemming from the severity of their condition.21 Therefore, we intentionally chose these subscales not for the purpose of detecting small changes that were clinically important, but for the purpose of assessing patients' functional states that were influenced by their illness-related negative experiences as well as their illnesses. The generic health status highly correlates with patients' limitations in terms of engaging in daily and social activities due to physical health, mental health or both, based on patients' objective and subjective ratings.24 Additionally, the SF-36 is a widely accepted valid health measure for estimating a specific disease burden, such as cardiac disease, asthma and diabetes. The RP subscale contains four items on role limitations due to physical problems, and it evaluates the extent to which physical capabilities limit daily activities. The RE subscale contains three items on role limitations due to emotional problems, and it evaluates the extent and degree to which emotional problems interfere with work or other activities. The SF subscale contains two items on social functioning, and it evaluates the amount of time and extent to which emotional problems interfere with family, friends and other social interactions. For each subscale, item scores were coded according to the scoring algorithm, and then summed, transferred onto a scale from 0 to 100 and converted to norm-based scoring, with a mean set to 50.24 Thereafter, we simply added up the total of each subscale score. We calculated the total mean score by dividing the total score by the total number of participants. We tested each subscale score (RP, RE, SF) individually, and then tested the total score of the subscales to see if participants' distributions were the same. Since the participants' distributions remained unchanged whether the subscale scores were tested individually or totally, we decided, for the sake of expediency, to use the total of the subscale scores as the levels of social participation in this study. All participants were assigned to one of two groups (high social participation/low social participation) based on whether their score was above or below the total mean score. Since the total of these subscale scores was normally distributed, the mean value of the total score was used as the threshold. Psychological and behavioural patterns of stigma In our previous qualitative study,21 we found that psychological and behavioural patterns of stigma varied depending on levels of illness-related self-esteem and attitudes towards social participation in patients with T2DM. For quantitative consistency, in this study we used the SES to measure self-esteem levels and three SF-36 subscales to measure social participation. As described above, participants were classified into different psychological and behavioural patterns of stigma based on each scale score: participants with high self-esteem were assigned to either group A (those with high social participation) or group B (those with low social participation; figure 1), and participants with low self-esteem were assigned to either group C (those with high social participation) or group D (those with low social participation; figure 1). Figure 1 Psychological and behavioural patterns of stigma in patients with type 2 diabetes—a hypothesis generated from results of a qualitative study. SF-36, 36-question Short Form Health Survey. Outcome measures Self-stigma The Self-Stigma Scale was used to assess patients' level of self-stigma.14 The reliability and validity of the scale's Japanese version (SSS-J) were reported previously.13 The 39-item SSS-J allows four responses on a Likert scale (strongly disagree, disagree, agree and strongly agree) scored 0, 1, 2 and 3, respectively. Total scores are treated as continuous and range from 0 to 117. Higher scores represent higher levels of self-stigma. In this study, the SSS-J had an internal consistency of 0.96. Patient activation The Patient Activation Measure (PAM-13) was used to assess patients' activation levels for self-care.25–27 Patient activation is a concept that entails a comprehensive approach to several elements related to activation, including the patients' knowledge, skills, confidence and behaviours needed to manage their illness. Previous studies have indicated that the PAM-13 predicts a range of comprehensive behaviours, not just healthy behaviours (eg, exercise, diet) but also disease-specific self-management behaviours (eg, keeping a written diary of glucose levels, taking diabetes medication as recommended by the physician).26 28 The PAM-13 is a clinically used, highly reliable and valid scale that consists of 13 questions on a Likert scale with five response categories scored 1 (strongly disagree), 2 (disagree), 3 (agree), 4 (strongly agree) or missing (not applicable), with a possible total score of 13–52. These scores were then converted to an interval scale (0–100). A high score corresponds to a positive attitude towards necessary behavioural changes during the course of treatment. The Japanese version of the PAM-13 for mental health was used without including the words ‘mental health’ as stipulated by the scale's developer.26 In this study, the scale had an internal consistency of 0.85. Glycated haemoglobin HbA1c was determined through blood samples taken that day. Statistical analyses Descriptive statistics were calculated using means and SDs or numbers and percentages based on the nature of the variables. To ensure the validity of the classification rationale of the four psychological and behavioural patterns of stigma in patients with T2DM, we calculated self-stigma levels using the SSS-J and then tested the respective self-stigma level within each of the four groups using a one-way analysis of covariance (ANCOVA) after controlling for age and sex. We then calculated means and 95% CIs of PAM-13 scores and HbA1c levels for each group. We conducted a one-way ANCOVA to test for differences in mean scores of PAM-13 as well as those of HbA1c levels among the four groups, controlling for age and sex. Bonferroni correction was used to correct for multiple comparisons. All analyses were performed using SPSS V.23.0 (SPSS Japan, Tokyo, Japan). Results We recruited 259 patients with T2DM through physicians, and obtained written informed consent from 218 patients (response rate 84.2%). Among these participants, 217 completed the questionnaire (1 patient declined). In the analysis, we excluded five participants who chose ‘strongly disagree’ for all 39 items of the SSS-J based on their strong responses and uncertainty over whether the scale could measure what it was originally intended to assess. We also excluded three participants who chose ‘strongly agree’ for all 13 items of the PAM-13 as advised by the scale's developer. Therefore, 209 participants were included in our final analysis. The percentage of missing data was 0 for all questionnaire items. Table 1 shows the sociodemographic and clinical characteristics of 209 participants. There were 168 male participants (80.4%) and 41 female participants (19.6%), with a mean age of 60.2±10.1 years. Mean duration of T2DM was 159.1±113.8 months, mean BMI was 26.3±5.2 kg/m2 and mean HbA1c was 7.3±1.2% (56±13.1 mmol/mol). Mean number of diabetes-related complications was 0.57±0.86, and 34.9% of participants received injection therapy (insulin or other injectable medications). Mean number of years of education was 13.9±2.3. Most participants were married (72.2%). We assessed the equivalence of sociodemographic and clinical variables (eg, education, marital status, BMI, duration of T2DM, diabetes-related complications) between the four groups; however, we did not find any differences between the four groups. Table 1 Sociodemographic and clinical characteristics of participants (n=209) Patient characteristics N (%) or mean (±SD) Sex  Male 168 (80.4)  Female 41 (19.6) Age (years) 60.2 (±10.1) BMI (kg/m2) 26.3 (±5.2) Duration of diabetes (months) 159.1 (±113.8) Injection therapy  Yes 73 (34.9)  No 136 (65.1) Number of diabetes-related complications (0–6)* 0.57 (±0.86) HbA1c (%) 7.3 (±1.2) HbA1c (mmol/mol) 56 (±13.1) Education (years) 13.9 (±2.3) Marital status  Married 151 (72.2)  Unmarried/divorced/bereaved 58 (27.8) *The DCI. BMI, body mass index; DCI, Diabetes Complications Index; HbA1c , glycated haemoglobin. Figure 2 shows mean self-stigma levels of different psychological and behavioural patterns of stigma. Using ANCOVA after controlling for age and sex, there was a significant difference in self-stigma levels between the four groups (F[3203]=15.70, p<0.001). We observed the highest mean self-stigma levels in group D followed in decreasing order by groups C, B and A. The lowest mean self-stigma levels was group A. Multiple comparisons using the Bonferroni test showed significant differences in self-stigma levels between groups A and D (p<0.001). The test also showed that there were no significant differences in self-stigma levels between groups B and C (p=1.0). Figure 2 Mean self-stigma levels (SSS-J) of different psychological and behavioural patterns of stigma (n=209). ANCOVA, analysis of covariance; SF-36, 36-question Short Form Health Survey; SSS-J, Japanese version of Self-Stigma Scale. Figure 3 shows mean PAM-13 scores and HbA1c levels of different psychological and behavioural patterns of stigma. Using ANCOVA after controlling for age and sex, there was a significant difference in mean PAM-13 scores between the four groups (F[3203]=9.34, p<0.001). Multiple comparisons using the Bonferroni test showed significantly lower PAM-13 scores in group D compared with groups A (p<0.001) and B (p=0.02). As for HbA1c levels, using ANCOVA after controlling for age and sex, differences in means between the four groups showed a moderate increase from group A to group D, using ANCOVA (F[3203]=2.36, p=0.07); however, there was no significant difference between groups A and D (p=0.09). Figure 3 Mean patient activation scores (PAM-13) and HbA1c levels of different psychological and behavioural patterns of stigma (n=209). ANCOVA, analysis of covariance; HbA1c, glycated haemoglobin; PAM-13, Patient Activation Measure; SF-36, 36-question Short Form Health Survey. Discussion In this study, we aimed to test the psychological and behavioural patterns of stigma (self-esteem and social participation) and their relationship to self-stigma, patient activation and glycaemic control. The highest self-stigma levels were observed in group D, the group that exhibited low self-esteem and low social participation. The results concerning the association between low self-esteem and self-stigma are consistent with established theoretical perspectives on self-stigma.13 19 20 On the other hand, the results concerning the association between low social participation and self-stigma could be a novel finding regarding the characteristics of self-stigma among patients with T2DM. Thus, those who would fall in group D and exhibit that particular psychological and behavioural pattern may experience self-stigma and poorer patient activation for self-care. The strengths of this study include its ability to delineate the psychological and behavioural patterns of self-stigma by the two dimensions (self-esteem and social participation) derived from our previous qualitative study, and then use objective measures through the Rosenberg SES and the SF-36 subscales to classify patients with T2DM into these groups. This study suggests that self-stigma includes both psychological states (self-esteem) and behavioural manifestations (social participation) among patients with T2DM. Without our previous analyses using qualitative data, we would not have been able to find a new dimension—the behavioural dimension—of self-stigma, or elucidate both the psychological manifestations (self-esteem) and behavioural manifestations (levels of social participation) of self-stigma in patients with T2DM. This study has several limitations. First, since this study is cross-sectional, it can only demonstrate what the psychological and behavioural patterns of self-stigma look like, and their relationship to patient activation for self-care; however, we do not know any causal relationships concerning each variable. For the next step, future research will be needed to investigate the causality of these variables to conceptualise self-stigma in patients with T2DM. Second, the amount of men in our sample outweighed the amount of women (80.4% of the participants were male). Although the prevalence of T2DM is higher among men (24.0%) than women (13.4%) in Japan,29 it is not known exactly why we had such a large number of male participants compared with female participants. Third, this study did not include patients who were examined by primary care physicians in the community. Therefore, these findings do not capture the entire picture of patients with varying degrees of severity of T2DM. To verify the effects of self-stigma on the attitudes of patients with T2DM towards self-care management, further studies with a more representative population will be needed. Fourth, in this study, we used only three SF-36 subscales (RP, RE and SF) as suitable approximations to assess the level of social participation both physically and emotionally, although the subscales may not directly assess social participation itself. A new questionnaire needs to be developed with reliability and validity to assess social participation more accurately. Our findings can have some important implications. Self-stigma can be observed by healthcare professionals during their daily practice. To help optimise the effectiveness of diabetes treatment, healthcare professionals could regularly assess whether patients are experiencing self-stigma and then help them develop positive personal images of their illness to enhance self-esteem. Additionally, for any patients who seem to limit their social participation regardless of the severity of the condition, healthcare professionals could closely examine and actively listen to patients' accounts of recent social activities and the reasoning behind their low social participation, and encourage them to find opportunities for social interaction as an important part of their treatment. The authors would like to acknowledge all the physicians who helped recruit patients for this study. They are also grateful to the many graduate students who helped collect the data. Similarly, the authors would like to express our appreciation to all of the study participants. Contributors: AK conceptualised and designed the study. AK coordinated the study; acquired, analysed and interpreted the data; and prepared the paper. HH helped to analyse and interpret the data. AK and HH held primary responsibility for data access. YF, SF, AI, YO, RS, TY, KU and TK made significant contributions to the critical interpretation of results with regard to important practical content. All authors read and approved the final version of the manuscript. Funding: This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Competing interests: None declared. Patient consent: Obtained. Ethics approval: The Research Ethics Committee of the University of Tokyo Graduate School of Medicine and Faculty of Medicine, and each participating facility. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Schabert J , Browne JL , Mosely K Social stigma in diabetes: a framework to understand a growing problem for an increasing epidemic . Patient 2013 ;6 :1 –10 . 10.1007/s40271-012-0001-0 23322536 2 Tak-Ying Shiu A , Kwan JJ , Wong RY Social stigma as a barrier to diabetes self-management: implications for multi-level interventions . J Clin Nurs 2003 ;12 :149 –50 . 10.1046/j.1365-2702.2003.00735.x 12519263 3 Scollan-Koliopoulos M , O'Connell KA , Walker EA Legacy of diabetes and self-care behavior . Res Nurs Health 2007 ;30 :508 –17 . 10.1002/nur.20208 17893932 4 Weiler DM , Crist JD Diabetes self-management in a Latino social environment . Diabetes Educ 2009 ;35 :285 –92 . 10.1177/0145721708329545 19204101 5 Singh H , Cinnirella M , Bradley C Support systems for and barriers to diabetes management in South Asians and Whites in the UK: qualitative study of patients’ perspectives . BMJ Open 2012 ;14 :e001459 . 6 Browne JL , Ventura A , Mosely K ‘I call it the blame and shame disease’: a qualitative study about perceptions of social stigma surrounding type 2 diabetes . BMJ Open 2013 ;3 :e003384 10.1136/bmjopen-2013-003384 7 Winkley K , Evwierhoma C , Amiel SA Patient explanations for non-attendance at structured diabetes education sessions for newly diagnosed type 2 diabetes: a qualitative study . Diabet Med 2015 ;321 :120 –8 . 10.1111/dme.12556 8 Hallgren EA , McElfish PA , Rubon-Chutaro J Barriers and opportunities: a community-based participatory research study of health beliefs related to diabetes in a US Marshallese community . Diabetes Educ 2015 ;41 :86 –94 . 10.1177/0145721714559131 25398722 9 Lee SM , Lim LC , Koh D Stigma among workers attending a hospital specialist diabetes clinic . Occup Med (Lond) 2015 ;65 :67 –71 . 10.1093/occmed/kqu150 25342711 10 Della LJ , Ashlock MZ , Basta TB Social constructions of stigmatizing discourse around type 2 diabetes diagnoses in Appalachian Kentucky . Health Commun 2016 ;31 :806 –14 . 10.1080/10410236.2015.1007547 26605947 11 Corrigan PW , Watson AC The paradox of self-stigma and mental illness . Clin Psychol Sci Pr 2002 ;9 :35 –53 . 10.1093/clipsy.9.1.35 12 Rüsch N , Angermeyer MC , Corrigan PW Mental illness stigma: concepts, consequences, and initiative to reduce stigma . Eur Psychiatr 2005 ;20 :529 –39 . 10.1016/j.eurpsy.2005.04.004 13 Kato A , Takada M , Hashimoto H Reliability and validity of the Japanese version of the Self-Stigma Scale in patients with type 2 diabetes . Health Qual Life Outcomes 2014 ;12 :179 10.1186/s12955-014-0179-z 25495723 14 Mak WWS , Cheung RYM Self-stigma among concealable minorities in Hong Kong: conceptualization and unified measurement . Am J Orthopsychiatry 2010 ;80 :267 –81 . 10.1111/j.1939-0025.2010.01030.x 20553520 15 Kato A , Fujimaki Y , Fujimori S Association between self-stigma and self-care behaviors in patients with type 2 diabetes: a cross-sectional study . BMJ Open Diabetes Res Care 2016 ;4 :e000156 10.1136/bmjdrc-2015-000156 16 Herzlich A Health and illness: a social psychological analysis . London : Academic Press , 1973 . 17 Radley A Making sense of illness: the social psychology of health and disease . Thousand Oaks : Sage , 1994 . 18 Rosenberg M Society and adolescent self-image . Princeton : Princeton University Press , 1965 . 19 Link BG , Struening EL , Neese-Todd S Stigma as a barrier to recovery: the consequence of stigma for the self-esteem of people with mental illness . Psychiatr Serv 2001 ;52 :1621 –6 . 10.1176/appi.ps.52.12.1621 11726753 20 Corrigan PW , Watson AC , Barr L The self-stigma of mental illness: implications for self-esteem and self-efficacy . J Soc Clin Psychol 2006 ;25 :875 –84 . 10.1521/jscp.2006.25.8.875 21 Kato A , Fujimaki Y , Fujimori S A qualitative study on the impact of internalized stigma on type 2 diabetes self-management . Patient Educ Couns 2016 ;99 :1233 –9 . 10.1016/j.pec.2016.02.002 27873575 22 Fincke BG , Clark JA , Linzer M Assessment of long-term complications due to type 2 diabetes using patient self-report: the diabetes complications index . J Ambul Care Manage 2005 ;28 :262 –73 . 10.1097/00004479-200507000-00010 15968219 23 Mimura C , Griffiths P Japanese version of the Rosenberg Self-esteem Scale: translation and equivalence assessment . J Psychosom Res 2007 ;62 :589 –94 . 10.1016/j.jpsychores.2006.11.004 17467414 24 Fukuhara S , Suzukamo Y Health-related quality of life-SF8 and SF36 . J Clin Exp Med 2005 ;213 :133 –6 . (SF-36v2TM Health Survey© 1992, 2000, 2003 Quality Metric Incorporated, Medical Outcomes Trust and Shunichi Fukuhara. All rights reserved. SF-36® is a registered trademark of Medical Outcomes Trust. (SF-36v2 Standard, Japanese ). 25 Fujita E , Kuno E , Kato D Development and validation of the Japanese version of the Patient Activation Measure 13 for Mental Health . Seishin Igaku (Clin Psychiatr) 2010 ;52 :765 –72 . 26 Hibbard JH , Mahoney ER , Stockard J Development and testing of a short form of the Patient Activation Measure (PAM) . Health Serv Res 2005 ;40 :1918 –30 . 10.1111/j.1475-6773.2005.00438.x 16336556 27 Hibbard JH , Stockard J , Mahoney ER Development of the Patient Activation Measure (PAM): conceptualizing and measuring activation in patient and consumers . Health Serv Res 2004 ;39 :1005 –26 . 10.1111/j.1475-6773.2004.00269.x 15230939 28 Hibbard JH , Mahoney ER , Stock R Do increases in patient activation result in improved self-management behaviors? Health Serv Res 2007 ;42 :1443 –63 . 10.1111/j.1475-6773.2006.00669.x 17610432 29 Mukai N , Doi Y , Ninomiya T Trends in the prevalence of type 2 diabetes and prediabetes in community-dwelling Japanese subjects: the Hisayama Study . J Diabetes Investig 2014 ;23 :162 –9 . 10.1111/jdi.12136

PMC005xxxxxx/PMC5372019.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01388010.1136/bmjopen-2016-013880Cardiovascular MedicineProtocol1506168316921699Northern Shanghai Study: cardiovascular risk and its associated factors in the Chinese elderly—a study protocol of a prospective study design Ji Hongwei Xiong Jing Yu Shikai Chi Chen Fan Ximin Bai Bin Zhou Yiwu Teliewubai Jiadela Lu Yuyan Xu Henry Zhang Yi Xu Yawei Department of Cardiology, Shanghai Tenth People's Hospital, Tongji University School of Medicine, Shanghai, ChinaCorrespondence to Dr Yi Zhang; yizshcn@gmail.com and Yawei Xu; yaweixu@aliyun.comHJ and JX contributed equally to this work. 2017 29 3 2017 7 3 e01388015 8 2016 31 1 2017 2 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction Cardiovascular (CV) diseases are the leading cause of death and disability in the world. Increasing lifespans and ageing populations also contribute to an increasing CV burden. However, in China, there were few well-designed cohort studies focusing on the elderly population, let alone an established CV risk score. The objective of this study is to establish a CV risk score based on a community-dwelling Chinese elderly population, determining the profile of the associated CV risk factors and target organ damages (TODs), so as to guide the later intervention. Methods and analysis The Northern Shanghai Study is an ongoing prospective community-based study. After enrolment, clinical examination, anthropometric measurement and a questionnaire will be administered to each participant at baseline and after every 2 years in the follow-up. Our tests and examinations include: blood/urine sample and biochemical measurements, office blood pressure recording, carotid ultrasonograph, echocardiograph, pulse wave velocity, pulse wave analysis, 4-limb blood pressure recording, body mass index, etc. Baseline measurement will also include the assessments on TODs and the conventional CV risk factors. In the follow-up, the incidence of CV events and mortality will be recorded. The Northern Shanghai Risk Score will be calculated, with considerations on CV risk factors and TODs. Ethics and dissemination This study was approved by the Shanghai Tenth People's Hospital Institutional Review Board. All participants signed a written consent form. Trial registration number NCT02368938; Pre-results. target organ damageChinese elderly ==== Body Strengths and limitations of this study This study is one of the largest ongoing prospective population studies to evaluate target organ damages (TODs) in the community-dwelling elderly Chinese, which is authorised and funded by the Shanghai municipal government. A systematic framework of a cardiovascular (CV) risk survey was conducted, with considerations on conventional CV risk factors, asymptomatic TODs, CV diseases, future events and mortality. Left ventricular ejection fraction was measured by the M-mode echocardiograph, but not by Simpson's method. The risk score is based on the elderly participants, who are already at significant risk due to their old age. However, collected data and the later risk score can also be used for validation in younger cohorts. Some measurements in this study need highly specialised equipment, which is hard to scale to a larger population. Background Cardiovascular diseases (CVDs), as ageing-related and chronic disorders, carry a high morbidity. It is one of the most common and deadliest diseases in the world.1 2 According to a report from the WHO, CVD is responsible for 17.5 million deaths annually worldwide.3 It was the leading cause of death and reduction in a human's expected lifespan.4 In 2012, about 3.5 million deaths in China were attributable to CVDs, which means there would be a CVD death every 10 s in China.5 Among all the disease-associated deaths in China, it was concluded that over 40% of them were due to CVDs.6 Hypertension is considered as the major contributor to CVDs.7 A report indicated that the prevalence of hypertension in Chinese adults increased from 18.8% in 2002 to 24.4% in 2012, while the control rate increased only from 6.1% to 9.3%.6 Furthermore, according to a recent large-scale survey in China with 205 167 men (41.0%) and 295 056 women (59.0%),8 32.5% of the cohort participants had hypertension with an overall control rate of only 4.2%. This great challenge is due in part to the absence of a domestic CV risk score in China. Given the perniciousness of CVDs, an established CV risk score is essential to guide prevention and therapy in China.9 10 In fact, in the USA and Europe, many well-known population studies with great professional achievements were conducted, and some mature CV risk score systems have already been established and applied efficiently, such as the Framingham Risk Score and the European SCORE Risk Charts.11 12 These studies pushed the transition from a poor understanding of CVDs to a more mature one going forward.13 The current risk scores in the USA and Europe are based on a mainly Caucasian general population. However, in China, things would be different. According to the World Population Ageing 2013, it will take China only 26 years to experience the population ageing, which means there is a rapid ageing trend in China. On the other hand, China also had the most rapid urbanisation in history. These two trends will interact in important ways with each other and will have a profound effect on Chinese CV health. Therefore, it would be inappropriate to apply those risk scores directly in the Chinese elderly. In Shanghai, the biggest urbanised city in China, the proportion aged over 60 years is 28.8% (Elderly population and cause of aging monitoring statistics of Shanghai in 2014 Accession Number: http://www.shmzj.gov.cn/Attach/Attaches/201506/20150610104009609.doc). In this respect, Shanghai could be a good representative of the future Chinese population, with the deep urbanisation and the geriatric population. We therefore selected community-based citizens in Shanghai as our target population. The characteristics and successful experience on the CV risk control in this Shanghai geriatric population, acting as an exemplary role, could be extrapolated to the future Chinese society. It means that effective interventions for the future Chinese would be designed 10–20 years sooner. Established CV risk prediction models are mainly based on the conventional risk factors such as age, sex, blood pressure, cholesterol, etc. However, if based only on the conventional risk factors, there is a fall in the predictive abilities of future risk in the older population.14 In the elderly, novel biomarkers are warranted to improve the risk stratification instead of relying on a model that is based only on established risk factors.15 Asymptomatic target organ damage (TOD), as an intermediate state between risk factors and clinical events, may be a good marker for risk stratification in the elderly.16 It might better represent exposure to risk factors than the risk factor itself.17 Therefore, we would like to add valuable TODs, together with conventional risk factors, into the risk assessment model in the elderly. As mentioned above, China is lacking in a well-established domestic CV risk score system at present, and prevention strategies as well as treatments for CVDs need significant improvement. There is an urgent desire in China to establish a CV risk score based on a Chinese population study, especially for the elderly. So we will perform a systematical framework of CV risk assessment for community-based elderly participants (>65 years) in the northern Shanghai area. The assessments include conventional CV risk factors, TODs and related diseases. Our objective is to establish a Chinese CV risk score, the Northern Shanghai Risk Score, to guide future risk assessments and interventions for the elderly in China. This paper is to describe the design and method plan for this study. Method Study design and sample size The Northern Shanghai Study is a prospective community-based ongoing study. This study was approved by the Shanghai Tenth People's Hospital Institutional Review Board and was conducted under financial support from the Shanghai municipal government (grant ID: 2013ZYJB0902 and 15GWZK1002). The preliminary sample size is expected to be 3000–4000 participants. Participant eligibility criteria The inclusion criteria include: (1) age 65 years or more; (2) informed consent should be signed voluntarily; (3) local residents from communities in northern Shanghai and (4) available for long-term follow-up. The individual should be excluded, if the individual: (1) was diagnosed with serious heart disease (NYHA≥IV) or end-stage renal disease (CKD ≥4 stage); (2) suffered from cancer or his/her life expectancy is <5 years; (3) had stroke within 3 months; (4) is not willing to participate in the clinical study; (5) has to quit the trial due to other diseases; (6) violates the protocol or (7) loses contact with the laboratory staff. Recruitment First, according to the Elderly population and cause of aging monitoring statistics of Shanghai in 2014, northern Shanghai region has the largest population of elderly adults in Shanghai, with a total population of 1.57 million and an elderly proportion of over 19%. Thus, northern Shanghai region, including Zhabei district and Putuo district, was selected from Shanghai. Second, we use a computer-generated list of communities, and 10 communities were randomly selected for the first-phase enrolment. Other communities in the list will be randomly selected for the later enrolment. According to the inclusion and exclusion criteria, we invite all the eligible older people (over 65 years) to participate in this study. The recruitment strategies include: (1) posting study recruitment files in the neighbourhood committees and community hospitals; (2) according to the health file, community hospitals recruit the potential participants by telephone; (3) hand out recruitment flyers directly to the potential participants. Contact number is included in all the recruitment files and flyers. Before data collection, the field staff will give a brief oral questionnaire according to the inclusion and exclusion criteria. When the eligible individual shows their interest in participating in this study, the individual will be sufficiently informed, and they will sign the consent form. Social, clinical and biological parameters Information is obtained from the standardised questionnaire, including gender, age, education level, smoking habits, drinking habits, history of diabetes, renal insufficiency and CVD. CVDs include chronic heart failure, peripheral vascular disease, hypertension, arrhythmia and previous CV event (the presence of a history of myocardial infarction (MI) and/or stroke and/or cardiac revascularisation with either angioplasty or coronary artery bypass graft (CABG)). Participants must disrobe and remove shoes and stand straight before their body height and body weight are measured. Waist circumference and hip circumference are measured by a flexible rule, with waistline and hipline referring to the smallest waist and the greatest circumferences, respectively. The body mass index is calculated by dividing body weight (kg) by the squared body height (m2). Venous blood samples are obtained after an overnight fast. Total cholesterol, high-density lipoprotein cholesterol and triglycerides are measured by standard methods,18 19 and the Friedewald formula is used to calculate the low-density lipoprotein (LDL) cholesterol (LDL-c).20 Other biological parameters like plasma/urine albumin and creatinine are measured by standard methods at local laboratories. The urine albumin-to-creatinine ratio (UACR) is also calculated. The serum and urine samples will be stored at −80°C. Prior to storage, the date, number of vials and recorder's name will be recorded. Office blood pressure measurement After an overnight fast, brachial blood pressure is measured in the morning with the participants' bladder empty, free of tobacco or caffeine for at least 30 min before the measurement. The blood pressure is measured in the sitting position, after resting for 5 min, using a semiautomatic oscillometric device (Omron Healthcare, Kyoto, Japan), according to the recommendations of the European Society of Hypertension.21 The average value is calculated for further analysis. Ultrasonography All the ultrasonography measurements, including echocardiography and carotid ultrasonography, are performed by a single experienced cardiologist, who is unaware of previous results. All measurements are performed with a MyLab 30 CV machine (ESAOTE SpA, Genoa, Italy), according to the American Society of Echocardiography (ASE) recommendations.22 The echocardiography is performed in the left decubitus position. Left ventricular (LV) internal diameter at end-diastole (LVIDd) and septal (SWTd) and posterior wall thickness at end-diastole (PWTd) are measured directly. The formula: is used to calculated the LV mass, based on modelling the LV as a prolate ellipse of revolution.22 The left ventricular ejection fraction (LVEF) is calculated by Teichholz's formula, and the left atrium (LA) size is measured in the parasternal long axis and apical four-chamber views. Left atrial volume is calculated using the ellipse model formula: In this equation, SA1 is the M-mode left atrial dimension in the parasternal short-axis view and SA2 and LA are measurements of short and long axes in the apical four-chamber view at ventricular end-systole.23 The heart diastolic function is also measured, including the peak E (early diastolic), peak A (late diastolic) velocities and the primary early diastolic velocities (Ea) with the PW and TDI Doppler. The primary early diastolic velocities (Ea) are measured by the lateral tissue Doppler signals. Carotid ultrasonography is evaluated at common carotid arteries of both sides using a 7.5 MHz transducer. Carotid artery intima-media thickness (CIMT) is measured on the left common carotid artery, 2 cm from the bifurcation, and is always performed on plaque-free arterial segments. Intima-media thickness (IMT) is measured manually. The border is determined from changes of density of the section which is perpendicular to the vessel wall. Common, internal and external carotid arteries are all scanned longitudinally and transversely to determine the presence of plaques. The plaque is defined as IMT of the internal carotid artery of more than 1.5 mm24 or a localised echo-structure encroaching into the vessel lumen with the arterial wall above 50% thicker than neighbouring sites. The measuring process is performed by the same sonographer as the echocardiography. Four-limb blood pressure measurement Four-limb blood pressures of participants are measured by VP-1000 (Omron Healthcare, Kyoto, Japan) automatically, performed by trained staff. Bilateral ankle brachial index (ABI), the ratio of the ankle systolic blood pressure (SBP) divided by the brachial SBP, could be read from the device and the lower ABI is applied for further analysis in subsequent studies. Pulse wave velocity Pulse wave velocity (PWV), which can be estimated by the SphygomoCor device (AtCor, Australia), is measured in a defined segment to assess the arterial stiffness.25 It is recommended that the arterial stiffness should be determined non-invasively by the measurement of carotid-femoral PWV (cf-PWV) (Class I; Level of Evidence A) as a golden standard.26 27 The measurement is performed with applanation tonometry (SphygmoCor, AtCor Medical, Australia), by two trained observers blinded to the other results according to the European Expert Consensus on Arterial Stiffness.28 Participants need to rest quietly in a temperature-controlled room for at least 10 min prior to the initial pulse pressure waveform measurements. The pulse analysis will be performed with sensors in the right radial, carotid and femoral arteries in a supine position. Recordings were made simultaneously with an ECG signal, which provided an R-timing reference. Simultaneously, the delay is estimated and the cf-PWV is calculated by the integral software automatically (m/s). The superficial distance covered by the pulse wave will be taken with a tape from the suprasternal notch to the carotid and femoral arteries at the sensor location.29 An operator index >80% is considered as a reliable measurement, in which the quality and reproducibility of the tonometry measurements are automatically tested. Pulse wave analysis Pulse wave analysis, which can be observed on the commonly used device—SphygomoCor device (AtCor, Australia),30 is measured to estimate central haemodynamic parameters.31 SphygmoCor is used to perform the applanation tonometry on a radial artery with the methodology previously described.30 After a 10 min rest in the supine position, the brachial BP is obtained with the SphygomoCor device. Then a radial waveform is recorded by one trained and experienced physician with a tonometry-based probe. The central waveform is estimated by the inbuilt software, automatically, with the help of a validated generalised transfer function. Eventually, the central waveform is calibrated by the calculated brachial mean and diastolic blood pressure (DBP), in order to obtain the central SBP and DBP. The SphygmoCor device provides a quality index, and only PWs with an operator index above 80 are accepted. The data are accepted only when a variation of heart rate is no greater than 5%. Electrocardiography The 12-lead electrocardiograph is recorded at 25 mm/s and 1 mV/cm standardisation with standard equipment after at least a 5 min rest. Electrocardiographic QRS wave voltage is detected in this study. Parameters are recorded and calculated, including the voltage of the S wave of the chest lead V1 (SV1), S wave of the lead V3 (SV3), R wave of the lead V5 (RV5), R wave of the lead aVL (RaVL) and the duration of QRS wave. Several indexes to distinguish the LV hypertrophy are as follows: Sokolow-Lyon-Rappaport index (SV1 or SV2+RV5 or RV6≥4.0 mv in men and SV1 or SV2+RV5 or RV6≥3.5mv in women), Cornell criterion (SV3+RaVL≥2.8mv in men and SV3+RaVL ≥2.8 mv in women) and Cornell Product ((SV3+RaVL)×QRS duration≥244mv·ms in men and (SV3+RaVL+0.6)×QRS duration≥244 mv·ms in women). Evaluation of peripheral artery involvement The ABI is used to evaluate the peripheral artery involvement.32 Brachial–ankle index and brachial–ankle PWV are assessed automatically by inbuilt software using a VP-1000 device (Omron, Japan). This measurement is performed in the morning without coffee or tobacco for at least 8 hours prior to measuring and in an ambient temperature of 22–24°C. Definition of TODs Generally, asymptomatic TODs include cardiac, arterial and renal TODs. LV hypertrophy is defined as LVMI ≥115 g/m2 (men) or LVMI ≥95 g/m2 (women).33 As for the arterial TODs, they are defined as increased CIMT (CIMT >0.9 mm) or peripheral artery disease (ABI<0.9). Chronic kidney diseases (creatinine clearance rate (CCR) <60 mL/min/1.73 m2) and microalbuminuria (UACR >30 mg/mmol) represent renal TODs. Specifically, LV diastolic dysfunction is present when ≥3 listed variables meet these cut-off values (septal e′, 7 cm/s, lateral e′, 10 cm/s, average E/e′ ratio, 14, LA volume index, 34 mL/m2).34 Clinical outcome The primary outcome is a composite of major adverse cardiovascular events (MACE), including CV death, non-fatal stroke, non-fatal MI or revascularisation (percutaneous coronary intervention (PCI) or CABG). Non-fatal stroke is defined as the new onset of neurological deficiency symptoms or signs lasting for at least 24 hours accompanied by evidence from either cranial CT or MRI. Non-fatal MI is defined by canonical chest pain symptoms and/or characteristic electrocardiographic changes with a rise of either troponin I >1.0 ng/mL or troponin T >0.1 ng/mL. Coronary revascularisation with either PCI or CABG is defined as a history of either stent implantation (PCI) or CABG. A death diagnosis is identified on the basis of a death certificate given by the related hospital. The end point is defined by a combination of a questionnaire, dropping-in follow-up and review of the computerised medical records of clinic visits and hospitalisations, conducted by a blinded end point evaluating committee. The secondary outcome includes: (1) subclinical organ damage which is defined as LV hypertrophy or decreased diastolic function or increased carotid IMT and/or the presence of plaque or hardening of the arteries or renal insufficiency (CKD3 period) or increased microalbuminuria; (2) newly diagnosed CV or cerebrovascular disease (including newly diagnosed hypertension, transient ischaemic attack, etc), or newly diagnosed renal insufficiency with proteinuria or newly diagnosed diabetes mellitus (table 1). Table 1 Baseline visit and patient follow-up Time-points Measure Baseline Every 2 years Every 5 years Consent form ● ● ● Baseline questionnaire (age, gender, smoking history, family history, medication history, symptoms and signs of HF) ● ● Follow-up questionnaire (including newly diagnosed cardiovascular or cerebrovascular events, kidney disease and DM) ● ● Body height, body weight, body mass index (BMI) and waist circumference and hip circumference ● ● Four-limb blood pressure measurement ● ● Office blood pressure measurement (3 times in a row) ● ● ● Venous blood biochemical parameters (blood glucose, blood lipid profile, serum creatinine and uric acid, pro-BNP, homocysteine) ● ● ● Urinalysis (urine microalbumin and urine creatinine) ● ● Blood and/or urine sample collection ● ● ● Electrocardiogram (rhythm, SV1+RV5) ● ● Vascular ultrasonography (bilateral carotid IMT) ● ● Echocardiography (LVM, LAV, LVEF, E/Ea, E/A) ● ● Determination of arterial elasticity (PWA, PWV) ● ● Evaluation of peripheral artery involvement ● ● ● Major adverse cardiovascular events ● ● Cardiovascular deaths ● ● All-cause deaths ● ● Statistical analytic approach for primary aim Survival curves are generated by the Cox proportional hazards regression model and survival among groups will be compared using the log-rank test. The receiver operating characteristic (ROC) curve is used to evaluate the effect of influential factors on the occurrence of MACEs. A two-sided significance level of 5% is defined as the level of statistical significance. The analyses are conducted with SAS software, V.9.3 (SAS Institute, Cary, North Carolina, USA). Development of the risk score at baseline and follow-up Individuals will be randomly selected to be exploratory and validation set. The score will be created on the exploratory set and tested on the validation set. Risk factors including TODs and conventional risk factors are selected. The TODs with a predictable value for cardiovascular events (CVE) are LV hypertrophy, arterial stiffening, carotid hypertrophy, lower limb atherosclerosis, microalbuminuria and renal function decline.35 The conventional risk factors for CVE include age, gender, smoking, obesity, diabetes, hypertension, blood glucose and lipid profile. The independent variables are defined and categorised as follows: LV hypertrophy: LV mass index >115 g/m2 for men and >95 g/m2 for women; arterial stiffening: cf-PWV>12 m/s; carotid hypertrophy: IMT ≥0.9 mm; lower limb atherosclerosis: ABI ≤0.9; microalbuminuria: UACR >30 mg/mmol; renal function decline: stage 2 CKD: eGFR 60–89 mL/min/1.73 m2 (MDRD); stage 3 or more CKD: eGFR <60 mL/min/1.73 m2 (MDRD); age is categorised into two groups: 65–80 years and over 80 years; obesity: BMI≥28.0 kg/m2. To estimate significant predictors of CVE, univariate analyses will be performed. In principle, only parameters with inflated coefficients are entered into the model. At baseline, considering that TODs are the basics of this model, multiple logistic regression (MLR) will be used to calculate β-coefficient of risk factors for CVE and to compare coefficients between TODs and conventional risk factors. Considering CVE as the dependent variable, variables significant at 5% will be included in MLR with stepwise backward elimination.36 At follow-up, the β coefficient in the Cox regression of each independent prognostic variable will be modified into an integral number to construct a prognostic score model (ie, exp (β)=HR).37 A p value of ≤0.05 is considered significant. In the scoring system, based on the magnitude of its regression coefficient, points will be assigned to each variable. Finally, by adding the score for each variable in the risk model, a sum score will be calculated for each participant. An ROC curve and area under the curve (AUC) will be assessed to stratify patients at a high risk of CVE. Sensitivity and specificity will be calculated for each cut-off score. The cut-off score with a maximum Youden index will be considered as the optimum. Validation of the risk score Since individuals are randomly selected to be exploratory and validation set, the performance of the risk score will be evaluated in the validation set as well as the entire sample. The predictive performance of the risk score will be evaluated with the AUCs in ROC curves, including sensitivity and specificity. During the evaluation, net reclassification indices will be used to measure the improvement of risk estimation by classifying individuals to a more correct category.38 Furthermore, the proportion of individuals who have a score above the optimal cut-off value in the risk score will be compared with those with a low risk score. Data entry and management of data files All data are entered into computerised database with SAS software, V.9.3 (SAS Institute, Cary, North Carolina, USA). Values that are out of range or represent errors of faulty logic are avoided by double check. Discussion Until now, the Northern Shanghai Study is one of the largest Chinese domestic population studies. We are aiming at building a CV risk score to guide the future risk assessments and interventions for the elderly Chinese. Since the burden from the chronic diseases is growing,9 10 our study will also contribute to Chinese CV health by establishing the CV profile of the Chinese ageing population. In the literature, risk predictive models have been established in various populations and in different settings.39 Though the ability to predict the occurrence of future events in old persons has been studied, there are few studies that have published a risk estimation system which can be used to calculate risks in this age group in clinical practice.40 Some well-established CV risk scores from famous population studies were widely adopted, such as the Framingham Risk Score and the European SCORE Risk Charts.11–13 In fact, most of them were conducted in the general population focusing on conventional risk factors such as smoking, blood pressure, lipid profile, glucose level, etc. However, for the elderly, the long-term exposure and accumulated microdamages from conventional risk factors have been converted into TODs. In this respect, just considering the conventional risk factors in the risk assessment strategy for the elderly may be inadequate.40 Taking a ‘70 years old chain smoker’ with severe atherosclerosis, for example, we prefer to reverse, terminate or at least control the process of atherosclerosis (TOD), instead of just advising him to quit his long-term formed smoking. For this patient, the CV risk from long-term exposure to smoking has turned into severe atherosclerosis. The intervention of smoking cessation would be less beneficial compared with the lipid-lowering therapy. For the Chinese elderly, some TODs are more likely to be reversible than the inveterate risk factors like smoking. Actually, many asymptomatic TODs have been proved to be modifiable by medications, even in the late stage. For example, angiotensin receptor blockade (losartan) has been validated for reversing cardiac hypertrophy.41 Therefore, we suggest the transition of the CV risk assessment from conventional risk factors (like age, gender and smoking) to the combination of asymptomatic TODs and risk factors in the elderly, which might be more compatible with the ageing population. In summary, we propose to establish a CV risk score system based on the TODs and conventional CV risk factors, focusing on the elderly. In this way, we can provide a more accurate CV assessment as well as a more effective guidance for treatment and intervention. Meanwhile, we may have a chance to provide the Chinese policymakers and opinion leaders with constructive suggestions regarding effective countermeasures to the national CV burden. Of note, we select Shanghai as the representative region, because the current proportion of those aged over 60 in Shanghai is 28.8%, which is similar to the estimated 28.1% of future China in 2040 (Elderly population and cause of aging monitoring statistics of Shanghai in 2014).42 At the baseline analysis, this study seeks to show the CV profile of about 4000 participants, and to determine subclinical TODs with conventional CV risks. We will analyse the change of the TOD indicators at every 2–5 years, and we will conduct new enrolment at the same time. Additionally, it is also expected that the subsequent follow-up studies in the Northern Shanghai Study will reveal the feasibility and necessity of establishing the Northern Shanghai Score. We believe that the CV risk score based on TODs and CV risk factors for the elderly Chinese can make a better prediction for future CV events, providing a more feasible intervention. Conclusion This protocol outlines the design and method of the Northern Shanghai Study. Results coming from this study will be used to construct the Northern Shanghai Risk Score, so as to guide the future assessments and interventions. Contributors: HJ, JX, SY, CC, XF, BB, YZho, JT, YL and YZha acquired the original data for this study. YZha and YX formulated the methods and designed the protocol. HJ and JX drafted the manuscript. HX helped us with the writing and language review. All authors contributed to revisions and approved the final version of the manuscript. Funding: This framework of cardiovascular risk assessment is conducted with financial support from the Shanghai municipal government (grant ID. 2013ZYJB0902 and 15GWZK1002). YZha was supported by the National Nature Science Foundation of China (grant ID: 81300239 and 81670377). Competing interests: None declared. Ethics approval: Shanghai Tenth People's Hospital Institutional Review Board. Provenance and peer review: Not commissioned; externally peer reviewed. ==== Refs References 1 Organization WH Global Atlas on cardiovascular disease prevention and control . World Health Organization , 2011 . 2 Mozaffarian D , Benjamin EJ , Go AS Heart disease and stroke statistics—2015 update: a report from the American Heart Association . Circulation 2015 ;131 :e29 –322 . 10.1161/CIR.0000000000000152 25520374 3 Organization WH GLOBAL STATUS REPORT on noncommunicable diseases . World Health Organization , 2014 . 4 Yang G , Wang Y , Zeng Y Rapid health transition in China, 1990–2010: findings from the Global Burden of Disease Study 2010 . Lancet 2013 ;381 :1987 –2015 . 10.1016/S0140-6736(13)61097-1 23746901 5 Xi B , Liu F , Hao Y The growing burden of cardiovascular diseases in China . Int J Cardiol 2014 ;174 :736 –7 . 10.1016/j.ijcard.2014.04.098 24794557 6 HSGRL L Report on cardiovascular diseases in China 2014 China : National Center for Cardiovascular Diseases , 2014 . 7 Organization WH A global brief on hypertension: silent killer, global public health crisis . World Health Day 2013 Report, 1–39 2013 . 8 Lewington S , Lacey B , Clarke R The burden of hypertension and associated risk for cardiovascular mortality in China . JAMA Intern Med 2016 ;176 :524 –32 . 10.1001/jamainternmed.2016.0190 26975032 9 Jaffe S 50 years of Medicare . Lancet 2015 ;386 :419 –20 . 10.1016/S0140-6736(15)61433-7 26251380 10 Rechel B , Grundy E , Robine JM Ageing in the European Union . Lancet 2013 ;381 :1312 –22 . 10.1016/S0140-6736(12)62087-X 23541057 11 Tsao CW , Vasan RS Cohort profile: the Framingham Heart Study (FHS): overview of milestones in cardiovascular epidemiology . Int J Epidemiol 2015 ;44 :1800 –13 . 10.1093/ije/dyv337 26705418 12 Conroy RM , Pyörälä K , Fitzgerald AP Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project . Eur Heart J 2003 ;24 :987 –1003 . 10.1016/S0195-668X(03)00114-3 12788299 13 Mahmood SS , Levy D , Vasan RS The Framingham Heart Study and the epidemiology of cardiovascular disease: a historical perspective . Lancet 2014 ;383 :999 –1008 . 10.1016/S0140-6736(13)61752-3 24084292 14 de Ruijter W , Westendorp RG , Assendelft WJ Use of Framingham risk score and new biomarkers to predict cardiovascular mortality in older people: population based observational cohort study . BMJ 2009 ;338 :a3083 10.1136/bmj.a3083 19131384 15 Zethelius B , Berglund L , Sundstrom J Use of multiple biomarkers to improve the prediction of death from cardiovascular causes . N Engl J Med 2008 ;358 :2107 –16 . 10.1056/NEJMoa0707064 18480203 16 Vernooij JW , van der Graaf Y , Nathoe HM Hypertensive target organ damage and the risk for vascular events and all-cause mortality in patients with vascular disease . J Hypertens 2013 ;31 :492 –9 . discussion 499–500 10.1097/HJH.0b013e32835cd3cd 23303394 17 van der Veen PH , Geerlings MI , Visseren FL Hypertensive target organ damage and longitudinal changes in brain structure and function: the Second Manifestations of Arterial Disease-Magnetic Resonance Study . Hypertension 2015 ;66 :1152 –8 . 10.1161/HYPERTENSIONAHA.115.06268 26503971 18 Grundy SM , Cleeman JI , Merz CN Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines . Circulation 2004 ;110 :227 –39 . 10.1161/01.CIR.0000133317.49796.0E 15249516 19 National Cholesterol Education Program Expert Panel on Detection Evaluation, and Treatment of High Blood Cholesterol in Adults . Third report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report . Circulation 2002 ;106 :3143 –421 .12485966 20 Friedewald WT , Levy RI , Fredrickson DS Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge . Clin Chem 1972 ;18 :499 –502 .4337382 21 O'Brien E , Asmar R , Beilin L Practice guidelines of the European Society of Hypertension for clinic, ambulatory and self blood pressure measurement . J Hypertens 2005 ;23 :697 –701 .15775768 22 Lang RM , Bierig M , Devereux RB Recommendations for chamber quantification: a report from the American Society of Echocardiography's Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology . J Am Soc Echocardiogr 2005 ;18 :1440 –63 .16376782 23 Zhang Y , Li Y , Liu M Cardiac structure and function in relation to cardiovascular risk factors in Chinese . BMC Cardiovasc Disord 2012 ;12 :86 10.1186/1471-2261-12-86 23035836 24 Polak JF , Pencina MJ , Pencina KM Sr Carotid-wall intima-media thickness and cardiovascular events . N Engl J Med 2011 ;365 :213 –21 .21774709 25 Townsend RR , Wilkinson IB , Schiffrin EL Recommendations for improving and standardizing vascular research on arterial stiffness: a scientific statement from the American Heart Association . Hypertension 2015 ;66 :698 –722 . 10.1161/HYP.0000000000000033 26160955 26 Chirinos JA Arterial stiffness: basic concepts and measurement techniques . J Cardiovasc Transl Res 2012 ;5 :243 –55 . 10.1007/s12265-012-9359-6 22447229 27 Laurent S , Cockcroft J , Van Bortel L Expert consensus document on arterial stiffness: methodological issues and clinical applications . Eur Heart J 2006 ;27 :2588 –605 . 10.1093/eurheartj/ehl254 17000623 28 Van Bortel LM , Laurent S , Boutouyrie P Expert consensus document on the measurement of aortic stiffness in daily practice using carotid-femoral pulse wave velocity . J Hypertens 2012 ;30 :445 –8 . 10.1097/HJH.0b013e32834fa8b0 22278144 29 Van Bortel LM Is arterial stiffness ready for daily clinical practice? J Hypertens 2006 ;24 :281 –3 . 10.1097/01.hjh.0000199805.03058.78 16508572 30 Agnoletti D , Millasseau SC , Topouchian J Pulse wave analysis with two tonometric devices: a comparison study . Physiol Meas 2014 ;35 :1837 –48 . 10.1088/0967-3334/35/9/1837 25154391 31 Wang KL , Cheng HM , Chuang SY Central or peripheral systolic or pulse pressure: which best relates to target organs and future mortality? J Hypertens 2009 ;27 :461 –7 .19330899 32 Fowkes FG , Rudan D , Rudan I Comparison of global estimates of prevalence and risk factors for peripheral artery disease in 2000 and 2010: a systematic review and analysis . Lancet 2013 ;382 :1329 –40 . 10.1016/S0140-6736(13)61249-0 23915883 33 Zhang Y , Kollias G , Argyris AA Association of left ventricular diastolic dysfunction with 24-h aortic ambulatory blood pressure: the SAFAR study . J Hum Hypertens 2015 ;29 :442 –8 . 10.1038/jhh.2014.101 25391758 34 Nagueh SF , Smiseth OA , Appleton CP Recommendations for the evaluation of left ventricular diastolic function by echocardiography: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging . Eur Heart J Cardiovasc Imaging 2016 ;17 :1321 –60 . 10.1093/ehjci/jew082 27422899 35 Violi F , Pastori D , Perticone F Relationship between low ankle-brachial index and rapid renal function decline in patients with atrial fibrillation: a prospective multicentre cohort study . BMJ Open 2015 ;5 :e008026 10.1136/bmjopen-2015-008026 36 Swartz MD , Yu RK , Shete S Finding factors influencing risk: comparing Bayesian stochastic search and standard variable selection methods applied to logistic regression models of cases and controls . Stat Med 2008 ;27 :6158 –74 . 10.1002/sim.3434 18937224 37 Case LD , Kimmick G , Paskett ED Interpreting measures of treatment effect in cancer clinical trials . Oncologist 2002 ;7 :181 –7 . 10.1634/theoncologist.7-3-181 12065789 38 Kerr KF , Wang Z , Janes H Net reclassification indices for evaluating risk prediction instruments: a critical review . Epidemiology 2014 ;25 :114 –21 . 10.1097/EDE.0000000000000018 24240655 39 Siontis GC , Tzoulaki I , Siontis KC Comparisons of established risk prediction models for cardiovascular disease: systematic review . BMJ 2012 ;344 :e3318 10.1136/bmj.e3318 22628003 40 Cooney MT , Selmer R , Lindman A Cardiovascular risk estimation in older persons: SCORE O.P . Eur J Prev Cardiol 2016 ;23 :1093 –103 . 10.1177/2047487315588390 26040999 41 Okin PM , Devereux RB , Jern S Regression of electrocardiographic left ventricular hypertrophy by losartan versus atenolol: the Losartan Intervention for Endpoint reduction in Hypertension (LIFE) Study . Circulation 2003 ;108 :684 –90 . 10.1161/01.CIR.0000083724.28630.C3 12885747 42 Affairs DoEaS, York PDUNN . World population ageing 2013 . United Nations Publications , 2013 .

PMC005xxxxxx/PMC5372020.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01498510.1136/bmjopen-2016-014985Health Services ResearchResearch1506170417251713From first symptoms to diagnosis of amyotrophic lateral sclerosis: perspectives of an Irish informal caregiver cohort—a thematic analysis Galvin Miriam 1Gaffney Rebecca 1Corr Bernie 2Mays Iain 1Hardiman Orla 12 1 Academic Unit of Neurology, Trinity Biomedical Sciences Institute, Trinity College Dublin, Dublin 2, Ireland 2 Department of Neurology, National Neuroscience Centre, Beaumont Hospital, Dublin 9, IrelandCorrespondence to Dr Miriam Galvin; galvinmi@tcd.ie2017 20 3 2017 7 3 e01498531 10 2016 23 12 2016 19 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives Informal caregivers play an integral part in the management of amyotrophic lateral sclerosis (ALS). The objective of this study was to explore the journey from first problem symptoms to diagnosis from the perspective of informal caregivers providing care to people with ALS. Design As part of a semistructured interview, information was collected on a range of caregiver demographic details, and from an open-ended question their experiences of the time of symptom onset to diagnosis. We carried out descriptive statistical analysis and thematic analysis of qualitative data. Setting and participants Home interviews with informal caregivers (n=74) of people with ALS attending the National ALS/Motor Neuron Disease Clinic at Beaumont Hospital, Dublin, Ireland. Results This was a largely female and spousal cohort of caregivers, living with the patient for whom they provided informal care. The majority of patients were men and were spinal onset. Caregivers described the time from first symptoms to diagnosis. Using a primarily inductive approach, the coding was data driven and the codes and themes derived from the content of these descriptions. Two main themes were identified (1) problem signs and symptoms (A) noticing and (B) reaction; (2) interaction with the health services. Conclusions Exploring the perspectives of caregivers from first problem symptoms to diagnosis provides valuable insights into the development of the condition, impediments to its recognition, help-seeking behaviours and interactions with healthcare services. The journey from early symptoms to diagnosis is important for future decision-making, affects readiness for caregiving and could negatively impact on caregiver health and well-being. The early acknowledgement by healthcare professionals of stressors along the journey to diagnosis, and appreciation of their possible impact on caregivers is important. The separate needs of caregivers should be assessed on a regular basis. Caregiverpre-diagnosishealth servicesamyotrophic lateral sclerosismotor neuron diseasesymptoms ==== Body Strengths and limitations of this study Informal caregivers describe the time from first symptoms to diagnosis of amyotrophic lateral sclerosis. The caregivers' perspectives provide valuable insights into the practical and psychosocial issues on the journey to diagnosis. Using an inductive approach the themes identified were derived from the data. Descriptions were provided in response to one open-ended question in a largely quantitative interview which limited the opportunity to expand on their experiences. It is important to explore experiences before diagnosis and the quality of these experiences through in-depth interviews with caregivers. Introduction Amyotrophic lateral sclerosis (ALS) also known as motor neuron disease (MND) is a progressive neurodegenerative condition which results in loss of physical function due to the degeneration of upper and lower motor neurons. Up to 50% of patients with ALS develop a degree of cognitive impairment.1 There are currently no effective disease-modifying therapies, management is symptomatic, and 70% of those affected die within 3 years of symptom onset.2 In the Republic of Ireland, there are ∼300 prevalent ALS cases and 110 new diagnoses each year.3 At least 80% of all patients within Ireland attend the National ALS/MND Clinic at Beaumont Hospital, Dublin. Informal caregivers play an integral part in the management of ALS with family and friends taking most of the responsibility. They are key figures in care provision, provide emotional and physical support for patients and often play a central role in clinical decision-making.4–6 Symptoms and site of onset ALS onset is generally classified as being bulbar, spinal or respiratory, with a proportion of patients presenting with cognitive and/or behavioural impairment. In the bulbar form of ALS, the muscles used for speaking, swallowing and breathing are affected. It is usually characterised by slurring of words, impaired tongue movements and difficulty swallowing. Patients with spinal/limb onset experience fine-motor disturbances and paralyses in the arms and legs, presenting symptoms usually include weakness and/or wasting in the leg and arm muscles and fasciculation. Respiratory onset is characterised by breathing difficulties and shortness of breath.7 Diagnosis A diagnosis of ALS is primarily based on the physician's interpretation of clinical symptoms and signs, in addition to investigations to exclude other causes.8 ALS diagnostic timelines have remained consistent over a 20-year period (1998–2008), time from first symptom to diagnosis is a median 12 months.9 Earlier diagnosis can enable more effective symptom management and care planning. Diagnostic delays in ALS are associated with clinical complexity, the patient either not recognising or denying early or intermittent symptoms, inefficient referral pathways, the relative rarity and the consequent lack of familiarity with the condition among general healthcare professionals (HCPs).8 General practitioners will only see one or two cases throughout their career.7 Delays in diagnosis can account for a significant proportion of total illness duration. This can represent a missed opportunity to start treatment with riluzole, the only disease-modifying therapy at present. It may also prohibit patients from enrolling in clinical trials at an early stage of disease, when the likelihood of benefiting from experimental treatments might be greater.10 Delays also impede referral to multidisciplinary clinics (MDCs), which can improve patient outcomes both with respect to survival and quality of life.3 11 Findings from an exploratory study noted a mean interval of 19 months (median 14.6) from first symptoms to arrival at an ALS MDC, and an average of four contacts with HCPs.12 Help seeking The general public has a poor understanding of ALS13 and it is often the case that patients and/or caregivers are slow to seek medical advice. Symptoms may be noticed but medical help is not sought if symptoms are not severe, with alternative reasons considered reasons for their difficulties. O'Brien et al14 found that the initial symptoms were regularly not recognised by patients, family members and health professionals, which delayed seeking medical attention and resulted in delayed diagnosis. Aim The aim of this analysis was to explore the journey from first problem symptoms to diagnosis from the perspective of informal caregivers providing care to people with ALS. Methods Participants Participants were consecutively recruited as primary informal caregivers of patients taking part in a longitudinal study of the patient and caregiver journey though ALS/MND, attending the specialist multidisciplinary National ALS/MND Centre at Beaumont Hospital, Dublin. Caregiver participants were identified by the person with ALS as his/her primary informal caregiver, providing unpaid care and assistance to them. Caregivers were approached by a research assistant (IM) during clinic visits, and were provided with information about the research study. This initial contact was followed up by telephone to establish their interest and consent to participate. Informed written consent was obtained from all participants at the time of interview. Eighty-eight caregivers were recruited to the study, 14 did not complete an open-ended question regarding the time from first symptoms to diagnosis. This analysis is based on data from 74 caregivers who responded to that question. Data collection During a pilot-tested, semistructured interview, information was collected on a range of demographic and socioeconomic factors. In an open-ended question, caregivers were asked about their experiences from the time of symptom onset to diagnosis. This question was formulated through group discussions with senior and experienced members of the clinical staff working as part of the multidisciplinary service for patients with ALS/MND. Patients’ clinical details were available through the National ALS Register, for which they had consented to inclusion of their codified clinical and demographic data. Interviews took place between May 2013 and November 2014. The face-to-face interviews with only the interviewer and participant present, lasted ∼1 hour and were conducted in the caregiver's own home by a male assistant psychologist (IM) or a female health services researcher (MG) both members of the research team. Responses to the open-ended questions, and any related field notes were recorded in written format by the interviewer. Data analysis Descriptive statistics summarised demographic, socioeconomic and clinical data. Thematic analysis, chosen as a theoretically flexible method, was used to identify patterns in the data and themes for analysis.15 From a critical realist perspective, and employing an inductive approach, the themes identified on a semantic level were closely linked to the data. Multiple coders (RG, IM and MG) took part in a multiphase process including initial coding, theme development, review and definition.15 Audit trails were developed using reflexive memos and codebooks. The codes generated and themes constructed were reviewed and credibility of findings was established based on clinical experience (BC). There was discussion on points of agreement/disagreement leading to consensual validation. Data analysis software NVivo V.10 (QSR International, NVivo qualitative data analysis software; QSR International Pty Ltd version 10, 2012) was used to collate and manage the qualitative data, record coding patterns and theme development. Results Caregiver characteristics: descriptive statistical analysis Table 1 summarises the characteristics of caregivers and the patients for whom they provided care. Table 1 Characteristics of caregivers and patients Caregivers (n=74) Age  Mean 55.7 (SD 12.82)  Range 25–76 years Sex  Male 23 31.1%  Female 51 68.9% Relationship to the patient  Spouse/partner 53 71.6%  Son/daughter 15 20.3%  Parent 2 2.7%  Sibling 3 4.1%  Friend 1 1.4% Lives with patient  Yes 60 81.1%  No 14 18.9% Principal economic status  Working for payment or profit 32 43.2%  Unemployed 4 5.4%  Looking after family/home 16 21.6%  Retired 19 25.7%  Unable to work due to permanent sickness or disability 3 4.1% In general, would you say your health is  Excellent 14 18.9%  Very good 20 27.0%  Good 26 35.1%  Fair 9 12.2%  Poor 5 6.8% Do you have any long-term illness, health problems or disability?  Yes 39 52.7%  No 35 47.3% Patients (n=74) Age  Mean 65.16 (SD 9.74)  Range 43–87 years Sex  Male 45 60.8%  Female 29 39.2% Site of onset  Bulbar 21 28.4%  Spinal 51 68.9%  Thoracic/respiratory 2 2.7% Time from symptom onset to diagnosis  Mean (months) 15.72 (SD 11.04)  Range 1–56 months  Median 12  IQR 8–22 A majority of the caregiver cohort was women 69% (n=51). The mean age was 55.7 years (SD 12.8 years) ranging from 25 to 76 years. Predominantly caregivers were family members, 72% (n=53) were a spouse/partner, 20% son or daughter, and 4% a sibling of a person with ALS. Eighty-one per cent (n=60) were living with the person with ALS. Forty-three per cent of caregivers were working for payment or profit at the time of interview, while 47% were either retired or looking after family/home. Over half (53%) indicated that they had health problems, long-term illness or disability, for example, asthma, blood pressure, diabetes and heart condition. Notwithstanding 81% rated that their own health was good, very good or excellent. Sixty-one per cent of patients were men, with a mean age of 65 years (SD 9.7 years), ranging from 43 to 87 years. Distribution of patients by site of onset showed the majority were spinal onset (69%), bulbar onset (28%) and thoracic/respiratory (3%; figure 1). The time from symptom onset to ALS diagnosis ranged from 1 month to 4.7 years (mean 15.7 months, SD 11.04), median 12 months (IQR 8–22), a mean 13.2 months (median 11) for bulbar onset, mean of 17 months (median 12) for spinal onset and respiratory onset with a mean of 10 months (median 10). Figure 1 Time (months) from first symptoms to diagnosis and site of onset. Using data from the National ALS/MND Patient Register, the time in months from first symptoms to diagnosis for each patient in ascending order, and the site of onset is illustrated in figure 1. Thematic analysis In an open-ended question, caregivers were asked to think about the time when the patient first started having problems up to diagnosis.Thinking of the time when [ ] first started having problems up to his/her diagnosis with MND, what was that time like for you? Through a thematic analysis two main themes and component subthemes were developed and are presented here: (1) problems and symptoms (A) noticing and (B) reaction and response and (2) interactions with healthcare services (figure 2). Figure 2 Themes and subthemes. These themes with selected supporting quotes are presented below, denoted by caregiver ID and relationship to patient (eg, #1cg, wife) and site of ALS onset (eg, spinal onset). Theme 1: problems and symptoms: noticing and reaction and response Noticing problems/symptoms Various problem signs and symptoms were described. There were instances of voice distortion, falling, twitching limbs, problems with slurred speech and eating. Respondents described noticing something was wrong and began attributing meaning to symptoms, for some this occurred gradually over time. Limb weakness meant problems with walking and daily tasks became noticeable:It started with a twitch in his arm, we never thought it was this. He said ‘I can't walk in a straight line. (56cg wife, spinal onset) Probably with the hands, not being able to open things or turn keys or cook. (98cg son, spinal-cognitive/behavioural onset) Some caregivers described not noticing problem signs which were brought to their attention by other family members and friends:I didn’t even notice it, my daughter noticed her speech was going. (48cg sister, bulbar onset) It wouldn’t be the family that noticed it would be family-in-law, wife or brother or sister-in law, slurring words, fatigue and open mouth. (92cg son, bulbar onset) Problems noticed were attributed by them to factors such as poorly fitting dentures, and conditions related to heart problems, arthritis, carpal tunnel syndrome, stress or stroke. Looking back, and from what they now know, caregivers could connect what were diverse symptoms at the time. Noticing symptoms, not noticing, misattribution to another condition and gradual evolution of symptoms was summed up by one caregiver:You could see a little bit of a problem, it was hard for me to notice, it was my sister in England who noticed; my sister thought she had a stroke; once she got the second set of teeth done we thought well there was something bad with her. Looking back she's been suffering for a few years, for example there's been twitching in her legs for 4–5 years or there was a fall from a ladder a year and a half ago. (94cg son, bulbar onset) Reaction and response When problems were noticed caregivers described a variety of psychoemotional and practical reactions and responses. The range of psychoemotional reactions included worry, denial, anxiety, uncertainty, annoyance, frustration, panic, fear, disbelief, suspicion, misattribution, annoyance and avoidance. More focused practical responses included searching for information and advice and attempts to link with healthcare services. These future caregivers were situated in their own domestic contexts, trying to cope with what was happening, balance other responsibilities and possible adjustment to their own lives:I was struggling with my own life as well. I was doing everything and anything to help and do things and prevent her from falling; so it got to the stage where I couldn’t leave her side; it was quite stressful, I was trying to hold a relationship, job and caring for mum…(99cg daughter, spinal onset) For some there was disbelief as they could not understand what was happening around them: “I thought ‘maybe I’m nuts’, my two friends checked, and said ‘you’re not seeing things'” (54cg wife, spinal onset) or that perhaps the patient was imagining symptoms. Conversely, other caregivers found themselves trying to convince other people, at times including the patient: “I think I was the only one that noticed the speech problems. I kept saying it and he [patient] said it was his teeth” (68cg wife, bulbar onset). There was confusion and uncertainty with the realisation that something was wrong but “we couldn't find out what.” Fear was commonly described, fear of the unknown or what symptoms could mean, fear of not being able to cope or manage to provide care in the future. There was also awareness that the problems were indicative of something serious. Some who suspected there was something seriously wrong described avoidance and denial in various forms in an effort to ‘block it out’ or redirect their thoughts and focus: “Running and exercising until you didn’t have to think about it was important for me to function” (49cg wife, spinal onset). There was a degree of desperate hoping and wishful thinking: “[patient] said ‘I hope it's not MS’, I said ‘I hope it's not MND’. By saying that I thought ‘well, it won’t be that because that could be the worst thing’” (42cg wife, spinal onset).One month before diagnosis I questioned the possibility [patient] had MND but ruled it out as friend's husband previously had condition and didn’t think it would strike twice. (26cg wife, bulbar onset) There were also practical problem-focused responses, as they sought a diagnosis, and were proactive in seeking advice and information, even if it was difficult to face what that could mean. Caregivers commonly encouraged the patient to seek medical advice, which was accompanied by irritation and frustration when the patient would not do so:He wasn't going to the doctor and I made him go. I knew there was something wrong. I didn't know what it was, I hadn't a clue to be honest. I used to get annoyed about him not going to the doctor. (70 g wife, spinal onset) Other people described the patient not complaining, and guarding his/her privacy as they chose not to disclose what was happening around this time:I didn't feel involved at all…The patient was going and seeing to himself and didn't feel the need to bother me. He had broken an arm and we thought it was that. I had little worry as I didn't know. (71cg wife, spinal onset) She kept a lot to herself and didn’t tell us about much of the tests. (91cg sister, bulbar onset) Theme 2: interaction with healthcare services Interaction with the healthcare services featured from the time of first symptoms up to diagnosis. This included attempts to convince the patient to seek medical advice, referrals across the health system and engagement with HCPs. Fear, worry, anger and frustration were again present in the responses, as caregivers described referral routes, tests and interventions and perceived misdiagnoses. In response to noticing symptoms caregivers sought information and advice. It sometimes proved difficult to get to see specialist physicians, while succeeding in obtaining an appointment with a specialist did not mean an end to the uncertainty, anxiety or fear. For many, it had been a struggle to persuade the patient to seek medical advice. When they had begun that process, caregivers and patients made their way through various tests and interventions, for example, back operation, and were referred to health practitioners such as physiotherapist, ear, nose and throat physician and neurologist, on the journey to a definitive diagnosis. Some respondents felt that they and the patients were not listened to, and they could not understand the medical decisions and information received, adding to the confusion and uncertainty experienced:The doctors said I was being overprotective and as a last resort they sent us to a speech and language therapist and the speech and language therapist referred us to Consultant Neurologist; now we had doctors do an MRI in [clinic name], they came back and said nothing was wrong. (100 cg husband, bulbar onset) There were perceived misdiagnoses of the presenting symptoms: “A lot was missed that has been spotted now. First went to physio, then neurologist about her migraines. The doctor said it wasn't neurological, it was mechanical. The neurologist apologised in the end” (55cg husband, spinal onset). According to respondents, other diagnoses made by HCPs included stroke, myasthenia gravis or ‘unknown’. With various diagnoses and none, caregivers found themselves in a difficult position as medical opinion offered did not seem to fit with their experiences:I was relieved she went to the doctor, but was sceptical when the doctors said she had a fine bill of health in the hospital. We thought it was a stroke as it was in the family. Those connections and evidence were there it just didn’t feel right… (92cg son, bulbar onset) The ways in which caregivers (and patients) experienced communication and received information from the HCPs, how tests and referrals were discussed, and diagnoses communicated was less than optimal for many. Respondents described various encounters when information was delivered and communicated to them by HCPs. The quotes below illustrate some of these experiences:The first doctor didn't highlight what he was doing, because we flew through a lot of tests we thought this was a good thing, one day I got a call from [patient] and he mentioned the doctor had asked an MND question and I was like ‘where did that come from?’. It could have been signposted a bit better. (7cg wife, spinal-cognitive onset) (Neurologist) did neurological tests. Stressful time. Left [patient] sitting in his underpants for like 10 minutes by consultant who took a phone call! (second neurologist)…said I would lay my bet on MND then went out the door. (12cg wife, spinal) Discussion The diagnosis of a terminal illness can have a dramatic impact on the lives of family members, many of whom will become caregivers. Caregivers' descriptions of the time from noticing the first problems to diagnosis are recollections of that time, related events and feelings. These recollections may differ from what occurred, nevertheless they are experiential and meaning-filled, and shape current and future experiences and the meanings attached to them. The majority of this cohort was women (69%), and were family members (99%) of the person with ALS. Forty-three per cent were working at the time of interview, and just over half indicated they had some health issues. Coming from individual precare contexts, caregivers described both noticing and failing to notice symptoms and engagement with health services. Responding to problems and symptoms and how they were interpreted included suspicion, denial, avoidance, fear, confusion, worry and help-seeking. While many caregivers noticed and may not have understood what was happening, others denied any observations. Emotional and avoidance-focused coping are generally considered to be maladaptive.16 Avoidance-based coping may be beneficial and act as a protective factor in some situations; however, it becomes a problem when applied continuously.17 Patients and caregivers seek more information following communication of ALS diagnosis.18 The internet, family, friends and media were used as sources for information, from outside the healthcare system, during the prediagnosis phase. There are physical, psychological, emotional and social challenges to the health and well-being of ALS caregivers. Caregiver stress leads to adverse outcomes, psychological distress and increased burden.19 20 The existence of elevated levels of anxiety from the prediagnosis phase should be incorporated into HCPs interpretation of caregiver burden.21 The findings from this study show psychological distress and burden present before diagnosis. The early acknowledgement by HCPs of stressors along the journey to diagnosis, and appreciation of their possible impact on the health and well-being of caregivers are crucial. Many caregivers described a failure of the HCP at the first point of contact to recognise the gravity of the symptoms. Previous research8 has indicated the presence of a ‘red flag’ system for clinicians, as symptoms (eg, changes in speech, gait disturbance) warrant prompt referral to a neurologist. It is recommended that patients with symptoms suggestive of ALS should be assessed as soon as possible by an experienced neurologist, early diagnosis and investigations performed with high priority. The experiences of caregivers in this study reflects a series of issues at both personal and professional levels—inexperienced physicians, diagnostic uncertainty, the patient's refusal to seek help, and individual coping styles and strategies such as denial and avoidance. In many cases, the family had already endured a prolonged and frustrating journey to come to the point of engagement with health services. Once in the system, problems were encountered in accessing specialist medical opinion in a timely way, caregivers felt they (and patients) ‘were passed around’ that HCPs did not listen, or failed to recognise the seriousness of the condition, or misdiagnosed the symptoms. The interpersonal dynamics and communication from some HCPs were experienced as disrespectful, perfunctory and negative by some caregivers. Previous research points to communication difficulties with HCPs with respect to obtaining information about diagnosis and health status, and a lack of empathy and an absence of compassion conveyed by medical professionals.4 14 18 Guidelines are available to assist in communication practices and educational programmes can be accessed to improve communication skills.8 14 22 23 Clinicians communicate their expertise and experience with managing ALS, while patients, caregivers and family communicate their experiences of living with it. Our findings suggest the continued need for improvement in information exchange and communication between HCPs and patients and caregivers. Caregivers play an important role in the healthcare ecosystem. Treating physicians should be aware of the importance of the caregiver–patient dyad and family environment in caring for patients with ALS. The ongoing and separate needs of caregivers should be assessed on a regular basis. It is important to recognise that the journey to diagnosis may impact the future caregiving course and subsequent engagement with health services. This exploratory qualitative substudy is part of a large longitudinal primarily quantitative study. The caregiver descriptions were provided in response to one open-ended interview question, with limited opportunity to expand on their experiences. It will be important to explore the time before diagnosis and the quality of their experiences through in-depth interviews with caregivers. Conclusion Exploring the perspectives of caregivers from first problem symptoms to diagnosis provides valuable insights into the development of the condition, impediments to recognition of early signs and symptoms, help-seeking behaviours and interactions with the HCPs and services. The prediagnosis trajectory of care has the potential to negatively affect future caregivers, and may impact on the subsequent caregiving course. Family caregivers are important in ALS and are influential in the patient–caregiver dyad. HCPs need to recognise that caregivers may already be in a psychologically distressed state at the point of diagnosis, which could place them at increased risk of experiencing distress and burden from caregiving. The authors would like to thank the caregivers who participated in this research, Mark Heverin and Alice Vajda Research Managers and Sile Carney Research Assistant at the Academic Unit of Neurology, TCD. Contributors: MG designed and developed the study, analysed and interpreted the data. MG and RG drafted the initial manuscript. BC was involved in the development of the research question and interpretation of the data. RG and IM assisted with analysis, and interpretation of data, IM and MG was involved in the acquisition and interpretation of data. MG and OH revised the drafts for intellectual content and edited the manuscript. All authors reviewed and approved the final draft. Funding: This research was supported by funding from the Irish Health Research Board Dublin as part of the HRB Interdisciplinary Capacity Enhancement Awards. Competing interests: None declared. Ethics approval: Beaumont Hospital Ethics (Medical Research) Committee (REC REF 12/84) and the Research Ethics Committee, Trinity College Dublin. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Elamin M , Bede P , Byrne S Cognitive changes predict functional decline in ALS: a population-based longitudinal study . Neurology 2013 ;80 :1590 –7 . 10.1212/WNL.0b013e31828f18ac 23553481 2 Rooney J , Byrne S , Heverin M Survival analysis of Irish amyotrophic lateral sclerosis patients diagnosed from 1995-2010 . PLoS ONE 2013 ;8 :e74733 10.1371/journal.pone.0074733 24098664 3 Rooney J , Byrne S , Heverin M A multidisciplinary clinic approach improves survival in ALS: a comparative study of ALS in Ireland and Northern Ireland . J Neurol Neurosurg Psychiatry 2015 ;86 :496 –501 . 10.1136/jnnp-2014-309601 25550416 4 Aoun SM , Connors SL , Priddis L Motor neurone disease family carers’ experiences of caring, palliative care and bereavement: an exploratory qualitative study . Palliat Med 2012 ;26 :842 –50 . 10.1177/0269216311416036 21775409 5 Aoun SM , Bentley B , Funk L A 10-year literature review of family caregiving for motor neurone disease: moving from caregiver burden studies to palliative care interventions . Palliat Med 2013 ;27 :437 –46 . 10.1177/0269216312455729 22907948 6 Rabkin JG , Albert SM , Rowland LP How common is depression among ALS caregivers? A longitudinal study . Amyotroph Lateral Scler 2009 ;10 :448 –55 . 10.3109/17482960802459889 19922139 7 McDermott CJ , Shaw PJ Diagnosis and management of motor neurone disease . BMJ 2008 ;336 :658 –62 . 10.1136/bmj.39493.511759.BE 18356234 8 Andersen PM , Abrahams S , Borasio GD EFNS guidelines on the Clinical Management of Amyotrophic Lateral Sclerosis (MALS)—revised report of an EFNS task force . Eur J Neurol 2012 ;19 :360 –E24 . 10.1111/j.1468-1331.2011.03501.x 21914052 9 Mitchell JD , Callagher P , Gardham J Timelines in the diagnostic evaluation of people with suspected amyotrophic lateral sclerosis (ALS)/motor neuron disease (MND)–a 20-year review: Can we do better? Amyotroph Lateral Scler 2010 ;11 :537 –41 . 10.3109/17482968.2010.495158 20565332 10 Paganoni S , Macklin EA , Lee A Diagnostic timelines and delays in diagnosing amyotrophic lateral sclerosis (ALS) . Amyotroph Lateral Scler Frontotemporal Degener 2014 ;15 :453 –6 . 10.3109/21678421.2014.903974 24981792 11 Van den Berg JP , Kalmijn S , Lindeman E Multidisciplinary ALS care improves quality of life in patients with ALS . Neurology 2005 ;65 :1264 –7 . 10.1212/01.wnl.0000180717.29273.12 16247055 12 Galvin M , Madden C , Maguire S Patient journey to a specialist amyotrophic lateral sclerosis multidisciplinary clinic: an exploratory study . BMC Health Serv Res 2015 ;15 :571 10.1186/s12913-015-1229-x 26700026 13 Davies Z , Turner MR Public awareness of motor neuron disease . Amyotroph Lateral Scler 2010 ;11 :490 –1 . 10.3109/17482961003716858 20672874 14 O'Brien MR , Whitehead B , Jack BA From symptom onset to a diagnosis of amyotrophic lateral sclerosis/motor neuron disease (ALS/MND): experiences of people with ALS/MND and family carers—a qualitative study . Amyotroph Lateral Scler 2011 ;12 :97 –104 . 10.3109/17482968.2010.546414 21208037 15 Braun V , Clarke V Using thematic analysis in psychology . Qual Res Psychol 2006 ;3 :77 –101 . 10.1191/1478088706qp063oa 16 Karekla M , Panayiotou G Coping and experiential avoidance: unique or overlapping constructs? J Behav Ther Exp Psychiatry 2011 ;42 :163 –70 . 10.1016/j.jbtep.2010.10.002 21315877 17 Kashdan TB , Barrios V , Forsyth JP Experiential avoidance as a generalized psychological vulnerability: comparisons with coping and emotion regulation strategies . Behav Res Ther 2006 ;44 :1301 –20 . 10.1016/j.brat.2005.10.003 16321362 18 Chiò A , Montuschi A , Cammarosano S ALS patients and caregivers communication preferences and information seeking behaviour . Eur J Neurol 2008 ;15 :55 –60 . 10.1111/j.1468-1331.2007.02000.x 18005051 19 Galvin M , Corr B , Madden C Caregiving in ALS—a mixed methods approach to the study of burden . BMC Palliat Care 2016 ;15 :81 10.1186/s12904-016-0153-0 27596749 20 Burke T , Elamin M , Galvin M Caregiver burden in amyotrophic lateral sclerosis: a cross-sectional investigation of predictors . J Neurol 2015 ;262 :2799 10.1007/s00415-015-7941-y 26566911 21 Lerum SV , Solbraekke KN , Frich JC Family caregivers’ accounts of caring for a family member with motor neurone disease in Norway: a qualitative study . BMC Palliat Care 2016 ;15 :22 10.1186/s12904-016-0097-4 26911713 22 Schellenberg KL , Schofield SJ , Fang SF Breaking bad news in amyotrophic lateral sclerosis: the need for medical education . Amyotroph Lateral Scler Frontotemporal Degener 2014 ;15 :47 –54 . 10.3109/21678421.2013.843711 24245652 23 Chiò A , Borasio GD Breaking the news in amyotrophic lateral sclerosis . Amyotroph Lateral Scler Other Motor Neuron Disord 2004 ;5 :195 –201 . 10.1080/14660820310017326 15799546

PMC005xxxxxx/PMC5372021.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01406610.1136/bmjopen-2016-014066EpidemiologyResearch1506169217241681Sport and scholastic factors in relation to smoking and smoking initiation in older adolescents: a prospective cohort study in Bosnia and Herzegovina Sekulic Damir 12Sisic Nedim 13Terzic Admir 45Jasarevic Indira 5Ostojic Ljerka 167Pojskic Haris 89Zenic Natasa 1 1 Faculty of Kinesiology, University of Split, Split, Croatia 2 University Department of Health Care Studies, Split, Croatia 3 University of Zenica, Zenica, Bosnia and Herzegovina 4 High School Hasan Kikic, Gradacac, Bosnia and Herzegovina 5 Faculty of Physical Education and Sport, University of Tuzla, Tuzla, Bosnia and Herzegovina 6 University of Mostar, Mostar, Bosnia and Herzegovina 7 Academy of Medical Sciences of Bosnia and Herzegovina, Sarajevo, Bosnia and Herzegovina 8 Department for Health Sciences, Mid Sweden University, Östersund, Sweden 9 Mid Sweden University, Swedish Winter Sports Research Centre, Östersund, SwedenCorrespondence to Professor Natasa Zenic; natasazenic@yahoo.com2017 22 3 2017 7 3 e01406629 8 2016 30 1 2017 2 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective Sport and scholastic factors are known to be associated with cigarette smoking in adolescence, but little is known about the causality of this association. The aim of this study was to prospectively explore the relationships of different sport and scholastic factors with smoking prevalence initiation in older adolescents from Bosnia and Herzegovina. Methods In this 2-year prospective cohort study, there were 872 adolescent participants (16 years at baseline; 46% females). The study consisted of baseline tests at the beginning of the third year (September 2013) and follow-up at the end of the fourth year of high school (late May to early June 2015). The independent variables were scholastic and sport-related factors. The dependent variables were (1) smoking at baseline, (2) smoking at follow-up and (3) smoking initiation over the course of the study. Logistic regressions controlling for age, gender and socioeconomic status were applied to define the relationships between independent and dependent variables. Results School absence at the baseline study was a significant predictor of smoking initiation during the course of the study (OR 1.4, 95% CI 1.1 to 1.8). Those who reported quitting sports at baseline showed an increased risk of smoking at the end of the study (OR 1.4, 95% CI 1.1 to 2.0) and of smoking initiation (OR 1.8, 95% CI 1.3 to 2.0). Adolescents who reported lower competitive achievements in sport were at a higher risk of (1) smoking at baseline (OR 1.5, 95% CI 1.1 to 2.1), (2) smoking at follow-up (OR 1.5, 95% CI 1.1 to 2.1) and (3) smoking initiation (OR 1.6, 95% CI 1.1 to 2.6). Conclusions In developing accurate antismoking public health policies for older adolescents, the most vulnerable groups should be targeted. The results showed that most participants initiated smoking before 16 years of age. Therefore, further investigations should evaluate the predictors of smoking in younger ages. cigaretteseducational achievementsportspubertyassociation ==== Body Strengths and limitations of this study The study lacks data on peer smoking and parental smoking, both of which can be associated with sport participation and educational achievement. Studied sport factors consisted of questions on formal sport participation, while some other important determinants of involvement in sport (non-formal physical exercising in fitness centres, self-exercising, etc) were not evaluated. The majority of participants started smoking before 16 years of age, and therefore the generalisability of the results regarding the predictors of smoking initiation is limited solely to adolescents who initiated smoking in late adolescence (16–18 years of age). This is one of the first studies to prospectively investigate the predictors of smoking in south-eastern Europe. The high retention rate (87% of the adolescents studied at baseline and follow-up) and low rate of missing data are important strengths of the study. Introduction Cigarette smoking is an important modifiable determinant of health, and preventing smoking initiation among adolescents eliminates the numerous health risks they would face as adult smokers.1 2 With more than 20% of adolescents who smoke cigarettes daily, Bosnia and Herzegovina is among the five European countries with the highest prevalence of smoking among adolescents, together with Austria, Croatia, Belgium and Hungary (all about 20% daily smokers).3 4 This high prevalence is mostly explained by the low prices of tobacco products, social acceptance of smoking in public and the lack of effective public health campaigns against smoking.5 Consequently, in the last couple of years, several cross-sectional investigations have explored the problem and found different sociodemographic, economic, community-specific, sport-related and scholastic factors to be associated with adolescent smoking in the country.6 7 Scholastic achievement (educational achievement) is one of the factors known to be associated with smoking in adolescence, with poor performance in school regularly observed in adolescents who smoke.5 8–10 However, the causality remains unknown. One possible explanation implies that smoking is the cause of poor performance in school because of the physiological mechanisms and the negative effects of smoking on cognitive function and learning capacities.11–13 Meanwhile, some authors are of the opinion that smoking should be observed as an effect, and not the cause, of educational failure.7 14 For example, children who fail academically are frequently in out-of-school situations where they are directly and/or indirectly exposed to individuals who smoke and are therefore at higher risk of smoking themselves.5 6 Indeed, social influences are known to be important with respect to a wide range of health behaviours, including smoking, and such peer influence on smoking is also logical.15 16 However, it is also possible that other factors, such as parental conflict and/or poor familiar control, result in educational failure and smoking. Additionally, ‘the theory of problem behaviour’ (ie, that the problem behaviours such as failure in school and smoking often appear in tandem because some people have a psychosocial tendency for unconventionality) has been used to explain the association between educational failure and smoking in adolescence.17 Regardless of the background, the cross-sectional design of the studies did not allow for the interpretation of the cause–effect relationship between educational achievement in high school and smoking status.5 7 18 Participation in sport is often considered as a potentially effective way of reducing the tendency of adolescents to smoke cigarettes.19–22 Indeed, when comparing groups of adolescent athletes versus non-athletes, there is a lower prevalence of smoking in those involved in sports.23–25 However, when sport participation was analysed more specifically, there were some conflicting findings with the association between sports and cigarette smoking.26–28 For example, in a recent study, authors found that adolescents who had stopped participating in a sport were at high risk of misusing substances, while lower competitive achievement in sports was found to be associated with a higher likelihood of cigarette smoking.3 5 27 Again, owing to the cross-sectional study design, the causality is not clear. Indeed, smoking could impair physical capacity, thus leading to poor sport performance (a low result) and consequent withdrawal from sport. On the other hand, it is also possible that adolescents first stopped participating in sports and then started to smoke.27 28 This investigation aimed to prospectively explore the potential relationships between scholastic and sport factors at the beginning of the third year of high school, and smoking and smoking initiation in the following 2 years (from 16 to 18 years of age on average) in adolescents from Bosnia and Herzegovina. Understanding the relationships studied here may help inform all responsible parties about the specific risks and benefits related to the studied covariates of smoking. Although there are other potential predictors of smoking, in this study we were specifically focused on scholastic and sport factors, as both groups of factors are regularly and independently monitored throughout the school system and a better understanding of the associations could help develop cost-effective and targeted preventive interventions. Methods Procedures and participants In this study, we aimed to prospectively investigate adolescents over their last 2 years of high school. At baseline, the examinees were 16 years old on average and were in their third year of high school. A multistage cluster sampling method was used to select the participants. First, we randomly selected one-third of the high schools in the territory of Zenica-Doboj Canton and Tuzla Canton, mostly because of their sociocultural environments as described below. Bosnia and Herzegovina is a multiethnic country, home to three constitutive ethnicities (Bosniaks, Serbs and Croats). Devastating wars that occurred in the early 1990s resulted in massive emigrations of minority ethnic groups (specifically for different parts of the country), and overall material devastation.29 For the two Cantons studied, prewar ethnic figures did not change drastically. Therefore, these two Cantons should be observed as two typical regions in Bosnia and Herzegovina. School size varied by just 10–15%, and therefore the schools were not stratified by size. In the second stage of sampling, half of all third year classes were selected at random from the selected schools, resulting in a sample size of 44 classes and a cohort of 1213 participants. After obtaining the necessary ethics approvals (see later text), study personnel explained the full procedure and study aims to potential participants and at least one parent/guardian in a regular school meeting. Consent was obtained from at least one parent, and none of the parents refused to let their child participate in the study. Two surveys were conducted, one (baseline) at the beginning of the third year of high school (September 2013), and the second (follow-up) at the end of high school (late May to early June 2015). Surveys were administered during school hours in groups of at least 15 examinees. Examinees were assured that their participation was voluntary and that they could leave any of the questions and/or the entire questionnaire blank. The study participants remained anonymous (no personal data were collected), but the participants were asked to use self-selected confidential codes for identification purposes in the repeated test. They were asked to use the last three digits of their email password as their code for identification (ie, these codes were easy to remember between testing waves while being simultaneously confidential). After completing the surveys, each participant placed the questionnaire in an envelope and then in a closed box. The next day, an investigator who was not present during the survey administration opened the boxes. The study fulfilled all ethical guidelines. After obtaining ethical approvals, the study was officially authorised by the Ministries of Education in Zenica-Doboj Canton and Tuzla Canton, the two areas of Bosnia and Herzegovina where the research was taking place. The study design and sampling is presented in figure 1. Figure 1 Sampling procedure, participants and non-responders (ie, absent from school on testing day, inconsistency in identification codes). Of the 1213 eligible students, 1059 (87%) had complete data both at baseline and in the follow-up study. Of the latter, 872 who identified as Bosnians were included in the study (72% of the total eligible). The analysis of attrition bias showed no significant differences in initial smoking status between adolescents who dropped out and those who remained in the study (χ2 2.11, p>0.05), but there were significantly more men than women who dropped out (χ2 8.00, p<0.01; see online supplementary table). 10.1136/bmjopen-2016-014066.supp1supplementary table Attrition bias analysis between responders and non-responders on a basis of smoking status and gender Variables To extend current knowledge and allow meaningful comparison with previous cross-sectional reports from countries belonging to former Yugoslavia, such as Bosnia and Herzegovina, the variables were collected using the Questionnaire of Substance Use, which was previously reported to be a reliable and valid measuring tool in similar samples of participants.6 7 27 In this study, we collected data on age (in years), gender, self-reported socioeconomic status (below average, average, above average), ethnicity (Bosniak, Serbian, Croatian, other), sport factors, scholastic factors and consumption of cigarettes. Sports factors consisted of questions about the participants' (1) involvement in sports (answers included: never been involved, quit, currently involved); (2) highest competitive achievement in sports (never competed/did not participate in sports, local ranked competitions, national and international ranked competitions) and (3) time of involvement in sports (never involved, <1, 2–5, 5+ years). Scholastic variables represented participants' academic achievement over the past semester (end of the second year of high school): (1) grade point average; (2) behavioural grade (both on a five-point scale ranging from excellent to poor) and (3) school absences (almost never, rarely, from time to time, often). Cigarette smoking was assessed on a four-point scale with the following responses: ‘no, I don't smoke’, ‘from time to time, but not daily’, ‘<10 cigarettes daily’, and ‘more than 10 cigarettes daily’. Participants were later classified as non-smokers (those who responded with ‘no, I don't smoke’) or smokers (the remaining three answers). Information on smoking initiation during the course of the study was obtained from each participant. Specifically, if the participant reported not smoking at baseline and responded differently when tested at follow-up, the initiation of smoking was indicated. Statistics For all variables, descriptive statistics (counts and percentages or means and SDs) were calculated. Depending on the characteristic of the variable, the differences between smokers and non-smokers were established by the Mann-Whitney test (for ordinal variables), or χ2 test (for categorical variables). Binary logistic regression was used to estimate the OR and the corresponding 95% CI of the following: (1) smoking status at baseline, (2) smoking status at the end of the study and (3) smoking initiation occurring during the course of the study by scholastic and sport covariates. The logistic analyses were additionally adjusted for gender, age and socioeconomic status. Results Table 1 presents the distribution of independent variables according to smoking status at baseline and follow-up. Overall, 28% of adolescents were identified as smokers at the beginning of their third year of high school, and 36% were smokers at the end of high school 20 months later. An increase in smoking prevalence over the observed period was particularly evident in women (from 27% to 38%, and 30% to 34.5% for women and men, respectively). At baseline and at follow-up, non-smokers achieved better grade point averages (Mann-Whitney Z values (MW) 6.03 and 6.36, p<0.01) and better behavioural grades (MW 7.76 and 7.71, p<0.01) and were less absent from school (MW 5.60 and 9.30, p<0.01, for baseline and follow-up, respectively) than smokers. Non-smokers achieved higher sports results than smokers at baseline (MW 2.34, p<0.01). Table 1 Baseline and follow-up characteristics with differences on a basis of smoking status (MW—Mann-Whitney Z values; χ2 test) Baseline Follow-up Smokers Non-smokers MW Smokers Non-smokers MW f (%) f (%) Z (p value) F (%) F (%) Z (p value) Experience in sport 1.54 (0.12) 0.93 (0.34)  Never been involved 180 (28.8) 56 (22.6) 174 (31.3) 62 (19.6)  Less than a year 132 (21.2) 56 (22.6) 104 (18.7) 84 (26.6)  2–5 years 164 (26.3) 72 (29) 136 (24.5) 100 (31.6)  >5 years 148 (23.7) 64 (25.8) 142 (25.5) 70 (22.2) Sport success/result 2.34 (0.02) 0.96 (0.33)  Never competed 344 (55.1) 114 (46) 308 (55.4) 150 (47.5)  Local rank 230 (36.9) 116 (46.8) 204 (36.7) 142 (44.9)  National/international 44 (7.1) 18 (7.3) 40 (7.2) 22 (7) Grade point average 6.03 (0.01) 6.36 (0.01)  Excellent 262 (42) 66 (26.6) 242 (43.5) 86 (27.2)  Very good 246 (39.4) 96 (38.7) 222 (39.9) 120 (38)  Average 106 (17) 76 (30.6) 86 (15.5) 96 (30.4)  Under average 6 (1) 4 (1.6) 2 (0.4) 8 (2.5)  Poor 4 (0.6) 6 (2.4) 4 (0.7) 6 (1.9) Behavioural grade 7.76 (0.01) 7.71 (0.01)  Excellent 538 (86.2) 180 (72.6) 500 (89.9) 218 (69)  Very good 46 (7.4) 26 (10.5) 24 (4.3) 48 (15.2)  Average 30 (4.8) 28 (11.3) 24 (4.3) 34 (10.8)  Under average 6 (1) 8 (3.2) 6 (1.1) 8 (2.5)  Poor 4 (0.6) 6 (2.4) 2 (0.4) 8 (2.5) School absence 5.60 (0.01) 9.30 (0.01)  Almost never 244 (39.1) 52 (21) 246 (44.2) 50 (15.8)  Rarely 244 (39.1) 102 (41.1) 206 (37.1) 140 (44.3)  From time to time 108 (17.3) 78 (31.5) 90 (16.2) 96 (30.4)  Often 28 (4.5) 16 (6.5) 14 (2.5) 30 (9.5) χ2 (p Value) χ2 (p Value) Gender 0.82 (0.36) 1.32 (0.25)  Male 138 (29.7) 326 (70.3) 160 (34.5) 304 (65.5)  Female 110 (27.0) 298 (73.0) 156 (38.2) 252 (61.8) Sport participation 0.46 (0.80) 3.79 (0.15)  Currently involved 128 (20.5) 56 (22.6) 118 (21.2) 66 (20.9)  Quit 206 (33) 80 (32.3) 170 (30.6) 116 (36.7)  Never been involved 290 (46.5) 112 (45.2) 268 (48.2) 134 (42.4) Socioeconomic status 5.66 (0.06) 2.58 (0.27)  Under average 10 (1.6) 4 (1.6) 10 (1.8) 4 (1.3)  Average 574 (92.0) 238 (95.7) 512 (92.1) 300 (95.0)  Below average 40 (6.4) 6 (2.4) 34 (6.1) 12 (3.8) Increased odds of smoking were observed in adolescents with a lower grade point average at baseline (baseline: OR 1.6, 95% CI 1.4 to 1.9; follow-up: 1.7, 1.4 to 1.9), poorer behavioural grades (baseline: 2.6, 2.0 to 3.5; follow-up: 2.3, 1.7 to 2.9) and more frequent absences from school, with the highest chances of being smokers for those children who reported that they were often absent from school (baseline: 4.4, 2.0 to 9.4; follow-up: 4.5, 2.1 to 9.4). School absence at baseline was a significant predictor of smoking initiation over the course of the study as children who reported that they were absent from time to time having the highest chances of starting smoking during the course of study (2.6, 1.5 to 4.8). Involvement in sports (sport participation) was not associated with smoking status at baseline, but those who reported quitting sports showed an increased risk of smoking at the end of the study (1.4, 1.0 to 1.9) and a higher risk of smoking initiation during the study (1.7, 1.1 to 2.9). Adolescents who were engaged in sports for <5 years showed a higher prevalence of smoking at the end of the study (<1 year: 2.7, 1.6 to 3.8; 2–5 years: 2.4, 1.6 to 3.6), and an increased risk of smoking initiation during the course of the investigation (<1 year: 2.7, 1.4 to 5.2; 2–5 years: 3.3, 1.6 to –6.2) than those who were never involved in sports. Finally, compared with peers who were never involved in sports, those who reported involvement in sports competitions but with lower competitive results were at a higher risk of the following: (1) smoking at baseline (1.5, 1.1 to 2.0), (2) smoking at the end of the study (1.5, 1.1 to 2.0) and (3) smoking initiation (1.6, 1.1 to 2.5; table 2). Table 2 The ORs for smoking at baseline, smoking at follow-up and smoking initiation over the course of the study Bivariate analyses Model 1 * Baseline characteristics Smoking at baseline Smoking at follow-up Smoking initiation Smoking at baseline Smoking at follow-up Smoking initiation Grade point average (grade)† 1.6 (1.4 to 1.9) 1.6 (1.4 to 1.9) 1.1 (0.9 to 1.4) 1.6 (1.4 to 1.9) 1.7 (1.4 to 1.9) 1.2 (0.9 to 1.4) Behavioural grade (grade)† 2.5 (1.9 to 3.3) 2.1 (1.6 to 2.7) 0.9 (0.6 to 1.3) 2.6 (2.0 to 3.5) 2.3 (1.7 to 2.9) 0.9 (0.6 to 1.3) School Absence  Almost never REF REF REF REF REF REF  Rarely 2.0 (1.4 to 2.8) 2.3 (1.7 to 3.2) 1.7 (1.0 to 2.8) 2.0 (1.4 to 2.7) 2.3 (1.7 to 3.2) 1.7 (1.1 to 2.8)  From time to time 2.7 (1.7 to 4.2) 3.5 (2.3 to 5.4) 2.5 (1.4 to 4.6) 2.7 (1.7 to 4.2) 3.6 (2.3 to 5.5) 2.6 (1.5 to 4.8)  Often 4.0 (1.9 to 8.6) 4.1 (2.0 to 8.6) 1.7 (0.5 to 5.0) 4.5 (2.1 to 95) 4.5 (2.1 to 9.4) 1.8 (0.6 to 5.5) Sport participation  Never been involved REF REF REF REF REF REF  Currently involved 1.1 (0.8 to 1.7) 1.2 (0.8 to 1.6) 1.2 (0.8 to 1.8) 1.1 (0.7 to 1.7) 1.2 (0.8 to 1.8) 0.9 (0.5 to 1.8)  Quit 1.0 (0.7 to 1.4) 1.4 (1.0 to 1.9) 1.7 (1.1 to 2.7) 1.0 (0.7 to 1.4) 1.4 (1.0 to 1.9) 1.7 (1.1 to 2.9) Experience in sport  Never been involved REF REF REF REF REF REF  Less than a year 1.3 (0.9 to 2.1) 2.3 (1.5 to 3.4) 2.6 (1.4 to 4.9) 1.4 (0.9 to 2.1) 2.7 (1.6 to 3.8) 2.7 (1.4 to 5.2)  2–5 years 1.4 (0.9 to 2.1) 2.1 (1.4 to 3.0) 2.8 (1.5 to 5.2) 1.4 (0.9 to 2.1) 2.4 (1.6 to 3.6) 3.3 (1.7 to 6.2)  >5 years 1.4 (0.9 to 2.1) 1.4 (0.9 to 2.0) 1.1 (0.6 to 2.3) 1.4 (0.8 to 2.2) 1.5 (0.9 to 2.5) 1.4 (0.7 to 2.9) Sport success/result  Never competed REF REF REF REF REF REF  Local rank 1.5 (1.1 to 2.1) 1.4 (1.1 to 1.9) 1.5 (1.0 to 2.2) 1.5 (1.1 to 2.0) 1.5 (1.1 to 2.0) 1.6 (1.1 to 2.5)  National/international 1.2 (0.7 to 2.2) 1.1 (0.6 to 1.9) 0.9 (0.4 to 2.4) 1.3 (0.7 to 2.2) 1.2 (0.7 to 2.1) 1.0 (0.4 to 2.4) * Adjusted for age, gender and socioeconomic status. †The higher value presents poorer scholastic achievement. Discussion This study aimed to prospectively investigate the potential relationships between scholastic and sport factors with smoking in older adolescents. The analyses revealed several important findings that should be highlighted. First, baseline scholastic factors were systematically associated with smoking, with poorer scholastic achievement in adolescents who reported smoking at baseline and follow-up. The absence from school at baseline was a predictor of smoking initiation in the following period. Quitting sports, poor competitive achievement and <5 years of participation in sports were shown to be specific risks for smoking and smoking initiation. Prior to discussing these findings, we will provide a brief overview of the established prevalence and trends in smoking. Smoking prevalence significantly increased from 28% to 36% during the course of the study. Consequently, ∼77% of adolescents who reported smoking at the end of high school (ie, 28/36) initiated smoking when they were younger than 16 years. Although participants self-reported their smoking status, which should be observed as a limitation of the study, the reported prevalence of smoking in this study is similar to previous reports of a 30–35% prevalence of adolescent smokers in Bosnia and Herzegovina and the wider territory of former Yugoslavia.3 27 30 Consequently, the self-reported data on smoking obtained here are plausible. This study is unique, as it is one of the first to prospectively investigate the factors associated with smoking in adolescents from south-eastern Europe. Therefore, the data on the relationships between scholastic and sport factors and smoking initiation are particularly interesting. The associations between scholastic factors and smoking at study baseline (ie, when participants were 16 years old) and the associations between scholastic factors and smoking at follow-up (ie, end of high school, 18 years of age) are similar. In both waves, higher odds of smoking are observed in adolescents with lower scholastic achievement. Our findings are therefore in accordance with the results of previous cross-sectional studies performed with adolescents of a similar age, which repeatedly reported lower educational achievement among children who smoke.7 18 27 Although there is a general consensus on the negative associations between smoking and performance in school, the mechanisms that lead to these associations are still controversial. Some authors highlight the negative effects of smoking on cognitive capacities, and consequently poorer learning capabilities, as a result of cigarette smoking.18 This explanation is strongly supported by the evident physiological mechanisms (ie, alterations in brain structure as a result of smoking).11–13 The main criticism of this explanation arises from the relatively short period of smoking in adolescents. Therefore, significant deterioration in cognitive capacities (and the resulting low academic achievement in adolescents) is less probable. As a result, it is suggested that lower academic achievement in adolescent smokers may actually be the cause, and not the effect, of smoking. Indeed, children who fail at school are frequently in ‘out-of-school situations’ and therefore in unique sociocultural environments in which they are more likely to initiate smoking.3 6 This theory thus focuses on social influence. In this study, we showed a negative relationship between ‘school absence’ at baseline (ie, 16 years of age) and smoking initiation in the following 2-year period. Therefore, our results actually support the theory of social influence as a probable explanation of the cause–effect relationship between academic failure and smoking in this age group. Briefly, the general associations between scholastic variables and smoking are clear, demonstrating that children who smoke perform poorly in school. However, it must be stressed that more frequent absences from school at the beginning of the third year of high school is a clear predictor of smoking initiation in the following 2 years. Of course, one can argue that scholastic achievement was not entirely objectively evaluated because the data were self-reported. Although this should be considered an important limitation of the study, we believe that the strict anonymity of the testing decreased the possibility that the participants responded dishonestly. Our study found a high risk for smoking initiation for those adolescents (1) who quit sport, (2) who reported low competitive success and (3) who had a relatively short period of involvement in sport. Generally, this is in accordance with very recent studies which have noted a higher prevalence of smoking in adolescents who quit sports and those who achieved poorer competitive results.7 28 Owing to the cross-sectional nature of studies, the cause-and-effect relationship between quitting sports and smoking had not been clearly identified. There is a possibility that smoking impairs physical capacities, and this could therefore result in poor sport performance and a lack of success, which would consequently result in withdrawal from sports.31 However, it is also possible that children first stop practising a sport and then start to smoke, as a result of the (negative) influence of their new sociocultural environment in which smoking is more prevalent.5 7 The results of our study support the latter explanation. Specifically, all three sport factors observed in this study were found to be significant predictors of smoking initiation in older adolescents. First, those who reported at study baseline that they had once practised sports and then quit were at a higher risk of starting to smoke during the course of the study (ie, between 16 and 18 years of age) than their peers who were never involved in sports. Additionally, a higher risk of initiating smoking was evident in adolescents who were involved in sports for <5 years and in those who practised sports but did not achieve significant competitive results. Most probably, the association of all three sport factors ((1) quitting sport, (2) short time of involvement and (3) low competitive result) with smoking initiation is generated by equal mechanism. Briefly, it is well known that better sport results are actually a direct or indirect consequence of a longer involvement in sport.32 33 At the same time, a lack of good sports results (ie, poor competitive achievement in sport) is one of the most important factors which results in withdrawing from a sport in adolescence.34 The adolescents aged 16–18 years who do not achieve competitive results regularly stop participating in sports at this particular age, mostly because they have become personally aware of their inferiority (ie, their lack of ability and/or skills).34 Meanwhile, it is known that individuals identify with particular groups of peers, and being a member of a specific social network or group influences individuals' values as well as their attitudes and the norms to which they are exposed.35 Therefore, it is likely that adolescents who quit sports started smoking as a way of adopting the norms of the ‘non-sporting’ society and of finding a place in a new social milieu. This study lacks qualitative data on the reasons for quitting sports and an objective evaluation of lower competitive achievement in sports. This is a clear limitation of the investigation, as we are not able to accurately explain the background of the relationship within the sport factors observed here. This study evidenced specific associations between scholastic and sport factors with smoking initiation in older adolescents, but we may not ignore the potential confounding effects of some covariates that were not observed in this study, such as those concerned with the home environment (ie, familiar factors). For example, it could be expected that both quitting sports and poor academic performance (ie, significant predictors of smoking initiation in our participants) might be a result of some family-related issues such as lack of parental monitoring and high parental conflict. Consequently, the lack of information on these issues could be highlighted as a study limitation. However, we were of the opinion that eventual knowledge of these confounding effects, although scientifically interesting, may not add much to our understanding of how to intervene. This is because children who have problems in their home environment, such as being in conflict with their parents and/or experiencing a lack of parental control, could not be tracked (ie, information on that manner was not obtainable). Meanwhile, all variables included in this study were easily obtainable throughout the education system, which allows identification of those children who are at specific risk for smoking initiation in late adolescence. The most important limitation comes from the fact that this study observed adolescents from 16 years of age, when many students had already started to smoke. Therefore, the generalisability of the findings is limited to older adolescents. Next, data on peer smoking and parental smoking were not collected. It is reasonable to expect that these two factors could be determinants of sport and educational factors, and consequently directly and/or indirectly influence the associations studied in this investigation. Also, we have evaluated ‘formal’ sport participation only, while some potentially important determinants of physical exercising (ie, self-exercising, fitness centres) were not observed. Finally, students were asked on scholastic achievement over the past semester, which is a relatively crude indicator of overall school success. Therefore, in future studies, more precise evaluation of physical activity and accurate depiction of changes in scholastic achievement (ie, positive or negative changes in scholastic success) are necessary. In order to objectively overview the findings, some specific contextual information on the sociocultural environments is necessary. Bosnia and Herzegovina is a country which is traditionally oriented towards tobacco consumption since the country was part of the former Ottoman Empire, and tobacco farming has been an important part of the economy in some cantons for more than 300 years (eg, particularly in the Herzegovina-Neretva Canton, which is the Mediterranean part of the country).6 As a result, smoking is socially accepted in public, and cigarettes are relatively cheap. Next, although smoking is prohibited in schools, such regulations are really only imposed for closed high school buildings. It is probably even more important that smoking is not prohibited in places of social gatherings (eg, pubs, cafe bars and disco clubs). Finally, although smoking is formally allowed only for those over 18 years, there is no ID control for purchasing cigarettes, while cigarette vending machines are also common. Therefore, although the generalisability of the findings is somewhat limited (mostly for Bosnian and Herzegovinian adolescents aged 16–18 years), the authors are of the opinion that the generalisability is to some extent possible for surrounding countries as well. This is because in many of the countries of former Yugoslavia, smoking is a socially accepted behaviour, there are no strict regulations against smoking in public, and tobacco products are relatively cheap.28 30 The following conclusions can be made. With 28% of adolescents who started smoking before they were 16 and an additional 8% who started smoking between 16 and 18 years of age, the prevalence of smoking is high. As expected from previous investigations, the smoking prevalence was higher in adolescents who achieved poor grades in school. This study expands on previous knowledge by demonstrating that school absences at the age of 16 are a predictor of smoking initiation over the next 2 years (ie, by the end of high school). Additionally, adolescents who reported quitting sports, those who were involved in sports for a relatively short time (ie, up to 5 years), and those who achieved low competitive success by the age of 16 were found to be at risk of starting smoking by the end of high school (ie, 18 years of age). Therefore, to develop accurate and problem-oriented public health policies against smoking in older adolescents, public health authorities should cooperate with school and sport organisations to target the most vulnerable groups of adolescents established in this study. Although it was not among the primary aims of the study, this was the first investigation which indirectly showed that the majority of adolescents from Bosnia and Herzegovina started smoking cigarettes before 16 years of age. Therefore, further investigations should evaluate the predictors of smoking in younger ages. Special thanks go to the Cantonal Ministries of Education which supported and approved the investigation. The authors are particularly grateful to all children who voluntarily participated in the study. Contributors: DS designed the study, performed the statistical analysis and discussed the data; NS, AT and IJ collected the data, overviewed previous research and drafted the manuscript; LO collected the data and discussed the public health issues of the investigation; HP overviewed the previous research and discussed the sport factors in relation to smoking; NZ discussed the data and participated in statistical analyses. All authors have read and approved the final version. Funding: This study and publication were partially financed by the University of Split, Faculty of Kinesiology, Split, Croatia. Competing interests: None declared. Patient consent: Obtained. Ethics approval: The Ethical Boards of University of Split, Faculty of Kinesiology, Split, Croatia and University of Mostar, School of Medicine, Mostar, Bosnia and Herzegovina approved the investigation. Additionally, the study was approved by the Cantonal Ministries of Education. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Data files are freely available here: https://www.dropbox.com/s/n7opixov3hrglj8/data.xlsx?dl=0. ==== Refs References 1 Lee KJ Current smoking and secondhand smoke exposure and depression among Korean adolescents: analysis of a national cross-sectional survey . BMJ Open 2014 ;4 :e003734 10.1136/bmjopen-2013-003734 2 Gendall P , Hoek J , Marsh L Youth tobacco access: trends and policy implications . BMJ Open 2014 ;4 :e004631 10.1136/bmjopen-2013-004631 3 Sekulic D , Ostojic M , Vasilj M Gender-specific predictors of cigarette smoking in adolescents: an analysis of sport participation, parental factors and religiosity as protective/risk factors . J Subst Use 2014 ;19 :89 –94 . 10.3109/14659891.2012.734544 4 The 2011 ESPAD report; substance use among students in 36 European countries; http://www.can.se/contentassets/8d8cb78bbd28493b9030c65c598e3301/the_2011_espad_report_full.pdf (accessed 20 July 2016 ). 5 Zenic N , Terzic A , Rodek J Gender-specific analyses of the prevalence and factors associated with substance use and misuse among Bosniak adolescents . Int J Environ Res Public Health 2015 ;12 :6626 –40 . 10.3390/ijerph120606626 26068091 6 Sekulic D , Ostojic M , Ostojic Z Substance abuse prevalence and its relation to scholastic achievement and sport factors: an analysis among adolescents of the Herzegovina-Neretva Canton in Bosnia and Herzegovina . BMC Public Health 2012 ;12 :274 10.1186/1471-2458-12-274 22480230 7 Zenic N , Ostojic L , Sisic N Examination of the community-specific prevalence of and factors associated with substance use and misuse among rural and urban adolescents: a cross-sectional analysis in Bosnia and Herzegovina . BMJ Open 2015 ;5 :e009446 10.1136/bmjopen-2015-009446 8 Latvala A , Rose RJ , Pulkkinen L Drinking, smoking, and educational achievement: cross-lagged associations from adolescence to adulthood . Drug Alcohol Depend 2014 ;137 :106 –13 . 10.1016/j.drugalcdep.2014.01.016 24548801 9 Wang M , Zhong JM , Fang L Prevalence and associated factors of smoking in middle and high school students: a school-based cross-sectional study in Zhejiang Province, China . BMJ Open 2016 ;6 :e010379 10.1136/bmjopen-2015-010379 10 Pennanen M , Vartiainen E , Haukkala A The role of family factors and school achievement in the progression of adolescents to regular smoking . Health Educ Res 2012 ;27 :57 –68 . 10.1093/her/cyr097 22052215 11 Starr JM , Deary IJ , Fox HC Smoking and cognitive change from age 11 to 66 years: a confirmatory investigation . Addict Behav 2007 ;32 :63 –8 . 10.1016/j.addbeh.2006.03.020 16650620 12 Whalley LJ , Fox HC , Deary IJ Childhood IQ, smoking, and cognitive change from age 11 to 64 years . Addict Behav 2005 ;30 : 77 –88 . 10.1016/j.addbeh.2004.04.014 15561450 13 Anstey KJ , von Sanden C , Salim A Smoking as a risk factor for dementia and cognitive decline: a meta-analysis of prospective studies . Am J Epidemiol 2007 ;166 :367 –78 . 10.1093/aje/kwm116 17573335 14 Bachman JG , O'Malley PM , Schulenberg JE The education-drug use connection: how successes and failures in school relate to adolescent smoking, drinking, drug use, and delinquency . Taylor & Francis , 2012 . 15 Simons-Morton BG , Farhat T Recent findings on peer group influences on adolescent smoking . J Prim Prev 2010 ;31 :191 –208 . 10.1007/s10935-010-0220-x 20614184 16 Filippidis FT , Agaku IT , Vardavas CI The association between peer, parental influence and tobacco product features and earlier age of onset of regular smoking among adults in 27 European countries . Eur J Public Health 2015 ;25 :814 –8 . 10.1093/eurpub/ckv068 25829499 17 Jessor R , Donovan JE , Costa FM Beyond adolescence: problem behaviour and young adult development . Cambridge University Press , 1994 . 18 So WY Association between cigarette consumption and academic achievement in Korean adolescents . Addict Res Theory 2014 ;22 :424 –30 . 10.3109/16066359.2013.874417 19 McGee CE , Trigwell J , Fairclough SJ Effect of a sport-for-health intervention (SmokeFree Sports) on smoking-related intentions and cognitions among 9–10 year old primary school children: a controlled trial . BMC Public Health 2016 ;16 :445 10.1186/s12889-016-3048-3 27229464 20 Veliz PT , Boyd CJ , McCabe SE Competitive sport involvement and substance use among adolescents: a nationwide study . Subst Use Misuse 2015 ;50 :156 –65 . 10.3109/10826084.2014.962049 25290659 21 Mattila VM , Raisamo S , Pihlajamäki H Sports activity and the use of cigarettes and snus among young males in Finland in 1999–2010 . BMC Public Health 2012 ;12 :230 10.1186/1471-2458-12-230 22439614 22 Guo H , Reeder AI , McGee R Adolescents’ leisure activities, parental monitoring and cigarette smoking—a cross-sectional study . Subst Abuse Treat Prev Policy 2011 ;6 :12 10.1186/1747-597X-6-12 21645407 23 Diehl K , Thiel A , Zipfel S How healthy is the behavior of young athletes? A systematic literature review and meta-analyses . J Sports Sci Med 2012 ;11 :201 –20 .24149192 24 Lisha NE , Sussman S Relationship of high school and college sports participation with alcohol, tobacco, and illicit drug use: a review . Addict Behav 2010 ;35 :399 –407 . 10.1016/j.addbeh.2009.12.032 20100638 25 Adachi-Mejia AM , Carlos HA , Berke EM A comparison of individual versus community influences on youth smoking behaviours: a cross-sectional observational study . BMJ Open 2012 ;2 :pii: e000767 10.1136/bmjopen-2011-000767 26 Grunbaum JA , Kann L , Kinchen SA Youth risk behavior surveillance—United States, 2001 . MMWR Surveill Summ 2002 ;51 :1 –62 . 27 Idrizovic K , Zenic N , Tahirajl E Cigarette smoking among 17–18 year old adolescents—prevalence and association with sociodemographic, familial, sport, and scholastic factors . Med Pr 2015 ;66 :153 –63 .26294308 28 Tahiraj E , Cubela M , Ostojic L Prevalence and factors associated with substance use and misuse among Kosovar adolescents; cross sectional study of scholastic, familial-, and sports-related factors of influence . Int J Environ Res Public Health 2016 ;13 :pii: E502 10.3390/ijerph13050502 29 Tuathail GO , O'Loughlin J After ethnic cleansing: return outcomes in Bosnia-Herzegovina a decade beyond war . Ann Assoc Am Geogr 2009 ;99 :1045 –53 . 10.1080/00045600903260671 30 Modric T , Zenic N , Sekulic D Substance use and misuse among 17- to 18-year-old Croatian adolescents: correlation with scholastic variables and sport factors . Subst Use Misuse 2011 ;46 :1328 –34 . 10.3109/10826084.2011.579677 21612340 31 Sekulic D , Tocilj J Pulmonary function in military divers: smoking habits and physical fitness training influence . Mil Med 2006 ;171 :1071 –5 . 10.7205/MILMED.171.11.1071 17153544 32 Sekulic D , Bjelanovic L , Pehar M Substance use and misuse and potential doping behaviour in rugby union players . Res Sports Med 2014 ;22 :226 –39 . 10.1080/15438627.2014.915839 24950111 33 Kondric M , Sekulic D , Petroczi A Is there a danger for myopia in anti-doping education? Comparative analysis of substance use and misuse in Olympic racket sports calls for a broader approach . Subst Abuse Treat Prev Policy 2011 ;6 :27 10.1186/1747-597X-6-27 21988896 34 Lee M Coaching children in sport: principles and practice . Taylor & Francis , 2002 . 35 Moore MJ , Werch CEC Sport and physical activity participation and substance use among adolescents . J Adolesc Health 2005 ;36 :486 –93 . 10.1016/j.jadohealth.2004.02.031 15901513

PMC005xxxxxx/PMC5372022.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01381110.1136/bmjopen-2016-013811Public HealthResearch150617241730My Home is My Marae: Kaupapa Māori evaluation of an approach to injury prevention Hayward Brooke 1Lyndon Mataroria 1Villa Luis 1Madell Dominic 1Elliot-Hohepa Andrea 2Le Comte Lyndsay 1 1 Ko Awatea, Counties Manukau Health, Auckland, New Zealand 2 OTS Consulting, Rotorua, New ZealandCorrespondence to Dr Dominic Madell; dom.madell@middlemore.co.nz2017 20 3 2017 7 3 e01381126 8 2016 9 1 2017 2 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective The objective of this study was to evaluate the New Zealand Accident Compensation Corporation's (ACC) ‘My Home is My Marae’ approach to injury prevention for whānau (families). Setting Over an 18 month period from November 2013 to June 2014, 14 ‘My Home is My Marae’ trials were conducted across the South Auckland and Far North regions of New Zealand. ACC engaged with local Māori providers of healthcare, education and social services to deliver the home safety intervention. Participants Participants of this evaluation were a purposive sample of 14 staff from six provider organisations in South Auckland and the Far North regions of New Zealand. Methods Kaupapa Māori theory-based evaluation and appreciative inquiry methodologies underpinned the evaluation. Interview participants led discussions about strengths and weaknesses of the approach, and partnerships with ACC and other organisations. The evaluation was also supported by pre-existing information available in project documentation, and quantitative data collected by Māori providers. Results Five key critical success factors of ‘My Home is My Marae’ were found from interviews: mana tangata (reputation, respect and credibility); manākitanga (showing care for people); kānohi-ki-te-kānohi (face-to-face approach); capacity building for kaimahi, whānau and providers and ‘low or no cost’ solutions to hazards in the home. Data collected for the Far North area showed that 76% of the hazards identified could be resolved through ‘low or no cost’ solutions. Unfortunately, similar data were not available for South Auckland. Conclusions Injury prevention or health promotion approaches that seek to engage with whānau and/or Māori communities would benefit from applying critical success factors of ‘My Home is My Marae’. AccidentInjuryPreventionKaupapa MāoriEvaluationAccident Compensation Corporationhttp://dx.doi.org/10.13039/100007598 ==== Body Strengths and limitations of this study Kaimahi (staff) shared the perspectives or experiences of whānau (family), but the evaluation did not involve the direct participation of whānau who had completed home safety audits alongside kaimahi. Provider organisations that participated in this evaluation were selected by ACC, and it is possible that provider organisations not involved in the evaluation could have had different experiences that were not captured. While ‘low or no cost’ solutions used in this programme were identified as potentially highly cost-effective for ACC, a specific cost–benefit analysis was not carried out. Incomplete quantitative data meant that it was not possible to draw robust conclusions regarding the reach of the programme and also presented difficulties in being able to compare different patterns in the hazards identified across the Far North and South Auckland regions. Introduction In New Zealand, as in other high-income countries, unintentional injuries represent a significant public health problem causing death, disability, financial and psychological cost.1 The WHO recognises the strong link between unintentional injuries and social deprivation, with more socially deprived groups, such as Māori, having the highest incidence of unintentional injuries.2 Adult Māori (aged 15–64 years) in New Zealand have higher rates of hospitalisation (1788.0 per 100 000 compared to 1104.5 per 100 000; RR 1.62, 95% CI 1.59 to 1.65) and mortality (42.8 per 100 000 compared to 18.7 per 100 000; RR 2.29. 95% CI 2.05 to 2.56) than adult non-Māori due to unintentional injury.3 Further, research by Mauri Ora Associates [ref. 4, p. 6] demonstrated that Māori are frequently unaware of the “services and benefits to which they are entitled” through the Accident Compensation Corporation (ACC) and have lower rates of access to health services, including ACC services, than non-Māori. Māori represent 14.6% of the total New Zealand population, and 11.6% of total ACC claims between 2004 and 2009 were from this ethnic group.5 The home is the site in which the majority of injuries leading to ACC claims occur.6 From 2014 to 2015, one in five people in South Auckland made a claim for an injury in the home.6 South Auckland, including Manukau City, Franklin and Papakura Districts, has a population of 524 505, with one in six people being Māori. Northland, which includes the Far North, Whangerei and Kaipara Districts, has a population of 151 692, with one in three people being Māori. From 2014 to 2015, one in three people in Northland made a claim for an injury that occurred in their home.6 Further, unintentional injuries are responsible for one out of three disabilities suffered by adult Māori.7 Subsequently, prevention of injuries in the home is a key priority area for ACC and they are working to achieve effective injury prevention among whānau and Māori communities through partnering with hapu/iwi and key Māori stakeholders (please see online supplementary Appendix 1 for a glossary of Māori words). ‘My Home is My Marae’ is delivered in partnership with local Māori health, social or community services, in South Auckland and Far North communities. 10.1136/bmjopen-2016-013811.supp1supplementary appendix ‘My Home is My Marae’ was informed by Māori models of health and what is known to work for Māori from previous injury prevention initiatives and available literature. The approach aims to reduce the risk, incidence or severity of injury in the home among Māori communities by empowering whānau with the knowledge and skills they need to identify hazards in their home environment, and to take continued action to minimise, isolate or eliminate hazards in their home. ‘My Home is My Marae’ aimed to work with the Māori working age population (25–64 years of age), who are known to be at most risk of a fall at home.6 The approach also aimed to be inclusive of all whānau, from tamariki to tūnohunohu, who reside at the same whare of those in the working age population. Over an 18 month period from November 2013 to June 2014, 14 ‘My Home is My Marae’ trials were conducted across the South Auckland and the Far North regions. ACC engaged with local Māori providers of healthcare, education and social services to deliver the home safety intervention, as it was considered that there would be a greater chance of success with messages coming via established, trusted and local channels. After commitment from the providers was gained, kaimahi were trained by ACC's injury prevention consultants to conduct home safety audits. This training involved equipping kaimahi with the knowledge, skills and resources to capacitate whānau to identify and reduce hazards that risk injury in the whare, through the implementation of ‘low or no cost’ solutions. ‘Low or no cost solutions’ included, for example, tidying up or reducing clutter such as shoes at the front or back door, tying electrical cables together, removing hazards from walkways in living spaces, moving poisonous household items out of the reach of young children and ensuring spills were cleaned up before anyone slipped. Using low or no cost solutions aimed to make addressing hazards in the home affordable for all whānau. Following training, kaimahi worked together with local whānau to conduct the safety audits in their homes, to raise awareness of hazards in the home and to assist them in keeping their whānau safe. This involved the use of standardised ACC collateral including home safety checklists and teaching aids. The main aspects that contributed to the home safety audit are depicted in figure 1. The approach also required that providers maintained whānau engagement and follow-up by returning 1-year later to review what changes had been made. Figure 1 Main aspects of ACC home safety audit. ACC, New Zealand Accident Compensation Corporation. After making small commitments to change in their whare, whānau were provided with a safety product to assist them to further reduce injury risks such as mould/lichen remover, rug grips/non-slip mats/shower mats/bath mats, non-slip paint for outdoor steps, cable grips or cord winders, step ladders, latches for windows and cupboards, smoke alarms or handrails. In total, kaimahi from provider organisations visited a total of 646 whare housing 2897 individuals. In South Auckland, 404 whare housing a total of 1882 individuals were visited (mean=4.7 people per household), and in the Far North 242 whare housing a total of 1015 individuals were visited (mean=4.7 people per household). The original logic model for ‘My Home is My Marae’ is presented in figure 2 and depicts ‘My Home is My Marae’ resources (‘inputs’), activities or deliverables and expected outcomes. Figure 2 Original logic model for ACC's ‘My Home is My Marae’ approach to injury prevention (provided February 2014). ACC, New Zealand Accident Compensation Corporation. A Kaupapa Māori evaluation of the ‘My Home is My Marae’ trials was completed by Ko Awatea at Counties Manukau Health (Auckland, New Zealand). The evaluation focused on identifying critical success factors of the approach, and strengths and weaknesses of these trials in the context of what is known to work well for Māori in injury prevention, and Māori models of health and well-being. Method The evaluation involved the participation of a purposive sample of 14 kaimahi from six provider organisations who shared their whakāro and experiences delivering ‘My Home is My Marae’ with the evaluation team. ACC selected providers that had started the trials earliest and therefore had the most well-established trials, and where there were established working relationships with ACC from prior contract work. These were (1) Papatūānuku Kōkiri Marae, (2) Tamaki Makaurau Māori Women's Welfare League, (3) Te Kura Kaupapa Māori a Rohe o Māngere, (4) Tumaitua Whānau Trust, (5) Te Hau Ora o Ngāpuhi (previously Kaikohe) and (6) Te Hau Āwhiowhio o Otāngarei Trust. Kaupapa Māori theory-based evaluation and appreciative inquiry methodologies underpinned the evaluation. Appreciative inquiry methodology is a ‘glass-half full’ approach that looks at the best of what already exists, to provide a foundation for thinking about how things could be in an ideal situation.8 The approach works on the principle that focussing on what is valued most allows rapid improvements to a situation to be made. With appreciative inquiry, the best aspects of the programme and how it functions are held uppermost in the minds of the evaluators. Kaupapa Māori evaluation approaches aim to normalise Māori worldviews, values, ways of doing things, customs and language across evaluation processes and outputs.9 Importantly, they also must aim to make a positive difference among Māori communities.10 Kerr11 summarised principles of kaupapa Māori research applied throughout this evaluation including control, challenge, culture, connection, change and credibility. These principles were demonstrated in practice by committing to whakawhanaungatanga through pōwhiri to start the evaluation, and throughout the evaluative process involving kōrero, hui and written communication with providers. Throughout the evaluation, evaluators maintained an open door policy with all evaluation participants who were encouraged to share, clarify and raise concerns as they needed. Evaluators travelled around South Auckland and Far North regions to ensure interviews could be conducted kānohi-ki-te-kānohi. Instead of having predetermined evaluation questioning, kaikōrero led the discussion in the areas of most importance and relevance to their experience. Areas of discussion included, for instance, what was unique about the ‘My Home is My Marae’ approach, what worked and did not work well for whānau, the success of the partnership approach with ACC and other organisations, their experience of programme resourcing and funding, their experience of co-constructing and then implementing trials and observed positive changes in communities brought about by the programme. Allowing kaikōrero to lead and direct kōrero was underpinned by our desire to value tino rangatiratanga – allowing kaikōrero to exercise control over relevant discussion areas and interview procedures. Mihimihi, whakawhanaungatanga, and kai were also an integral part of the interview process. Time was dedicated by the interview facilitator and kaikōrero to accommodate for this before the kōrero started. Māori interviewers, translators and writers enabled extensive use of te reo Māori throughout the interview and written outputs, while also ensuring Māori systems and values were normative in evaluation reporting. A Kaupapa Māori consultant assisted in the design and delivery of the evaluation throughout its duration, to contextualise and analyse provider activities and to provide a framework for the discussion of findings. Hui with evaluation participants provided an opportunity for them to challenge preliminary findings and further ensure that Māori values and priorities were at the fore of the evaluation. A kōrero record was translated (still using te reo Māori to capture cultural terms) and sent to each kaikōrero to review, clarify, amend or share missed information, to ensure the record accurately reflected what they wished to communicate. This allowed time for participants to reflect on the kōrero and add any information they felt was relevant. Kōrero records were thematically analysed with the aid of NVivo software applying an inductive process. Inductive thematic analysis is a widely applied method for the analysis of qualitative data that involves identifying underlying data-driven patterns or themes in narrative or written materials.12 This process involved the evaluation team first familiarising themselves with the records through reading, and in discussion with the kōrero facilitator, who offered valuable insights about the interview process, Māori worldviews and the areas of discussion that kaikōrero were most passionate about. Kōrero records were then coded by one evaluator, and themes developed by three evaluators together with evaluation participants and others as detailed below. We sought to make meaning from the many experiences and perspectives of kaikōrero by forming patterns of ‘truth’.13 As provider experiences were community specific, there was not always consensus regarding what worked, what did not and what is the best way forward, and therefore finding consensus was not a focus of analyses. Potential themes were presented to representatives from all provider organisations, ACC's Community injury prevention consultants and ACC's senior research advisor in hui for review and critique. Feedback from this hui was then used to further define and refine key themes. In addition to the knowledge gained through interviews, the evaluation was also informed by pre-existing information available in project documentation, including letters of agreement, project plans, an intervention logic model (previous presented as figure 2) and existing monitoring and evaluation reports. These documents were provided by ACC and consisted largely of qualitative information. The review of documentation was used to assist in identifying to what extent delivery of the approach aligned with what is already known from the literature about what works best for Māori. The evaluative processes included reviewing the original logic model to assist ACC in updating the model. Logic model review findings are not included in this article. Project documents provided by ACC to the evaluation team were also thematically analysed using a deductive or theory-driven process, whereby existing theoretical models were used to frame and interpret data. This enabled the evaluation team to determine how the ‘My Home is My Marae’ approach aligned with existing models of Māori healthcare, and expectations that Māori have of ACC. Quantitative data were also provided by ACC in Excel documents, which detailed the number of whare visited, age of whānau present during the home safety audit, total number of whānau living in the home, hazards identified and solutions to hazards suggested. This was analysed through basic statistical analyses such as calculating the total number of whare visited, total number of whānau included, total number of hazards identified and per cent of hazard types, locations and solutions applied. Results Five critical success factors underpinning ‘My Home is My Marae’ were developed from analysis of the kōrero: (1) Mana tangata, (2) manākitanga, (3) kānohi-ki-te-kānohi, (4) capacity building for kaimahi, whānau and providers and (5) ‘low cost/ no cost’ solutions to hazards in the home, which are discussed below. Mana tangata and manākitanga Across kaikōrero, a powerful and consistent message emerged: the importance of having the right people at ACC and in the community to support and deliver ‘My Home is My Marae’. Having the ‘right’ people facilitated (1) buy-in from provider organisations to deliver the programme in local communities, (2) engagement and credibility with local communities and (3) the integrity of a ‘by and for Māori’ approach to injury prevention. ‘My Home is My Marae’ was led by ACC's injury prevention consultants who are kānohi kitea and deeply respected by staff from provider organisations. Provider commitment to delivering the ‘My Home is My Marae’ programme was secured by these consultants as a direct result of the mana tangata, or reputation, respect and credibility, of these individuals in Māori communities. The injury prevention consultants acted as key conduits between ACC and provider organisations, managing kōrero between ACC and providers, delivering training and offering their commitment and support to kaimahi from provider organisations who delivered ‘My Home is My Marae’. More importantly, however, the injury prevention consultants were well known and connected to local providers and communities in which the programme was delivered. As illustrated by the following quotes, for all providers, the mana tangata of the consultants—Hineamaru and Sandra—was pivotal in their decision to be a part of the journey in delivering ‘My Home is My Marae’:The key is the relationships- and Hineamaru is the key for us and if it wasn't for Hineamaru then we wouldn't have taken part in this programme… ACC is lucky to have Hineamaru. Everything we did we did through Sandra. Sandra was the conduit between ACC and us. We liked that it was Sandra, she was a great go to person, she understood us as Māori. If it was anyone other than Sandra to come and speak to us we would have said ‘no, it's not worth it’…Sandra is trusted by us and she completed the training which was absolutely vital, she came back to the marae and did extra training—so that's one-to-one going that extra distance. That person is accountable to her marae, her whānau and to the [provider organisation]. Kaikōrero advocated that it is also vital to have “the right mix of providers and the right type of people chosen to go into the homes” to support whānau engagement. As described by one kaikōrero, the ‘My Home is My Marae’ approach works because “The relationships and partnerships with whānau and community are at our level—they are ours.” Kaimahi observed a deep suspicion of government agencies among Māori communities, some of whom are “still fighting for the return of land—their distrust of government agencies [is high].” As such, ACC was not perceived by providers or whānau as a socially or culturally appropriate vehicle for the delivery of messages about home safety to whānau:The perception of whānau is that the main agenda of ACC is to get whānau back into work after suffering injuries. Therefore, whānau are resistant to a relationship with ACC, whereas whānau have relationship with us as providers and as members of their community. This distrust of ACC may also stem from poor previous experiences with claims processes4 and meant that kaimahi had to take time for whakawhanaungatanga: meeting, connecting, explaining the kaupapa and assuring whānau that kaimahi were there for the right reasons. The providers enabled whānau engagement because kaimahi were local Māori, carrying local knowledge, speaking the right language, were personally connected to whānau through whakapapa and their residence in local communities and had the passion and integrity to deliver messages to whānau in a way that is mana enhancing; showing whānau that they are valued and cared for. Kaimahi offered the programme unique characteristics that could not be replicated or reproduced directly by Māori who were not kānohi kitea, or by ACC as an organisation. These characteristics were manākitanga, local knowledge, community connection through whakapapa and local language:If you don't know the whānau then you will get the door shut in your face! So having the relationships and trust with our whānau is important. We knew them through whānau and social circles and that's why the programme was acceptable. Kaimahi have to be skilled in engaging with our local people. You have to speak their language and be a local. They will ask if you are a local, and if you say no, then you will be lucky if they talk to you. They want a connection and we have had to work at building this connection for many years. As providers we all have the passion so we thrived doing this programme. Kānohi-ki-te-kānohi A unique part of the ‘My Home is My Marae’ approach was that it took place in the whare of whānau. This is a significant departure from previous approaches undertaken by ACC and providers and is fundamental to the approach being ‘kaupapa Māori’ as it allows for face-to-face engagement and whakawhanaungatanga with whānau: “This is not just about the project, this is about creating conversations and talking with our people within the home.” While reflecting on how health or injury prevention messages were delivered to whānau previously, one kaikōrero commented that:[Before] it was health promotion stuff, standing at a sports event, handing out panui, having wānanga. Whānau will be picky and come and not retain anything at all. It wasn't very effective. It was a waste of resources. You never really had the opportunity to engage with the whānau and see what it actually looked like at their whare. Pamphlets were only given—that's what it used to be like. In contrast, ‘My Home is My Marae’ gave kaimahi the opportunity to engage with whānau more meaningfully kānohi-ki-te-kānohi in their whare. Kaimahi were grateful for and humbled by this opportunity and recognised the value in connecting with whānau to create opportunities for further work in promoting their health and well-being:The project gave us a reason to go into the homes and meet with whānau. We cannot just go into the homes and look around and inspect as this is disrespectful. This project has enabled us to go into the homes and talk with our [whānau] and say we are doing this project, [we need] your help and this is the understanding of this project and why we are here. However, not all whānau were receptive to the face-to-face approach within their whare. Some required assurance from kaimahi or preferred to engage within different community settings beyond the home such as the marae. Reportedly, however, kānohi-ki-te-kānohi was generally of preference to whānau and was particularly advantageous when engaging with whānau who needed assistance to complete surveys or checklists. The approach was also of particular value to whānau when someone close to them had experienced an injury in the past:The whānau enjoyed having us in their whare and were open to the project due to someone in their whānau (especially their kaumatua) having a fall. The kōrero was the whānau were extremely grateful to have this opportunity because the dangers or the awareness of dangers in the home have been looked at for their homes. Kānohi-ki-te-kānohi repeatedly allowed for (1) kaimahi and provider organisations to better engage and connect with whānau, (2) kaimahi to develop a deeper understanding of the circumstances, home safety hazards and other health issues occurring in the homes of whānau and subsequently to (3) be more responsive to the unique needs and circumstances of whānau, and to hazards in the home (increased responsiveness). Capacity building for kaimahi and whānau Building capacity among kaimahi and whānau was a key strength of the ‘My Home is My Marae’ approach and has occurred in two key ways: increasing the capacity of kaimahi through a train-the-trainer or tuakana-teina approach to training; and empowering whānau to address hazards through changes in their knowledge about hazards in the home. Training was delivered to kaimahi to complete home safety audits collaboratively with whānau. Kaikōrero indicated that training was highly valuable in providing kaimahi with the knowledge they needed. During the training, kaimahi “looked at the principles of the programme and had to familiarise ourselves with its resources. We also had to become competent and confident within ourselves and the kaupapa first.” This illustrates the professional and spiritual journey kaimahi took to learn and apply their new knowledge to make homes a safer place for whānau. The spiritual aspects of their work became more pronounced as kaimahi entered the homes of whānau and became aware of the many hazards and other challenges whānau may encounter in their daily lives. Observing poverty was something that kaimahi carried spiritually as they sought to support and empower whānau with needs that often fell beyond the scope and resources of ‘My Home is My Marae’. The tuakana-teina model supports sustainability of the ‘My Home is My Marae’ approach not only by equipping kaimahi with knowledge around reducing hazards in the home, but also by building leadership and capacity among whānau and rangatahi. As explained by one kaikōrero:The good thing about it is that we are not only talking about the awareness with just the mothers and fathers, it's with the kids—the mokopuna, koroua, and kuia. So this has shown the togetherness and connectedness of the whānau and it has created a wider awareness and a greater involvement of the dangers within the home. Observing the impact their work had on whānau was inspiring for kaimahi. Although measurement of changes in knowledge or awareness among whānau was out of scope of this evaluation, kaimahi reported observing that the approach created awareness among whānau about hazards in the home:Whānau didn't realise the potential consequences of the hazards in their homes because it had never been explained to them. They never thought about hazards because it is the ‘norm’ in whānau homes (that is, it is normal to have these ‘hazards’ in the home as they are not perceived nor recognised as hazards). So this programme was great in that sense that it explained what hazards are and created this awareness for whānau. It was an eye opener for whānau about potential hazards in their homes. The upskilling of whānau promoted tino rangatiratanga in protecting whakapapa. This capacity building of whānau also meant that the providers could “call on whānau that we already have engaged to do the mahi and be facilitators of the programme. They are applying this teaching and these practices within their homes.” The approach aimed to be inclusive of all whānau members, from tamariki to tūnohunohu, who could take action to minimise, eliminate, isolate or reduce hazards. ‘Low or no cost’ solutions to hazards in the home ‘My Home is My Marae’ is predicated on implementing ‘low or no cost’ solutions to hazards in the home. Comments from whānau captured in project documentation stated that “financial costs were the greatest barrier to change” (Tamaki Makaurau Māori Women's Welfare League, 2016). The ability for whānau to reduce hazards in their home with little or no financial cost was a key strength of this approach, particularly when addressing hazards in low-income households. Hazard auditing in Far North whare showed that 76% of the hazards identified and recorded in the whare of whānau could be resolved through ‘low or no cost’ solutions (368 of 481 hazards). 23% of the hazards encountered could not be resolved through low or no cost solutions (113 hazards). These unresolved hazards required a high-cost solution such as plumbing and electrical work (16% or 79 of 481 hazards), other unspecified solutions (6% or 30 of 481 hazards) or had no solution identified (1% or 4 of 481 hazards). Unfortunately, these data were not available for South Auckland. Discussion This evaluation explored critical success factors of ACC's ‘My Home is My Marae’ injury prevention approach and identified mana tangata and manākitanga, kānohi-ki-te-kānohi, capacity building and the implementation of ‘low or no cost’ solutions as integral to the approach and engagement of whānau and/or Māori communities. The evaluation also aimed to understand these factors in the context of what is known to work well for Māori in injury prevention, and Māori models of health and well-being. Having the ‘right people’ to support and deliver ‘My Home is My Marae’ extended from strong Māori leadership at ACC who facilitated buy-in from provider organisations, to frontline staff who were locally informed, present and connected. As trainers, advocates and key conduits between ACC and Māori providers, Māori leaders at ACC are a key support system for kaimahi. Coggan et al14 emphasise the importance of strong Māori support networks and cultural competency of project co-ordinators. Their evaluation of the Waitakere Community Injury Prevention Project highlighted the pivotal role the Māori Coordinator played in building strong support networks for the project at the local marae. The current evaluation showed that without key Māori representation at ACC, providers would have been reluctant to participate. The networks, relationships and trust that the injury prevention consultants were able to leverage to secure provider involvement in ‘My Home is My Marae’ were a key strength of this approach. However, this same strength also presents a challenge for the future growth of ‘My Home is My Marae’. Increased Māori representation at ACC is needed to enable succession planning, growth and increased sustainability of the approach. Beyond Māori leadership, kaimahi—equipped with their care and passion for communities, and local knowledge, language and connection—delivered home safety messages with credibility. A previous evaluation of the Ngāti Porou Community Injury Prevention Project highlighted the importance of staff having strong community ties.15 The project involved working with a rural Māori community in the North Island to deliver road safety campaigns, alcohol and drug programmes and family violence initiatives. It connected with Māori (kānohi-ki-te-kānohi) in hui at various local settings such as marae, Māori immersion schools and sports clubs and achieved a significant increase in awareness of injury prevention among participating whānau: (pre 17% and post 25%, p>0.05). Post-intervention, whānau were more likely to agree that alcohol-related injuries are preventable (pre 2%, post 10%). Evaluators positioned the establishment of strong community ties as pivotal to the project's success.15 Through local Māori providers, the delivery of home safety messages for ‘My Home is My Marae’ reflects a ‘for and by Māori’ approach that enhances whānau engagement and importantly assists in (re)building ACC's relationship and reputation with Māori communities. This is particularly significant given the distrust of ACC among Māori communities.4 A systematic review by Klassen et al16 showed that injury prevention programmes are generally more effective when they are tailored to address unique community characteristics, such as ethnicity and socioeconomic status. The ‘My Home is My Marae’ approach achieves this through localisation of Māori providers with strong community ties, but provides nationally standardised ACC resources (eg, checklist and teaching aids). This evaluation identified an opportunity for injury prevention initiatives to authorise tailoring of resources to address unique community needs, effectively (1) increasing provider autonomy to develop or amend resources, and further (2) supporting increased visibility of provider organisations (eg, through incorporation of provider organisation branding). Brewin and Coggan15 also attributed the success of the Ngāti Porou Community Injury Prevention programme to how the project addressed Māori aspirations. While ‘My Home is My Marae’ creates strong community ties through the contracting of frontline delivery to local Māori providers, the approach fails to connect with/address higher level Māori aspirations for improved well-being and reduced health inequities experienced by Māori communities. A focus on ‘low or no cost’ solutions enables whānau to address minor household hazards in an affordable way. However, broader social causes impacting the state of housing and health and well-being of whānau are not addressed within the scope and resources for ‘My Home is My Marae’. Observations of poverty and poor housing states experienced by some whānau are carried on the wairua of kaimahi whom, beyond their contracted roles, continue to act as advocates for whānau to have high costs hazards and/or substandard housing addressed. Typically, this requires the attention and involvement of stakeholders outside the whānau (such as landlords) and support for whānau to navigate government agencies. Through home safety auditing, kaimahi aimed to increase whānau knowledge of hazards in their whare so that these can be effectively reduced, isolated or eliminated (‘capacity building’). This multifaceted approach to intervention addresses behavioural (knowledge and awareness of whānau to reduce or eliminate hazards) and environmental (changes made in whare) dimensions. This reflects, to some extent, a more holistic approach to health that better aligns with Māori approaches and understandings of health and well-being. Towner and Doswell's research1 was not specific to whānau, but highlighted the importance of stimulating cultural change through a mix of environmental and behavioural intervention for injury prevention. That is, community-based injury prevention efforts should aim to create/build knowledge, but also act on knowledge or create environmental changes. A New Zealand study17 that measured falls at home requiring medical treatment (per person, per year) attributed environmental changes in homes, including the fastening of “handrails along outside steps and internal stairs, grab rails for bathrooms, outside lighting, edging for outside steps, and slip-resistant surfacing for outside areas such as decks and porches” (p. 231), to a 26% reduction in the rate of injuries caused by falls at home per year. Cherrington and Masters18 provided a review of Māori models of health in previous work prepared for injury prevention at ACC and identified three models—Te Whare Tapa Whā, Ngā Pou Mana and Te Wheke—which are regarded as emulating tikanga practices. Cherrington and Masters identified common features of models and frameworks to be a focus on holistic approaches and “acknowledgement of wairuatanga (spirituality) and whānau (family)” [ref. 18, p. 2]. Drawing from these Māori models of health, ‘My Home is My Marae’ takes a preventative approach that is whānau inclusive (reflecting the hauora of the whānau), whānau empowering (improving knowledge and awareness of whānau to reduce, eliminate or isolate hazards in their whare), environmental (making changes in the home—te oa turoa) and physical (reducing the risk or incidence and/or severity of injury in the home—te taha tinana). The aim of empowering and motivating whānau is particularly aligned to dimensions of te taha whānau (family well-being) and whanaungatanga (the extended whānau and social interactions). From previous research, we concluded that the ‘My Home is My Marae’ approach aligns well with what is known to work for Māori in injury prevention for several aspects including leveraging strong community ties, being whanau inclusive, creating and sharing awareness/knowledge, maintaining strong Māori leadership and delivering a multifaceted intervention which simultaneously addresses environmental and behavioural dimensions (educating and acting). Key challenges and opportunities for the future development of ‘My Home is My Marae’ identified in this article include increased Māori leadership at ACC, resourcing to address Maori aspirations of ‘My Home is My Marae’ and tailoring of resources to address unique community needs. There were some limitations to our Kaupapa Māori evaluation of ‘My Home is My Marae’ trials. In many instances, kaimahi shared the perspectives or experiences of whānau, but the evaluation did not involve the direct participation of whānau who had completed home safety audits alongside kaimahi. In addition, provider organisations that participated in this evaluation were selected by ACC and it is possible that provider organisations not involved in the evaluation could have had different experiences that were not captured. While low or no cost solutions used in this programme were identified as potentially highly cost-effective for ACC, a specific cost–benefit analysis was not carried out—this may be a direction for future research. Finally, incomplete quantitative data in Excel spreadsheets meant that it was not possible to draw robust conclusions regarding the reach of the programme and also presented difficulties in being able to compare different patterns in the hazards identified across Far North and South Auckland regions. Conclusions ‘My Home is My Marae’ is a multifaceted approach to intervention that addresses behavioural (knowledge and awareness of whānau to reduce or eliminate hazards) and environmental (changes made in whare) dimensions. The Māori leadership and mana tangata of ACC's injury prevention consultants acting as conduits between ACC and provider organisations, and securing provider's engagement, were key strategic factors for success. Programme delivery by local Māori organisations provided the opportunity to integrate injury prevention in other health promotion activities by these organisations; facilitating a holistic rather than isolated response to whānau needs. ‘My Home is My Marae’ reflects a holistic approach to injury prevention which (to varying extents) aligns with Māori tikanga and Māori models of health and well-being—specifically in that the approach is whānau inclusive (reflecting the hauora of the whānau), whānau empowering (improving knowledge and awareness of whānau to reduce, eliminate or isolate hazards in their whare), environmental (making changes in the home—te oa turoa) and physical (reducing the risk or incidence and/or severity of injury in the home—te taha tinana). Injury prevention or health promotion approaches that seek to engage with whānau and/or Māori communities would benefit from realising critical success factors of ‘My Home is My Marae’. The authors express a deep appreciation to kaikōrero from provider organisations who contributed to this evaluation kōrero—from Management to kaimahi who deliver the approach in the whare of whānau. Your passion and enthusiasm to support whānau is an inspiration to the evaluation team and an asset to ACC and Māori communities. They also thank whānau who so courageously let kaimahi into their homes to tackle hazards head on and make changes for the safety of our whānau—kuia, kororua, pākeke, tamariki and mokopuna. Heoi anō kia ACC mō o koutou tino tautoko I tēnei kaupapa whakahirahira “Ko tōku kainga, tōku marae”. Kia ora anō ACC mō o koutou tautoko I tēnei waitara me te whakanui I te kaupapa. No reira, he mihi mahana anō kia tātou katoa. Contributors: This article has been principally developed by the Research and Evaluation Office of Ko Awatea, Counties Manukau Health: BH, DM, LL and LV. ML also contributed to the article development. AE-H of OTS Consulting was the Kaupapa Māori consultant for this evaluation. BH managed the evaluation process including the development of participant documents, data analysis, evaluation reporting and partnerships with stakeholders. ML was the interviewer for the evaluation and also contributed to the development of information sheets, co-analysed interview data and reviewed evaluation reporting. LV developed the evaluation framework in partnership with ACC and reviewed evaluation reporting. DM drafted the paper and analysed home audit data. AE-H provided cultural knowledge and support as a kaupapa Maori expert and ensured integrity of information sheets, evaluation processes and evaluation reporting. LL reviewed all evaluation documents. Funding: The kaupapa Māori evaluation of ‘My Home is My Marae’ was funded and supported by the Accident Compensation Corporation (ACC), Wellington, New Zealand. Views and/or conclusions in this report are drawn from the analyses of evaluation data completed by Ko Awatea's Research and Evaluation Office (Counties Manukau Health) and may not reflect the position of ACC. Competing interests: The authors declare that they have no competing interests. Ethics approval: Ethical approval to conduct this evaluation was granted by The ACC Ethics Committee. The ACC Ethics Committee is made up of a mix of ACC staff and external representatives who review ACC's research methods to ensure the rights and interests of ACC clients and stakeholders are protected. All participating kaikōrero provided written informed consent to participate in this evaluation. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: ‘My Home is My Marae’ data are not available from Ko Awatea. Interested parties should refer to ACC with data requests. New Zealand injury data are publically available from http://www.acc.co.nz/about-acc/statistics/. ==== Refs References 1 Towner E , Dowswell T Community-based childhood injury prevention interventions: what works? Health Promot Int 2002 ;17 :273 –84 . 10.1093/heapro/17.3.273 12147642 2 Laflamme L , Sethi D , Burrows S Addressing the socioeconomic safety divide: a policy briefing . Copenhagen : WHO Regional Office for Europe , 2009 :38 . 3 Ministry of Health . Tatau Kahukura: Māori Health Chart Book 2010 . 2nd edn Wellington, New Zealand : Ministry of Health , 2010 :77 . 4 Māuri Ora Associates . Māori experience of ACC: Mauri Ora associates final report for Department of Labour . Auckland, New Zealand : Mauri Ora Associates , 2010 :92 . 5 Wyeth EH , Derrett S , Hokowhitu B Indigenous injury outcomes: life satisfaction among injured Māori in New Zealand three months after injury . Health Qual Life Outcomes 2013 ;11 :120 10.1186/1477-7525-11-120 23866834 6 Accident Compensation Corporation (ACC) . Injury statistics tool [internet]. Wellington, New Zealand : ACC , 2015 [cited 1 Dec 2016]. http://www.acc.co.nz/about-acc/statistics/injury-statistics-tool/index.htm 7 Statistics New Zealand, Office for Disability Issues . Disability and Māori in New Zealand in 2006: results from the New Zealand Disability Survey . Wellington, New Zealand : Statistics New Zealand , 2010 :36 . 8 Bushe GR Appreciative inquiry: theory and critique . In: Boje D , Burnes B , Hassard J , eds. The Routledge companion to organizational change . Oxford, UK : Routledge , 2011 :87 –103 . 9 Smith L Decolonizing methodologies: research and indigenous peoples . London, UK : Zed Books , 1999 . 10 Te Ropu Whariki . The evaluation hikoi: a Maori overview of programme evaluation . Auckland, New Zealand : Massey University , 2009 . 11 Kerr S Kaupapa Māori theory-based evaluation . Evaluation J Australasia 2012 ;12 :6 –18 . 12 Boyatzis RE. Transforming qualitative information: thematic analysis and code development . London : Sage , 1998 :184 . 13 Guba E , Lincoln Y Effective evaluation . San Francisco : Jossey Bass , 1983 . 14 Coggan C , Patterson P , Brewin M Evaluation of the Waitakere community injury prevention project . Inj Prev 2000 ;6 :130 –4 . 10.1136/ip.6.2.130 10875670 15 Brewin M , Coggan C Evaluation of the Ngāti Porou community injury prevention project . Ethn Health 2004 ;9 :5 –15 . 10.1080/1355785042000202754 15203462 16 Klassen TP , MacKay JM , Moher D Community-based injury prevention interventions . Future Child 2000 ;10 :83 –110 . 10.2307/1602826 10911689 17 Keall MD , Pierse N , Howden-Chapman P Home modifications to reduce injuries from falls in the Home Injury Prevention Intervention (HIPI) study: a cluster-randomised controlled trial . Lancet 2015 ;385 :231 –8 . 10.1016/S0140-6736(14)61006-0 25255696 18 Cherrington L , Masters B Literature review of Māori models of health and indigenous injury prevention and health promotion interventions . Prepared for Ronald Karaitiana, Cultural Strategies Manager, Injury Prevention Wellington, New Zealand : Accident Compensation Corporation , 2005 .

PMC005xxxxxx/PMC5372023.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01531610.1136/bmjopen-2016-015316Renal MedicineResearch150617281696Three-year outcomes after acute kidney injury: results of a prospective parallel group cohort study Horne Kerry L 1Packington Rebecca 1Monaghan John 2Reilly Timothy 3http://orcid.org/0000-0003-0351-8326Selby Nicholas M 14 1 Department of Renal Medicine, Derby Teaching Hospitals NHS Foundation Trust, Derby, UK 2 Department of Chemical Pathology, Derby Teaching Hospitals NHS Foundation Trust, Derby, UK 3 Department of Informatics, Derby Teaching Hospitals NHS Foundation Trust, Derby, UK 4 Centre for Kidney Research and Innovation, School of Medicine, University of Nottingham, Nottingham, UKCorrespondence to Dr Nicholas M Selby; nicholas.selby@nottingham.ac.uk2017 29 3 2017 7 3 e01531625 11 2016 24 1 2017 23 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives Using a prospective study design, we aimed to characterise the effect of acute kidney injury (AKI) on long-term changes in renal function in a general hospital population. Participants Hospitalised patients with AKI (exposed) and hospitalised patients without AKI (non-exposed), recruited at 3 months after hospital admission. Design Prospective, matched parallel group cohort study, in which renal function and proteinuria were measured at 3 months, 1 year and 3 years. Setting Single UK centre. Clinical end points Clinical end points at 3 years were comparison of the following variables between exposed and non-exposed groups: renal function, prevalence of proteinuria and albuminuria and chronic kidney disease (CKD) progression/development at each time point. CKD progression was defined as a decrease in the estimated glomerular filtration rate (eGFR) of ≥25% associated with a decline in eGFR stage. Results 300 exposed and non-exposed patients were successfully matched 1:1 for age and baseline renal function; 70% of the exposed group had AKI stage 1. During follow-up, the AKI group had lower eGFR than non-exposed patients at each time point. At 3 years, the mean eGFR was 60.7±21 mL/min/1.73 m2 in the AKI group compared with 68.4±21 mL/min/1.73 m2 in the non-exposed group, p=0.003. CKD development or progression at 3 years occurred in 30 (24.6%) of the AKI group compared with 10 (7.5%) of the non-exposed group, p<0.001. Albuminuria was more common in the AKI group, and increased with AKI severity. Factors independently associated with CKD development/progression after AKI were non-recovery at 90 days, male gender, diabetes and recurrent AKI. Conclusions AKI is associated with deterioration in renal function to 3 years, even in an unselected population with predominantly AKI stage 1. Non-recovery from AKI is an important factor determining long-term outcome. AKIAcute Kidney InjuryCKDChronic Kidney DiseaseAlbuminuriaE-alerts ==== Body Strengths and limitations of this study In contrast to much of the previous work examining long-term outcomes after acute kidney injury (AKI), prospective design with matched cohorts and standardised sampling times reduces chances of residual confounding when compared with retrospective studies. General hospitalised patients with AKI have not been well studied previously, despite making up the majority of those affected by AKI. This is a pilot study, designed to test methodology and determine initial results prior to a larger scale study; therefore, sample size is limited to 300. Prospective study design prevents determination of whether patients had pre-existing proteinuria. Introduction Acute kidney injury (AKI) is common, affecting up to 20% of hospitalised patients.1 In addition to well-established short-term risks, there is growing recognition that AKI is associated with adverse long-term consequences.2 Large retrospective analyses of administrative data sets have demonstrated associations between AKI and subsequent increased risk of mortality, chronic kidney disease (CKD) progression and cardiovascular complications. Despite these findings, there remains a need for large-scale prospective observational studies. Prospective studies provide more robust evidence with standardised follow-up assessments that reduce the risk of ascertainment bias, as well as more detailed patient endotyping that lessens the risk of confounding. In addition, it is crucial that studies are generalisable to the wider AKI population, rather than focussing on AKI in specific circumstances. AKI has many different causes and can affect individuals in a wide variety of healthcare settings, and it is plausible that this varied population could have diverse outcomes. Furthermore, detailed and accurate characterisation of individuals with AKI is required to determine factors associated with good and poor outcomes. This includes better description of the natural history of renal impairment following AKI and the point at which CKD development or progression occurs. This evidence gap is being addressed through large-scale prospective observational studies currently in progress. The AKI Risk in Derby (ARID) study and The ASsessment, Serial Evaluation, and Subsequent Sequelae in Acute Kidney Injury (ASSESS-AKI) study3 aim to investigate long-term effects of AKI in unselected generalisable hospitalised populations. We have previously reported evidence of renal impairment and proteinuria persisting to 1 year after AKI from the ARID pilot study;4 we now present the 3-year follow-up results. Methods Study design and methodology is reported in full elsewhere.4 In brief, a prospective parallel group pilot cohort study with intergroup matching was performed between September 2011 and October 2012 at the Royal Derby Hospital to confirm feasibility of the recruitment and data collection methodology for the main ARID main study (ISRCTN25405995). The Royal Derby Hospital is a 1139-bed teaching hospital that provides all major medical and surgical specialties excepting cardiothoracic surgery and has a tertiary referral nephrology unit. A single chemical pathology laboratory covers all inpatient areas and referring primary care centres. Approvals for the study were obtained from Derbyshire Research Ethics Committee and the National Information Governance Board. All patients provided written, informed consent. Participants were patients who sustained AKI at any point during a hospital admission (exposed group) or emergency admissions who did not sustain AKI (non-exposed group). Potential participants were identified through automated screening of results from an electronic AKI detection system as previously described.4 5 AKI was defined by modified Acute Kidney Injury Network (AKIN) criteria (the most current at the time of study inception). Baseline creatinine was the most recent stable serum creatinine from up to 12 months previously; patients were excluded if no baseline value was available. All AKI episodes were subsequently adjudicated by a member of the research team to confirm that AKI rather than CKD progression had been detected. Exposed and non-exposed patients were matched 1:1 for baseline estimated glomerular filtration rate (eGFR) stage (eGFR>60 mL/min/1.73 m2, eGFR stage 3A, 3B or 4) and age ±5 years. Exclusion criteria were age <18 or >85 years, death during admission, receiving palliative care and language barrier preventing remote consent (<1% of local population being non-English speaking). Serum creatinine, eGFR (calculated using CKD Epidemiology Collaboration Equation) and urinary protein excretion were measured at 3 months, 1 year and 3 years after index blood test; for the exposed group, this was day of AKI onset and for the non-exposed group, this was first blood test in admission. Participants were asked not to eat cooked meat for at least 12 hours before giving a blood sample and were asked to provide an early morning urine specimen. Samples were handled separately from routine clinical samples with rapid transfer and analysis within 7 hours in the central hospital laboratory. Hospital admission data, Charlson score, coded comorbidities and mortality were extracted from the hospital patient administration system. Aetiology of AKI was determined by manual review of electronic patient records. Cross referencing with the local renal database was used to track initiation of long-term renal-replacement therapy (RRT). Clinical end points at 3 years were comparison of the following variables between the exposed and non-exposed groups: renal function, prevalence of proteinuria and albuminuria and CKD progression/development at each time point. CKD progression was defined as a decrease in eGFR of ≥25% associated with a decline in eGFR stage. This definition was used in individuals with known CKD at baseline and those with baseline eGFR>60 mL/min/1.73 m2. Proteinuria was defined as PCR>15 mg/mmol and albuminuria as ACR>3.0 mg/mmol. Mortality and a composite renal end point of doubling of serum creatinine, initiation of RRT or eGFR <15 mL/min/1.73 m2 were also recorded. Statistical analysis Parametric data are expressed as mean±SD and non-parametric as median (IQR). The χ2 test was used to compare categorical data. The t-test or Mann-Whitney test was used to compare parametric or non-parametric data, respectively. Binary logistic regression analysis was used to test significant univariate associations with CKD progression. p≤0.05 was considered statistically significant. Statistical analysis was performed with SPSS V.22. Results Participant disposition A total of 2036 patients (975 AKI and 1061 non-exposed patients) met inclusion criteria and were invited to participate. Three hundred and sixty-seven consented to take part and 300 were successfully matched 1:1 according to age and baseline renal function. Figure 1 is a study flow diagram, showing the numbers of participants at each stage of the study. Figure 1 Study flow diagram. Eight people were lost to follow-up after matching. One person withdrew their initial consent, two were no longer interested, one could not participate because of ill health and one for family reasons, one failed to submit samples and we were unable to contact two people. Patient demographics Demographic information is summarised in table 1. There were no significant differences between the exposed and non-exposed groups in age, baseline eGFR stage, eGFR and serum creatinine. Coded comorbidities were similar, though a higher proportion of the exposed group had cancer than non-exposed (eight in the exposed group compared with none in the non-exposed group, p=0.003). Table 1 Baseline characteristics of acute kidney injury and non-exposed groups Non-exposed (n=150) AKI (n=150) p Value Median age (years) 71 (IQR 63–79) 72 (IQR 63–77) 0.6 Number of male patients 85 (57%) 88 (59%) 0.7 Median baseline Charlson score 0 (IQR 0–1) 1 (IQR 0–2) 0.2 Mean baseline creatinine (µmol/L) 93±32 98±33 0.2 Mean baseline eGFR (mL/min/1.73 m2) 71±21 68±19 0.1 Proportion of patients per eGFR stage  eGFR>60 103 (68.7%) 99 (66.0%) 0.7  3A 31 (20.7%) 38 (25.3%)  3B 12 (8.0%) 11 (7.3%)  4 4 (2.7%) 2 (1.3%) Diabetes mellitus 22 (15%) 28 (20%) 0.3 Renal disease 12 (8%) 19 (13%) 0.1 Cancer 0 (0%) 8 (5.6%) 0.003 Previous MI 15 (10.0%) 15 (10.0%) 0.93 Previous stroke 2 (1.3%) 0 0.49 Heart failure 6 (4.0%) 6 (4.0%) 1.0 Peripheral vascular disease 3 (2.0%) 4 (2.7%) 0.67 Pulmonary disease 14 (9.3%) 13 (8.7%) 0.9 One hundred and five (70%) in the exposed group had AKI stage 1, 24 (16%) had stage 2 and 21 (14%) had stage 3 AKI. Only one person required dialysis during their acute admission. The majority of AKI was community-acquired, defined as AKI detected on the first day of hospital admission; this was seen in 103 patients (74%). The most common cause of AKI was multifactorial in 72 (48%) cases; 33 (46%) of these patients were in the perioperative period. Twenty patients (13%) had AKI due to sepsis alone, 12 (8%) due to hypovolaemia and 11 (7%) had urinary tract obstruction. The remaining 24% had a variety of aetiologies including AKI in the setting of acute cardiac disease, contrast-induced AKI and intrinsic renal diseases (in 11 patients (7%)). Renal function eGFR was significantly lower in the exposed group when compared with the non-exposed group at each point of follow-up, and this difference persisted to 3 years (figure 2). At 3 years after AKI, the mean eGFR was 60.7±21 mL/min in the exposed group compared with 68.4±21 mL/min in the non-exposed group, p=0.003. A higher proportion of those in the AKI group had CKD progression between baseline and each follow-up point. At 3 years, the number of AKI patients with CKD progression was 30 (24.6%), compared with 10 (7.5%) of the non-exposed group, p<0.001. Of these, CKD development or progression was already evident at 3 months in 14 (46.7%) exposed patients compared with none of the non-exposed, p=0.007. There was no difference between the proportions of the exposed and non-exposed groups that had CKD progression between 3 months and 3-year follow-up. Similar associations were seen when only the 105 matched pairs with stage 1 AKI were studied; CKD progression from baseline to 3 years was seen in 21 (24.1%) AKI patients compared with 7 (7.4%) non-exposed patients, p=0.002. Figure 2 Mean estimated glomerular filtration rate (eGFR) at each assessment point in non-exposed (no acute kidney injury (AKI)) compared with matched participants with AKI. Estimated GFR measurements are included for inhospital time points to illustrate the effect of AKI; however, the absolute eGFR values at these time points should be interpreted with caution as patients are not in steady state. *p<0.001, **p=0.001, ***p=0.003. Error bars: 95% CI. There were 29 deaths in the 3-year follow-up period; 17 (11.3%) in the AKI group and 12 (8.0%) in the non-exposed group, p=0.3. Reaching the composite renal end point was uncommon, occurring in three AKI and four non-exposed patients, p=0.7. A greater proportion of the exposed group had at least one further episode of AKI during the follow-up period; this was seen in 45 (30%) exposed group compared with 28 (18.7%) non-exposed, p=0.02. Twenty-three (15.3%) of the exposed group were under follow-up at 3 months; this fell to six (7.4%) of the exposed group at 3 years. Proteinuria Albuminuria (ACR≥3 mg/mmol) and proteinuria (PCR≥15 mg/mmol) were relatively common, but to a greater extent in the exposed group at each time point (table 2). At 3 years, 46 (40.7%) of the exposed group had albuminuria compared with 26 (21.1%) non-exposed, p=0.001. Proteinuria was predominantly low grade; at 3 years, the median ACR was 1.7 mg/mmol in exposed and 0.75 mg/mmol in non-exposed groups (p=0.003) and the median PCR was 13 mg/mmol in the exposed group and 10 mg/mmol in the non-exposed group (p=0.006). The ACR/PCR ratio was low in the exposed and non-exposed groups, with median values of <0.4 at all time points. Table 2 Number (%) of acute kidney injury and non-exposed groups with albuminuria (ACR ≥3 mg/mmol) and proteinuria (PCR ≥15 mg/mmol) at each time point 3 months 1 year 3 years AKI (n=150) Non-exposed (n=150) AKI (n=137) Non-exposed (n=147) AKI (n=125) Non-exposed (n=138) Albuminuria number (%) 48 (36.6)* 30 (20.5) 40 (31.5) 36 (25.4) 46 (40.7)*** 26 (21.1) Proteinuria number (%) 71 (50.7)* 53 (36.3) 50 (38.8)** 31 (22.1) 53 (46.5)** 36 (29.8) Significance of comparisons between the AKI group and non-exposed group: *p<0.05, **p≤0.01, ***p≤0.001. There was a graded relationship between AKI severity and proportion of participants with albuminuria (table 3). The proportion of individuals with albuminuria was lowest in non-exposed, and increased as the severity of AKI increased. The same pattern was seen when participants without diabetes were analysed. Table 3 Number (%) of participants with albuminuria at 3 years, stratified by acute kidney injury severity, in the whole cohort and in non-diabetics Non-exposed AKI 1 AKI 2 AKI 3 Whole cohort 26 (21.1%) 31 (38.3%) 5 (29.4%) 10 (66.7%) Non-diabetics 19 (18.1%) 20 (32.8%) 5 (33.3%) 5 (55.6%) Risk factors for CKD progression in those with AKI Associations with CKD progression after AKI were investigated through a subset analysis of only the 194 individuals who sustained AKI, matched and unmatched participants. Comparisons were made between those who did and did not have CKD progression at 3 years. The median age of this subgroup was 71 years (IQR 15) and mean baseline eGFR 63.7±19.6 mL/min, with 84 (43.3%) having pre-existing known CKD (predominantly CKD G3a). The majority (59.3%) were men and 41 (22.5%) had diabetes at baseline. Again, AKI stage 1 was the most common; 138 (71.1%) had stage 1, 31 (16.0%) had stage 2 and 25 (12.9%) had stage 3. At 3 years, CKD progression had occurred in 39 participants (25.3%). The pattern of eGFR over time in those who did and did not have CKD progression at 3 years is shown in figure 3. The majority of those who did not show CKD development or progression at 3 years had recovery of renal function at 3 months after AKI. Complete recovery at 3 months (defined as creatinine <1.1×baseline creatinine) was seen in 85 (74%) of those who did not have CKD progression at 3 years when compared with 10 (25.6%) who did, p<0.001. Univariate and multivariate associations with CKD progression at 3 years are shown in table 4. On binary logistic regression, male gender, greater decline in eGFR at 3 months and diabetes remained independently associated with CKD progression. When decline in eGFR at 3 months was removed from the model, age and diabetes remained independent predictors. Analysis of associations with composite renal end point was limited by small numbers. This occurred in only six individuals (3.1%); four of these individuals had CKD 4 at baseline and five sustained AKI stage 3. Figure 3 Mean estimated glomerular filtration rate (eGFR) over time in individuals after acute kidney injury (AKI) with and without chronic kidney disease progression. Estimated GFR measurements are included for inhospital time points to illustrate the effect of AKI; however, the absolute eGFR values at these time points should be interpreted with caution as patients are not in steady state. Error bars: 95% CI. *p<0.001, **p=0.002. Table 4 Univariate and multivariate associations with chronic kidney disease progression at 3 years Clinical variable Univariate association OR (95% CI) Multivariate association OR (95% CI) Age 1.04 (1.00 to 1.08)* 1.03 (0.98 to 1.08) Male gender 2.57 (1.15 to 5.76)* 2.88 (1.06 to 7.79)* Diabetes 3.65 (1.61 to 8.26)* 5.23 (1.84 to 14.90)* Baseline eGFR (mL/min) 1.00 (0.98 to 1.02) 1.01 (0.98 to 1.04) Delta eGFR at 3 months (quartiles) mL/min  ≤−8.20 14.13 (3.01 to 66.29)* 28.03 (5.37 to 146.31)*  −8.19 to −2.17 1.60 (0.29 to 8.91) 3.73 (0.65 to 21.48)  −2.16 to 5.21 2.68 (0.50 to 14.50) 4.48 (0.8 to 25.13)  ≥5.22 1 (ref) 1 (ref) Severe AKI (AKI stage 2 or 3) 1.26 (0.57 to 2.80) Recurrent AKI 1.59 (0.73 to 3.45) Hospital-acquired AKI 1.7 (0.75 to 3.88) Albuminuria at 3 months 2.09 (0.93 to 4.71) Proteinuria at 3 months 1.39 (0.66 to 2.92) *p<0.005. Sixty-four (33%) had at least one further AKI during the 3-year follow-up period. Recurrent AKI was associated with CKD progression during the follow-up period; 10 (22.7%) of those with recurrent AKI had CKD progression between 3 months and 3 years after the initial AKI episode, compared with 10 (9%) of those without recurrent AKI (p=0.02). This association remained an independent predictor when corrected for age, baseline eGFR and gender (OR 3.48, CI 1.28 to 9.46). Discussion Our results demonstrate that long-term reductions in renal function and proteinuria are associated with AKI. This builds on our previous work, and we now show that this association persists to 3 years following AKI, even in a general hospital population with predominantly AKI stage 1. Epidemiological evidence consistently demonstrates an association between AKI and adverse long-term consequences, including CKD development and progression.2 The nature of the association is thought to be bidirectional; CKD is a risk factor for AKI, but also it is biologically plausible that AKI can cause CKD. Animal models demonstrate that the processes of CKD progression can persist and self-perpetuate after an acute insult.6 In addition, studies blocking fibrogenic signalling cascades following animal models of AKI demonstrate a reduction in subsequent development of tubulointerstitial fibrosis.7 In clinical studies, causality is supported by the association of AKI and subsequent CKD development in children, who do not have the same baseline burden of comorbidity which can confound adult studies.8 9 In addition, examination of the trajectory of eGFR decline in CKD patients suggests that this is not linear, and may therefore be punctuated by acute episodes of deterioration which change the pattern of decline.10 Our results are consistent with this evidence, confirming the association between AKI and subsequent CKD progression in an unselected group of AKI survivors in a general hospital setting. Importantly, this association was seen in those individuals who sustained stage 1 AKI, demonstrating the relevance to even those with less severe renal injury. Although we observed a low combined risk of progression to end-stage kidney failure (ESKD) in this group over a 3-year follow-up period, the incidence of CKD progression was much higher. It is possible that over a longer follow-up period, a greater proportion of these patients would progress to ESKD, although further study is required to substantiate that. Nevertheless, the high incidence of CKD following AKI may have other consequences; higher cardiovascular risk as previously reported11 and in our cohort, a higher risk of recurrent AKI. This is coherent with other descriptions of AKI and CKD forming a bidirectional relationship12 but also opens up the possibility of recurrent AKI as a potentially modifiable risk factor in some patients. Although the risk of CKD progression is increased after AKI, not all individuals are affected. Identifying those at greatest risk is important to target resources appropriately and to reassure individuals at low risk of adverse outcome. Factors identified previously to increase the risk of CKD progression after AKI include AKI severity,13 recurrent AKI,14 increased age15 and diabetes.13 The factors associated with CKD progression that we identified are generally consistent with this, though one of the strongest associations identified was in those individuals with the greatest eGFR decline at 3 months compared with pre-AKI measurements. This suggests that non-recovery of AKI is an important predictor of long-term outcome. This is consistent with a study by D’Hoore et al,16 who demonstrated through analysis of eGFR trajectories that CKD progression was largely determined by non-recovery of AKI. However, others have described an ongoing risk of CKD and mortality, even in individuals with recovery of renal function to <1.10%×baseline at hospital discharge.17 This discrepancy could partly reflect limitations of creatinine-based measures of renal function which can be affected by changes in body mass during acute illness and may not accurately reflect residual renal function at discharge, as well as differences in timing of assessment. However, it is also clear that risk prediction on an individual basis is complex. Novel biomarkers and risk stratification scores validated using representative general hospital populations are needed to improve this. Identification of high-risk patients is important as this cohort may benefit from additional interventions. Potential approaches for future intervention may be based on optimisation of CKD management, in particular pharmacotherapy for control of albuminuria and blood pressure. Indeed, elevation in blood pressure has been demonstrated for 2 years after AKI in previously normotensive adults.18 This may reflect changes in renal salt handling following injury, though prescribing changes after AKI may also be clinically important. Proteinuria and albuminuria following AKI are less commonly reported, but our results suggested that both are common and persistent. Similar observations were seen in the Randomised Evaluation of Normal versus Augmented Levels of RRT (RENAL) trial which showed a prevalence of microalbuminuria or macroalbuminuria of 42.1% at 90 days after episodes of severe AKI in a critical care setting.19 Similarly, in paediatric populations, post-AKI proteinuria has been demonstrated.20 Albuminuria is a risk factor for AKI, and could predate the development of AKI. This makes interpretation of proteinuria challenging, as studies which recruit at the time of AKI lack baseline measurement. Proteinuria assessment is not commonly undertaken as screening in the general population, meaning incidental baseline measurements are likely to have been undertaken on the basis of clinical need and be subject to indication bias. Alternatively, albuminuria could also result from renal damage sustained at the time of AKI. Animal work demonstrates that the tubules have a role in albumin reabsorption,21 so it is biologically plausible that tubular injury could lead to albuminuria. We have shown previously how AKI in a well-characterised CKD population does lead to increased albuminuria.22 Our current work shows that high prevalence of proteinuria is sustained to 3 years after AKI, and there is a graded relationship of AKI severity with prevalence of albuminuria. In addition, the high incidence of albuminuria in those subgroups less likely to have baseline albuminuria (such as non-diabetics) would be consistent with proteinuria resulting from parenchymal damage sustained at the time of AKI. Finally, the low ACR/PCR ratio of <0.4 seen in most participants would be consistent with tubulointerstitial injury.23 Our study does have some limitations. This was a pilot study, powered to test feasibility of the methodology and examine for a signal in our proposed study population. This prevented certain analyses, such as the effect of AKI aetiology, which need a larger sample size and this will be investigated during the main study. A further limitation is that this is a single-centre study and we lack baseline proteinuria measurement. Key strengths are that this is a prospective study with good matching between exposed and non-exposed, and a high degree of individual patient and AKI episode characterisation. Conclusion Our results demonstrate that even less severe AKI can have significant long-term effects on renal function and proteinuria in a general hospital population. Although few progress to end-stage renal disease within 3 years, individuals are at increased risk of subsequent AKI, and in the longer term, CKD progression and its attendant cardiovascular risk. Ongoing large-scale studies will better describe the natural history of AKI and hopefully lead to validated risk predication models to identify individuals at greatest risk. Contributors: KLH is responsible for data collection, analysis and interpretation of data, manuscript drafting, revision and final approval. RP, JM and TR are responsible for data collection and handling, manuscript revision and final approval. NMS is responsible for study conception and design, data collection, data analysis and interpretation, manuscript drafting, revision and final approval. Funding: This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Competing interests: NMS has previously received speaker honoraria from Amgen and Baxter. Patient consent: Obtained. Ethics approval: Derbyshire Research Ethics Committee. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Susantitaphong P , Cruz DN , Cerda J World incidence of AKI: a meta-analysis . Clin J Am Soc Nephrol 2013 ;8 :1482 –93 . 10.2215/CJN.00710113 23744003 2 Coca SG , Singanamala S , Parikh CR Chronic kidney disease after acute kidney injury: a systematic review and meta-analysis . Kidney Int 2012 ;81 :442 –8 . 10.1038/ki.2011.379 22113526 3 Go AS , Parikh CR , Ikizler TA The assessment, serial evaluation, and subsequent sequelae of acute kidney injury (ASSESS-AKI) study: design and methods . BMC Nephrol 2010 ;11 :22 10.1186/1471-2369-11-22 20799966 4 Horne KL , Packington R , Monaghan J The effects of acute kidney injury on long-term renal function and proteinuria in a general hospitalised population . Nephron Clin Pract 2014 ;128 :192 –200 . 10.1159/000368243 25472765 5 Selby NM , Crowley L , Fluck RJ Use of electronic results reporting to diagnose and monitor AKI in hospitalized patients . Clin J Am Soc Nephrol 2012 ;7 :533 –40 . 10.2215/CJN.08970911 22362062 6 Venkatachalam MA , Griffin KA , Lan R Acute kidney injury: a springboard for progression in chronic kidney disease . Am J Physiol Renal Physiol 2010 ;298 :F1078 –94 . 10.1152/ajprenal.00017.2010 20200097 7 Wynn TA Fibrosis under arrest . Nat Med 2010 ;16 :523 –5 . 10.1038/nm0510-523 20448575 8 Mammen C , Al Abbas A , Skippen P Long-term risk of CKD in children surviving episodes of acute kidney injury in the intensive care unit: a prospective cohort study . Am J Kidney Dis 2012 ;59 :523 –30 . 10.1053/j.ajkd.2011.10.048 22206744 9 Greenberg JH , Coca S , Parikh CR Long-term risk of chronic kidney disease and mortality in children after acute kidney injury: a systematic review . BMC Nephrol 2014 ;15 :184 10.1186/1471-2369-15-184 25416588 10 Li L , Astor BC , Lewis J Longitudinal progression trajectory of GFR among patients with CKD . Am J Kidney Dis 2012 ;59 :504 –12 . 10.1053/j.ajkd.2011.12.009 22284441 11 Omotoso BA , Abdel-Rahman EM , Xin W Acute kidney injury (AKI) outcome, a predictor of long-term major adverse cardiovascular events (MACE) . Clin Nephrol 2016 ;85 :1 –11 . 10.5414/CN108671 26636328 12 Chawla LS , Eggers PW , Star RA Acute kidney injury and chronic kidney disease as interconnected syndromes . N Engl J Med 2014 ;371 :58 –66 . 10.1056/NEJMra1214243 24988558 13 Chawla LS , Amdur RL , Amodeo S The severity of acute kidney injury predicts progression to chronic kidney disease . Kidney Int 2011 ;79 :1361 –9 . 10.1038/ki.2011.42 21430640 14 Thakar CV , Christianson A , Himmelfarb J Acute kidney injury episodes and chronic kidney disease risk in diabetes mellitus . Clin J Am Soc Nephrol 2011 ;6 :2567 –72 . 10.2215/CJN.01120211 21903988 15 Schmitt R , Coca S , Kanbay M Recovery of kidney function after acute kidney injury in the elderly: a systematic review and meta-analysis . Am J Kidney Dis 2008 ;52 :262 –71 . 10.1053/j.ajkd.2008.03.005 18511164 16 D'Hoore E , Neirynck N , Schepers E Chronic kidney disease progression is mainly associated with non-recovery of acute kidney injury . J Nephrol 2015 ;28 :709 –16 . 10.1007/s40620-015-0181-5 25700932 17 Jones J , Holmen J , De Graauw J Association of complete recovery from acute kidney injury with incident CKD stage 3 and all-cause mortality . Am J Kidney Dis 2012 ;60 :402 –8 . 10.1053/j.ajkd.2012.03.014 22541737 18 Hsu CY , Hsu RK , Yang J Elevated BP after AKI . J Am Soc Nephrol 2016 ;27 :914 –23 . 10.1681/ASN.2014111114 26134154 19 Gallagher M , Cass A , Bellomo R Long-term survival and dialysis dependency following acute kidney injury in intensive care: extended follow-up of a randomized controlled trial . PLoS Med 2014 ;11 :e1001601 10.1371/journal.pmed.1001601 24523666 20 Askenazi DJ , Feig DI , Graham NM 3–5 year longitudinal follow-up of pediatric patients after acute renal failure . Kidney Int 2006 ;69 :184 –9 . 10.1038/sj.ki.5000032 16374442 21 Russo LM , Sandoval RM , McKee M The normal kidney filters nephrotic levels of albumin retrieved by proximal tubule cells: retrieval is disrupted in nephrotic states . Kidney Int 2007 ;71 :504 –13 . 10.1038/sj.ki.5002041 17228368 22 Horne KL , Shardlow A , Taal MW Long term outcomes after acute kidney injury: lessons from the ARID Study . Nephron 2015 ;131 :102 –6 . 10.1159/000439066 26340262 23 Smith ER , Cai MMX , McMahon LP The value of simultaneous measurements of urinary albumin and total protein in proteinuric patients . Nephrol Dial Transplant 2012 ;27 :1534 –41 . 10.1093/ndt/gfr708 22193048

PMC005xxxxxx/PMC5372024.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01328910.1136/bmjopen-2016-013289DermatologyProtocol1506168716941710Comparative effectiveness of different wound dressings for patients with partial-thickness burns: study protocol of a systematic review and a Bayesian framework network meta-analysis Jiang Qiong Chen Zhao-Hong Wang Shun-Bin Chen Xiao-Dong Fujian Burns Institute, Union Hospital, Fujian Medical University, Fuzhou, ChinaCorrespondence to Dr Xiao-Dong Chen; xiaodongchenfj@mail.china.com2017 22 3 2017 7 3 e0132892 7 2016 31 10 2016 20 12 2016 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction Selecting a suitable wound dressing for patients with partial-thickness burns (PTBs) is important in wound care. However, the comparative effectiveness of different dressings has not been studied. We report the protocol of a network meta-analysis designed to combine direct and indirect evidence of wound dressings in the management of PTB. Methods and analysis We will search for randomised controlled trials (RCTs) evaluating the wound-healing effect of a wound dressing in the management of PTB. Searches will be conducted in MEDLINE, EMBASE, the Cochrane Central Register of Controlled Trials, the Cochrane Wounds Group Specialised Register and CINAHL. A comprehensive search strategy is developed to retrieve articles reporting potentially eligible RCTs. Besides, we will contact the experts in the field and review the conference proceedings to locate non-published studies. The reference lists of articles will be reviewed for any candidate studies. Two independent reviewers will screen titles and abstracts of the candidate articles. All eligible RCTs will be obtained in full text to perform a review. Disagreement on eligibility of an RCT will be solved by group discussion. The information of participants, interventions, comparisons and outcomes from included RCTs will be recorded and summarised. The primary outcome is time to complete wound healing. Secondary outcomes include the proportion of burns completely healed at the end of treatment, change in wound surface area at the end of treatment, incidence of adverse events, etc. Ethics and dissemination The result of this review will provide evidence for the comparative effectiveness of different wound dressings in the management of PTB. It will also facilitate decision-making in choosing a suitable wound dressing. We will disseminate the review through a peer-review journal and conference abstracts or posters. Trial registration number PROSPERO CRD42016041574; Pre-results. partial-thickness burnswound dressingsystematic review protocolnetwork meta-analysis ==== Body Strengths and limitations of this study The comparative effectiveness of different wound dressings in the management of partial-thickness burn has not been studied; our network meta-analysis will be the first to answer the question. We will conduct this network meta-analysis in a Bayesian framework and test robustness of the result with multiple meta-regressions and sensitivity analyses. We will include trials assessing the effectiveness of traditional wound dressings that were treated as standard wound care. These trials are usually published before the year 2000, the quality of which may be generally low. Introduction Partial-thickness burns (PTBs) are burn wounds commonly seen in emergency rooms. PTBs are characterised as blister, swelling and redness on the skin. These skin damages affect the epidermis and structures beneath the epidermis such as blood vessels, hair follicles and nerves.1 PTB is usually categorised as superficial PTB or deep PTB. Superficial PTBs involve damage to the papillary dermis, characterised as intact blisters, moderate oedema, a moist surface under the blisters, a bright pink or red colour. Deep PTBs involve damage to both the papillary and reticular dermis, characterised as broken blisters, substantial oedema, a wet surface, waxy white colour.2 The difference between superficial and deep PTB is whether the burns extend through all skin layers, and deep PTBs do. In the USA alone, there are 486 000 burn injuries (43% caused by fire or flame, 34% by scald, 9% by contact, 4% by electrical burns, 3% by chemical burns and 7% by other reasons) each year.3 More than 60% of the acute hospitalisations are caused by burn injuries.4 Patients suffer pain, scars and mood disturbances like distress, anxiety or depression from PTB. Healthcare costs for burns are high; a median cost of $44 024 is needed for one patient.5 Although direct data on cost for PTB are not available, they were assumed to be costly.6 Therefore, efficient and effective wound care for PTB management is urgently needed. PTB management consists of wound preparation, wound cover and postwound care.7 Wound dressings, covering the wound to accelerate wound healing and protect the wound from infection, include modern dressings like hydrocolloids, hydrofibre, silicones, alginates and polyurethane and traditional dressings like paraffin gauze and silver sulfadiazine (SSD). In the wound dressings, modern dressings are reported to achieve better burn healing than traditional dressings like silver SSD, although quality of the overall evidence is low.8 Additionally, the relative efficacy and toxicity of modern dressings have not been studied. Silver SSD, one of the traditional dressings, has been reported as the gold standard for PTB management.9 However, SSD is under criticism for causing wounds to dry up and not supporting optimal healing.10 Unlike SSD, modern dressings have the advantages of keeping a moist environment around burn wounds and effectively protecting the wound from exposure to pathogenic bacterium.8 Each of these modern dressings has its own features in achieving optimal healing. For example, hydrocolloids, one of the modern dressings, contain gelatin, pectin and sodium carboxymethylcellulose in an adhesive polymer matrix. When the polymer matrix of the hydrocolloid dressings contacting wound exudate would form a gel, so they facilitate autolytic debridement of wounds.11 Polyurethane films are adhesive-coated sheets applied directly to burn wounds. The advantage of polyurethane lies in that it is permeable to water vapour, oxygen and carbon dioxide but not to water or pathogenic bacterium. However, polyurethane is not suitable for wounds with heavy exudate.12 The other modern dressings have their own advantages, which leave the physicians and patient with a selection dilemma. The goal of burn management is to achieve rapid wound healing, pain relief, rehabilitation with minimal scars and optimal functional ability. Therefore, questions are raised: Which are the best wound dressings in achieving rapid healing, in relieving pain, in retaining optimal functional ability or in leaving minimal scars? To answer these questions, we need results from pairwise comparisons of different wound dressings, so a network meta-analysis is warranted. The network meta-analysis combines direct (head-to-head comparisons of different dressings) and indirect (we could simulate dressing A vs B if they share the same comparator C) evidence.13 The analysis is usually done with the Bayesian Markov Chain Monte Carlo (MCMC) method, since it has multiple advantages, such as producing results for all comparisons of interest within a connected network, calculating the probability that each drug is the best treatment and adjusting for correlations within multiarm trials.14 In summary, we will perform a Bayesian network meta-analysis to study the comparative effectiveness of different wound dressings in the management of PTB, providing a reference basis for decision-making. Methods This systematic review and network meta-analysis has been registered (http://www.crd.york.ac.uk/PROSPERO/, reference number: CRD42016041574). Figure 1 gives an overview of the study process. It is anticipated to be finished in December 2016. The review is financially supported by the Key Program of National Clinical Specialty Discipline Construction of China and the Key Clinical Specialty Discipline Construction Program of Fujian. It is also sponsored by the Science and Technology Key Project of Fujian Province, China (reference number: 2014y0056). This study is designed and reported according to the standards of quality for reporting systematic review and meta-analysis protocols (PRISMA-P).15 Figure 1 Flow chart of the systematic review. Eligibility criteria Types of studies We will search for randomised controlled trials (RCTs) evaluating the effectiveness of a wound dressing for wound healing in patients with PTB, since RCTs are recognised as the gold standard in evaluating the effectiveness of an intervention. We will include RCTs comparing wound dressings with no intervention, placebo dressings or other wound dressings (treated as standard wound cover). RCTs with a cross-over design will be excluded, since PTB has rapid evolution (superficial PTBs usually heal within 2 weeks16). Types of participants Patients who have burns classified as PTB (including superficial and deep PTB) will be included in this review. According to the guideline for management of PTBs,7 patients with burns that are wet, painful, blistering, red, white or pink will be included. Patients with burns that are dry, painless, grey, white or brownish will be excluded, for these burns may be full-thickness burns. We will also include patients with burns classified as grade II burns, which used to be another name of PTB.17 We will include both paediatric and adult patients who receive treatments in primary, secondary or tertiary care settings. Patients who need referral to wound-care specialists will be excluded, since using wound dressings alone plays a little role in the treatments for these patients. Additionally, we will include patients with fresh PTBs (within 72 hours after injury), since PTBs beyond 72 hours may be infected and thus need special wound care in addition to wound dressings. Adolescents (age under 18 years) or elders (aged over 65 years) with the size of the burn area surpassing 10% of the total body surface area (TBSA) or adults (aged between 18 and 65 years) with the size surpassing 15% of TBSA will be excluded, since these patients also need special wound care managed by specialists.7 18 19 Patients with burns caused by electric or chemistry or with burns located on the face, neck, hands, feet, armpits, popliteal region or genitals will be excluded, since these burns need special wound care in addition to local wound dressings.7 20 21 Concomitant diseases in the endocrine and immune systems (eg, diabetes) should be reported in RCTs recruiting patients aged over 50 years; otherwise, these RCTs will be excluded. Diseases in these two systems will influence time to heal PTB.22 Types of interventions Both traditional and modern wound dressings will be assessed in this review. Traditional wound dressings refer to paraffin gauze and silver SSD, which are used less often now, because they may cause wounds to dry up and lead to increased risk of infection.7 9 23 However, since they have been conventionally used as standard wound care in the past 30 years,24 we will include them as reference comparators. Modern wound dressings, such as hydrocolloid, hydrogel, hydrofibre, silicones, bioengineered skin substitutes, alginates and polyurethane, will be included for assessment. These modern dressings, compared with traditional dressings, have several advantages of keeping a moist wound environment to facilitate healing, providing an effective barrier to reduce the risk of infection and maintaining maximum contact with the wound to relieve pain.25 Additionally, they are easy to use and remove, which guarantees a minimal extent of pain during attaching or removing dressings. Duration of wound dressings or frequency of changing wound dressings will not be limited. Types of outcomes We will include RCTs evaluating the effects of wound dressings with the following outcomes: time to complete wound healing, proportion of wounds achieving complete closure at the end of treatment (3–4 weeks), change in wound surface area at the end of treatment (3–4 weeks), patient's satisfaction with the attachment and removal of dressings, proportion of patients needing burn care from specialists or surgery, or incidence of adverse events. Search methods We will electronically search in the following database: Ovid MEDLINE (from 1966 to 2016), EMBASE (from 1980 to 2016), the Cochrane Central Register of Controlled Trials (from inception to 2016), the Cochrane Wounds Group Specialised Register (from inception to 2016) and EBSCO CINAHL (from 1980 to 2016). A comprehensive search strategy has been developed (table 1). We will also search clinicaltrials.gov for RCTs that were registered but not reported (from inception to 2016), and we will contact the sponsors or principal investigators of these trials to ask for data. Besides, we will contact the experts in the field and review the conference proceedings to locate non-published studies. The reference lists of articles will be reviewed for any candidate studies. Table 1 Search strategy for MEDLINE (via OVID) N Search terms 1 randomized controlled trial.pt 2 controlled clinical trial.pt 3 randomized.ab 4 placebo.ab 5 randomly.ab 6 trial.ab 7 groups.ab 8 or/1–7 9 exp animals/not humans.sh 10 8 not 9 11 exp Bandages, Hydrocolloid 12 (hydrocolloid$ or hydrofibre or hydrofiber).tw 13 exp alginate-polyethylene glycol acrylate 14 (alginate$ or seasorb or sorbalgon or sorbsan or tegagen).tw 15 exp Hydrogel 16 (hydrogel$ or hydrosorb or novogel or purilon or sterigel).tw 17 exp occlusive dressings 18 foam dressing$.tw 19 (retention tape or hypafix).tw 20 (paraffin gauze).tw 21 (biosynthetic substitute$).tw 22 (antimicrobial dressing$ or acticoat).tw 23 or/11–22 24 exp burns 25 (burn or burns or burned).tw 26 (partial thickness burn$).tw 27 or/24–26 28 10 and 23 and 27 Identification of studies Two reviewers (QJ and S-BW) will independently screen titles and abstracts of the retrieved articles for eligible RCTs. If it is impossible to determine the eligibility of the eligible RCTs through titles and abstracts, we will obtain full-text copies for further evaluation. Disagreement on eligibility of a RCT will be solved by group discussion and arbitrated by a third reviewer (X-DC). Risk of bias assessment Risk of bias of included RCTs will be assessed with the Cochrane risk of bias assessment tool. We will use this assessment tool to determine internal validity of the RCTs in six domains: sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting and other issues. In assessing other issues, we will focus on baseline imbalance and source of financial support. The baseline imbalance may cause overestimation or underestimation of an experimental intervention.26 Receiving financial support from a commercial company that produces and provides active wound dressings may cause overestimation of the effect of this wound dressing. Each of the six domains will be classified as low risk of bias, high risk of bias or unclear risk of bias. Details for this classification are described in the Cochrane handbook.27 Any discrepancy in the risk of bias assessment will be resolved by group discussion and arbitrated by a third reviewer (X-DC). The overall strength of the body of evidence will be assessed using Grades of Recommendation, Assessment, Development, and Evaluation (GRADE).28 Data extraction Standardised extraction forms will be developed to record the following information: study characteristics (first author, publication year, study sites, number of participants, open label/single blind/double blind, study duration, source of financial support), patient's characteristics (age, gender, diabetes (yes/no), wound infected (yes/no), TBSA, body mass index, cause of burn, duration of burn, burn depth), interventions and comparisons (name of experimental or control interventions, duration of treatment, frequency of wound changes, special wound care (yes/no), healthcare facilities where the participants receive wound dressings (primary/secondary/tertiary care)) and outcomes (definition of an outcome, intention-to-treat analysis (yes/no), main results and variables for calculating effect size). Extracted data will be entered into an electronic system developed with the EpiData EntryClient (V.2.0.7.22). To lower the risk of entry error, double data entry and cross-check will be performed in the system. Statistical analysis We will summarise characteristics of the included RCTs and present direct and indirect comparisons between different wound dressings. We will check clinical heterogeneity in the included RCTs through tables and visual plots. Examination of the clinical heterogeneity will focus on patients' baseline characteristics, healthcare facilities where patients receive wound dressings and duration of follow-up. Statistical heterogeneity will also be investigated with the I2 statistics. Different wound dressings will first be categorised as different classes like hydrocolloid, hydrogel, silicones, etc. We will test the agreement in effect size in a specific class, using the I2 value of 50% as a cut-off point. I2>50% indicates significant heterogeneity in the RCTs testing one of the wound dressing classes. We will second test the heterogeneity in individual wound dressings when possible. If the included RCTs show clinically or statistically significant heterogeneity, we will give a narrative review. Otherwise, we will proceed to a traditional meta-analysis, calculating the effect size of different classes of wound dressings compared with placebo or active controls. For continuous outcomes, standardised mean difference (SMD) will be calculated; for dichotomous outcomes, ORs will be computed. We will synthesise SMD or OR with the DerSimonian-Laird method (random-effects model). A Bayesian network meta-analysis will be performed to combine direct and indirect evidence. In this network meta-analysis, measures of relative effect sizes will be ORs in the dichotomous outcomes and SMDs in the continuous outcomes. Both ORs and SMDs will be reported with their 95% credible intervals (95% CI). The 95% CI could be elucidated as a 95% probability that the true OR or SMD falls in the reported range. The network meta-analysis will be performed with a random-effect model using WinBUGS 1.43 (MRC Biostatistics Unit, Cambridge University, UK). Each analysis will be based on non-informative priors for calculation of effect sizes and 95% CI. We will use the surface under the cumulative ranking (SUCRA) to rank the wound dressing classes. The SUCRA presented as percentages compare each wound dressing with an intervention presumed as the best treatment without uncertainty, so a larger SUCRA means a more effective wound dressing class. Transitivity and consistency are the two key elements representing the credibility of a network meta-analysis. The assumption of transitivity is that the distribution of effect modifiers is not different in each pairwise comparison in the network. To account for transitivity in the network, we will assess participant's characteristics (infected or non-infected wounds, patients with or without diabetes, smoker or non-smoker, with or without long-term use of steroids), study designs (duration of follow-up and risk of bias) and interventions (duration of treatment and wound dressings given in primary or secondary or tertiary care). Consistency between direct and indirect comparisons will be assessed. The result of direct comparisons will be acquired through the aforementioned traditional meta-analysis, whereas the result of indirect comparisons will be obtained through excluding those studies with head-to-head comparisons. The consistency will be evaluated with the I2 statistics, with an I2<25% indicating mild inconsistency, an I2 between 25% and 50% showing moderate inconsistency, and an I2>50% representing severe inconsistency. Given the feature of Bayesian statistical analysis, p values will not be estimated and reported for each comparison. Several sensitivity analyses will be performed on the primary outcome to investigate the reasons for potential heterogeneity. The analyses include exclusion of: trials with a high risk of bias (assessed by the Cochrane risk of bias tool); trials including patients with diabetes (patients with diabetes present with a longer time to wound healing); trials including smokers; trials including patients with long-term use of steroids; trials comparing different usages of the same wound dressings (usages that are different in frequency of changing dressings or ingredients); trials with missing SD or 95% CI; trials that are not analysed on an intention-to-treat basis; trials with negative findings (in which the experimental wound dressings or active comparators are not superior to placebo controls). Multiple meta-regressions will be performed to study the impact of sponsorship (whether the sponsors are involved with manufacture of experimental wound dressings or active comparators); mean age of the included participants (ageing leads to delayed epithelialisation, so elders may need longer healing time than adolescents); subtypes of PTB (superficial or deep PTB; patients with deep PTB may theoretically need a longer time in wound healing than those with superficial PTB), blinding method (open label, single blind, double blind), risk of bias (high, unclear, low), TBSA and duration of PTB. If a significant impact of these factors is found, we will perform subgroup analyses according to these factors. Dealing with the missing data We will contact authors who reported trials with missing data to ask for original data. If the original data are not available, we will try to calculate the missing data through other variables given in the articles. For example, we will estimate SD through 95% CI or SE. Discussion This network meta-analysis will summarise the direct and indirect evidence aiming to provide a ranking of the conservative treatment for PTB. The results of this network meta-analysis may help the patients with PTB and their physicians select the best option. To the best of our knowledge, this will be the first network meta-analysis conducted to determine the optimal conservative treatment for PTB. This study protocol is designed to meet the PRISMA-P standards.15 It is designed without knowing the study data or results from the existing published literature. The results of this meta-analysis will help the decision-makers come to their own conclusions regarding which is the best wound dressing for patients with PTB. This work is financially supported by the key Program of National Clinical Specialty Discipline Construction of China and key Clinical Specialty Discipline Construction Program of Fujian, China and also sponsored by the science and technology key project of Fujian Province, China (NO: 2014Y0056). Contributors: X-DC contributed to the conception and design of the study protocol. The search strategy was developed by Z-HC. QJ and S-BW will screen the title and abstract of retrieved studies and extract data from included trials. X-DC will check the accuracy and completeness of data entry. Z-HC will synthesise and analyse the extracted data. All authors drafted this protocol and approved it for publication. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: If this systematic review has results, the authors would like to share the original data with the public. ==== Refs References 1 Klasen HJ A historical review of the use of silver in the treatment of burns. II. Renewed interest for silver . Burns 2000 ;26 :131 –8 . 10.1016/S0305-4179(99)00116-3 10716355 2 Hettiaratchy S , Papini R Initial management of a major burn: II—assessment and resuscitation . BMJ 2004 ;329 :101 –3 . 10.1136/bmj.329.7457.101 15242917 3 American Burn A Burn incidence and treatment in the United States: 2015 . Chicago : American Burn Association , 2015 . 4 Association AB . Burn incidence and treatment in the United States: 2013 fact sheet 2013 http://www ameriburn org/resources_factsheet php 2015 5 Hop MJ , Polinder S , van der Vlies CH Costs of burn care: a systematic review . Wound Repair Regen 2014 ;22 :436 –50 . 10.1111/wrr.12189 25041616 6 Goei H , Hop MJ , van der Vlies CH Return to work after specialised burn care: a two-year prospective follow-up study of the prevalence, predictors and related costs . Injury 2016 ;47 :1975 –82 . 10.1016/j.injury.2016.03.031 27085837 7 Alsbjörn B , Gilbert P , Hartmann B Guidelines for the management of partial-thickness burns in a general hospital or community setting—recommendations of a European working party . Burns 2007 ;33 :155 –60 . 10.1016/j.burns.2006.07.025 17280913 8 Wasiak J , Cleland H , Campbell F Dressings for superficial and partial thickness burns . Cochrane Database Syst Rev 2013 ;(3) :CD002106 10.1002/14651858.CD002106.pub4 23543513 9 Atiyeh BS , Costagliola M , Hayek SN Effect of silver on burn wound infection control and healing: review of the literature . Burns 2007 ;33 :139 –48 . 10.1016/j.burns.2006.06.010 17137719 10 Hoogewerf CJ , Van Baar ME , Hop MJ Topical treatment for facial burns . Cochrane Database Syst Rev 2013 ;(1) :CD008058 10.1002/14651858.CD008058.pub2 23440823 11 Thomas S Hydrocolloid dressings in the management of acute wounds: a review of the literature . Int Wound J 2008 ;5 :602 –13 . 10.1111/j.1742-481X.2008.00541.x 19134061 12 Boateng JS , Matthews KH , Stevens HN Wound healing dressings and drug delivery systems: a review . J Pharm Sci 2008 ;97 :2892 –923 . 10.1002/jps.21210 17963217 13 Lumley T Network meta-analysis for indirect treatment comparisons . Stat Med 2002 ;21 :2313 –24 . 10.1002/sim.1201 12210616 14 Dias S , Welton NJ , Sutton AJ NICE DSU technical support document 2: a generalised linear modelling framework for pairwise and network meta-analysis of randomised controlled trials . London, UK : National Institute for Health and Clinical Excellence , 2011 . 15 Shamseer L , Moher D , Clarke M Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015: elaboration and explanation . BMJ 2015 ;349 :g7647 10.1136/bmj.g7647 16 Schmauss D , Rezaeian F , Finck T Treatment of secondary burn wound progression in contact burns—a systematic review of experimental approaches . J Burn Care Res 2015 ;36 :e176 –89 .25094011 17 Kearns RD , Holmes JH , Cairns BA Burn injury: what's in a name? Labels used for burn injury classification: a review of the data from 2000–2012 . Ann Burns Fire Disasters 2013 ;26 :115 –20 .24563636 18 Travis TE , Moffatt LT , Jordan MH Factors impacting the likelihood of death in patients with small TBSA burns . J Burn Care Res 2015 ;36 :203 –12 . 10.1097/BCR.0000000000000113 25094012 19 Jackson PC , Hardwicke J , Bamford A Revised estimates of mortality from the Birmingham Burn Centre, 2001–2010: a continuing analysis over 65 years . Ann Surg 2014 ;259 :979 –84 . 10.1097/SLA.0b013e31829160ca 23598383 20 Alemayehu H , Tarkowski A , Dehmer JJ Management of electrical and chemical burns in children . J Surg Res 2014 ;190 :210 –13 . 10.1016/j.jss.2014.03.009 24698499 21 Kornhaber R , Wilson A , Abu-Qamar MZ Adult burn survivors’ personal experiences of rehabilitation: an integrative review . Burns 2014 ;40 :17 –29 . 10.1016/j.burns.2013.08.003 24050979 22 Barsun A , Sen S , Palmieri TL A ten-year review of lower extremity burns in diabetics: small burns that lead to major problems . J Burn Care Res 2013 ;34 :255 –60 . 10.1097/BCR.0b013e318257d85b 22929524 23 Vermeulen H , Westerbos SJ , Ubbink DT Benefit and harm of iodine in wound care: a systematic review . J Hosp Infect 2010 ;76 :191 –9 . 10.1016/j.jhin.2010.04.026 20619933 24 Marone P , Monzillo V , Perversi L Comparative in vitro activity of silver sulfadiazine, alone and in combination with cerium nitrate, against staphylococci and gram-negative bacteria . J Chemother 1998 ;10 :17 –21 .9531070 25 Queen D , Orsted H , Sanada H A dressing history . Int Wound J 2004 ;1 :59 –77 . 10.1111/j.1742-4801.2004.0009.x 16722898 26 Roberts C , Torgerson DJ Understanding controlled trials: Baseline imbalance in randomised controlled trials . BMJ 1999 ;319 :185 10.1136/bmj.319.7203.185 10406763 27 Higgins J , Green S Cochrane handbook for systematic reviews of interventions version 5.1. 0 . The Cochrane Collaboration , 2011 , 2013 . 28 Atkins D , Best D , Briss PA Grading quality of evidence and strength of recommendations . BMJ 2004 ;328 :1490 –4 . 10.1136/bmj.328.7454.1490 15205295

PMC005xxxxxx/PMC5372025.txt

==== Front RMD OpenRMD OpenrmdopenrmdopenRMD Open2056-5933BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR 28405471rmdopen-2016-00035810.1136/rmdopen-2016-000358Osteoarthritis1506ReviewMRI-based hip cartilage measures in osteoarthritic and non-osteoarthritic individuals: a systematic review Aguilar Hector N 1Battié Michele C 2http://orcid.org/0000-0001-5314-2297Jaremko Jacob L 11 Faculty of Medicine and Dentistry, Department of Radiology and Diagnostic Imaging, University of Alberta, Edmonton, Alberta, Canada2 Faculty of Rehabilitation Medicine, Department of Physical Therapy, University of Alberta, Edmonton, Alberta, CanadaCorrespondence to Dr Jacob L Jaremko; jjaremko@ualberta.ca2017 22 3 2017 3 1 e0003583 9 2016 27 2 2017 1 3 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Osteoarthritis is a common hip joint disease, involving loss of articular cartilage. The prevalence and prognosis of hip osteoarthritis have been difficult to determine, with various clinical and radiological methods used to derive epidemiological estimates exhibiting significant heterogeneity. MRI-based methods directly visualise hip joint cartilage, and offer potential to more reliably define presence and severity of osteoarthritis, but have been underused. We performed a systematic review of MRI-based estimates of hip articular cartilage in the general population and in patients with established osteoarthritis, using MEDLINE, EMBASE and SCOPUS current to June 2016, with search terms such as ‘hip’, ‘femoral head’, ‘cartilage’, ‘volume’, ‘thickness’, ‘MRI’, etc. Ultimately, 11 studies were found appropriate for inclusion, but they were heterogeneous in osteoarthritis assessment methodology and composition. Overall, the studies consistently demonstrate the reliability and potential clinical utility of MRI-based estimates. However, no longitudinal data or reference values for hip cartilage thickness or volume have been published, limiting the ability of MRI to define or risk-stratify hip osteoarthritis. MRI-based techniques are available to quantify articular cartilage signal, volume, thickness and defects, which could establish the sequence and rate of articular cartilage changes at the hip that yield symptomatic osteoarthritis. However, prevalence and rates of progression of hip osteoarthritis have not been established in any MRI studies in the general population. Future investigations could fill this important knowledge gap using robust MRI methods in population-based cross-sectional and longitudinal studies. Magnetic Resonance ImagingOsteoarthritisArthritis ==== Body Key messages We performed a systematic review of MRI-based estimates of hip articular cartilage in the general population and in patients with established osteoarthritis, using MEDLINE, EMBASE and SCOPUS current to June 2016; 11 studies were found appropriate for inclusion, but they were heterogeneous in osteoarthritis assessment methodology and composition. Overall, the studies consistently demonstrate the reliability and potential clinical utility of MRI-based estimates. However, no longitudinal data or reference values for hip cartilage thickness or volume have been published, limiting the ability of MRI to define, risk-stratified hip osteoarthritis, or satisfactorily answer basic epidemiological questions about hip osteoarthritis. Hip cartilage volume and thickness have been shown to differ by measurement technique, femoral head size, body mass index and sex; more work is necessary to determine the optimal techniques and quantification approaches (automated or otherwise) to establish basic reference values. Longitudinal MRI-based studies that can establish the sequence and rate of articular cartilage changes at the hip that yield clinically and radiologically evident osteoarthritis are needed to aid in understanding normal variation and hip osteoarthritis pathogenesis. Introduction Osteoarthritis (OA), the most common disease of joints, generally involves slow progressive loss of articular cartilage. The associated musculoskeletal sequelae are a significant source for disability and reduced quality of life,1–3 which have been on the rise globally over the past several decades.1 Two of the most common and most disabling sites for OA are the hip and knee joints. Despite OA ranking as the most frequent pathological process at the hip joint,2 considerably fewer studies of OA are aimed at the pathophysiology of hip as compared with the knee. The risk factors for the development of hip OA are similar to those for knee OA, particularly repetitive physical stress, obesity, genetics and advancing age,3–9 although the relationship between increasing bodyweight or body mass index (BMI) and hip OA is complex and may be non-linear.10–15 In any study of prevalence, risk stratification or treatment of hip OA, it is crucial to settle on an accurate disease definition. Historically, this has been done using clinical or radiological criteria, but MRI has potential to improve this process since it allows direct visualisation of cartilage and other articular structures. Numerous robust, multicentre, prospective studies have been conducted to assess a wide variety of end points related to the development of OA at the knee,16–23 but we were surprised to find a relative lack of literature on the prevalence and progression of OA at the hip. We primarily focused on quantitative measures for cartilage volume and thickness which can be measured from routine MRI sequences. For reference, typical findings of superior joint space narrowing (JSN), subchondral cyst formation and marginal osteophytes in two patients with hip OA are demonstrated using standard T1 fat-suppressed and T2-weighted sequences in figure 1. We have also included brief discussions on promising advanced techniques such as T2 mapping and relaxometry, which require specialised MRI protocols. To the best of our knowledge, these advanced methods have not yet been applied to community screening. Figure 1 Typical findings of osteoarthritis at the hip using routine T1-weighted and T2-weighted MRI. (A) Sixty-three years M, T1FS and T2-weighted sequences showing superior joint space narrowing, labral tearing (blue arrow) and an acetabular subchondral cyst (arrow head) adjacent to the labral tear. (B) Fifty-nine years M, T1FS and T2-weighted sequences showing small femoral osteophytes (red arrows), a femoral head–neck junction cyst (arrow head), superior labral tearing (blue arrow) and mild joint space narrowing. M, male; T1FS, T1 fat-suppressed. With respect to MRI-based assessments of hip articular cartilage volume and thickness, we were interested to learn what measures are currently used, whether there are established normative values for these in the general population and, if so, whether hip OA can be defined in terms of variation from these normal values. Furthermore, we wondered whether the rate of change over time for these normative values had been investigated. We, therefore, conducted a systematic review of MRI-based estimates of hip articular cartilage variation and hip OA prevalence and progression in the general population. Methods We searched MEDLINE (1946 to present), EMBASE (1974 to present) and SCOPUS (1960 to present) current to June 2016 using combinations of the search terms, such as ‘hip’, ‘hip joint’, ‘femur’, ‘femoral head’, ‘cartilage’, ‘hyaline cartilage’, ‘articular cartilage’, ‘volume’, ‘thickness’. The screening strategy and results are summarised in the flow diagram in figure 2 (as per the PRISMA 2009 guidelines).24 The full electronic search strategy used for combining MEDLINE and EMBASE searches is demonstrated in the online supplementary appendix. Results were limited to peer-reviewed studies published in English and relating to human participants. We limited studies to ‘primary’ hip OA, as opposed to hip OA resulting from other pathologies (femoroacetabular impingement, developmental hip dysplasia, etc) since primary OA may be a specific entity resulting largely from complex systemic and genetic factors.12 25–27 Figure 2 PRISMA 200924 flow diagram. Potentially relevant published studies were retrieved from database searches, and reduced to the final number (11) for synthesis, as shown. 10.1136/rmdopen-2016-000358.supp1supplementary appendix The initial search produced 1171 non-duplicate titles, which were screened by two reviewers (HNA and JLJ). Common reasons for exclusion at this stage were emphasis on alternate pathologies (developmental dysplasia of the hip, femoroacetabular impingement, etc) and on joints other than hips. After eliminating clearly irrelevant titles, a total of 246 abstracts were further scrutinised, yielding 70 studies for full-text review. Manual searches of the bibliographies for relevant studies yielded an additional five results for full-text review (total 75). At this stage, studies were commonly excluded for an in vitro/cadaveric specimen approach, a primary focus on mathematical or technical aspects of MRI, or for emphasising physicochemical assessments of cartilage in vitro. A further four studies were excluded because the full text was inaccessible, despite efforts to contact the corresponding authors. Finally, 11 publications satisfied the inclusion criteria for this qualitative review. From these studies, we extracted demographic information, MRI sequences used, the primary study end points with respect to MRI measures, the OA defining criteria used to categorise patients and the available prevalence estimates. Results and discussion The 11 relevant works are summarised in table 1. Table 1 MRI-based assessments of hip cartilage or volume in symptomatic (OA) or asymptomatic populations Authors Demographics MRI details Image analysis and reliability Hip cartilage measure(s) Defining criteria for OA Incidence and/or prevalence estimates Cicuttini et al, 200028 n=6: 3M, 3F. Ages 24–65 years. 1.5T, fat-sat T1, side not specified, 3D volume calculation from sagittal images FHCV determined estimation of isotropic voxel size by trilinear interpolation, manual contouring, and data resampling. FHCV estimated by summing pertinent voxels by one user. Intraobserver CoV for FHCV was 6.6%. ICC for FHCV was 0.94. FHCV (1800–7800 mm3) None None Zhai et al, 200529 n=151: participants from TASOACS. 79 M, 72 F. Mean age 63 years 1.5T, 3D fat-sat T1 GRE, right hip, sagittal images Same technique as above Intraobserver and interobserver reliabilities CoV 2.5% and 4.4% FHCV (M: 5900±1000 and F: 4700±800 mm3) FHCT (M: 1.6±0.2 mm F: 1.7±0.2 mm) Altman Radiographic OA in 46% of M and 56% of F. No prevalence estimates for MRI OA Naish et al 200630 n=6: all female. Ages 22–34 years 1.5T, 3D gradient echo, right hip, sagittal images WJCV segmentation to subvoxel accuracy using a semiautomated method. Intraobserver CoV for volume of knee cartilage 1.8% (data not shown). Further detail in Gougoutas et al 200431 WJCT (2.3±0.13 mm) None None Carballido-Gamio et al, 200832 n=7: 5 asymptomatic, 2 with radiographic OA. Mean age 26.6±7.4 years for normal, 54 and 61 years for OA participants 3T, T1ρ and T2 relaxometry, either hip, sagittal images WJCV and WJCT segmentation by one user with a semiautomated technique. Correlation coefficient of volume and thickness estimates compared with saline displacement >0.95. Further detail in Carballido-Gamio et al 200533 WJCV (6263 mm3) WJCT (3.12 mm) K-L scale. K-L≤2 mild, K-L>2 advanced None Khan et al, 201334 n=151: participants from TASOACS. 79M, 72F. Ages 50–81 years 1.5T, fat-sat T1 GRE, right hip, 3D volume calculation from sagittal images Identical methods to Zhai et al, 200529 FHCV (5297 mm3) Altman None Ahedi et al, 201435 n=243: participants from TASOACS. M/F not specified. Mean age 64 years Field strength not specified, STIR, right hip, plane not specified BMLs identified as areas of increased signal intensity on STIR adjacent to the subchondral bone. One user assessed maximum area of the lesion by manual contouring. ICC was 0.98. Semiquantitative assessment of hip cartilage defects Altman 77% had hip cartilage defects, not related to age, sex, or BMI Teichtahl et al, 201410 n=160: participants from MCCS. n=141 non-OA, 19 OA. 56F and 58M. Mean age 66.8 and 59.2 years, respectively 3T, 3D GRE fat-sat T2, PD, spin echo, dominant hip, sagittal images Identical methods to Zhai et al, 2005.29 ICC was 0.99. FHCV (OA: 1763 mm3 Non-OA: 3343 mm3) K-L scale Prevalence of BML and cartilage defects in OA and non-OA participants Teichtahl et al, 201511 n=141: participants from MCCS. 62 M, 79 F. Non-OA. Mean age 66.8 years 3T, 3D GRE fat-sat T2, PD, spin echo, dominant hip, sagittal images Identical methods to Zhai et al, 2005.29 ICC was 0.99. FHCV (M; 3891 mm3 F: 2867 mm3) None Prevalence of hip cartilage defects in non-OA participants Chandra et al, 201636 n=24; healthy asymptomatic volunteers (as per clinical examination, subjective scoring, and radiological evaluation (HOAMS).37 12M, 12F. Ages 23–34 years 3T, single hip for each participant (side not specified), 3D fat-sat PD 3D-SPACE, multiecho spin-echo T2 map in the sagittal plane Fully automated T2 assessment of femoral and acetabular cartilage vs manual segmentation. Automated method claimed to avoid measurement reliability/reproducibility issues WJCV not explicitly stated; figure 5 demonstrates range of ∼4000–12 000 mm3 HOAMS None Ramme et al, 201638 n=20. Hip pain or hip OA. 5M, 15F. Ages 23–74 years 3T, n=10 scanned with 3D true FISP, n=10 scanned with 3D GRE Manual segmentations of the proximal femur and acetabular cartilage GRE and true FISP MR sequences, one observer considered gold standard. Third observer was automated rater. ICC for automated process vs gold standard rater was for GRE 0.286 and 0.614 for true FISP WJCV estimated by an expert (8290–18 880 mm3), a physician (8420–21 330 mm3), and an automated computer algorithm (2300–9580 mm3) None None Gallo et al, 201639 n=54 participants. Longitudinal study on hip OA and FAI. FAI participants excluded. Exclusion criteria included knee OA (K-L>2) and hip K-L=4. 25F, 29M. Mean age 46.5±13.2 years 3T, unilateral hip MR at baseline and 18 months for hip with greater K-L score. Semiquantitative: multiaxial intermediate-weighted fat-sat FSE. Quantitative: sagittal T1ρ/T2 mapping 3D SPGR. Fat-sat 3D MERGE Two expert observers applied SHOMRI scoring system to evaluate the presence of articular cartilage lesions using the three FSE MRI series. SHOMRI has been shown to have high intra-rater and inter-rater reliability (ICC>0.91). Further detail in Lee et al 201440 Semiquantitative assessment of cartilage defects as per SHOMRI.40 Hip OA progression per K-L scale. T1ρ and T2 relaxation times K-L scale At baseline, 16 participants (29.6%) had mild or moderate hip OA (K-L=2, 3). At 18-month follow-up, 9/54 (16.7%) demonstrated progression of hip OA. 3D, three-dimensional; BML, bone marrow lesion; F, female; FAI, femoroacetabular impingement; fat-sat, fat-suppressed; FH, femoral head; FHCT, femoral head cartilage thickness; FHCV, femoral head cartilage volume; FSE, fast spin-echo; GRE, gradient-recalled echo; HOAMS, Hip Osteoarthritis MRI Scoring System; ICC, intraclass correlations; JSN, Joint Space Narrowing; K-L, Kellgren-Lawrence;41 M, male; MCCS, Melbourne Collaborative Cohort Study; MERGE, multiple echo recombined gradient echo; OA, osteoarthritis; PD, proton density; SHOMRI, scoring hip osteoarthritis with MRI; SPGR, segmented spoiled gradient-recalled acquisition; STIR, short T1 inversion recovery sequence; TASOACS, Tasmanian Older Adult Cohort; THR, total hip replacement; WJ, whole-joint; WJCT, whole-joint cartilage thickness; WJCV, whole-joint cartilage volume; WOMAC, Western Ontario and McMaster Universities Osteoarthritis index.42. The studies were heterogeneous in terms of composition, and three of the studies were limited by samples of <10 participants.28 30 32 The studies having more than 100 participants10 11 29 34 35 were derived from the larger databases of the Tasmanian Older Adult Cohort Study (TASOACS) or the Melbourne Collaborative Cohort Study (MCCS). Ten studies included male and female participants, with one of the smaller studies limited to women only.30 Eight of 10 studies relied on routine MRI sequences for cartilage assessment, and 2 emphasised advanced compositional sequences aimed at imaging the cartilage proteoglycan/collagen network.32 39 An advanced discussion of these specialty techniques for assessing articular cartilage is beyond the scope of the current review, and thus reader is referred elsewhere for further details.43 44 Instead of providing quantitative hip cartilage measurements, two studies provided semiquantitative data regarding hip cartilage signal changes, description or quantification of hip cartilage defects, and/or quantification of other MR cartilage parameters such as relaxometry.35 39 OA defining criteria were not used in five studies.11 28 30 38 One study36 used an MRI-based scoring system for OA (Hip Osteoarthritis MRI Scoring System, HOAMS37), whereas the remaining studies used radiographic grading of hip OA as per the Kellgren-Lawrence (K-L) score or Altman atlas. Heterogeneous prevalence data regarding various MRI or radiographic findings at the hip joint were published in five studies.10 11 29 35 39 We were particularly interested in information and data available from the papers related to: (1) MRI-based hip cartilage measures used, (2) normative, quantitative hip cartilage values and associated factors, (3) presence and prevalence of OA based on hip cartilage measures, and (4) rate of change of hip articular cartilage or progression of hip OA as evidenced by changes in articular cartilage. MRI-based hip articular cartilage measurement methods Accurate and reliable measurement of thin curved cartilage plates at the hip is challenging. Two studies used a fully automated segmentation approach36 38 with the remainder using a combination of manual and semiautomated segmentation. Two groups using automated computer-aided segmentation36 38 were aimed at validating this technique and evaluating its performance when combined with different MRI sequences for optimal cartilage detection. Among the non-automated approaches, the most popular method involved application of a trilinear interpolation routine to retrieve isotropic voxels, and thereafter manual segmentation of the femoral head cartilage by manual disarticulation followed by reconstruction into a three-dimensional (3D) volume with summation of pertinent voxels.10 11 28 29 34 A fairly similar semiautomated method was facilitated by a different software algorithm30 and a third approach used techniques including edge detection and edge enhancement to identify voxels containing cartilage.32 Hip cartilage volume measurements on MRI have been directly validated ex vivo in 10 explanted femoral head specimens, where whole-joint cartilage volumes (WJCV) ranged from 1800 to 7800 mm3 by MRI and 1600 to 8100 mm3 by surgical dissection, with measurement error 600 mm3, similar to the observed SD of MRI measurements.28 Normative values for hip cartilage volume and thickness, and associated factors Teichtahl et al studied the dominant hip on 3T MRI in 141 community participants recruited from the MCCS who had never been diagnosed with OA and had no symptoms of hip OA, and 19 separately recruited participants meeting American College of Rheumatology (ACR) criteria for hip OA, including K-L grade >1. This study confirmed that the OA participants had significantly reduced femoral head cartilage volumes (FHCV) compared with their non-OA counterparts, with mean values 1763±321 vs 3343±808 mm3, p<0.001.10 In addition, cartilage defects and bone marrow lesions were more prevalent in the OA participants after adjusting for age, gender and BMI. Further work by the same investigators on presumably the same cohort of 141 non-OA participants showed that FHCV was significantly and substantially higher in men than women (mean 3891±636 vs 2867±451 mm3, respectively). In women only, increasing BMI correlated negatively with FHCV and with increased cartilage defects. Interestingly, increased fat-free mass was beneficial in terms of FHCV for both genders.11 These relations were demonstrated after adjusting for age and femoral head bone area, but unfortunately the authors did not report the correlations between FHCV and age. Hip OA prevalence could not be estimated by this study design. These results confirmed earlier cross-sectional study findings involving the TASOACS cohort, in which significantly larger WJCV was found in 79 men (5900±1000 mm3) vs 72 women (4700±800 mm3).29 This finding could be at least partially attributable to a significantly larger femoral head size in men (men: 18.6±2.0 cm2, women: 14.1±1.5 cm2). As in the Teichtahl et al10 11 studies, higher BMI correlated significantly to decreased hip cartilage volumes in this study (although increased radiographic JSN did not).29 Also, radiographic JSN, but not the presence of osteophytes, correlated with cartilage volume loss, such that each increase in radiographic K-L grade (ie, worsening JSN) resulted in a mean 13% reduction in FHCV, and a 9% reduction in femoral head cartilage thickness. Of the study participants, 46% of men and 56% of women had radiographic OA (K-L grade >1), with no equivalent prevalence estimates for MRI features provided.29 A 2008 feasibility study by Carballido-Gamio et al32 for T1ρ and T2 relaxometry measures for hip articular cartilage used 3T MRI in five healthy participants, one with ‘mild’ OA (K-L grade 1–2) and another with ‘severe’ OA (K-L grade 3–4). WJCV in the five non-OA participants was 5075–7370 mm3, with a mean value 6260 mm3 and a coefficient of variation (CoV) of 2.2%. WJCVs were not reported for the patients with OA. A third cross-sectional study in the TASOACS cohort used 3D fat-suppressed T1 gradient-recalled echo sequences to quantify and correlate MRI-measured cartilage volumes at the hip and knee with each other and with radiographic JSN as per the Altman atlas. The FHCV was relatively high in this study, with a median estimate of 5227 mm3, and this correlated more strongly with total knee cartilage volume than with cartilage volume in a specific knee compartment or with radiographic JSN at either the hip or knee after adjusting for age, sex, height and weight.34 Unfortunately, variation of FHCV with factors such as age was not analysed in that report. The above estimates largely derived from participants with clinical and radiological evidence for hip OA share a significant overlap with the estimates from Chandra et al36 and Ramme et al38 which were conducted on asymptomatic and otherwise healthy individuals. Both groups used a fully automated segmentation approach to provide estimates for WJCV of ∼4000–12 000 and 2300–9580 mm3, respectively, though the former estimates were not stated in the publication text, but rather can be gleaned from a figure contained there. The latter of these two reports acknowledge the use of MR data for which the image thickness exceeded the ideal (<1 mm) for optimal software performance, which may account for some of the discrepancy between the two studies. Still, there remains a significant overlap in WJCV estimates across OA and non-OA individuals, and it is not clear whether such differences are the result of varying approaches to obtaining cartilage measurements, or to what extent morphometric parameters, such as femoral head size, might account for this. Some have taken the approach of estimating the mean whole-joint cartilage thickness (WJCT) rather than volume. Carballido-Gamio et al32 determined average cartilage thickness of 2.8–3.3 mm (mean 3.1 mm) in five normal participants, with CoV 2.2%, and 2.8 mm in a patient with mild OA and 3.4 mm in another with severe OA. Another study used semiautomated segmentation in six healthy women (ages 22–34 years), to visualise WJCT as it varies across the hip.30 The overall mean WJCT was 2.3±0.13 mm thinner than in the Carballido-Gamio et al's32 study and differed from an in vitro measurement in a previous report.45 It remains unclear whether this difference is due to technical issues or differences between the samples. By applying traction to separate the acetabular and femoral head cartilages, others have successfully used automatic segmentation methods to estimate acetabular cartilage thickness at 0.76–3.21 mm (mean 1.60±0.50) in a small sample (n=4), with good agreement to direct ex vivo and semiquantitative measurements.46 Intuitively, it would seem that cartilage thickness across individuals might be less affected than cartilage volume with respect to variation in patient height and weight, which presumably could reflect variations in femoral head and acetabular sizes; however, no data are available, to the best of our knowledge, to clarify this. In summary, hip cartilage has been quantified by several groups using a variety of methods primarily focused on FHCV and WJCV. The reported cartilage volumes vary substantially by method used, sex, BMI, age and OA disease status. Furthermore, it may be important to adjust for joint size for meaningful measurements. Cartilage thickness was measured by fewer authors and varied substantially. Hip OA disease definition, prevalence and rate of progression Owing to cross-sectional study designs, varied participant populations and limited sample sizes, the studies available do not provide sufficient information to enable clear definitions for OA in terms of cartilage volume or thickness. Furthermore, there is a dearth of longitudinal data, from which to estimate hip OA incidence and rate of progression. Unfortunately, even the two Teichtahl et al10 11 studies of the MCCS with 10–15 years of follow-up only used MRI at the final time point, thus no longitudinal data were available. The only longitudinal study among those fitting our search criteria examined T1ρ and T2 relaxation as potential imaging biomarkers for detection of hip OA progression.39 The Scoring Hip Osteoarthritis with MRI (SHOMRI)40 scoring system was used to evaluate the presence of hip cartilage lesions in multiple anatomic subregions using fast spin echo (FSE) images. Baseline cartilage scores for 54 participants were compared with cartilage scores at 18 months. Stratification into disease progression or non-progression was based on increasing SHOMRI scores at follow-up. At baseline, 3D segmented spoiled gradient-recalled acquisition sequences were acquired for T1ρ and T2 relaxation parameters to evaluate their prognostic value for disease progression. At baseline, 16 of 54 participants (29.6%) had mild or moderate hip OA (K-L—2, 3), and at 18-month follow-up, 9 of 54 participants (16.7%) demonstrated progression. The only statistically significant association was for greater BMI with acetabular cartilage lesion progression. The mean baseline SHOMRI scores for femoral cartilage were 2.9 for femoral lesion progressors and 1.3 for non-progressors. The average SHOMRI acetabular cartilage scores were 1.6 and 0.9 for progressors and non-progressors, respectively. Baseline T1ρ and T2 relaxation times were significantly different between progressors and non-progressors in the femoral, but not acetabular, cartilage, even after adjusting for patient demographic factors and the K-L score. The results suggest that T1ρ and T2 relaxation times could be helpful in terms of deriving a disease definition. However, as acknowledged by the authors, it remains difficult to assess whether the observed MRI-based progression rate observed (16.7%) is in agreement with radiographic rates of progression, especially given the relatively young and healthy study population and short follow-up interval. One of the previously mentioned cross-sectional studies of 243 participants from the TASOACS database35 did provide modest prevalence data. Hip OA severity was graded as per the Altman atlas, patient pain graded as per the Western Ontario and McMaster Universities Osteoarthritis index (WOMAC) score and MRI hip cartilage defects were assessed by semiquantitative grading. At least one defect (femoral or acetabular) was found in 189 (77%) of the participants. The presence of these lesions did not correlate with advancing age, sex or increasing BMI, but was associated with other imaging findings, including more severe radiographic features of OA, as well as hip pain in men, but not in women. Overall, we could find no studies assessing the expected rate of change in hip cartilage thickness or volume over time. At the knee, studies have found rates of cartilage thickness loss ranging from 0.4% to 1.9% per year depending on the study cohort.47–50 Conclusion OA is a very complex whole-joint disease, involving many features. Here, we have focused on the available MRI methods of assessing one such feature—cartilage loss, and acknowledge this narrow focus as a limitation of the current work. This systematic review demonstrated that while reliable techniques for measuring hip cartilage volume, thickness and defects are now available on MRI, these approaches have not yet been used to satisfactorily answer basic epidemiological questions about hip OA. Some early promising work using T1ρ and T2 relaxometry demonstrated the potential for prognostication based on these cartilage mapping parameters. It remains unclear, however, how these MRI biomarkers relate to the basic measures of cartilage thickness or volume, or the rate of change thereof. Considering that hip cartilage volume and thickness have been shown to differ by measurement technique, femoral head size, BMI and sex, more work is necessary to determine the optimal techniques and quantification approaches (automated or otherwise) to establish basic reference values. The exact MRI sequences to be assessed for this purpose remain debatable. However, guidelines for an approach to assessment of the hip joint using MRI for clinical trials are now available for reference.51 It would seem that multiaxial, high field strength (3T), high resolution (≤1.5 mm thick slices), fat-suppression techniques could be best suited for this purpose. Further cross-sectional studies will also help in this endeavour, but longitudinal MRI-based studies that can establish the sequence and rate of articular cartilage changes at the hip that yield clinically and radiologically evident OA are also needed. We performed a systematic search of three large databases (MEDLINE, SCOPUS, EMBASE) and found no longitudinal studies assessing changes to hip cartilage volume or thickness on MRI in any cohort. We acknowledge that some relevant search results may have escaped our search strategy; however, on balance, there is a clear need to perform such longitudinal studies, which will be crucial to understanding normal variation and hip OA pathogenesis. The authors thank Linda Slater, Public Services Manager at the John W Scott Health Sciences Library at the University of Alberta, for her help in constructing and executing the database searches. Contributors: HNA, JLJ and MCB were involved in conception and design, analysis and interpretation of the data, critical revision of the article, and final approval of the article. HNA was involved in collection and assembly of data, and drafting of the article. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Vos T , Flaxman AD , Naghavi M Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010 . Lancet 2012 ;380 :2163 –96 . 10.1016/S0140-6736(12)61729-2 23245607 2 Hoaglund FT , Shiba R , Newberg AH Diseases of the hip. A comparative study of Japanese Oriental and American white patients . J Bone Jt Surg Am Vol 1985 ;67 :1376 –83 . 3 Litwic A , Edwards MH , Dennison EM Epidemiology and burden of osteoarthritis . Br Med Bull 2013 ;105 :185 –99 . 10.1093/bmb/lds038 23337796 4 Juhakoski R , Heliövaara M , Impivaara O Risk factors for the development of hip osteoarthritis: a population-based prospective study . Rheumatology (Oxford) 2009 ;48 :83 –7 . 10.1093/rheumatology/ken427 19056801 5 Scher DL , Belmont PJ Jr, Mountcastle S The incidence of primary hip osteoarthritis in active duty US military service members . Arthritis Care Res 2009 ;61 :468 –75 . 10.1002/art.24429 6 Felson DT , Zhang Y An update on the epidemiology of knee and hip osteoarthritis with a view to prevention . Arthritis Rheum 1998 ;41 :1343 –55 . 10.1002/1529-0131(199808)41:8<1343::AID-ART3>3.0.CO;2-9 9704632 7 Saase JL van Romunde LK et al Epidemiology of osteoarthritis: Zoetermeer survey. Comparison of radiological osteoarthritis in a Dutch population with that in 10 other populations . Ann Rheum Dis 1989 ;48 :271 –80 .2712610 8 Tsezou A Osteoarthritis year in review 2014: genetics and genomics . Osteoarthritis Cartilage 2014 ;22 :2017 –24 . 10.1016/j.joca.2014.07.024 25456297 9 Evangelou E , Kerkhof HJ , Styrkarsdottir U A meta-analysis of genome-wide association studies identifies novel variants associated with osteoarthritis of the hip . Ann Rheum Dis 2014 ;73 :2130 –6 . 10.1136/annrheumdis-2012-203114 23989986 10 Teichtahl AJ , Wang Y , Smith S Structural changes of hip osteoarthritis using magnetic resonance imaging . Arthritis Res Ther 2014 ;16 :466 10.1186/s13075-014-0466-4 25304036 11 Teichtahl AJ , Wang Y , Smith S Early cartilage abnormalities at the hip are associated with obesity and body composition measures—a 3.0T MRI community-based study . Arthritis Res Ther 2015 ;17 :107 10.1186/s13075-015-0618-1 25897761 12 Glyn-Jones S , Palmer AJR , Agricola R Osteoarthritis . Lancet 2015 ;386 :376 –87 . 10.1016/S0140-6736(14)60802-3 25748615 13 Wang Y , Wluka AE , Simpson JA Body weight at early and middle adulthood, weight gain and persistent overweight from early adulthood are predictors of the risk of total knee and hip replacement for osteoarthritis . Rheumatology (Oxford) 2013 ;52 :1033 –41 . 10.1093/rheumatology/kes419 23362222 14 Grotle M , Hagen KB , Natvig B Obesity and osteoarthritis in knee, hip and/or hand: an epidemiological study in the general population with 10 years follow-up . BMC Musculoskelet Disord 2008 ;9 :132 10.1186/1471-2474-9-132 18831740 15 Lievense AM , Bierma-Zeinstra SM , Verhagen AP Influence of obesity on the development of osteoarthritis of the hip: a systematic review . Rheumatology (Oxford) 2002 ;41 :1155 –62 . 10.1093/rheumatology/41.10.1155 12364636 16 Roemer FW , Kwoh CK , Hannon MJ What comes first? Multitissue involvement leading to radiographic osteoarthritis: magnetic resonance imaging-based trajectory analysis over four years in the osteoarthritis initiative . Arthritis Rheumatol 2015 ;67 :2085 –96 . 10.1002/art.39176 25940308 17 Joseph GB , McCulloch CE , Nevitt MC A reference database of cartilage 3 T MRI T2 values in knees without diagnostic evidence of cartilage degeneration: data from the osteoarthritis initiative . Osteoarthritis Cartilage 2015 ;23 :897 –905 . 10.1016/j.joca.2015.02.006 25680652 18 Guermazi A , Eckstein F , Hayashi D Baseline radiographic osteoarthritis and semi-quantitatively assessed meniscal damage and extrusion and cartilage damage on MRI is related to quantitatively defined cartilage thickness loss in knee osteoarthritis: the Multicenter Osteoarthritis Study . Osteoarthritis Cartilage 2015 ;23 :2191 –8 . 10.1016/j.joca.2015.06.017 26162806 19 Reichenbach S , Yang M , Eckstein F Does cartilage volume or thickness distinguish knees with and without mild radiographic osteoarthritis? The Framingham Study . Ann Rheum Dis 2010 ;69 :143 –9 . 10.1136/ard.2008.099200 19193659 20 Cotofana S , Benichou O , Hitzl W Is loss in femorotibial cartilage thickness related to severity of contra-lateral radiographic knee osteoarthritis? Longitudinal data from the Osteoarthritis Initiative . Osteoarthritis Cartilage 2014 ;22 :2059 –66 . 10.1016/j.joca.2014.09.020 25262648 21 Hunter DJ , Niu JB , Zhang Y Premorbid knee osteoarthritis is not characterised by diffuse thinness: the Framingham Osteoarthritis Study . Ann Rheum Dis 2008 ;67 :1545 –9 . 10.1136/ard.2007.076810 18218668 22 Eckstein F , Wirth W , Nevitt MC Recent advances in osteoarthritis imaging—the osteoarthritis initiative . Nat Rev Rheumatol 2012 ;8 :622 –30 . 10.1038/nrrheum.2012.113 22782003 23 Hunter DJ , Beavers DP , Eckstein F The Intensive Diet and Exercise for Arthritis (IDEA) trial: 18-month radiographic and MRI outcomes . Osteoarthritis Cartilage 2015 ;23 :1090 –8 . 10.1016/j.joca.2015.03.034 25887362 24 Moher D , Liberati A , Tetzlaff J Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement . PLoS Med 2009 ;6 :e1000097 10.1371/journal.pmed.1000097 19621072 25 Rogers J , Shepstone L , Dieppe P Is osteoarthritis a systemic disorder of bone? Arthritis Rheum 2004 ;50 :452 –7 . 10.1002/art.20136 14872487 26 Felson DT , Lawrence RC , Dieppe PA Osteoarthritis: new insights. Part 1: the disease and its risk factors . Ann Intern Med 2000 ;133 :635 –46 .11033593 27 Felson DT , Lawrence RC , Hochberg MC Osteoarthritis: new insights. Part 2: treatment approaches . Ann Intern Med 2000 ;133 :726 –37 .11074906 28 Cicuttini F , Forbes A , Morris K Determining the volume of hip cartilage by magnetic resonance imaging . Radiography 2000 ;6 :79 –82 . 10.1053/radi.2000.0239 29 Zhai G , Cicuttini F , Srikanth V Factors associated with hip cartilage volume measured by magnetic resonance imaging: the Tasmanian Older Adult Cohort Study . Arthritis Rheum 2005 ;52 :1069 –76 . 10.1002/art.20964 15818695 30 Naish JH , Xanthopoulos E , Hutchinson CE MR measurement of articular cartilage thickness distribution in the hip . Osteoarthr Cartil 2006 ;14 :967 –73 . 10.1016/j.joca.2006.03.017 16713719 31 Gougoutas AJ , Wheaton AJ , Borthakur A Cartilage volume quantification via Live Wire segmentation1 . Acad Radiol 2004 ;11 :1389 –95 . 10.1016/j.acra.2004.09.003 15596377 32 Carballido-Gamio J , Link TM , Li X Feasibility and reproducibility of relaxometry, morphometric. and geometrical measurements of the hip joint with magnetic resonance imaging at 3T . J Magn Reson Imaging 2008 ;28 :227 –35 .18581346 33 Carballido-Gamio J , Bauer JS , Lee K-Y Combined image processing techniques for characterization of MRI cartilage of the knee . Conf Proc IEEE Eng Med Biol Soc 2005 ;3 :3043 –6 . 10.1109/IEMBS.2005.1617116 17282885 34 Khan HI , Aitken D , Zhai G Association between hip and knee cartilage measured using radiographs and magnetic resonance imaging: the Tasmanian older adult cohort study . Rheumatology (Oxford) 2013 ;52 :2009 –15 . 10.1093/rheumatology/ket243 23901132 35 Ahedi H , Aitken D , Blizzard L A population-based study of the association between hip bone marrow lesions, high cartilage signal, and hip and knee pain . Clin Rheumatol 2014 ;33 :369 –76 . 10.1007/s10067-013-2394-0 24196987 36 Chandra SS , Surowiec R , Ho C Automated analysis of hip joint cartilage combining MR T2 and three-dimensional fast-spin-echo images . Magn Reson Med 2016 ;75 :403 –13 . 10.1002/mrm.25598 25644241 37 Roemer FW , Hunter DJ , Winterstein A Hip Osteoarthritis MRI Scoring System (HOAMS): reliability and associations with radiographic and clinical findings . Osteoarthr Cartil 2011 ;19 : 946 –62 . 10.1016/j.joca.2011.04.003 21550411 38 Ramme AJ , Guss MS , Vira S Evaluation of automated volumetric cartilage quantification for hip preservation surgery . J Arthroplasty 2016 ;31 :64 –9 . 10.1016/j.arth.2015.08.009 26377376 39 Gallo MC , Wyatt C , Pedoia V T1ρ and T2 relaxation times are associated with progression of hip osteoarthritis . Osteoarthritis Cartilage 2016 ;24 :1399 –407 . 10.1016/j.joca.2016.03.005 [cited 2016 Jun 18]; Available from: http://linkinghub.elsevier.com/retrieve/pii/S1063458416010621 .26973330 40 Lee S , Nardo L , Kumar D Scoring hip osteoarthritis with MRI (SHOMRI): a whole joint osteoarthritis evaluation system . J Magn Reson Imaging 2015 ;41 :1549 –57 . 10.1002/jmri.24722http://www.scopus.com/inward/record.url?eid=2-s2.0-84906033530&partnerID=40&md5=d9f6fc0006be9289474cf58843de1f4c 25139720 41 Kellgren JH , Lawrence JS Radiological assessment of osteo-arthrosis . Ann Rheum Dis 1957 ;16 :494 –502 . 10.1136/ard.16.4.494 13498604 42 Bellamy N , Klestov A , Muirden K Perceptual variation in categorizing individuals according to American College of Rheumatology classification criteria for hand, knee, and hip osteoarthritis (OA): observations based on an Australian Twin Registry study of OA . J Rheumatol 1999 ;26 :2654 –8 .10606378 43 Guermazi A , Alizai H , Crema MD Compositional MRI techniques for evaluation of cartilage degeneration in osteoarthritis . Osteoarthr Cartil 2015 ;23 :1639 –53 . 10.1016/j.joca.2015.05.026 26050864 44 Potter HG , Black BR , Chong le R New techniques in articular cartilage imaging . Clin Sports Med 2009 ;28 :77 –94 . 10.1016/j.csm.2008.08.004 19064167 45 Adam C , Eckstein F , Milz S The distribution of cartilage thickness within the joints of the lower limb of elderly individuals . J Anat 1998 ;193 :203 –14 . 10.1046/j.1469-7580.1998.19320203.x 9827636 46 Abraham CL , Bangerter NK , McGavin LS Accuracy of 3D dual echo steady state (DESS) MR arthrography to quantify acetabular cartilage thickness . J Magn Reson Imaging 2015 ;42 : 1329 –38 . 10.1002/jmri.24902 25851109 47 Carnes J , Stannus O , Cicuttini F Knee cartilage defects in a sample of older adults: natural history, clinical significance and factors influencing change over 2.9 years . Osteoarthr Cartil 2012 ;20 :1541 –7 . 10.1016/j.joca.2012.08.026 22960091 48 Roemer FW , Felson DT , Wang K Co-localisation of non-cartilaginous articular pathology increases risk of cartilage loss in the tibiofemoral joint—the MOST study . Ann Rheum Dis 2013 ;72 :942 –8 . 10.1136/annrheumdis-2012-201810 22956600 49 Buck RJ , Wirth W , Dreher D Frequency and spatial distribution of cartilage thickness change in knee osteoarthritis and its relation to clinical and radiographic covariates—data from the osteoarthritis initiative . Osteoarthr Cartil 2013 ;21 :102 –9 . 10.1016/j.joca.2012.10.010 23099212 50 Ding C , Cicuttini F , Scott F Association of prevalent and incident knee cartilage defects with loss of tibial and patellar cartilage: a longitudinal study . Arthritis Rheum 2005 ;52 : 3918 –27 . 10.1002/art.21474 16320339 51 Gold GE , Cicuttini F , Crema MD OARSI clinical trials recommendations: hip imaging in clinical trials in osteoarthritis . Osteoarthr Cartil 2015 ;23 :716 –31 . 10.1016/j.joca.2015.03.004 25952344

PMC005xxxxxx/PMC5372027.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01311710.1136/bmjopen-2016-013117Diabetes and EndocrinologyProtocol150618431709Smartphone application for women with gestational diabetes mellitus: a study protocol for a multicentre randomised controlled trial Borgen Iren 1Garnweidner-Holme Lisa Maria 1Jacobsen Anne Flem 2Bjerkan Kirsti 3Fayyad Seraj 4Joranger Pål 1Lilleengen Anne Marie 1Mosdøl Annhild 5Noll Josef 4Småstuen Milada Cvancarova 1Terragni Laura 1Torheim Liv Elin 1http://orcid.org/0000-0002-9873-4023Lukasse Mirjam 1 1 Faculty of Health Sciences, Oslo and Akershus University College of Applied Sciences, Oslo, Norway 2 University of Oslo, Oslo, Norway 3 Division of Medicine, Section of Dietetics, Oslo University Hospital, Oslo, Norway 4 University Graduate Centre, Kjeller, Norway 5 The Norwegian Knowledge Centre for the Health Services, Oslo, NorwayCorrespondence to Iren Borgen; iren.borgen@hioa.no2017 27 3 2017 7 3 e01311723 6 2016 30 11 2016 17 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction The promotion of a healthy diet, physical activity and measurement of blood glucose levels are essential components in the care for women with gestational diabetes mellitus (GDM). Smartphones offer a new way to promote health behaviour. The main aim is to investigate if the use of the Pregnant+ app, in addition to standard care, results in better blood glucose levels compared with current standard care only, for women with GDM. Methods and analysis This randomised controlled trial will include 230 pregnant women with GDM followed up at 5 outpatient departments (OPD) in the greater Oslo Region. Women with a 2-hour oral glucose tolerance test (OGTT) ≥9 mmol/L, who own a smartphone, understand Norwegian, Urdu or Somali and are <33 weeks pregnant, are invited. The intervention group receives the Pregnant+ app and standard care. The control group receives standard care only. Block randomisation is performed electronically. Data are collected using self-reported questionnaires and hospital records. Data will be analysed according to the intention-to-treat principle. Groups will be compared using linear regression for the main outcome and χ2 test for categorical data and Student's t-test or Mann-Whitney-Wilcoxon test for skewed distribution. The main outcome is the glucose level measured at the 2-hour OGTT 3 months postpartum. Secondary outcomes are a change in health behaviour and knowledge about GDM, quality of life, birth weight, mode of delivery and complications for mother and child. Ethics and dissemination The study is exempt from regional ethics review due to its nature of quality improvement in patient care. Our study has been approved by the Norwegian Social Science Data Services and the patient privacy protections boards governing over the recruitment sites. Findings will be presented in peer-reviewed journals and at conferences. Trial registration number NCT02588729, Post-results. gestational diabetes mellitussmartphone applicationdietphysical activityrandomized controlled trialpregnancy ==== Body Strengths and limitations of this study Tailored health information through the Pregnant+ app in Norwegian, Urdu or Somali for women with gestational diabetes mellitus. Automatic transfer of blood glucose levels from the glucometer to the app via Bluetooth Low Energy. Privacy protection of participants' information through local storage instead of cloud service. No blinding of participants, staff and researchers but blinding of those analysing samples and the statistician. Introduction Gestational diabetes mellitus (GDM), defined as glucose intolerance first identified in pregnancy, is an increasing problem among pregnant women worldwide.1 The prevalence is ranging from 1.7% to 20% depending on diagnosis criteria and population characteristics.1 On the basis of cases reported to the Norwegian Medical Birth Registry (NMBR), the prevalence of GDM in Norway is ∼2%.2 However, in a cohort study where all pregnant women were screened for GDM, the prevalence was 13% overall, with 11% in ethnic Norwegians and 14.6% in groups of non-European origin.3 It is therefore likely that GDM is underdiagnosed in the general population. The majority of pregnant women with GDM recover from their glucose intolerance once the pregnancy is over.4 However, research suggests that a history of GDM increases the risk for developing type 2 diabetes mellitus (T2DM) later in life.5 6 Well-known risk factors for GDM include maternal obesity, advanced maternal age, ethnicity, family history of diabetes and a previous history of GDM.5 7 8 Women with GDM have an increased risk for pre-eclampsia, induction of labour, birth injuries, postpartum haemorrhage and caesarean section.9 GDM increases the infant's risk of macrosomia, birth injuries, neonatal hypoglycaemia and stillbirth.8 9 In addition, the Hyperglycaemia and Adverse Pregnancy Outcome study (HAPO) suggests an increased risk of adverse maternal and infant outcomes with increasing levels of hyperglycaemia.10 In Norway, pregnant women with a 2-hour oral glucose tolerance test (OGTT) ≥9 mmol/L receive additional healthcare at a specialised Outpatient Department (OPD).11 New national Norwegian diabetic guidelines are being developed and expected to be implemented by 2017. Current care in Norway includes giving information about a healthy diet, physical activity and monitoring of blood glucose levels and observing the fetus through cardiotocography (CTG) and ultrasound.12 Health information is commonly given verbally, often accompanied by leaflets. Anecdotal evidence indicates that information on non-western food items in different languages is limited. During the restricted time of clinic visits, information about healthy eating and physical activity competes with other components of care and other information. Health information via an app is easily and constantly available. Automatic transfer of blood glucose measurements to the app provides a novel way to monitor blood glucose levels. Nearly 80% of the adult population in Norway has a smartphone.13 Pregnant women are mostly young adults who are familiar with the use of electronic devices such as smartphones. In a previous review of health behaviour interventions and use of apps, the authors found that apps are highly accepted by mobile phone users and may be a suitable way of providing health interventions. Even though previous smartphone-based randomised controlled trials (RCTs) show promising results for the self-management of diabetes and lifestyle factors,14 15 more research is needed to define the exact efficacy of apps on health outcomes.16 Since the prevalence of GDM is higher among certain groups of women from Asia and Africa compared with Norwegians, health information needs to be culture-sensitive, easy understandable and meet the individual's needs.3 17–19 A systematic review indicated better blood sugar control with culturally tailored counselling to ethnic minority patients with diabetes compared with standard care.20 We have developed an app for women with GDM, the Pregnancy+ app.21 This app is available in Norwegian, Urdu and Somali and consist of linguistically and culturally adapted information. These two groups of non-Western immigrants in Norway, Pakistani and Somali were selected due to their high number of annual births and high risk for GDM.3 17 The Pregnant+ app automatically transfers blood glucose measurements to the smartphone and provides a graphic overview indicating if the levels are satisfactory. Additionally, the app provides information about a healthy diet and physical activity. Our Pregnant+ app is meant for daily use by women, and not solely during consultation with health professionals. The main aim of the study is to assess whether the use of the Pregnancy+ app in addition to standard care results in better glucose levels at routine OGTT, 3 months postpartum, compared with standard care only. Secondary outcomes are birth weight, mode of delivery and complications for mother and child. Methods Study design We will perform a multicentre RCT. This protocol includes the elements elaborated in the SPIRIT (Standard Protocol Items: Recommendations for Interventional Trials) checklist.22 We recruit women attending the diabetic OPD at the Oslo University Hospital, Vestre Viken Hospital Trust and Akershus University Hospital. Study design and data collection are shown in the flow chart (figure 1). Figure 1 Design of the randomised controlled trial. Participants, recruitment and blinding Midwives and general practitioners refer pregnant women with recognised risk factors for GDM to a 2-hour OGTT at weeks 24–28 of gestation. On the basis of results of the OGTT, women are referred to the specialist diabetic OPD at the hospital. Participants in this study are selected among these referred women. The inclusion and the exclusion criteria for the participants are provided in table 1. Table 1 The inclusion and the exclusion criteria for the study population Inclusion criteria Exclusion criteria <33 weeks pregnant Diabetes Type 1 or 2 A 2-hour OGTT ≥9 mmol/L Twin pregnancy Above 18 years old Lactose and gluten intolerance Owns a smartphone (iPhone or Android) Understands Norwegian, Urdu or Somali We aim to recruit in total 230 women with GDM in the study. Women who decline to participate in this study are asked about their reason for declining. Answering options are: moving during pregnancy, no interest in the study, too time-consuming or other reasons. The hospital staff at the OPD hand out the information brochure about the study and subsequently invite women to participate. This information brochure is included in the online supplementary material. Women who agree to participate sign two consent forms, one copy for themselves and one for the study administration. The consent form is included in the online supplementary material. Once written consent is obtained, questionnaire 1 (Q1) is filled out on an electronic tablet at the hospital. After completion of Q1, a computer-based program randomly allocates women to either receive access to the app and standard care or the group with standard care only. Randomisation is on a 1:1 basis with balanced variable blocks of 4. Women and the staff at the OPD were not blinded to the allocation. However, the staff analysing the blood samples will not be aware of which group the women belong to and neither will the staff providing care during labour. The statistician performing the analysis for the main outcome will be blinded to the allocation of the women until the data are analysed and primary results for the intervention group and control group are available. 10.1136/bmjopen-2016-013117.supp1supplementary material 10.1136/bmjopen-2016-013117.supp2supplementary material Standard care consists of information about a healthy diet, including limitation of sugar-rich products, increase in the amount of whole grain and vegetables and frequent small meals. The midwives and the diabetic nurse give standard care during the consultations. Pregnant women with GDM are advised to be regularly physically active during pregnancy. They are taught how to perform measures with the glucometer. Regular ultrasound and CTG are part of standard care for women with GDM. The intervention with the Pregnant+ app Our intervention consists of an app for pregnant women, the Pregnant+ app. The detailed development of the app is described in a previous publication.21 The Pregnant+ app aims to give women tailored information to support their management of GDM by adapted dietary intake and physical activity and feedback on blood glucose levels. The development of the Pregnant+ app was guided by the theory of the Health Belief Model (HBM). The HBM emphasises that the awareness of risk factors and perceived health benefits are essential components in behaviour change.23 The HBM has previously been used in a pregnant population to promote health benefits during pregnancy.24 The app contains four main icons: Blood glucose, Food and beverages, Physical activity and Diabetes information.21 Women can automatically (via Bluetooth Low Energy (BLE)) or manually transfer the blood glucose values from the glucometer to the smartphone. When choosing the icon Blood glucose, the real-time blood glucose values appear in a graph and a green or red smiley indicates a good or too high blood glucose level, respectively. Pregnant women have the opportunity to print the blood glucose values at the hospital and to discuss them with the healthcare providers. The icon Food and beverages gives women general and culturally tailored information about healthy food choices. In our app, women can select a Norwegian, Somali or Pakistani food culture, also demonstrated by pictures. The Pregnant+ app has an approved link to food recipes from The Norwegian Diabetes Foundation. The icon Physical activity gives women suggestions for physical activity, the opportunity to register them and write down personal goals. Diabetes information consists of a diabetic lexicon, a frequently asked questions section (FAQ) and general information on GDM and its treatment. The participants in the intervention group can download the Pregnant+ app at the hospital or at home. Women in the standard group may also download the Pregnant+ app but will only get access to a very limited version, with a link to the official websites of The Norwegian Directorate of Health and Norwegian Diabetes Foundation. The Wireless Network and Information Security Group in Norway has contributed to the development of the Pregnant+ app ensuring users’ privacy. All the information on the smartphone is stored on the pregnant women's smartphone and not in a cloud. The feasibility study From April 2015 to July 2015, we performed a feasibility study at all five recruitment sites. The main aim of the feasibility study was to experience the speed of the recruitment, test of technical equipment like the blood sugar glucometer, the ability to answer the questionnaires on an iPad and the time used for including the women at the OPD. The feasibility study allowed us to improve some of the technical support needed for the study and implement recruitment routines at the hospitals. To increase the recruitment rate, we changed the inclusion criteria for participation from <32 to 33 weeks and decided to also include women with previous GDM. The experiences from the feasibility study allow us to inform women more correctly about the average time required to fill out the questionnaires. Outcome measurements Primary outcome The primary outcome of this study is the 2-hour blood glucose level measured at the routine OGTT (75 g) performed approx. 3 months postpartum. Secondary outcomes and data collection The secondary outcomes include a change in diet and physical activity from baseline to 36 weeks of gestation as measured by a modification of the Fit for Delivery questionnaire and the Pregnancy Physical Activity Questionnaire (PPAQ).25 26 Quality of life will be measured by a short version of the Edinburgh Postnatal Depression Scale (EDS-5) and health-related quality of life (EQ-5D-SL) during pregnancy and postpartum.27 In addition, we will measure women’s knowledge of diabetes and one of the secondary outcomes is the blood glucose levels during pregnancy. We collect the blood glucose levels during pregnancy, not via the app but through questionnaire 2 and from the medical records registered at OPD visits. The main reason for not collecting these values via the app is due to the privacy concern and that the data from the control group are not available via the app. Other secondary outcomes are maternal complications during pregnancy and at birth such as: pre-eclampsia defined as raised blood pressure of ≥140/90 and significant proteinuria, induction of labour including prostaglandin, artificial rupture of membranes and oxytocin, operative instrumental birth defined as forceps, vacuum extraction, elective and emergency caesarean section, third and/or fourth degree sphincter rupture according to the national delivery ICD-10 code and postpartum haemorrhage defined as vaginal bleeding ≥500 and ≥1000 mL.28 Newborn health is measured by Apgar score <7 at 5 min, birth weight measured in grams, transfer to intensive care within 3 hours of birth and exclusive breast feeding at discharge. To estimate the cost-effectiveness of the intervention, we will do a patient-level cost-utility analysis.29 30 Health benefit will be measured by using health-related quality of life (EQ-5D-SL) during pregnancy and postpartum. To estimate costs, we collect data regarding changes in resources used in the health sector, time used for the intervention among women, expenses for the women and sick leave. The resource use related to the health sector will be based on data from the questionnaire (number of visits to maternity ward, GP, health centre, hospitals for treatments, etc) and literature. Time used on the intervention will be calculated by estimating the time used by the pregnant women to learn how to use of the app and time spent by the health professionals to teach them. Out-of-pocket expenses for the women will mainly be estimated by the travel cost related to handing over the app and training in the use of this. To estimate sick leave, we use the data from the questionnaire. The unit costs will mainly be based on marked prices, the reimbursement systems in Norway and the literature. The results will be reported as incremental cost-effectiveness ratios and its CI, scatter plot in the cost-effectiveness plane and cost-effectiveness acceptability curves.31 32 Additionally, we will perform sensitivity analysis for changes in unit cost. Background variables such as civil status, age, income, weight prior to pregnancy (also collected through three questionnaires), education and mother tongue will be collected at baseline. The first and the second questionnaires, before 33 weeks and at 36 weeks, respectively, are filled out using an electronic tablet at the OPD. The third questionnaire is filled out at ∼3 months postpartum, via a link to a web page. From the medical records, we collect the following data: blood glucose levels, complication during pregnancy and mode of delivery. Participants' and health professionals' experiences with study participation will be evaluated in qualitative studies. Sample size calculation Sample size calculation is based on the following assumptions: power of 80%, a significance level of 5% and allowing for an attrition rate of 25%. A total of 230 participants, 115 participants per group, are required to detect a 10% difference between the intervention and control group, based on the assumption that the intervention group has a 2-hour glucose level at 3 months postpartum of 7.5 mmol/L (SD of 1.8 mmol/L). This power calculation was based on the results from several other RCT studies, investigating the effect of lifestyle changes to prevent T2DM.33–35 With a participation rate of 75%, it will take ∼20 months to recruit 230 women. Data analysis Data will be analysed according to the intention-to-treat principle. Primary outcome: glucose level as a continuous variable. This outcome will be analysed using linear regression with glucose as the dependent variable and type of intervention (intervention vs controls) as the independent variable. Further, the model will be adjusted for possible confounders as age and ethnicity. In addition, glucose levels will be dichotomised as normal versus high and this outcome will be modelled using logistic regression. Secondary outcomes included blood glucose levels during pregnancy taken from the medical records. These will be examined in the same way as the primary outcome. Other secondary outcomes are women's knowledge of diabetes. For each correct answer, the responder will be given a score. The total sum will be computed and recorded at two time points, one before and the other after the delivery. Changes over time and between groups (intervention vs controls) will be modelled using mixed models for repeated measurements. Age and ethnicity will be included as fixed factors and thus controlled for as possible confounders. Finally, the number of women with complications in both groups will be computed and the proportion will be compared using χ2 test. The scales for dietary habits and physical activity will be analysed as indicated by the authors.25 26 The main outcome is 2-hour blood glucose level measured as a continuous variable ∼3 months postpartum, using linear regression. Crude differences between groups (intervention vs control) will be assessed using χ2 test for categorical data. Continuous data will be analysed using the t-test (when normally distributed) or Mann-Whitney-Wilcoxon test (when the distribution is skewed). Additionally, glucose levels will be dichotomised and logistic regression models used. The results will be expressed as ORs. All statistical models will be adjusted for ethnicity (Norwegian, Pakistani and Somali) as a possible confounder. Descriptive statistics will be used to present the characteristics of the intervention and control group, providing frequencies (numbers) and proportion (percentages) for categorical variables. Continuous variables will be described with mean and SD. Missing data in this study will not be replaced. Analyses will be performed using SPSS statistics V.22.0 (IBM SPSS Statistics for Windows, IBM Corp., Armonk, New York, USA) and Stata V.9. Adverse events The Pregnant+ app is additional to standard care. The app was developed together with healthcare providers from each of the recruitment sites and with user involvement.21 The healthcare providers approved the content which is largely in agreement with their local recommendations as national guidelines are under development. We therefore consider the risk of potential adverse events as minimal. If a woman loses her smartphone, the blood glucose values are still stored in the glucometer. The software of the app was regularly updated to be in line with new versions of mobile operating systems. However, we assume that these software updates will not affect the usage of the app. All participants will be under close medical observation at their hospital. Events of harms can therefore be detected at an early stage and relevant care will be given. In the feasibility study, we tested the glucometer, Diamond Mini, model DM30b from FORA and compared it with hospital measurements and found a good agreement between the measured levels. Any harmful events occurring will be reported by the project leader to the leader of the Department of Nursing and Health Promotion at Oslo and Akershus University College of Applied Sciences, who is responsible for the study. Harmful events and the appropriate response to these will in addition be discussed within the research group. Data management Data from this study will be stored encrypted on an electronic server at the University Graduate Centre at Kjeller, Norway. The data from the glucometer, personal goals and notes registered in the Pregnant+ app will be on women's smartphones only. These data are not collected for the study. All the data will be treated anonymously. App usage data are collected, with specific information about the clicks per page and the duration of the visit of each page. Data from the medical records will be collected by the hospital staff and noted down on an outcome form. The written consent forms, together with the outcome forms, will initially be stored in a locked cupboard at the hospital and then be transferred by a PhD student to the Oslo and Akershus University College of Applied Sciences where they will again be stored in a locked cupboard. The personal data which identify the women will be deleted in the beginning of 2018, but anonymised information will be used for follow-up studies, depending on future funding. Dissemination In our study, we will follow the Word Medical Association Declaration of Helsinki (http://www.wma.net).36 Women meeting the inclusion criteria will be informed that neither the decision to participate nor declining to participate will influence their care. Participation is voluntary and women can withdraw whenever they like. Hospital staff have been made aware of this. The intervention, the Pregnant+ app, is a non-invasive method. The FORA Diamond Mini glucometer is used in this study and has ISO certificate approval. The glucometer has a BLE function allowing automatic transfer of blood glucose levels from the device to the smartphone. This study has been considered by the Norwegian Regional Committees for Medical and Medical Health Research Ethics, REK and by the Norwegian Social Science Data Services and the patient privacy protections boards governing over the recruitment sites. Important modifications to the protocol are updated on the ClinicalTrials.gov website and disseminated to all relevant parties. We have a contractual agreement with key collaborators determining access to trial data and authorship. Data services Sensitive information such as: participant’s name, project ID-number, phone number, country of birth from The Recruitment form and information from the medical records are stored in a locked cupboard at the hospitals. Everybody involved in recruitment at each hospital has signed a confidentiality form. Results from the study will be presented in scientific journals and at national and international congresses. Discussion There is vast evidence that GDM is associated with detrimental effects for the health of pregnant women and their offspring.8 9 In addition, the HAPO study showed an increased risk of adverse maternal and infant outcomes with increasing levels of hyperglycaemia.10 Thus, all efforts preventing hyperglycaemia during pregnancy are important. Essential elements of GDM treatment are improvement in diet and increased physical activity. Health professionals constantly search for better ways of encouraging healthy behaviour. Since verbal information may be easily forgotten and leaflets can be lost, the use of apps in health interventions appears appropriate.16 37 Compared with leaflets and verbal information, apps are more flexible and have more variable modes of communication (text, pictures, sound, interactivity) and functionality (response on blood sugar levels). The Pregnant+ app includes information about a healthy diet, physical activity and gestational diabetes, which is always near at hand on the woman's smartphone. One important component of the Pregnant+ app is the automatic transfer of the blood glucose values from the glucometer to the smartphone. The blood glucose values are visually presented in either a graph or a table. This overview and information if the blood glucose values are too high allows women to gain better control over their blood glucose levels.21 In addition, the availability of essential information on GDM, hospital routine and relevant phone numbers may decrease worries, distress and hospital visits. We developed the Pregnant+ app in Norwegian and subsequently translated and customised it for women speaking Somali or Urdu, as they are among the identified risk groups.3 12 The Pregnant+ app was developed for everyday use and may be used during consultations, for example, to refer to blood glucose levels or advice given.21 Although most of the women with GDM recover from the condition after birth, several studies have shown an increased risk of developing type 2 later in life.38 Research suggests that pregnant women with the highest blood glucose levels during pregnancy have the highest risk of developing type 2 diabetes in the future.38 Therefore, strategic interventions to reduce hyperglycaemia for women with GDM may prevent T2DM among women with GDM or recent GDM.5 A recently published article demonstrated that a smartphone with an interactive blood-glucose-management was convenient and acceptable for women with GDM.39 In another study with a smartphone application for GDM, the authors found the app to be useful in managing GDM.40 Women adopting a healthy lifestyle will positively influence their health during pregnancy and later in life. The authors would like to thank the women who participated in the feasibility study and the ongoing RCT study. The authors thank the health professionals at the outpatient departments and bachelor's and master's students from the Faculty of Health Sciences for recruitment and data collection. The authors thank the Clinic of Innovation at Oslo University for supporting this study. Contributors: ML, AM, IB, LMG-H, AML and LET planned the design of the study. ML, MCS and IB performed the power calculation and planned the analysis of the data. LMG-H, ML, JN, IKB and IB developed the Pregnant+ app intervention. ML, LMG-H, KB, LET, AM and IB developed and SF programmed the questionnaires. IB, JN, LMG-H and the hospital staff at the five OPDs conducted the feasibility study. JN and SF worked with technical support. IB, ML, LMG-H and AFJ wrote the draft and all authors have reviewed and approved the final manuscript. Funding: This study is funded by the Norwegian Research Council (http://www.forskningsradet.no; identifier: 228517). Competing interests: None declared. Ethics approval: This study has been evaluated by the Norwegian Regional Committees for Medical Health Research Ethics South East (REK) but is exempt from regional ethics review due to its nature of quality improvement in patient care (id-number: 2014/5068). The study has been approved by the Norwegian Social Science Data Services (id-number: 2014/38942) and the patient privacy protections boards governing over the recruiting sites. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: This protocol is available for the public. ==== Refs References 1 Galtier F Definition, epidemiology, risk factors . Diabetes Metab 2010 ;36 (Pt 2 ):628 –51 . 10.1016/j.diabet.2010.11.014 21163426 2 Norwegian, Institute, of, Public, Health . Diabetes in Norway-Norwegian Institute of Public Health Rapport 2014 http://www.fhi.no/eway/default.aspx?pid=239&trg=Content_7242&Main_6157=7239:0:25,8904&MainContent_7239=7242:0:25,8906&Content_7242=7244:110410::0:7243:1:::0:0#eHandbook1104107 (accessed 22 Feb 2016 ). 3 Jenum KA , Richardsen KR , Berntsen S Gestational diabetes, insulin resistance and physical activity in pregnancy in a multi-ethnic population—a public health perspective . Norsk Epidemiologi 2013 ;23 :45 –54 . 10.5324/nje.v23i1.1602 4 Skupień J , Cyganek K , Małecki MT Diabetic pregnancy: an overview of current guidelines and clinical practice . Curr Opin Obstet Gynecol 2014 ;26 :431 –7 . 10.1097/GCO.0000000000000111 25268167 5 Bellamy L , Casas JP , Hingorani AD Type 2 diabetes mellitus after gestational diabetes: a systematic review and meta-analysis . Lancet 2009 ;373 :1773 –9 . 10.1016/S0140-6736(09)60731-5 19465232 6 Lauenborg J , Hansen T , Jensen DM Increasing incidence of diabetes after gestational diabetes: a long-term follow-up in a Danish population . Diabetes Care 2004 ;27 :1194 –9 . 10.2337/diacare.27.5.1194 15111544 7 Kim C Gestational diabetes: risks, management, and treatment options . Int J Womens Health 2010 ;2 :339 –51 . 10.2147/IJWH.S13333 21151681 8 Schneider S , Hoeft B , Freerksen N Neonatal complications and risk factors among women with gestational diabetes mellitus . Acta Obstet Gynecol Scand 2011 ;90 :231 –7 . 10.1111/j.1600-0412.2010.01040.x 21306307 9 Metzger BE , Lowe LP , Dyer AR Hyperglycemia and adverse pregnancy outcomes . N Engl J Med 2008 ;358 :1991 –2002 . 10.1056/NEJMoa0707943 18463375 10 Lowe LP , Metzger BE , Dyer AR , HAPO Study Cooperative Research Group . Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study: associations of maternal A1C and glucose with pregnancy outcomes . Diabetes Care 2012 ;35 : s574 –80 . 10.2337/dc11-1687 11 Henriksen T , Thordarson H Svangerskapsdiabetes. Veileder i Fødselhjelp , 2014 http://legeforeningen.no/Fagmed/Norsk-gynekologisk-forening/Veiledere/Veileder-i-fodselshjelp-2014/Diabetes-i-svangerskapet/ (accessed 22 Feb 2016 ). 12 Helsedirektoratet . http://www.helsebiblioteket.no/retningslinjer/diabetes/12.diabetes-og-graviditet/12.3-svangerskapsdiabetes (accessed 12 Feb 2016 ). 13 Medie Norge . Andelen Som Har Smartphone . Oslo, Norway : Medienorge , 2014 (In Norwegian ). 14 Holmen H , Torbjørnsen A , Wahl AK A mobile health intervention for self-management and lifestyle change for persons with type 2 diabetes, part 2: one-year results from the Norwegian randomized controlled trial RENEWING HEALTH . JMIR Mhealth Uhealth 2014 ;2 :e57 10.2196/mhealth.3882 25499872 15 Glynn LG , Hayes PS , Casey M Effectiveness of a smartphone application to promote physical activity in primary care: the SMART MOVE randomised controlled trial . Br J Gen Pract 2014 ;64 :e384 –91 . 10.3399/bjgp14X680461 24982490 16 Payne HE , Lister C , West JH Behavioral functionality of mobile apps in health interventions: a systematic review of the literature . JMIR Mhealth Uhealth 2015 ;3 :e20 10.2196/mhealth.3335 25803705 17 Macaulay S , Dunger DB , Norris SA Gestational diabetes mellitus in Africa: a systematic review . PLoS ONE 2014 ;9 :e97871 10.1371/journal.pone.0097871 24892280 18 Kreuter MW , McClure SM The role of culture in health communication . Annu Rev Public Health 2004 ;25 :439 –55 . 10.1146/annurev.publhealth.25.101802.123000 15015929 19 Shiavo R Health communication: from theory to practice . 1st edn San Francisco, CA, USA : Jossey-Bass , 2007 . 20 Attridge M , Creamer J , Ramsden M Culturally appropriate health education for people in ethnic minority groups with type 2 diabetes mellitus . Cochrane Database Syst Rev 2014 ;(9) :CD006424 10.1002/14651858.CD006424.pub3 25188210 21 Garnweidner-Holme LM , Borgen I , Garitano I Designing and developing a mobile smartphone application for women with gestational diabetes mellitus followed-up at diabetes outpatient clinics in Norway . Healthcare (Basel) 2015 ;3 :310 –23 . 10.3390/healthcare3020310 27417764 22 Chan AW , Tetzlaff JM , Altman DG SPIRIT 2013 statement: defining standard protocol items for clinical trials . Ann Intern Med 2013 ;158 :200 –7 . 10.7326/0003-4819-158-3-201302050-00583 23295957 23 Rosenstock IM The health belief model and preventive health behavior . Health Educ Monogr 1974 ;2 :354 –86 . 10.1177/109019817400200405 24 Pearce EE , Evenson KR , Downs DS Strategies to promote physical activity during pregnancy: a systematic review of intervention evidence . Am J Lifestyle Med 2013 ;7 :38–50 10.1177/1559827612446416 25 Fit for fødsel questionnaire . http://www.biomedcentral.com/1471-2458/13/132 (accessed 22 Feb 2016 ). 26 Çırak Y , Yılmaz GD , Demir YP Pregnancy physical activity questionnaire (PPAQ): reliability and validity of Turkish version . J Phys Ther Sci 2015 ;27 :3703 –9 . 10.1589/jpts.27.3703 .26834336 27 Eberhard-Gran M , Eskild A , Samuelsen SO A short matrix version of the Edinburgh Depression Scale . Acta Psychiatr Scand 2007 ;116 :195 –200 . 10.1111/j.1600-0447.2006.00934.x 17655561 28 International Statistical Classification of Diseases and Related Health Problems . 10th Revision, ICD-10 Version , 2016 http://apps.who.int/classifications/icd10/browse/2016/en (accessed 25 Nov 2016 ). 29 Drummond M , Sculpher M , Claxton K Methods for the economic evaluation of health care programmes . 4th edn Oxford University Press , 2015 . 30 Glick HA , Doshi JA , Sonnad SS Economic evaluation in clinical trials . Oxford : Oxford University Press , 2007 . 31 Briggs A Handling uncertainty in economic evaluation and presenting the results . In: Drummond M , McGuire A , eds. Economic evaluation in health care . Oxford : Oxford University Press , 2001 :172 –214 . 32 Briggs AH , Wonderling DE , Mooney CZ Pulling cost-effectiveness analysis up by its bootstraps: a non-parametric approach to confidence interval estimation . Health Econ 1997 ;6 :327 –40 . 10.1002/(SICI)1099-1050(199707)6:4<327::AID-HEC282>3.0.CO;2-W 9285227 33 Telle-Hjellset V , Råberg Kjøllesdal MK , Bjørge B The InnvaDiab-DE-PLAN study: a randomised controlled trial with a culturally adapted education programme improved the risk profile for type 2 diabetes in Pakistani immigrant women . Br J Nutr 2013 ;109 :529 –38 . 10.1017/S000711451200133X 22717286 34 Koh D , Miller YD , Marshall AL Health-enhancing physical activity behaviour and related factors in postpartum women with recent gestational diabetes mellitus . J Sci Med Sport 2010 ;13 :42 –5 . 10.1016/j.jsams.2008.10.003 19119066 35 Oldroyd JC , Unwin NC , White M Randomised controlled trial evaluating lifestyle interventions in people with impaired glucose tolerance . Diabetes Res Clin Pract 2006 ;72 :117 –27 . 10.1016/j.diabres.2005.09.018 16297488 36 Word Medical Association Declaration of Helsinki . http://www.wma.net (accessed 28 Apr 2016 ). 37 Årsand E , Frøisland DH , Skrøvseth SO Mobile health applications to assist patients with diabetes: lessons learned and design implications . J Diabetes Sci Technol 2012 ;6 :1197 –206 . 10.1177/193229681200600525 23063047 38 Kim C , Newton KM , Knopp RH Gestational diabetes and the incidence of type 2 diabetes: a systematic review . Diabetes Care 2002 ;25 :1862 –8 . 10.2337/diacare.25.10.1862 12351492 39 Hirst JE , Mackillop L , Loerup L Acceptability and user satisfaction of a smartphone-based, interactive blood glucose management system in women with gestational diabetes mellitus . J Diabetes Sci Technol 2015 ;9 :111 –15 . 10.1177/1932296814556506 25361643 40 Jo S , Park HA Development and evaluation of a smartphone application for managing gestational diabetes mellitus . Healthc Inform Res 2016 ;22 :11 –21 . 10.4258/hir.2016.22.1.11 26893946

PMC005xxxxxx/PMC5372028.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01498610.1136/bmjopen-2016-014986EpidemiologyProtocol150616921706Predicting late-onset sepsis by routine neonatal screening for colonisation by gram-negative bacteria in neonates at intensive care units: a protocol for a systematic review Harder Thomas Seidel Juliane Eckmanns Tim Weiss Bettina Haller Sebastian Robert Koch Institute, Berlin, GermanyCorrespondence to Dr Thomas Harder; HarderT@rki.de2017 29 3 2017 7 3 e0149863 11 2016 31 1 2017 14 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction Hospitals conduct extensive screening procedures to assess colonisation of the body surface of neonates by gram-negative bacteria to avoid complications like late-onset sepsis. However, the benefits of these procedures are controversially discussed. Until now, no systematic review has investigated the value of routine screening for colonisation by gram-negative bacteria in neonates for late-onset sepsis prediction. Methods and analysis We will conduct a systematic review, considering studies of any design that include infants up to an age of 12 months. We will search MEDLINE and EMBASE (inception to 2016), reference lists and grey literature. Screening of titles, abstracts and full texts will be conducted by two independent reviewers. We will extract data on study characteristics and study results. Risk of bias will be assessed using Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) and Quality in Prognosis Studies (QUIPS) tools. Subgroup analyses are planned according to characteristics of studies, participants, index tests and outcome. For quantitative data synthesis on prognostic accuracy, sensitivity and specificity of screening to detect late-onset sepsis will be calculated. If sufficient data are available, we will calculate summary estimates using hierarchical summary receiver operating characteristics and bivariate models. Applying a risk factor approach, pooled summary estimates will be calculated as relative risk or OR, using fixed-effects and random-effects models. I-squared will be used to assess heterogeneity. All calculations will be performed in Stata V14.1 (College Station, Texas, USA). The results will be used to calculate positive and negative predictive value and number needed to be screened to prevent one case of sepsis. Grading of Recommendations Assessment, Development and Evaluation (GRADE) will be used to assess certainty in the evidence. The protocol follows the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) guideline. Ethics and dissemination This study will not require ethical approval since it is not carried out in humans. The systematic review will be published in an open-access peer-reviewed journal. Trial registration number CRD42016036664. Systematic reviewProtocolPrognostic accuracyneonatal sepsisscreeningEuropean Centre for Disease Prevention and Controlhttp://dx.doi.org/10.13039/5011000008052012/040; 2014/008 ==== Body Strengths and limitations of this study This systematic review will provide a comprehensive overview on the available evidence regarding the value of routine screening for colonisation by gram-negative bacteria in neonates for late-onset sepsis prediction. Subgroup analysis will allow investigating the particular role of setting, birth characteristics, sampling strategy and cointerventions for test performance and predictive values Limitations of the systematic review will arise from the limitations of the included studies, particularly regarding consideration and reporting of confounders in the publications. Introduction Epidemiological and clinical background At neonatal intensive care units (NICUs), late-onset sepsis due to gram-negative pathogens is an important cause of neonatal morbidity and mortality.1 The majority of sepsis episodes (>80%) occurs in preterm neonates.2 Depending on individual factors, setting and species of bacteria, between 11% and 46% of very low birth weight infants (VLBW; <1500 g) are affected.3 Susceptibility to infection is strongly associated with low gestational age and low birth weight.4 5 Already in the 1970s, data were published indicating that infants at NICUs colonised with gram-negative bacteria were at increased risk of developing infections subsequently.6 Consecutively, a number of studies investigated the value of routine surface cultures for the prediction of sepsis.7–10 Some hospitals conduct extensive and costly screening procedures to assess the colonisation of non-sterile locations of the body surface of neonates by gram-negative bacteria to avoid complications like sepsis. In Germany, routine screening for a selection of pathogens is recommended by the German Committee on Hospital Infections and Hygiene (KRINKO). However, the benefits of these screening procedures are controversially discussed. Moreover, since microbiological screening is introduced as part of a bundle of measures (eg, isolation, enhanced barrier nursing), it is often challenging to measure the particular effect of screening. Until now, no systematic review has been published that has investigated the prognostic value of routine screening for colonisation by gram-negative bacteria in this at-risk group for the prediction of late-onset sepsis. Here, we present and explain the protocol for a respective systematic review that will be conducted as part of the piloting phase of the Project on a Framework for Rating Evidence in Public Health (PRECEPT).11 Prognostic/diagnostic test accuracy and risk factors According to the Cochrane handbook for systematic reviews of diagnostic test accuracy studies, prognostic accuracy studies are using test information to identify patients who will develop an outcome later on (see http://methods.cochrane.org/sdt/handbook-dta-reviews). In this sense, studies that are using screening for gram-negative bacteria to predict sepsis are prognostic accuracy studies. In such studies, the result of a test is compared with the (clinical) outcome. This differs from the approach of diagnostic test accuracy studies where the test result is compared with the result of a reference or ‘gold standard’ test (figure 1). Therefore, prognostic accuracy is not a surrogate for patient-important outcomes, as in diagnostic test accuracy studies.12 This approach has consequences for the design of the studies to be considered. In contrast to diagnostic test accuracy studies where cross-sectional study designs are common practice, cohort studies (prospective or retrospective) are needed to obtain measures of prognostic accuracy. Figure 1 Diagnostic versus prognostic test accuracy. A complementary approach to the analysis of the same data is to conceptualise a positive screening test as the presence of a risk (or prognostic) factor and to calculate the relative risk of developing the outcome. However, it is important to consider that the presence of a high risk ratio (or OR), which is often used to identify prognostic factors for a certain outcome, does not indicate that the respective risk factor performs well in predicting this outcome.13–15 Ware showed that a risk factor strongly associated with a hypothetical outcome (OR 3.58) might have a sensitivity as low as 13% for predicting this outcome. Using the same data, he demonstrated that an OR of 228 would be needed to reach a sensitivity of 80%.15 Therefore, it may not be concluded that a risk factor which is strongly associated with the outcome provides a basis for an effective preventive measure. Concepts for systematic reviews of prognostic studies Various approaches exist regarding the systematic assessment and data synthesis of prognostic studies.16 During recent years, it has become more and more accepted that systematic reviews in this field should focus on measures of association between the predictive/prognostic factor and the outcome, such as risk ratio, OR or HR, as well as comprise measures of prognostic accuracy like sensitivity and specificity (eg, see17). Liu et al18 proposed to distinguish between systematic reviews of screening tests and those of diagnostic and prognostic studies. For screening and diagnosis, they suggested assessing sensitivity and specificity, whereas for questions related to prognosis the use of HRs was proposed. The Agency for Healthcare Research and Quality (AHRQ) suggests using a particular framework for systematic reviews of a prognostic test.19 In that paper, Rector et al conclude that it may be informative to assess the accuracy of a prognostic test by calculating sensitivity, specificity and predictive values. However, these authors emphasise that it is critical to consider the time interval between the test and the occurrence of the outcome.19 In our own systematic review, we will compute measures of prognostic accuracy and measures of relative risk and compare the results of these calculations to each other. Risk of bias Given the particularities of systematic reviews of prognostic accuracy studies, the question arises whether an established risk-of-bias tool exists that captures common sources of bias in this study design. A number of authors applied tools that were originally designed to address risk of bias in diagnostic test accuracy studies.17 20 21 Currently, the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool is the most advanced and widely used tool for the assessment of risk of bias in diagnostic accuracy studies.22 QUADAS-2 comprises four domains: patient selection, index test, references standard and flow and timing. In each domain, questions related to risk of bias and concerns regarding applicability are included. However, as explained above, there are apparent differences in study design between diagnostic and prognostic accuracy studies. At least two sources of bias can be identified which are important in prognostic accuracy studies but are not relevant in diagnostic accuracy studies: Attrition bias: Owing to the prospective character of the study design, loss to follow-up of study participants in the time interval between the conduct of the screening test and the detection of the outcome might create attrition bias. Depending on whether or not rates of loss to follow-up differ between participants with positive and negative screening test results (differential vs non-differential loss to follow-up), sensitivity and specificity will change in point estimate or CI. Confounding: Confounding will occur if interventions are delivered to study participants depending on the result of the screening test. This may influence the probability of developing the outcome. Again, estimates of sensitivity and specificity might be affected. Theoretically, it is possible that domain four of the QUADAS-2 tool (‘flow and timing’) sufficiently captures attrition bias as well as confounding in the time interval between screening test and outcome assessment. If this appears not to be the case, we may test whether the additional application of a risk of bias tool for risk factor/prognostic studies such as the Quality in Prognosis Studies (QUIPS) tool23 is of additional value. General objective To assess the usefulness and value of routine screening for colonisation by gram-negative bacteria performed in NICUs as predictive measures for late-onset sepsis. Research question This systematic review will focus on the following primary research questions: What is the prognostic value (in terms of sensitivity and specificity) of routine screening for colonisation by gram-negative bacteria in neonates at intensive care units for the prediction of late-onset sepsis? Is colonisation by gram-negative bacteria in neonates at intensive care units a risk factor for later development of late-onset sepsis? Methods This systematic review protocol follows the recommendations of the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) guideline.24 A copy of the completed PRISMA-P checklist is attached to this protocol (see online supplementary appendix 1). This systematic review is registered in the Prospective Register of Systematic Reviews (Reg. No. CRD42016036664). 10.1136/bmjopen-2016-014986.supp1supplementary appendix Eligibility criteria Study designs Studies of any design will be considered. No restrictions will be made regarding publication language or publication status. Participants Studies that include infants up to an age of 12 months who are still in a NICU will be considered, irrespective of the gestational age, birth weight and geographical region where the study has been conducted. Study setting Studies that were performed in NICUs will be considered. Search strategy Database search We will search MEDLINE and EMBASE from inception to 2016, using the DIMDI (Deutsches Institut für Medizinische Dokumentation und Information) platform. The planned search strategy is shown in box 1. Box 1 Search strategy of the systematic review #1 neonat* #2 newborn* #3 infant* #4 colonisation #5 ‘mucosal site*’ #6 ‘mucosal sample*’ #7 ‘mucosal culture*’ #8 ‘superficial culture*’ #9 ‘surveillance culture*’ #10 aspirate #11 ‘predictive value’ #12 sensitivity #13 specificity #14 sepsis #15 ‘body fluid*’ #16 ‘systemic inflammatory response syndrome’ #17 ‘routine culture’ #18 ‘skin culture’ #19 ‘surface culture’ #20 ‘bacterial colonisation’ #21 ‘microbiological screening’ #22 swab* #23 #1 OR #2 OR #3 #24 #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 #25#14 OR #15 OR #16 #26 #23 AND #24 AND #25 Reference lists These searches will be supplemented by ‘snowballing’, that is, searching for additional studies in the reference lists of identified original studies and reviews. Grey literature We will search for grey literature using the Grey Matters Light checklist of the Canadian Agency for Drugs and Technologies in Health (CADTH) (http://www.cadth.ca/resources/finding-evidence/grey-matters-practical-search-tool-evidence-base-medicine). Study selection The study selection process will involve the following steps: Screening of titles and abstracts. Screening of full texts. At both steps, screening will be conducted by two independent reviewers. Potential disagreement will be resolved by discussion or by involving a third reviewer. We will construct a flow chart to document the selection process. A list of excluded studies will be prepared, along with reasons for exclusion. Data extraction and management From the included studies, we will extract data on study characteristics and study results. We will construct a data extraction form and pilot test it prior to the start of the review process. Microsoft Office Excel will be used to construct specific extraction forms. One researcher will perform data extraction while a second researcher will independently check for accuracy and details. The following data will be extracted from the original studies: General study characteristics: complete reference of the study (author, year of publication, title, journal, citation details) date of study place setting (hospital, department, unit, ward) study design funding source Patient/population characteristics: inclusion criteria exclusion criteria gestational age at birth birth weight age at screening sex ethnicity length of follow-up (time interval between index test and outcome assessment) comorbidities central line use need for surgery Index test characteristics: description of sampling device sampling time point(s) sampling location(s) (eg, umbilicus, tracheal, rectal, etc) sampling intervals (if repetitive) processing of specimen detected bacteria (species, characterisation) Outcome: definition of sepsis detected bacteria (species, characterisation) (Co)-interventions antibiotic use isolation hand hygiene Prognostic accuracy measures: true positives true negatives false positives false negatives Measures of association (risk factor approach): unadjusted relative risk (or OR) adjusted relative risk (or OR) confounders considered in adjusted analysis. Risk of bias assessment Following the guidance of the PRECEPT framework,11 25 we will use the QUADAS-2 tool to assess risk of bias in the included individual studies which report measures of prognostic accuracy.22 Table 1 shows the main components of the tool. The results of the risk of bias assessment will be documented in a separate table for each study alongside the items of QUADAS-2. For studies reporting on prognostic measures in terms of a risk factor (or prognostic study), we will use the QUIPS tool.23 We will construct bar charts as suggested by Van't Hooft et al21 to report summary results of the risk of bias assessments. Table 1 Structure of the QUADAS-2 tool22 Domain Risk of bias Concerns regarding applicability 1. Patient selection Was a consecutive or random sample of patients enrolled? Yes/no/unclear Is there concern that the included patients do not match the review question? Concern: low/high/unclear Was a case–control design avoided? Yes/no/unclear Did the study avoid inappropriate exclusions? Yes/no/unclear Could the selection of patients have introduced bias? Risk: low/high/unclear 2. Index test(s) Were the index test results interpreted without knowledge of the results of the reference standard? Yes/no/unclear Is there concern that the index test, its conduct, or interpretation differs from the review question? Concern: low/high/unclear If a threshold was used, was it prespecified? Yes/no/unclear Could the conduct or interpretation of the index test results have introduced bias? Risk: low/high/unclear 3. Reference standard* Is the reference standard likely to correctly classify the target condition? Yes/no/unclear Is there concern that the target condition as defined by the reference standard does not match the review question? Concern: low/high/unclear Were the reference standard results interpreted without knowledge of the results of the index test? Yes/no/unclear Could the reference standard, its conduct, or its interpretation have introduced bias? Risk: low/high/unclear 4. Flow and timing Was there an appropriate interval between the index test and reference standard? Yes/no/unclear     Did all patients receive a reference standard? Yes/no/unclear Did patients receive the same reference standard? Yes/no/unclear Were all patients included in the analysis? Yes/no/unclear Could the patient flow have introduced bias? Risk: low/high/unclear *Here equivalent to outcome. Subgroup analyses We will extract detailed information on study participants, definitions and settings to enable stratified analysis. In particular, we aim at stratifying the results of the systematic review and meta-analysis, respectively, according to the following variables: General study characteristics: geographic region (Europe vs North America, etc) developed country versus developing country study period (<1970, 1971–1980, 1981–1990, 1991–2000, 2001–2010, >2010) Patient/population characteristics: gestational age (<37 weeks vs ≥37 weeks; <32 weeks vs ≥32 weeks; <26 weeks vs >26 weeks) birth weight (<1000 g vs ≥1000 g; <1500 g vs ≥1500 g; <2500 g vs ≥ 2500 g) length of follow-up (time interval between index test and outcome assessment) Index test characteristics: sampling time point(s) sampling location(s) (umbilicus vs tracheal, etc) species: single species; groups (multidrug-resistant; difficult to treat) Outcome characteristics: different definitions of sepsis Study setting clinical routine study outbreak investigation type of ward Study design. Statistical analysis Prognostic accuracy approach: For quantitative data synthesis on prognostic accuracy, we will construct 2×2 tables to calculate sensitivity and specificity for each included study. If sufficient comparable data from more than one study are available, we will perform meta-analysis. To account for the correlation between sensitivity and specificity, we will calculate summary estimates using hierarchical summary receiver operating characteristics models26 as well as bivariate models.27 Results will be displayed graphically using summary receiver operating characteristic (SROC) plots. We will investigate sources of heterogeneity, using subgroup analysis. Risk factor approach: For quantitative data synthesis using the risk factor approach, pooled summary estimates will be calculated as relative risk or OR with 95% CIs, using fixed-effects and random-effects models. I-squared will be used to assess heterogeneity. If ≥10 studies per outcome are available, publication bias will be assessed by inspection of funnel plots and applying Begg's and Egger's test. All calculations will be performed in STATA. The results of the meta-analysis will be used to calculate positive predictive value, negative predictive value and number needed to be screened to prevent one case of sepsis. Certainty in the evidence (GRADE) We will use two complementary approaches to assess the certainty in the evidence (formerly: quality of the evidence) according to the methodology suggested by the Grading of Recommendations Assessment, Development and Evaluation (GRADE) working group. Prognostic accuracy approach: We will adopt the GRADE approach to diagnostic accuracy test reviews for the purpose of our systematic review on prognostic test accuracy. The certainty in the evidence will be assessed for true positives (TP), true negatives (TN), false positives (FP) and false negatives (FN), as suggested by GRADE.12 In brief, the application of GRADE will be conducted as follows: For each body of evidence on diagnostic studies, all studies start as ‘high’. ‘True positives’, ‘true negatives’, ‘false positives’ and ‘false negatives’ are defined as outcomes. Risk of bias is assessed by the QUADAS-2 tool, and evidence quality can be downgraded, if necessary. Thereafter, the other GRADE criteria for downgrading quality of evidence (inconsistency, indirectness, imprecision, publication bias) are applied, according to the approach published by the GRADE working group.12 Risk factor approach: We will use the GRADE approach to risk factor/prognostic factor studies. The certainty in the evidence will be assessed for the outcome late-onset sepsis according to the GRADE methodology28 as follows: For each body of evidence, certainty in the evidence is initially rated as ‘high’, irrespective of study design. Risk of bias is assessed by the appropriate risk of bias tool, and evidence certainty can be downgraded, if necessary. Thereafter, the other GRADE criteria for downgrading quality of evidence (inconsistency, indirectness, imprecision, publication bias) are applied. Upgrading of the quality of evidence is possible, according to the criteria introduced by GRADE. Reporting of this review The systematic review will be reported according to the PRISMA guidelines. The PRISMA checklist will be published with the report. Dissemination of findings The resulting systematic review will be published in a peer-reviewed journal as an open access paper. This systematic review will be performed as part of the piloting phase of the PRECEPT project. PRECEPT is funded by the European Centre for Disease Prevention and Control (ECDC; tenders no. 2012/040; 2014/008). Contributors: TH, JS and SH developed the concept of this protocol. TH wrote the first draft. JS, BW, TE and SH provided important intellectual input to revise the draft protocol. All authors approved the final manuscript as submitted. TH is the guarantor of this protocol. Funding: European Centre for Disease Prevention and Control; grant number (2012/040; 2014/008). Disclaimer: The funder had no role in the development and writing of this protocol. Ethics approval: This study will not require ethical approval since it is not carried out in humans. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. ==== Refs References 1 Smith A , Saiman L , Zhou J Concordance of gastrointestinal tract colonization and subsequent bloodstream infections with gram-negative Bacilli in very low birth weight infants in the neonatal intensive care unit . Pediatr Infect Dis J 2010 ;29 :831 –5 . 10.1097/INF.0b013e3181e7884f 20539251 2 Vergnano S , Menson E , Kennea N Neonatal infections in England: the NeonIN surveillance network . Arch Dis Child Fetal Neonatal Ed 2011 ;96 :F9 –F14 . 10.1136/adc.2009.178798 20876594 3 Alshaikh B , Yusuf K , Sauve R Neurodevelopmental outcomes of very low birth weight infants with neonatal sepsis: systematic review and meta-analysis . J Perinatol 2013 ;33 :558 –64 . 10.1038/jp.2012.167 23328927 4 Cordero L , Rau R , Taylor D Enteric gram-negative bacilli bloodstream infections: 17 years’ experience in a neonatal intensive care unit . Am J Infect Control 2004 ;32 :189 –95 . 10.1016/j.ajic.2003.07.004 15175611 5 Stoll BJ , Hansen N , Fanaroff AA Late-onset sepsis in very low birth weight neonates: the experience of the NICHD Neonatal Research Network . Pediatrics 2002 ;110 :285 –91 . 10.1542/peds.110.2.285 12165580 6 Sprunt K , Leidy G , Redman W Abnormal colonization of neonates in an intensive care unit: means of identifying neonates at risk of infection . Pediatr Res 1978 ;12 :998 –1002 . 10.1203/00006450-197810000-00010 724305 7 Dobson SR , Isaacs D , Wilkinson AR Reduced use of surface cultures for suspected neonatal sepsis and surveillance . Arch Dis Child 1992 ;67 :44 –7 . 10.1136/adc.67.1_Spec_No.44 1536585 8 Fox GP , Clarke TA , Matthews TG Are routine superficial cultures worth while in neonatal practice? Br Med J (Clin Res Ed) 1988 ;296 :1443 10.1136/bmj.296.6634.1443 9 Lau YL , Hey E Sensitivity and specificity of daily tracheal aspirate cultures in predicting organisms causing bacteremia in ventilated neonates . Pediatr Infect Dis J 1991 ;10 :290 –4 . 10.1097/00006454-199104000-00005 2062624 10 Slagle TA , Bifano EM , Wolf JW Routine endotracheal cultures for the prediction of sepsis in ventilated babies . Arch Dis Child 1989 ;64 :34 –8 . 10.1136/adc.64.1_Spec_No.34 2923483 11 Harder T , Abu Sin M , Bosch-Capblanch X Towards a framework for evaluating and grading evidence in public health . Health Policy 2015 ;119 :732 –6 . 10.1016/j.healthpol.2015.02.010 25863647 12 Schünemann HJ , Oxman AD , Brozek J Grading quality of evidence and strength of recommendations for diagnostic tests and strategies . BMJ 2008 ;336 :1106 –10 . 10.1136/bmj.39500.677199.AE 18483053 13 Wald NJ , Hackshaw AK , Frost CD When can a risk factor be used as a worthwhile screening test? . BMJ 1999 ;319 :1562 –5 . 10.1136/bmj.319.7224.1562 10591726 14 Pepe MS , Janes H , Longton G Limitations of the odds ratio in gauging the performance of a diagnostic, prognostic, or screening marker . Am J Epidemiol 2004 ;159 :882 –90 . 10.1093/aje/kwh101 15105181 15 Ware JH The limitations of risk factors as prognostic tools . N Engl J Med 2006 ;355 :2615 –17 . 10.1056/NEJMp068249 17182986 16 Dretzke J , Ensor J , Bayliss S Methodological issues and recommendations for systematic reviews of prognostic studies: an example from cardiovascular disease . Syst Rev 2014 ;3 :140 10.1186/2046-4053-3-140 25466903 17 Young YR , Sheu BF , Li WC Predictive value of plasma brain natriuretic peptide for postoperative cardiac complications--a systemic review and meta-analysis . J Crit Care 2014 ;29 :696 e1 –10 . 10.1016/j.jcrc.2014.03.022 18 Liu Z , Yao Z , Li C A step-by-step guide to the systematic review and meta-analysis of diagnostic and prognostic test accuracy evaluations . Br J Cancer 2013 ;108 :2299 –303 . 10.1038/bjc.2013.185 23695015 19 Rector TS , Taylor BC , Wilt TJ Chapter 12: systematic review of prognostic tests . J Gen Intern Med 2012 ;27 (Suppl 1 ):S94 –101 . 10.1007/s11606-011-1899-y 22648680 20 Lurati Buse GA , Koller MT , Burkhart C The predictive value of preoperative natriuretic peptide concentrations in adults undergoing surgery: a systematic review and meta-analysis . Anesth Analg 2011 ;112 :1019 –33 . 10.1213/ANE.0b013e31820f286f 21372274 21 Van't Hooft J , van der Lee JH , Opmeer BC Predicting developmental outcomes in premature infants by term equivalent MRI: systematic review and meta-analysis . Syst Rev 2015 ;4 :71 10.1186/s13643-015-0058-7 25982565 22 Whiting PF , Rutjes AW , Westwood ME QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies . Ann Intern Med 2011 ;155 :529 –36 . 10.7326/0003-4819-155-8-201110180-00009 22007046 23 Hayden JA , van der Windt DA , Cartwright JL Assessing bias in studies of prognostic factors . Ann Intern Med 2013 ;158 :280 –6 . 10.7326/0003-4819-158-4-201302190-00009 23420236 24 Moher D , Shamseer L , Clarke M Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement . Syst Rev 2015 ;4 :1 10.1186/2046-4053-4-1 25554246 25 Harder T , Takla A , Rehfuess E Evidence-based decision-making in infectious diseases epidemiology, prevention and control: matching research questions to study designs and quality appraisal tools . BMC Med Res Methodol 2014 ;14 :69 10.1186/1471-2288-14-69 24886571 26 Rutter CM , Gatsonis CA A hierarchical regression approach to meta-analysis of diagnostic test accuracy evaluations . Stat Med 2001 ;20 :2865 –84 . 10.1002/sim.942 11568945 27 Reitsma JB , Glas AS , Rutjes AW Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews . J Clin Epidemiol 2005 ;58 :982 –90 . 10.1016/j.jclinepi.2005.02.022 16168343 28 Iorio A , Spencer FA , Falavigna M Use of GRADE for assessment of evidence about prognosis: rating confidence in estimates of event rates in broad categories of patients . BMJ 2015 ;350 :h870 10.1136/bmj.h870 25775931

PMC005xxxxxx/PMC5372029.txt

==== Front BMJ Open Diabetes Res CareBMJ Open Diabetes Res CarebmjdrcbmjdrcBMJ Open Diabetes Research & Care2052-4897BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjdrc-2016-00032410.1136/bmjdrc-2016-000324Epidemiology/Health Services Research15061867Association between secondhand smoke and obesity and glucose abnormalities: data from the National Health and Nutrition Examination Survey (NHANES 1999–2010) Kermah Dulcie 1Shaheen Magda 1Pan Deyu 1http://orcid.org/0000-0001-9555-7626Friedman Theodore C 2 1 Office of Research, Charles R. Drew University of Medicine and Science, Los Angeles, California, USA 2 Division of Endocrinology, Metabolism, and Molecular Medicine, Department of Internal Medicine, Charles R. Drew University of Medicine and Science, Los Angeles, California, USACorrespondence to Dr Theodore C Friedman; theodorefriedman@cdrewu.edu2017 21 3 2017 5 1 e00032413 9 2016 22 12 2016 5 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective The objective of this study is to investigate the relationship between cotinine level-confirmed secondhand smoke (SHS) exposure and glycemic parameters and obesity. Research design and methods We examined a cohort of 6472 adults from the National Health and Nutrition Examination Surveys, 1999–2010. Serum cotinine levels and self-reported data on smoking were used to determine smoking status. The outcome variables were body mass index (BMI) and glycemic status (HbA1c), Homeostasis Model Assessment of Insulin Resistance (HOMA-IR), and fasting plasma glucose (FPG). Descriptive, bivariate, and multivariate analyses were conducted. Results Using cotinine level-confirmed smoking status, 1794 (27.4%) of the sample were current smokers, 1681 (25.0%) were former smokers, 1158 (17.8%) were secondhand smokers, and 1839 (29.8%) were non-smokers. In a generalized linear model after controlling for potential confounding variables, secondhand smokers had higher adjusted levels of HOMA-IR, FPG, and BMI compared with non-smokers (p<0.05). Adjustment for BMI demonstrated that some, but not all, of the detrimental effects of SHS on glycemic parameters are mediated by the increased body weight of secondhand smokers. Conclusions We conclude that SHS is associated with obesity and worsening glycemic parameters. More studies are needed to show a causal relationship between SHS and glycemic parameters and to understand the mechanisms involved in the association. NicotineSmokingEpidemiologyObesity ==== Body Significance of this study What is already known about this subject? Studies have shown a positive association between cigarette smoking and the incidence of diabetes mellitus (DM). A recent meta-analysis examining prospective cohort studies on passive smoking (by history) showed a significant relative risk of developing DM. What are the new findings? Cotinine level-confirmed secondhand smoke (SHS) is associated with a higher rate of Homeostasis Model Assessment of Insulin Resistance (HOMA-IR), hemoglobin A1c, fasting plasma glucose as well as obesity after correcting for confounders. Adjustment for body mass index demonstrated that some, but not all of the detrimental effects of SHS on glycemia are mediated by the increased body weight of secondhand smokers. The percentage of participants who were classified as secondhand smokers using our definition decreased between 1999–2000 and 2009–2010, which demonstrates that smoke-free laws are effective and may help decrease the number of Americans who develop obesity and DM as a result of being exposed to SHS. How might these results change the focus of research or clinical practice? Our study showing the association between cotinine-verified secondhand smokers and worsening of glycemic parameters as well as obesity prompts the need for reliable and cost-effective methods for interventions to prevent secondhand smoking including stronger smoke-free laws. Future research studies are needed to show a causal relationship between SHS and glycemic parameters and to understand the mechanisms involved in the association. Introduction In 2011, 43.8 million people or 19.0% of all adults (aged 18 years and older) in the USA smoked cigarettes.1 During 2000–2004, an estimated 443 000 persons in the USA died prematurely each year as a result of the health consequences of smoking or exposure to secondhand smoke (SHS),2 making it the single most preventable cause of death. SHS is classified by the US Environmental Protection Agency (EPA) as a ‘known human carcinogen’.3 Cigarette smoking and type 2 diabetes mellitus (DM) are major public health concerns and both are risk factors for cardiovascular disease, with an increased mortality rate among female smokers who have DM.4 Studies have shown the association between cigarette smoking and an increased risk of developing type 2 DM,5 an association that is surprising as smokers are leaner than non-smokers and obesity is positively associated with DM.6 SHS is responsible for about 46 000 deaths per year from heart disease in current non-smokers.7 According to the US Department of Health Services, exposure to SHS causes lung cancer in non-smoking adults.8 Exposure to SHS can be tested by measuring the amount of cotinine (a breakdown product of nicotine) in a non-smoker's blood, saliva, or urine.9 Studies have shown the association of cotinine with an increased level of hemoglobin A1c (HbA1c) (a measurement of glycemic exposure) in subjects without DM.10 Multiple studies have shown a positive association between cigarette smoking and the incidence of DM,5 and a recent meta-analysis examining prospective cohort studies on passive smoking (by history) showed a significant relative risk of developing DM;11 the relationship between cotinine-verified SHS and DM is less established. As the development of DM is clearly related to obesity,6 we also wanted to examine the relationship between SHS and obesity and if obesity influences the relationship between SHS and glycemic parameters. In our study, we tested the association between SHS, confirmed by serum cotinine levels, and glycemic parameters and obesity. Research design and methods Study design We used data from the National Health and Nutrition Examination Survey (NHANES), a cross-sectional survey conducted by the National Center for Health Statistics, using a stratified multistage probability sample to obtain a representative sample of the total civilian, non-institutionalized US population.12 Since 1999, the NHANES has released data at 2-year intervals. NHANES collected questionnaire data during a face-to-face home interview followed by a physical examination and additional interviews administered in a specially equipped mobile examination center where blood samples were drawn, from which serum cotinine and glycemic levels were determined. We combined data using six successive waves of NHANES data (1999–2010) for our analyses. Our analytical sample was 6472 adults 20 years and older (figure 1) who had complete data on recent self-reported cigarette smoking status and had serum cotinine levels measured. Participants with missing data on any of the study variables were excluded from the analysis. We excluded 627 subjects who were taking insulin or other antiglycemic medications, as these medications would not allow us to determine the true effect of primary or SHS on glycemic outcomes. Participants who used other nicotine products such as pipes, snuff, patch, gum, cigars, or chewed tobacco, either alone or with cigarettes were also excluded from the analysis, as we were concerned about confounding effects of these products (figure 1). Figure 1 Flow chart depicting six successive waves of NHANES cycle (1999–2010). NHANES, National Health and Nutrition Examination Survey. Definition of variables and laboratory measurements Obesity Body weight and height were measured according to standard techniques during the examinations and body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters.13 Participants with BMI 25–29.9 kg/m2 were classified as overweight and those with BMI≥30 kg/m2 were classified as obese.14 Glycemic status Glycemic status was assessed by measures based on Homeostasis Model Assessment of Insulin Resistance (HOMA-IR),15 fasting plasma glucose (FPG, mg/dL), and HbA1c (%). FPG was measured in participants who were examined in the morning session after an 8–24-hour fast using the hexokinase enzymatic reference method (Roche Diagnostics, Indianapolis, IN). Serum insulin was measured by means of a radioimmunoassay (Pharmacia Diagnostics, Uppsala, Sweden). All measurements were performed at the University of Missouri—Columbia School of Medicine Department of Child Health, Diabetes Reference Laboratory, Columbia, Missouri (David Goldstein, MD, Director). HOMA-IR was determined according to the following equation: fasting plasma glucose (mmol/L) × fasting plasma insulin (µU/mL)/22.5).15 HbA1c was measured on all participants using the boronate affinity high performance liquid chromatography (HPLC) system. We categorized normal values for HOMA-IR as <2.2, for HbA1c as <6.5%, and for FPG as <126 mg/dL.16 17 Cotinine level-confirmed smoking status Cotinine is a metabolite of nicotine that is used as a marker for active and passive smoking. Serum cotinine was measured by an isotope dilution-high performance liquid chromatography/atmospheric pressure chemical ionization tandem mass spectrometry (ID HPLC-APCI MS/MS). The racial/ethnic cut-offs for cotinine categorization by Benowitz et al18 were used to categorize individuals as non-smokers (NS) (cotinine <0.05 ng/mL, all races/ethnicities), secondhand smokers (SHSers) (cotinine 0.05–5.91 ng/mL for non-Hispanic blacks; 0.05–4.84 ng/mL for non-Hispanic whites, and 0.05–0.83 ng/mL for Hispanics), and current smokers (CS) (cotinine ≥5.92 ng/mL, 4.85 ng/mL, and 0.84 ng/mL for non-Hispanic blacks, non-Hispanic whites, and Hispanics, respectively). As Benowitz et al18 did not define a cut-off for Hispanics, we used the Mexican-American cut-off for Hispanics. We categorized smoking status according to two methods, that is, one based on the objective cotinine levels as defined above and a second method based on a combination of cotinine status and response to specific questions. In the second method, CS were defined as those who smoked at least a 100 cigarettes in lifetime AND responded ‘Everyday’ or ‘Some days’ to ‘Do you now smoke’. Former smokers (FS) were defined as those who smoked at least a 100 cigarettes in lifetime AND responded ‘Not at all’ to Do you now smoke” AND also responded ‘No’ to smoking cigarettes in last 5 days. SHSers were defined as (1) those who responded ‘no’ to ‘Smoking at least 100 cigarettes in lifetime’ AND ‘No’ to smoking cigarettes in last 5 days AND had cotinine <0.05 ng/mL AND answered ‘Yes’ to the question ‘Were you exposed to cigarettes at home/work?’ OR (2) those who responded ‘No’ to ‘Smoking at least 100 cigarettes in lifetime’ AND ‘No’ to smoking cigarettes in last 5 days AND had cotinine ≥0.05 ng/mL. NS were defined as those who responded ‘No’ to ‘Smoking at least 100 cigarettes in lifetime’ AND ‘No’ to smoking cigarettes in last 5 days AND had cotinine <0.05 ng/mL. Confounding variables Age at interview in years was used as a continuous variable. Gender was categorized as males and females. Self-reported race/ethnicity was categorized as non-Hispanic white, non-Hispanic black, Hispanic, and other race. Physical activity was categorized as being active if respondent answered ‘Yes’ to either the question ‘Over the past 30 days, did you do moderate activities for at least 10 minutes that cause only light sweating or a slight to moderate increase in breathing or heart rate? Some examples are brisk walking, bicycling for pleasure, golf, and dancing’ OR the question ‘Over the past 30 days, did you do any vigorous activities for at least 10 minutes that caused heavy sweating, or large increases in breathing or heart rate? Some examples are running, lap swimming, aerobics classes or fast bicycling’. Participants were considered non-active if they answered ‘No’ to both questions mentioned above. Alcohol consumption was surveyed and categorized as ≤1, 2–3, 4–5, and ≥6 drinks per week.10 Education level was reported as number of years of school attended and was categorized into less than high school (less than 9th grade, 9–11 grade, 12th grade with no diploma), high school (high school graduate/general education degree or equivalent) and greater than high school (some college or associate arts degree, college graduate or above). Poverty income ratio (PIR) is an income-to-needs variable measuring the ratio of household income to the US poverty threshold for each respondent's family size and composition. PIR was used to determine whether the participant's household income was < or ≥200% of the federal poverty level (FPL).19 The year of the survey cycle used in the analyses was every 2 years from 1999 to 2010. There were six cycles. Statistical analyses Descriptive statistics was used to characterize the subjects (mean±SE) for continuous variables, and unweighted number and weighted percent for categorical variables in table 1. We used the log base (e) of BMI, HOMA-IR, FPG, and HbA1c for the analyses as the original variables were not normally distributed. To test the statistical differences between the smoking status-related groups in the categorical variables of the sample characteristics, χ2 test for categorical variables was used. For the normally distributed log-transformed continuous variables, we used generalized linear model (GLM) and present the unadjusted values in table 1. p<0.05 was considered statistically significant. Table 1 NHANES 1999–2010 sample characteristics (N=6472) Variables Total N (%) Cotinine levels N (%) Cotinine level-confirmed smoking status N (%) High cotinine 1878 (29.0) Medium cotinine 1759 (27.2) Low cotinine 2835 (43.8) Current smokers 1794 (27.4) Former smokers 1681 (25.0) Secondhand smokers 1158 (17.8) Non-smokers 1839 (29.8) Gender  Male 3352 (50.0) 1092 (31.4)* 953 (30.0)* 1307 (38.6) 1045 (30.4)† 938 (25.7)† 604 (19.2)† 765 (24.8)  Female (reference) 3120 (50.0) 786 (25.2) 806 (24.5) 1528 (50.3) 749 (24.5) 743 (24.3) 554 (16.3) 1074 (34.8) Race/ethnicity  White (reference) 3495 (74.6) 952 (27.0) 924 (26.8) 1619 (46.2) 926 (26.5) 1098 (27.9) 507 (15.7) 964 (29.9)  Black 1094 (9.0) 441 (40.5)* 392 (35.0)* 261 (24.5) 418 (38.3)† 161 (11.6)† 317 (30.5)† 198 (19.6)  Hispanic 1238 (7.6) 299 (24.7) 267 (22.1)* 672 (53.2) 271 (22.7)† 293 (19.9)† 196 (17.4) 478 (39.9)  Other Hispanic 399 (4.0) 106 (27.7) 111 (27.3) 182 (45.0) 103 (26.8) 78 (17.1)† 90 (24.0) 128 (32.1)  Other 246 (4.8) 80 (32.0) 65 (27.8) 101 (40.2) 76 (30.3) 51 (18.9) 48 (21.4) 71 (29.5) Alcohol consumption (drinks per day)  ≤1 (reference) 2316 (35.3) 334 (13.9) 593 (25.3) 1389 (60.8) 314 (13.3) 674 (27.6) 390 (16.5) 938 (42.6)  2–3 2658 (43.4) 784 (28.0)* 772 (28.7)* 1102 (43.2) 739 (26.9)† 747 (26.9)† 500 (18.6)† 672 (27.7)  4–5 741 (11.2) 354 (49.0)* 201 (28.7)* 186 (22.3) 341 (47.7)† 135 (17.8)† 143 (19.9)† 122 (14.6)  ≥6 757 (10.1) 406 (56.7)* 193 (25.9)* 158 (17.4) 400 (56.6)† 125 (15.7)† 125 (16.3)† 107 (11.4) Physical activity  Active (reference) 3834 (65.6) 911 (23.5) 1096 (28.0) 1827 (48.5) 866 (22.6) 1048 (25.9) 733 (18.8) 1187 (32.7)  Non-active 2638 (34.4) 967 (37.5)* 663 (25.8) 1008 (36.7) 928 (36.7)† 633 (23.3)† 425 (15.8) 652 (24.2) CRP  <0.5 (reference) 4997 (79.5) 1429 (27.9) 1354 (27.1) 2214 (45.0) 1366 (27.3) 1268 (24.1) 912 (17.9) 1451 (30.7)  ≥0.5 1475 (20.5) 449 (29.7) 405 (27.7 621 (42.6) 428 (28.0) 413 (28.3)† 246 (17.4) 388 (26.2) Federal poverty level  <200% 2494 (27.1) 1042 (42.8)* 654 (28.4)* 798 (28.8) 1006 (41.9) 495 (17.4) 459 (20.2) 534 (20.5)  ≥200% (reference) 3978 (72.9) 836 (22.9) 1105 (26.8) 2037 (50.3) 788 (22.1)† 1186 (27.8) 699 (16.9)† 1305 (33.2) Education  Less than high school 1507 (14.4) 632 (47.4)* 385 (26.8)* 490 (25.8) 593 (45.2)† 352 (20.2)† 262 (18.2)† 300 (16.4)  High School 1489 (22.7) 549 (39.4)* 435 (30.0)* 505 (30.6) 529 (38.5)† 344 (21.8)† 286 (19.4)† 330 (20.3)  More than high school (reference) 3476 (62.9) 697 (19.9) 939 (26.3) 1840 (53.7) 672 (19.4) 985 (27.2) 610 (17.1) 1209 (36.3) Year of survey cycle  1999–2000 653 (3.2) 193 (29.9) 229 (36.6)* 231 (33.5) 186 (30.3)† 176 (27.5)† 142 (22.3)† 149 (19.9)  2001–2002 1148 (5.9) 350 (31.9)* 311 (27.9)* 487 (40.1) 339 (31.5)† 309 (25.0)† 204 (18.9)† 296 (24.6)  2003–2004 979 (20.7) 290 (30.0)* 299 (31.0)* 390 (39.0) 275 (29.1)† 265 (24.4)† 192 (20.6)† 247 (25.9)  2005–2006 1047 (23.0) 305 (30.6)* 308 (28.3)* 434 (41.1) 287 (29.3)† 273 (25.6)† 203 (18.0)† 284 (27.0)  2007–2008 1236 (23.3) 377 (28.6)* 318 (28.5)* 541 (42.9) 361 (28.0)† 308 (23.8) 211 (18.5)† 356 (29.7)  2009–2010 (reference) 1409 (24.0) 363 (23.3) 294 (20.3) 752 (56.4) 346 (22.3) 350 (25.7) 206 (13.5) 507 (38.4) Variables Unadjusted mean+SE High cotinine Medium cotinine Low cotinine Current smokers Former smokers Secondhand smokers Non-smokers Age (years) 45 (0.3) 40 (0.4)* 43 (0.6)* 48 (0.5) 40 (0.4)† 53 (0.5)† 40 (0.5)† 45 (0.5) BMI (kg/m2) 28 (0.1) 26.6 (0.1) 28.2 (0.1)* 27.4 (0.2) 26.5 (0.1)† 28.0 (0.2)† 28. 3(0.2)† 27.2 (0.2) HOMA-IR (mg/iU) 2.9 (0.1) 1.9 (0.1) 2.3 (0.1)* 2.1 (0.1) 1.9 (0.1) 2.3 (0.1)† 2.3 (0.1)† 2.0 (0.1) HbA1c (%) 5.4 (0.01) 5.3 (0.01) 5.3 (0.02) 5.4 (0.01) 5.3 (0.01) 5.4 (0.02)† 5.3 (0.02) 5.3 (0.01) FPG (mg/dL) 99.2 (0.3) 97.2 (0.4) 98.8 (0.4) 98.2 (0.4) 96.9 (0.4) 100.9 (0.6)† 98.3 (0.5)† 96.8 (0.4) *Statistically significant at p<0.05 compared with low cotinine. †Statistically significant at p<0.05 compared with non-smokers. BMI, body mass index; FPG, fasting plasma glucose; HbA1c, hemoglobin A1c; HOMA-IR, Homeostasis Model Assessment. A GLM was used to examine the association between the outcome variables (log-transformed BMI, HOMA-IR, FPG, and HbA1c) and smoking status, controlling for age, gender, race/ethnicity, physical activity, education, FPL, year of survey cycle, and alcohol consumption. We estimated the adjusted mean and 95% CIs of log BMI, HbA1c, HOMA-IR, and FPG among the three smoking groups with non-smokers as the reference group. We adjusted for physical activity and alcohol consumption because it has been shown that smokers are less physically active20 and consume more alcohol than never smokers.21 We also examined the association between the outcome variables (log-transformed BMI, HOMA-IR, FPG, and HbA1c) and cotinine groups, controlling for age, gender, race/ethnicity, physical activity, education, FPL, year of survey cycle, and alcohol consumption. We estimated the adjusted mean and 95% CIs of log BMI, HbA1c, HOMA-IR, and FPG among the cotinine groups with low cotinine as the reference group. The above analyses were repeated in table 3 with BMI added as an additional covariate in the model. We performed the Sobel-Goodman mediation tests22 in STATA V.14.2 to test the extent of how BMI influences the effect of SHS on glycemic parameters. Table 3 Adjusted mean (95% CI) of HOMA-IR, HbA1c, and FPG according to cotinine levels, and cotinine level-confirmed smoking status for participants ≥20 years old (N=6472) with adjustment for BMI *High cotinine 1878 (29.0%) *Medium cotinine 1759 (27.2%) Low cotinine 2835 (43.8%) †Current smoker 1794 (27.4%) †Former smoker 1681 (25.0%) †Secondhand smoker 1158 (17.8%) Non-smoker 1839 (29.8%) HOMA-IR (mg/iU) 2.02 (1.94 to 2.10) 2.17 (2.08 to 2.25)‡ 2.05 (1.97 to 2.13) 2.00 (1.92 to 2.08) 2.16 (2.05 to 2.26)‡ 2.15 (2.04 to 2.27) 2.02 (1.92 to 2.11) HbA1c (%) 5.39 (5.37 to 5.42)‡ 5.35 (5.33 to 5.38)‡ 5.32 (5.30 to 5.33) 5.39 (5.37 to 5.42)‡ 5.33 (5.30 to 5.36) 5.37 (5.34 to 5.40)‡ 5.31 (5.29 to 5.33) FPG (mg/dL) 98.3 (97.5 to 99.0) 98.5 (97.7 to 99.4) 97.7 (97.0 to 98.4) 98.1 (97.3 to 98.9) 98.3 (97.4 to 99.3) 99.0 (98.0 to 99.9)‡ 97.4 (96.7 to 98.1) Data are presented as mean (95% CI). Adjusted for age, sex, races, BMI, C reactive protein, alcohol, physical activity, education, poverty level, and year of survey cycle. Values in bold indicate changed statistical significance compared with table 2. *Reference is low cotinine. †Reference is non-smoker. ‡Statistically significant at p<0.05. BMI, body mass index; FPG, fasting plasma glucose; HbA1c, hemoglobin A1c; HOMA-IR, Homeostasis Model Assessment of Insulin Resistance. We used multiple logistic regression models to obtain the adjusted OR and SEs for the categorized outcome variables (BMI 25–29.9 kg/m2, BMI≥30 kg/m2, HOMA-IR≥2.2, FPG≥126 mg/dL, and HbA1c≥6.5%) after controlling for age, gender, race/ethnicity, physical activity, education, FPL, year of survey cycle, and alcohol consumption. Data were presented as adjusted OR and SE and p<0.05 was considered statistically significant. All analyses were performed using SAS software V.9.4 (SAS Institute, Cary, North Carolina, USA) and the survey module of STATA software V.14.2 taking into consideration the complex sampling design and the sampling weights. Sample weights provided by the National Center for Health Statistics (NCHS) were used to correct for differential selection probabilities and to adjust for non-coverage and non-response. Table 2 Adjusted mean (95% CI) of BMI, HOMA-IR, HbA1c, and FPG according to cotinine levels, and cotinine level-confirmed smoking status for participants ≥20 years old (N=6472) without adjusting for BMI *High cotinine 1878 (29.0%) *Medium cotinine 1759 (27.2%) Low cotinine 2835 (43.8%) †Current smoker 1794 (27.4%) †Former smoker 1681 (25.0%) †Secondhand smoker 1158 (17.8%) Non-smoker 1839 (29.8%) BMI (kg/m2) 26.3 (26.0 to 26.6)‡ 28.2 (27.9 to 28.5)‡ 27.6 (27.2 to 27.9) 26.2 (25.9 to 26.6)‡ 27.9 (27.5 to 28.2) 28.3 (27.9 to 28.7)‡ 27.5 (27.1 to 27.9) Waist circumference (cm) 93.17 (92.35 to 94.00)‡ 97.32 (96.60 to 98.04)‡ 95.29 (94.41 to 96.16) 92.91 (92.01 to 93.80)‡ 96.78 (95.80 to 97.76)‡ 97.32 (96.27 to 98.38)‡ 94.87 (93.88 to 95.85) HOMA-IR (mg/iU) 1.85 (1.75 to 1.94)‡ 2.31 (2.20 to 2.42)‡ 2.09 (2.00 to 2.18) 1.82 (1.73 to 1.92)‡ 2.24 (2.13 to 2.35)‡ 2.31 (2.16 to 2.47)‡ 2.04 (1.94 to 2.15) HbA1c (%) 5.38 (5.35 to 5.40)‡ 5.37 (5.34 to 5.39)‡ 5.32 (5.30 to 5.34) 5.38 (5.35 to 5.40)‡ 5.34 (5.31 to 5.36) 5.38 (5.35 to 5.41)‡ 5.31 (5.29 to 5.34) FPG (mg/dL) 97.64 (96.88 to 98.40) 99.00 (98.11 to 99.90)‡ 97.84 (97.11 to 98.57) 97.41 (96.63 to 98.20) 98.62 (97.69 to 99.55) 99.51 (98.51 to 100.51)‡ 97.46 (96.64 to 98.29) Data are presented as mean (95% CI). Adjusted for age, sex, races, C reactive protein, alcohol, physical activity, education, poverty level, and year of survey cycle. *Reference is low cotinine. †Reference is non-smoker. ‡Statistically significant at p<0.05. BMI, body mass index; FPG, fasting plasma glucose; HbA1c, hemoglobin A1c; HOMA-IR, Homeostasis Model Assessment of Insulin Resistance. Results Of the 6472 participants with complete data used in the study, using cotinine level-confirmed smoking status as defined by the cotinine method above, 2835 (43.8%) had low cotinine and were considered NS, 1759 (27.2%) had medium cotinine and were considered SHSers, and 1878 (29.0%) had high cotinine and were considered were CS (table 1). We could not define FS using this method. Using the second method of survey results combined with cotinine levels, 1839 (29.8%) were NS, 1158 (17.8%) were SHSers, 1681 (25.0%) were FS, and 1794 (27.4%) were CS. Table 1 shows the sample characteristics and the outcome variables by the two methods of cotinine level-confirmed smoking status. In general, the percentage of current smokers was similar for the cotinine group defined by cotinine level and by survey/cotinine definition. Gender, race/ethnicity, alcohol consumption, physical activity, FPL, education, study cycle, age, BMI, and HOMA-IR and FPG varied significantly by cotinine level-confirmed smoking status (p<0.05) (table 1). Among the male subjects, 30% were SHSers by cotinine level and 19.2% by survey/cotinine definition, while among female subjects, 24.5% were SHSers by cotinine level and 16.3% by survey/cotinine definition. Blacks had the highest percentage of being SHSers; 35% by cotinine level and 30.5% by survey/cotinine definition compared with all other race/ethnic groups. Among participants with higher consumption of alcohol (six or more drinks per day), 25.9% were SHSers by cotinine level and 16.3% by survey/cotinine definition, whereas 56.7% were CS by cotinine level and 56.6% by survey/cotinine definition. Among participants with low income (FPL<200%), 28.4% were SHSers by cotinine level and 20.2% by survey/cotinine definition and among those with FPL≥200%, 26.8% were SHSers by cotinine level and 16.9% by survey/cotinine definition. The percentage of participants who were classified as SHSers defined by cotinine level decreased from 29.9% in 1999–2000 to 22.3% in 2009–2010 (F(5,87)=3.77, p=0.004) and by survey/cotinine definition decreased from 22.3% in 1999–2000 to 13.5% in 2009–2010 (F(5,83)=8.54, p=0.0001). The percentage of participants who were classified as NS defined by cotinine level increased from 33.5% in 1999–2000 to 56.4% in 2009–2010 (F(5,87)=8.96, p=0.0001) and by survey/cotinine definition increased from 19.9% in 1999–2000 to 38.4% in 2009–2010 (F(5,87)=8.95, p=0.0001). Using the survey/cotinine definition and unadjusted means (to describe the sample), SHSers and CS were younger and FS were older than NS and using cotinine levels, CS and SHSers were younger than NS. Using both definitions, SHSers had a higher BMI and HOMA-IR than NS (table 1). By the survey/cotinine definition, SHSers had a higher FPG than NS. By both definitions, CS had a lower BMI than NS, but similar HOMA-IR, HbA1c, and FPG as NS. By the survey/cotinine definition, FS had higher HOMA-IR, HbA1c, and FPG compared with NS. Table 2 shows the adjusted mean and 95% CI of BMI, HOMA-IR, HbA1c, and FPG by the GLM using the cotinine levels, and survey/cotinine definitions of smoking groups after controlling for age, gender, race/ethnicity, physical activity, education, FPL, year of survey cycle, and alcohol consumption. Using cotinine levels, the adjusted mean BMI in the medium cotinine group (SHS) (28.2 kg/m2, 95% CI 27.9 to 28.5) was higher compared with the low cotinine group (NS) (27.6 kg/m2, 95% CI 27.3 to 27.9 kg/m2; p=0.02). The adjusted mean HOMA-IR was found to be higher in the group with medium cotinine levels (SHSers) (2.31, 95% CI 2.20 to 2.42; p=0.004) and lower in the group with high cotinine levels (CS) (1.85, 95% CI 1.75 to 1.94; p=0.002) compared with the group with low cotinine levels (NS) (2.09, 95% CI 2.0 to 2.18). The group with high cotinine levels (CS) had higher levels of HbA1c (5.38%, 95% CI 5.35% to 5.40%; p<0.001) as did the medium cotinine group (SHSers) (5.37%, 95% CI 5.34% to 5.39%; p<0.001) than the group with low cotinine levels (NS) (5.32%, 95% CI 5.30% to 5.34%). Higher levels of FPG were found in the group with medium cotinine levels (SHSers) (99.0 mg/dL, 95% CI 98.1 to 99.9; p=0.04), but not in the group with high cotinine levels (CS) (97.6 mg/dL, 95% CI 96.9 to 98.4; p=NS), compared with the group with low cotinine levels (NS) (97.8 mg/dL, 95% CI 97.1 to 98.6 mg/dL). Using the survey/cotinine definition, the CS group had lower adjusted BMI (26.2 kg/m2, 95% CI 25.9 to 26.6 kg/m2; p<0.0001) and the SHSers group had a higher adjusted BMI (28.29 kg/m2, 95% CI 27.9 to 28.7 kg/m2; p=0.008) than the NS (27.5 kg/m2, 95% CI 27.1 to 27.89 kg/m2). Adjusted HOMA-IR was higher in the SHSers group (2.31 mg/iU, 95% CI 2.16–2.47 mg/iU; p=0.004) and the FS group (2.24 mg/iU, 95% CI 2.13 to 2.35 mg/iU; p=0.004) and lower in the CS group (1.82 mg/iU, 95% CI 1.73 to 1.92 mg/iU; p=0.004) compared with the NS group (2.04 mg/iU, 95% CI 1.94 to 2.15 mg/iU). The adjusted mean HbA1c was similar among the SHSers group (5.38%, 95% CI 5.35% to 5.41%, p=0.001) and CS group (5.38%, 95% CI 5.35% to 5.40%, p=0.001) but higher compared with the NS group (5.31%, 95% CI 5.29% to 5.34%). FPG in the SHSers group (99.5 mg/dL, 95% CI 98.5 to 100.5 mg/dL; p=0.002) was higher compared with the NS (97.5 mg/dL, 95% CI 96.6 to 98.3 mg/dL). Table 3 shows the effect of adjusting for BMI, in addition to the above variables, on relationship between smoking status and glycemic indices. After controlling for BMI, HOMA-IR in the high cotinine group and CS and SHSer group, and FPG in the medium cotinine group were no longer statistically significant. We tested the extent of how BMI influences the effect of medium cotinine (table 3, third column) on glycemic parameters and found that the proportion of total effect that is mediated by BMI is 0.62 on HOMA-IR, 0.73 on FPG, and 0.53 on HbA1c. The effect of SHS (table 3, seventh column) that is mediated by BMI is 0.69 on HOMA-IR, 0.41 on FPG, and 0.42 on HbA1c. Figure 2 shows the use of a multivariate logistic regression to examine the association between the outcomes and the smoking status determined by the survey/cotinine definition controlling for the confounding variables (age, gender, race/ethnicity, physical activity, education, FPL, year of survey cycle, and alcohol consumption). SHSers had higher adjusted odds of FPG≥126 mg/dL (OR=2.2, SE=0.6; p=0.01) and HOMA-IR≥2.2 (OR=1.4, SE=0.1; p=0.001) relative to NS. CS had lower odds of BMI≥30 kg/m2 (OR=0.60, SE=0.1; p<0.01) and BMI 25–29.9 kg/m2 (OR=0.66, SE=0.1; p< 0.01) compared with NS. SH had greater odds of BMI≥30 kg/m2 (OR=1.5, SE=0.2; p=0.01) and BMI 25–29.9 kg/m2 (OR=1.3, SE=0.1; p=0.01) compared with NS. FS also had higher adjusted odds of BMI≥30 kg/m2 (OR=1.4, SE=0.2; p=0.01), but was not significant for BMI 25–29.9 kg/m2 (OR=1.2, SE=0.1; p=0.1). FS also had higher adjusted odds of HOMA-IR≥2.2 (OR=1.3, SE=0.1; p=0.01). The adjusted odds of HbA1c≥6.5% was not statistically different for all groups compared with NS. Figure 2 Adjusted OR and SE for BMI≥30 kg/m2, BMI 25–29.9 kg/m2, HbA1c≥6.5%, HOMA-IR≥2.2, and FPG≥126 mg/dL in CS, FS, and SHS using the survey/cotinine definition compared with the reference group of NS after adjusting for age, sex, races, alcohol, physical activity, education, FPL, year of survey cycle using multiple logistic regression. *p<0.05. BMI, body mass index; CS, current smokers; FPG, fasting plasma glucose; FS, former smokers; HbA1c, hemoglobin A1c; HOMA-IR, Homeostasis Model Assessment of Insulin Resistance; SHS, secondhand smokers. Figure 3 shows the use of a multivariate logistic regression to examine the association between the outcomes (BMI, HOMA, HbA1c, and FPG) and the smoking status determined by cotinine levels controlling for the same confounding variables. CS (high cotinine level) had a lower adjusted OR of having a BMI≥30 kg/m2 (OR=0.57, SE=0.1; p<0.001) and BMI 25–29.9 kg/m2 (OR=0.67, SE=0.1; p< 0.001) relative to NS (with a low cotinine level). SHS (medium cotinine level) had a higher adjusted OR of having a BMI≥30 kg/m2 (OR=1.4, SE=0.2; p=0.009) and BMI 25–29.9 kg/m2 (OR=1.3, SE=0.1; p=0.004) relative to NS (with a low cotinine level). SHS (medium cotinine level) had a higher adjusted OR of having a HOMA-IR≥2.2 (OR=1.3, SE=0.1; p<0.0001) relative to NS (with a low cotinine level). The adjusted ORs of having an HbA1c≥6.5% or FPG≥126 mg/dL were not statistically different for all groups compared with the NS (low cotinine level). Figure 3 Adjusted OR and SE for BMI≥30 kg/m2, BMI 25–29.9 kg/m2, HbA1c≥6.5%, HOMA-IR≥2.2, and FPG≥126 mg/dL in those with high cotinine levels (CS) and medium cotinine levels (SHS) compared with the reference group of low cotinine levels (NS) after adjusting for age, sex, races, alcohol, physical activity, education, FPL, year of survey cycle using multiple logistic regression. *p<0.05. BMI, body mass index; FPG, fasting plasma glucose; HbA1c, hemoglobin A1c; HOMA-IR, Homeostasis Model Assessment of Insulin Resistance. Discussion This is the first study, to the best of our knowledge, that has examined the relationship between secondhand smoking and glycemic parameters and obesity using both recent self-reported cigarette smoking status and race/ethnicity-corrected cotinine values to define secondhand smoking. Our most significant finding is that using both methods of defining SHSers, we found that SHS exposure was associated with higher adjusted levels of HOMA-IR, FPG, and HbA1c as well as higher BMI, while CS is associated with a lower BMI compared with non-smokers. Adjustment for BMI demonstrated that some, but not all of the detrimental effects of SHS on glycemia are mediated by the increased weight of SHSer. The effects of SHS on HOMA-IR, a glycemic parameter, especially related to obesity, was especially reduced when adjusting for BMI. The present findings are consistent with the results of a recent meta-analysis examining prospective cohort studies on passive smoking that showed a significant relative risk of developing DM.11 This meta-analysis assessed four publications that all defined passive smoking by history and found a pooled relative risk (RR) of 1.28 (95% CI 1.14 to 1.44). The CARDIA study,23 a prospective cohort study, is noteworthy in that they did confirm SHS exposure by measuring serum cotinine levels and found that those with passive exposure to tobacco smoke verified by cotinine levels had an intermediate risk for developing glucose intolerance between current smokers and never smokers without passive smoke exposure. This study did not examine the incidence of obesity. Also, the CARDIA study was done among black and white young adults aged 18–30 years, whereas our study looked at all race/ethnicity and adults over 20 years. Several cross-sectional studies have shown an association between passive smoking with the development of type 2 DM.23–26 Lajous et al showed an association between SHS and DM in a prospective setting, but the study was done only in female Caucasians, thus the results cannot be generalized to men or non-Caucasian populations.25 Furthermore, this study used self-report SHS exposure that was not verified by cotinine levels. Zhang et al26 found that there was a positive relationship between exposure to passive smoke and active smoking and the risk of developing type 2 DM. This study, which confirms our findings of the association between secondhand smoking and DM, was a prospective study with high follow-up rates, although DM was only by self-report. In our study, we used biochemical parameters to define DM and other glycemic parameters. The study by Zhang et al26 included mainly white females, thus the results and conclusions cannot be generalized to the entire population since the study was not on a national representative sample. Our study used the NHANES database and thus our results can be generalized to the non-institutionalized population. In addition, passive smokers in study by Zhang et al26 were based on self-report. A prospective study by Hayashino et al27 showed that self-reported exposure to SHS in the workplace and current active smoking at baseline were positively associated with an increased risk of developing DM, even after controlling for possible confounders consistent with the findings of our study. Studies of users of nicotine gum have higher rates of insulin resistance28 and therefore, it is believed that the nicotine in cigarettes is the component that is associated with a higher rate of DM. We found that SHSers who are involuntarily exposed to cigarette smoking, had higher rates of two conditions, DM and obesity, that are associated with high mortality and morbidity. It is possible that the hand-to-mouth action of cigarettes, or the effect of cigarettes on taste buds, conditions that are not present for SHSers, may lead to CS being leaner than SHSers. Although one can argue that CS are also exposed to SHS, we are not surprised that CS are leaner than SHSers for several reasons. While SHSers are exposed to less smoke than CS, we do not know the dose relationship between amount of smoke and its effects on DM. SHS is produced at different temperatures and different reducing conditions, so some toxic substances may occur in different concentrations in SHS compared to primary smoke. Finally, the filter of cigarettes may decrease some of the negative effects of smoke in CS, but not SHSers.3 29 30 We found a statistical difference in genders between SHSers and NS that could be due to exposure of male subjects to more environmental smoke compared with female subjects. Males may be working in industries or in bars where more people smoke. There was a higher percentage of SHSers among blacks compared to whites that could be due to the disparities in occupations, with blacks being disproportionately employed in laborer and factory jobs. Although the national average of those in the laborer and factory jobs is 27%, the percent of blacks in those jobs is 41%.31 According to the Praxis Project, an analysis of municipal tobacco control ordinances in the USA found that communities with significant numbers of people of color are less likely to have any municipal tobacco control ordinances in place when compared with communities with fewer people of color.32 Non-Hispanic blacks metabolize cotinine slower than other racial/ethnic groups resulting in a higher cotinine level for the same amount of tobacco used.33 We found that there was a higher percentage of SHSers among those who had four or more alcohol drinks per day compared with NS. Those who drink more and thus visit bars and pubs are more exposed to SHS compared to those who have fewer drinks per day. According to the US Department of Health and Human Services, levels of SHS in restaurants and bars were found to be two to five times higher than in residences with smokers and two to six times higher than in office workplaces.34 Most interestingly, using both our definitions of smoking status, there was a decline in the rates of SHS in each 2-year iteration of the NHANES survey. This confirms a recent report using higher cut-offs of cotinine than we did to define SHSers and found that SHS smoke exposure in non-smokers declined from 52.5% during 1999–2000 to 25.3% during 2011–2012.35 Our data showed that SHS exposure declined from 22.3% during 1999–2000 to 13.5% during 2009–2010. We, as well as Homa et al,35 posit that this decline is related to introduction of statewide indoor smoking restriction laws during this time. According to the American Nonsmokers' Rights Foundation (http://www.no-smoke.org), in 2016, across the USA, 22 590 municipalities, representing 82.1% of the US population, are covered by a 100% smoke-free provision in non-hospitality workplaces, and/or restaurants, and/or bars, by either a state, commonwealth, territorial, or local law. Forty-one states and the District of Columbia have local laws in effect that require non-hospitality workplaces and/or restaurants and/or bars to be 100% smoke-free. The decline in rates of SHSers between 1999–2000 and 2009–2010 demonstrate that smoke-free laws are effective and may help decrease the number of Americans who develop obesity and DM as a result of being exposed to SHS. However, with several states without smoke-free laws, our data support the need for more states to pass smoke-free legislation. Laws prohibiting smoking in multiunit housing, including public housing, would further decrease the consequences of SHS. Although it is unknown if there will be lag time between the decline in SHS exposure and the potential decline in detrimental glycemic parameters and obesity, the effect of smoke-free laws on other health parameters is rapid, with immediate improvements in heart disease and improvements in lung cancer in about 2 years.36 Thus, we are optimistic that there will be a decline in SHS-associated DM and obesity in the near future that will continue with further reduction in SHS. Strengths and limitations Our study had several limitations. Since the NHANES data are a cross-sectional study, we cannot assess a causal relationship, although our results are consistent with several prospective cohort studies,23 24 26 37 which did show that SHS exposure led to the development of glucose intolerance and DM. Also, although we controlled for major confounders and found a robust association, it is possible that other unknown confounders could account for the association between secondhand smoking and DM and obesity. Cotinine levels decrease with time following the last exposure, so measurement of cotinine only reflects recent exposure. We defined CS, SHSers, and NS in two manners as discussed in the Research design and methods section using survey/cotinine definition and the objective race-ethnicity-adjusted measurement of serum cotinine levels. For the most part, the characterization by survey/cotinine definition was consistent with that by cotinine levels. However, it is possible that participants did not answer the surveys correctly (recall bias) as studies have shown that self-report underestimates smoking status.38 Those defined as SHSers by cotinine could be CS who did not smoke in the few days before the sample was collected, never smokers exposed to SHS, or FS exposed to SHS. Glycemic abnormalities and obesity develop over time, so using current cotinine levels may not reflect what happened decades ago when the pathogenesis of these diseases began. We attempted to overcome this difficulty by examining FS using survey/cotinine definition and found that these individuals had glycemic and obesity outcomes that were intermediate between CS and SHSers. FS can be currently exposed to SHS. We only examined the association between smoking status and body weight; in future studies, we may examine the association between smoking status and fat distribution, as visceral obesity is the key determinant of the predilection to develop DM. Conclusion In summary, because the USA has witnessed widespread increases in the prevalence of obesity and DM over the past few decades leading to high health costs, our study showing the association between cotinine-verified SHSers and worsening of glycemic parameters as well as obesity prompts the need for reliable and cost-effective methods for interventions to prevent secondhand smoking including stronger smoke-free laws. However, the only true solution for the health consequences of SHS exposure is a complete smoking ban. The authors acknowledge Thomas Yoshikawa, MD, of the professional development core of the Charles R. Drew University Accelerating Excellence in Translational Science (AXIS) for help with editing the manuscript. Contributors: TCF, MS, and DK had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. DK, MS, DP, and TCF were involved in study concept and design, interpretation of data, drafting of the manuscript and statistical analysis. DK, MS, and TCF were involved in the critical revision of the manuscript for important intellectual content. TCF obtained funding. Funding: Funding for this project came from NIH-NIMHD grant number U54MD007598 (formerly U54RR026138), MIDARP/DIDARP grant number R24DA017298, NIMHD grant number S21 MD000103 and NIH/NCATS grant number UL1TR000124. Disclaimer: The funding sources had no role in the writing of the manuscript or the decision to submit it for publication as the corresponding author, TCF, had full access to all the data in the study and had the final responsibility for the decision to submit for publication. Competing interests: All authors have completed the ICMJE uniform disclosure form at http://www.icmje.org/coi_disclosure.pdf and declare: no support from any organization for the submitted work; no financial relationships with any organizations that might have an interest in the submitted work in the previous 3 years; no other relationships or activities that could appear to have influenced the submitted work. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Original data sets are available on request. ==== Refs References 1 Centers for Disease Control, Prevention . Current cigarette smoking among adults—United States, 2011 . MMWR Morb Mortal Wkly Rep 2012 ;61 :889 –94 .23134971 2 Center for Disease Control . Smoking-attributable mortality, years of potential life lost, and productivity losses—United States, 2000–2004 . MMWR CDC Surveill Summ 2008 ;57 :1226 –8 . 3 US EPA. Respiratory health effects of passive smoking (also known as exposure to secondhand smoke or Environmental Tobacco Smoke-ETS) . US Environmental Protection Agency , 1992 . 4 Al-Delaimy WK , Willett WC , Manson JE Smoking and mortality among women with type 2 diabetes: The Nurses’ Health Study cohort . Diabetes Care 2001 ;24 :2043 –8 .11723080 5 Willi C , Bodenmann P , Ghali WA Active smoking and the risk of type 2 diabetes: a systematic review and meta-analysis . JAMA 2007 ;298 :2654 –64 . 10.1001/jama.298.22.2654 18073361 6 Sparrow D , Borkan GA , Gerzof SG Relationship of fat distribution to glucose tolerance. Results of computed tomography in male participants of the Normative Aging Study . Diabetes 1986 ;35 :411 –5 .3956878 7 California Environmental Protection Agency OoEHHA . Proposed Identification of Environmental Tobacco Smoke as a Toxic Air Contaminant: Part B Health Effects 2005 https://www.arb.ca.gov/regact/ets2006/app3exe.pdf 8 US Department of Health and Human Services . The Health Consequences of Involuntary Exposure to Tobacco Smoke: a Report of the Surgeon General . Atlanta : US Department of Health and Human Services, Centers for Disease Control and Prevention, Coordinating Center for Health Promotion, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health , 2006 . 9 National Toxicology Program . Report on Carcinogens: Eleventh Edition . US Department of Health and Human Services, Public Health Service, National Toxicology Program , 2005 . 10 Clair C , Bitton A , Meigs JB Relationships of cotinine and self-reported cigarette smoking with hemoglobin A1c in the US: results from The National Health and Nutrition Examination Survey, 1999–2008 . Diabetes Care 2011 ;34 :2250 –5 . 10.2337/dc11-0710 21836101 11 Wang Y , Ji J , Liu YJ Passive smoking and risk of type 2 diabetes: a meta-analysis of prospective cohort studies . PLoS One 2013 ;8 :e69915 10.1371/journal.pone.0069915 23922856 12 The National Health and Nutrition Examination Survey Analytic and Reporting Guidelines . Hyattsville, MD : Centers for Disease Control and Prevention (CDC). National Center for Health Statistics (NCHS) , 2004 http://www.cdc.gov/nchs/about/major/nhanes/nhanes2003–2004/analytical_guidelines.htm (accessed 24 Oct 2014 ). 13 Lohman TG , Roche AF , Martorell R Anthropometric standardization reference manual. Abridged ed . Champaign, IL : Human Kinetics Books , 1988 . 14 Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults: executive summary. Expert Panel on the Identification, Evaluation, and Treatment of Overweight in Adults . Am J Clin Nutr 1998 ;68 :899 –917 .9771869 15 Matthews DR , Hosker JP , Rudenski AS Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man . Diabetologia 1985 ;28 :412 –9 .3899825 16 Rayburn WF Diagnosis and classification of diabetes mellitus: highlights from the American Diabetes Association . Diabetes Care 2010 ;33 (Suppl 1 ):S62 –69 . 10.2337/dc10-S062 20042775 17 Haffner SM , Miettinen H , Stern MP The homeostasis model in the San Antonio Heart Study . Diabetes Care 1997 ;20 :1087 –92 .9203442 18 Benowitz NL , Bernert JT , Caraballo RS Optimal serum cotinine levels for distinguishing cigarette smokers and nonsmokers within different racial/ethnic groups in the United States between 1999 and 2004 . Am J Epidemiol 2009 ;169 :236 –48 . 10.1093/aje/kwn301 19019851 19 Seligman HK , Laraia BA , Kushel MB Food insecurity is associated with chronic disease among low-income NHANES participants . J Nutr 2010 ;140 :304 –10 . 10.3945/jn.109.112573 20032485 20 Deruiter WK , Faulkner G , Cairney J Characteristics of physically active smokers and implications for harm reduction . Am J Public Health 2008 ;98 :925 –31 . 10.2105/AJPH.2007.120469 18381990 21 Bobo JK , Husten C Sociocultural influences on smoking and drinking . Alcohol Res Health 2000 ;24 :225 –32 .15986717 22 Sobel ME Asymptotic confidence intervals for indirect effects in structural equation models . Sociol Methodol 1982 ;13 :290 –312 . 23 Houston TK , Person SD , Pletcher MJ Active and passive smoking and development of glucose intolerance among young adults in a prospective cohort: CARDIA study . BMJ 2006 ;332 :1064 –9 . 10.1136/bmj.38779.584028.55 16603565 24 Kowall B , Rathmann W , Strassburger K Association of passive and active smoking with incident type 2 diabetes mellitus in the elderly population: the KORA S4/F4 cohort study . Eur J Epidemiol 2010 ;25 :393 –402 . 10.1007/s10654-010-9452-6 20369275 25 Lajous M , Tondeur L , Fagherazzi G Childhood and Adult Secondhand Smoke and Type 2 Diabetes in Women . Diabetes Care 2013 ;36 :2720 –5 . 10.2337/dc12-2173 23757428 26 Zhang L , Curhan GC , Hu FB Association between passive and active smoking and incident type 2 diabetes in women . Diabetes Care 2011 ;34 :892 –7 . 10.2337/dc10-2087 21355099 27 Hayashino Y , Fukuhara S , Okamura T , HIPOP-OHP Research Group . A prospective study of passive smoking and risk of diabetes in a cohort of workers: the High-Risk and Population Strategy for Occupational Health Promotion (HIPOP-OHP) study . Diabetes Care 2008 ;31 :732 –4 . 10.2337/dc07-1905 18235051 28 Eliasson B , Taskinen MR , Smith U Long-term use of nicotine gum is associated with hyperinsulinemia and insulin resistance . Circulation 1996 ;94 :878 –81 .8790020 29 Chan-Yeung M , Dimich-Ward H Respiratory health effects of exposure to environmental tobacco smoke . Respirology 2003 ;8 :131 –9 .12753526 30 National Cancer Institute . Health effects of exposure to environmental tobacco smoke: the report of the California Environmental Protection Agency . Bethesda, MD : National Institutes of Health, National Cancer Institute, US Department of Health and Human Services , 1999 :12 –3 . 31 US Census Bureau CPS: March 2000. Major Occupation Group of the Employed Civilian Population, Tables 10.1 & 11 . 32 Secondhand Smoke Exposure, Clean Indoor Air Policies, and Communities of Color, The PRAXIS Project , 2003 https://www.thepraxisproject.org 33 Perez-Stable EJ , Herrera B , Jacob P III Nicotine metabolism and intake in black and white smokers . JAMA 1998 ;280 :152 –6 .9669788 34 US Department of Health and Human Services . Report on Carcinogens, Tenth Edition 2002 . National Toxicology Program . 35 Homa DM , Neff LJ , King BA Vital signs: disparities in nonsmokers’ exposure to secondhand smoke—United States, 1999–2012 . MMWR Morb Mortal Wkly Rep 2015 ;64 :103 –8 .25654612 36 Hahn EJ Smokefree legislation: a review of health and economic outcomes research . Am J Prev Med 2010 ;39 :S66 –76 . 10.1016/j.amepre.2010.08.013 21074680 37 Ko KP , Min H , Ahn Y A prospective study investigating the association between environmental tobacco smoke exposure and the incidence of type 2 diabetes in never smokers . Ann Epidemiol 2011 ;21 :42 –7 . 10.1016/j.annepidem.2010.10.006 21130368 38 Sasaki S , Braimoh TS , Yila TA Self-reported tobacco smoke exposure and plasma cotinine levels during pregnancy–a validation study in Northern Japan . Sci Total Environ 2011 ;412–413 : 114 –18 .

PMC005xxxxxx/PMC5372031.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01493110.1136/bmjopen-2016-014931Medical Education and TrainingProtocol1506170916951694170217101719Innovative approach for self-management and social welfare of children with cystic fibrosis in Europe: development, validation and implementation of an mHealth tool (MyCyFAPP) Calvo-Lerma Joaquim 12Martinez-Jimenez Celia P 34Lázaro-Ramos Juan-Pablo 5Andrés Ana 2Crespo-Escobar Paula 1Stav Erlend 6Schauber Cornelia 7Pannese Lucia 8Hulst Jessie M 9Suárez Lucrecia 10Colombo Carla 11Barreto Celeste 12de Boeck Kris 13Ribes-Koninckx Carmen 1on behalf of MyCyFAPP 1 Instituto de Investigación Sanitaria La Fe. Avenida Fernando Abril Martorell, Valencia, Spain 2 Instituto de Ingeniería de Alimentos para el Desarrollo, Universitat Politècnica de València, Valencia, Spain 3 University of Cambridge, Cancer Research UK Cambridge Institute, Cambridge, UK 4 Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Cambridge, UK 5 Soluciones Tecnológicas para la Salud y el Bienestar, Paterna, Spain 6 STIFTELSEN SINTEF, Trondheim, Norway 7 YOUSE GmbH. Kyreinstraße, München, Germany 8 Imaginary SRL. Piazza Caiazzo, Milano, Italy 9 Erasmus Medical Center, Sophia Children's Hospital, Rotterdam, The Netherlands 10 Servicio Madrileño de Salud—Hospital Universitario Ramón y Cajal, Colmenar Viejo, Spain 11 Università degli Studi di Milano. Fondazione IRCCS Ca’ Granda, Ospedale Maggiore Policlinico, Milan, Italy 12 Associação Portuguesa para a Investigação e Desenvolvimento da Faculdade de Medicina, Lisbon, Portugal 13 Department of Paediatrics, University Hospital of Leuven, University of Leuven, Leuven, BelgiumCorrespondence to Professor Carmen Ribes-Koninckx; ribes_car@gva.es2017 16 3 2017 7 3 e0149314 11 2016 26 1 2017 8 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction For the optimal management of children with cystic fibrosis, there are currently no efficient tools for the precise adjustment of pancreatic enzyme replacement therapy, either for advice on appropriate dietary intake or for achieving an optimal nutrition status. Therefore, we aim to develop a mobile application that ensures a successful nutritional therapy in children with cystic fibrosis. Methods and analysis A multidisciplinary team of 12 partners coordinate their efforts in 9 work packages that cover the entire so-called ‘from laboratory to market’ approach by means of an original and innovative co-design process. A cohort of 200 patients with cystic fibrosis aged 1–17 years are enrolled. We will develop an innovative, clinically tested mobile health application for patients and health professionals involved in cystic fibrosis management. The mobile application integrates the research knowledge and innovative tools for maximising self-management with the aim of leading to a better nutritional status, quality of life and disease prognosis. Bringing together different and complementary areas of knowledge is fundamental for tackling complex challenges in disease treatment, such as optimal nutrition and pancreatic enzyme replacement therapy in cystic fibrosis. Patients are expected to benefit the most from the outcomes of this innovative project. Ethics and dissemination The project is approved by the Ethics Committee of the coordinating organisation, Hospital Universitari La Fe (Ref: 2014/0484). Scientific findings will be disseminated via journals and conferences addressed to clinicians, food scientists, information and communications technology experts and patients. The specific dissemination working group within the project will address the wide audience communication through the website (http://www.mycyfapp.eu), the social networks and the newsletter. NUTRITION & DIETETICSHorizon 2020http://dx.doi.org/10.13039/501100007601643806 ==== Body Strengths and limitations of this study Innovative evidence-based method for pancreatic enzyme replacement therapy adjustment and self-management by means of a mobile application. Multidisciplinary team of experts for an integrative and co-designed patients-directed approach. Envisaged medium-term to long-term market uptake of the resulting mobile health application. Limited but statistically significant number of patients from five European countries will be included in the clinical validation. Introduction Cystic fibrosis (CF) is the most common life-threating autosomal inherited disease in Europe, with over 38 000 cases of CF currently registered in Europe.1 Along with pulmonary dysfunction and recurrent lung infections, the majority of patients (85%) suffer from lifelong pancreatic insufficiency (PI), which leads to maldigestion of foods and malabsorption of nutrients, especially lipids. In fact, pancreatic enzyme deficiency is occurring in ∼50% of infants by the age of 2 with a further 28% of the cases developing PI in early childhood.2 These malfunctions secondarily cause malnutrition, fat-soluble vitamin deficiencies and gastrointestinal symptoms. There is high-grade evidence that maintaining normal growth and nutrition adds 10 years more to the median survival since a close relationship between pulmonary function and nutritional status has been repeatedly ascertained.2–4 Malnutrition and growth stunting can only be avoided by accurate pancreatic enzyme replacement therapy (PERT) and close nutritional follow-up, as well as by early nutritional support and intervention. Nowadays, PERT consists of oral supplements containing a mixture of pancreatic enzymes—amylases, proteases and especially lipases—that have to be taken with every meal, while nutritional therapy relies on a high-energy and high-fat diet.5–9 However, at present, there is a lack of evidence-based methods to adjust PERT dosing and there are few handy tools or resources adequately available to promote a balanced and adapted diet (figure 1).5 10 11 Figure 1 Overview of current nutritional therapies in cystic fibrosis and the tools needed for successfully achieving a good nutritional status, quality of life and disease prognosis. Current recommendations for PERT dose adjustment rely on low level of evidence12 and counsel a number of units of lipase per gram of lipids. This means that in every meal, fat content should be known by the patient to estimate the corresponding PERT dose. The only way to achieve this would be by roughly estimating fat content from nutritional information databases and those should be easily available for patients. This approach is challenging for the patients and imprecise to maintain satisfactory levels of fat absorption. In this regard, clinical trials aimed at elucidating maldigestion in CF have led to inconsistent conclusions.5 Therefore, the demand for an evidence-based criterion for PERT adjustment has been highlighted,5 10 13 and the corresponding development of new innovative tools is imperative. Dietary lipids need to be accessible to digestive enzymes so that digestion and absorption can occur. The food matrix is dissociated through the digestion process, thus allowing the release of the embedded lipids and the access of the enzymes (lipases) to their substrates (lipids).5 14 Recent advances in food science research revealed that the different food structures modulate fatty acids release during digestion and their final metabolic fate.15–17 In addition, pancreatic lipase exhibits different hydrolytic activity depending on the intramolecular structure of the lipids.14 18 19 Therefore, lipolysis may cause different kinetics of release of absorbable fatty acids. This can be translated into different enzymatic dosage depending on the inherent-to-food characteristics, so nutrition and dietary habits play a key role in PERT effectiveness.20 21 Moreover, the lack of appropriate tools and resources for nutritional management can impair quality of life and lead to a lack of treatment adherence. For instance, if an incorrect nutritional behaviour or an inadequate PERT dosage occurs, the most likely scenario is that it will occur repeatedly and, in the majority of the cases, it will not be detected and corrected until the next contact at the CF unit. This could lead to long periods of omissions and/or wrong decisions. Consequently, the small daily actions related to nutrition that contribute to the overall disease prognosis would not be optimally used to improve the health status. Hence, nutritional treatment in CF can be considered as one of the ideal targets of mobile health (mHealth) and patients' self-management. In fact, CF is one of the most representative examples in which patients' monitoring and self-management can lead to a great improvement in the evolution and prognosis of the disease. Among other priorities in health, the current European Union's Research and Innovation Programme, Horizon 2020, strongly supports that current and future lines of research and technological development should be focused on this area (http://www.ec.europa.eu). In this framework, the MyCyFAPP Project (http://www.mycyfapp.eu) has been granted to develop an innovative approach focused on paediatric children with CF, self-management of nutrition and PERT by means of a mobile application (APP) linked to a web-based professional management tool. The objective of the present work is to describe the overall approach and study design of the MyCyFAPP Project as an example of multidisciplinary research and innovation project in mHealth.  Methods The consortium The Consortium was established in 2015 with the signature of the grant agreement with the European Commission. The multidisciplinary research team comprises nutritionists–dieticians, paediatric gastroenterologists and pulmonologists, food engineers, IT experts, game developers, software developers, psychologists, sociologists, biologists and patients' representatives. We have brought together our expertise to ensure the successful development of the project through a holistic and integrative approach of the different and complementary areas of knowledge and experts included. There are 12 organisations involved: 6 clinical partners linked to their corresponding research institutes or foundations, 3 small–medium enterprises (SMEs) related to mHealth, 1 Information and Communications Technology (ICT) Research Institute, 1 Food Technology Research Institute and the European Federation of Patients with CF (table 1). Table 1 List of participating organisations in the MyCyFAPP Project Country Organisation Type of activities Spain Instituto de Investigación Sanitaria La Fe Non-profit organisation pursuing the fostering and promotion of excellent research, scientific and technological knowledge and the translation to the productive sector. It manages research activities of Hospital La Fe, where the regional CF unit is the reference. Spain Soluciones TSB R&D and innovation SME focused on knowledge-intensive solutions for healthcare and well-being. Germany YOUSE GmbH Interdisciplinary SME working on increasing the usability and user experience of products and services Italy Imaginary SRL Experienced SME in creativity and innovation backed by solid technical competence and an understanding of the commercial potential of serious games and gamification Norway STIFTELSEN SINTEF Research organisation with expertise within user-centred design, software architecture, software development methods, mobile and social computing and evaluation of technology Spain Universitat Politècnica de València—Instituto de Ingeniería de Alimentos para el Desarrollo University Research Institute focused on food engineering. It applies its strong experience in industrial food processing to the area of the digestive food processing, involved in numerous collaborative projects between the industry and academia. Belgium University of Leuven The CF reference centre is based at the University Hospital of Leuven and has a strong research focus since many years. Portugal Associação Portuguesa para a Investigação e Desenvolvimento da Faculdade de Medicina It is the funding body that supports medical research in the Hospital de Santa Maria. The CF team conforms the reference unit in the country. Italy Università degli studi di Milano Research group linked to the Ospedale Maggiore Policlinico with a wide experience in CF multicentre projects, which is the largest CF reference unit in the region. The Netherlands Erasmus Medical Center, Sophia Children’s Hospital Rotterdam The hospital embraces the reference CF unit for children in the region. The medical team has a commitment to science and research integrity and therefore is actively involved in research projects. Spain Servicio Madrileño de Salud. Hospital Universitario Ramón y Cajal The hospital is one of the reference CF units for children in the region. The medical team has a broad experience in clinical trials and research in the field of CF Belgium Cystic Fibrosis Europe It is the representation of the Patients Organisations in Europe, which is actively involved in dissemination of CF activities and has been playing a key role in EU research projects. CF, cystic fibrosis; EU, European Union; R&D, research and development; SME, small–medium enterprise; TSB, Tecnológicas para la Salud y el Bienestar. Funding The MyCyFAPP Project is funded by the European Union through the Horizon 2020 Research and Innovation Programme (PHC-26-2014: self-management of health and disease: citizen engagement and mHealth) under grant agreement number 643806. Study design The 4-year long project (1 January 2015 to 31 December 2018) is constructed on nine inter-related work packages (WP; figure 2). Four multidisciplinary WP (1, 2, 3, 4) set the ground and generate the necessary knowledge and resources to develop the APP. A central technical WP (5) integrates the information in the development of the different software tools. These tools are thereafter tested for impact through a European multicentre clinical trial (WP6) and once the ICT tool is validated, another WP (8) takes care to bring the tool to the market by following different business models. Along the whole project, a specific WP (7) ensures the dissemination of the project to the very wide spectrum of audiences and another is devoted to the coordination of the consortium and the management of the implementation. Figure 2 General overview and inter-relation of WP. CF, cystic fibrosis; PERT, pancreatic enzyme replacement therapy; WP, work packages. WP underpinning the project European study on dietary habits in children with CF (WP1) One of the first actions of the project aims at obtaining information related to nutritional habits and dietary assessment of children with CF in the participating countries. It is used to establish the current nutritional habits of children with CF, PERT dosage, nutritional status and dietary assessment as a ground setting. The final milestone is then the generation of educational tools and resources for a customised nutritional self-management of the disease and patients' empowerment. In vitro assessment of enzyme requirements for foods and dishes (WP2) In parallel to the development of the European survey, we have set up a methodology to simulate in vitro digestion of a wide range of foods and meals under standardised CF gastrointestinal conditions. It allows for characterising inherent-to-food factors (chemical composition, molecular structure of lipids, food matrix) and gastrointestinal conditions (composition of digestive fluids and pH of the digestive environment), which affect fatty acids release and enzyme activity. The ultimate goal is to apply these results for determining the optimal PERT doses for foods and meals. They conform a key database supporting the mathematical algorithm. Development of the PERT dose predictive model (WP3) We conduct a pilot study with the enrolled children with CF. They follow a fixed menu consisting of a selection of foods and fixed enzyme doses according to the in vitro studies (theoretical optimal dose, TOD). Analyses of fat in stools reveal the degree of effectiveness of the predicted dose in each individual. Biostatistical modelling of the results determines an individual correction factor (ICF) calculation that will be able to correct the in vitro dose, for any other meal (even one not tested in the pilot study). Thus, from WP2, the TOD estimates the requirements of PERT considering food characteristics. Then, from WP3, the ICF will adjust the TOD according to patients' individual characteristics. These two key elements conform the predictive model, which calculates for each patient an individual optimal dose. User requirements specification for CF self-management (WP4) User requirements describe how software solutions work in a certain context of use; how the end users will benefit from it; how the application is managed and maintained; and how it is technically and organisationally deployed. As already mentioned, MyCyFAPP is an ecosystem of APPs, as well as a number of tools and components devoted to support the execution of those APPs. It is critical to gather a multidisciplinary team (developers, clinical partners, psychologists, experts in user experience and acceptance, paediatric and adult end users and patients' associations) to define in detail what the mobile applications will do, and how the clinical processes implemented through the web professional tool will be perceived by the users, both children and caregivers. MyCyFAPP has selected a methodology for the identification of user requirements, called ‘co-creation’, with the goal of maximising the opportunities for further adoption. A series of activities including interviews, focus groups and hands-on workshops to establish the needs and preferences regarding the APP usage will be conducted. We establish five focus groups (three patients and two parents): patients aged >16 years, patients aged 12–16 years, patients <12 years, parents of patients aged 12–16 years and parents of patients younger than 12 years. The APP will have different functions according to the role and responsibility of the target group in the self-management. The results will be translated into tailored interfaces and will be easily accessible and user-friendly for the different target populations. Software development of APP and health professional management tool (WP5) The results from WP4 are translated into technical specifications, and finally to software mobile and web applications. To this purpose, the system architecture, technical specifications, integration plan and software testing strategy are defined. Finally, after software development for full CF self-management, the implementation and integration of the algorithm developed in WP3 and the other resources developed in WP1 are conducted. At that point, the overall system will be delivered for the clinical trial in WP6. Impact assessment through a European multicentre clinical trial (WP6) We will carry out a European multicentre clinical trial to assess the impact derived from the usage of the APP on children's quality of life (especially related to nutrition and gastrointestinal symptoms), nutritional status and healthcare usage. A cohort of 200 patients will be recruited. The sample size was estimated using Monte Carlo simulations assuming normally distributed variables, and aiming for a precision of ±10% for each variable. A validation step is crucial for implementing MyCyFAPP in the usual clinical practice and transferring the self-management utility to patients with CF. Training and dissemination (WP7) This WP embraces a double scope. Training activities are aimed at achieving patient's engagement in self-management of their own disease so that specific workshops and webinars are scheduled prior to the start of the clinical trial addressing both patients and health professionals. Dissemination pursues the project's awareness, through all media channels, among the key stakeholders: patients and their families, patients' associations, health authorities, professionals from the different disciplines involved in the project, the industry and the general public. Overall, it targets the successful implementation of MyCyFAPP. Exploitation actions (WP8) This WP takes care of the exploitation of the final product and the Intellectual Property Rights (IPR) protection plans envisaged in the project. Specific actions include the identification of business models for the exploitation of the project's outcomes, the definition and execution of the strategy for exploitation and the coordination of the exploitation activities with disseminations to maximise the impact and awareness of the project. Coordination and management (WP9) It is aimed at orchestrating all the activities and partners of the project towards the successful implementation of the action and the reach of the goals and milestones. Expected results The MyCyFAPP Project pursues a final scenario where children with CF and their families and the health professionals can jointly and barriers-free manage the treatment of the disease. On the one side, patients and families count on the APP to self-manage nutrition and PERT and, on the other, health professionals use the professional tool to supervise and monitor patients' progress, ensuring feedback between the two parts when needed. This process is possible thanks to the specifically developed procedures and tools (features) that are addressed in the framework of the project from a rigorous scientific approach, responding to the current gaps on the resources needed but not available for a successful nutritional therapy (figure 3). Figure 3 Summary of the project: generated tools (A), expected final scenario at the end of the project (B) and desired outcomes (C). PERT, pancreatic enzyme replacement therapy. Tools and resources for MyCyFAPP Throughout the first WPs of the project, we conduct research that results in the generation of the needed tools and resources for the APP (figure 3A). The ‘mathematical predictive model’ of the optimal dose of enzymes is the main feature, tackling the currently existing gap to successfully adjust PERT. It is fed by the ‘theoretical PERT doses database’ including the optimal dose to digest a particular food or meal plus the ICF of each patient. It becomes functional when the users indicate the foods consumed and the amounts. A full and ‘interactive nutritional recommendations handbook’ is also available in the APP supporting children's dietary habits towards avoiding and correcting nutritional imbalances and reaching the recommendations. A ‘food and symptoms record’ is automatically generated and stored from the data introduced by the patients into the APP. This feature works thanks to the calculation algorithms and the ‘foods databases’, which include specific foods and meals/recipes according to the survey on nutritional habits and the complete nutritional profile information. The record allows for consulting at any time patients' progress in terms of nutritional composition of their diets, their symptoms and the actions they have performed in the system. ‘Educational games’ are developed in order to convey educational content of the recommendations handbook to the youngest children who cannot consult it. Games also have versions for older patients, these being aimed at consolidating the knowledge learnt by the other features. Finally, ‘alerts and messages’ systems smooth the usability of the APP between the two sides of MyCyFAPP—the patients and the clinical teams—making the experience profitable and appealing. Other specific features will be incorporated in the management system to enable health professionals to play their role: the professional tool. This module contains several features, such as a patients' dashboard displaying a summary of each patient—energy intake, percentage of nutrients, symptoms, number of depositions, etc—from where patients' profiles (especially focused on nutrition) can be accessed. Then ‘adjustment of parameters’ allows for making a more focused follow-up and to set up goals, and the ‘care plan management module’ is to define the overall strategy for patient. Complementarily, an education content management module and a report module are in charge of creating a report to be sent to the patients describing how they fit to their personalised plan. Through an iterative process with partners and final users, updates and corrections are periodically applied. Thus, the final set of features and tools will be decided along the project. The ultimate goal is to motivate the users to adhere to the plan with positive messages when needed, and proposing new challenges. Final scenario When the APP is ready to use (figure 3B), patients introduce the food products or dishes and the APP indicates in real time the optimal PERT dose for the particular meal and considering the ICF of the patient. This, at the same time, generates in real time a food record and its automatic nutritional report. Complementary patients are already taught and skilled to build up their menus according to the dietary recommendations and, when needed, they are offered to consult for suggestions or practical tips. Some of the functions enabled by the interaction between the patients and the clinical teams include the periodic check of the daily results of the nutritional profile of the diet. The software is programmed to alert patients and medical teams in case of a deviation from general or personalised recommendations (eg, percentage of lipids does not reach the threshold this week). If a deviation is identified as relevant—according to the definition of a risk and the plan for the patient—the health professionals can be notified, through the professional web tool, and are then responsible to decide which correction procedure has to apply (eg, consult educational resource number 1.3). For some situations, however, the software is programmed to automatically pop up reaction messages. Thus, the overall aim is to provide feedback and assistance to the patients outside the scheduled face-to-face visits. Of note, the aforedescribed situation is thoroughly assessed through a multicentre clinical trial that will allow for the identification of errors and the features and procedures showing room for improvement. Therefore, updates and modifications can be applied before upgrading the system to the final and fully functional version. If success in the clinical validation occurs, MyCyFAPP can be able to reach the market by following the defined exploitation plan. Desired outcomes Overall, we expect that the mHealth solution contributes to reach the project's goals: an evidence-based method for PERT adjustment, reaching nutritional goals and close nutritional follow-up. The desired outcomes derived from its long-term usage are a triple improvement: quality of life specifically related to gastrointestinal symptoms, nutritional status and disease prognosis (figure 3C). Conclusion Through MyCyFAPP, we have brought together highly experienced professionals from various European countries with different areas of knowledge to jointly address the challenges faced by adequate nutrition and PERT in the management of CF. We mainly tackle two gaps within the project: first, we develop from scratch the required tools for effective PERT and nutritional therapy; second, we make the tools available to patients, enabling effective adherence to the disease treatment through self-management but still, when needed, maintaining a close and dynamic interaction with the medical teams throughout the mHealth tool. The beneficiaries of the projects' results comprise patients, caregivers, families and healthcare professionals. MyCyFAPP is designed in a tailored way and clinically tested for CF self-management and monitoring. Additionally, MyCyFAPP has a pivotal role as a decision support system and provides a solution to the current gaps in the treatment. The participating SMEs and business models will ensure the commercial exploitation of the results, the market uptake and the MyCyFAPP distribution for the benefit of the patients. We envisage a prominent impact on nutritional status, quality of life and overall disease prognosis in the near future.When people ask me to provide an example of how patients, caregivers, researchers, a Foundation, NIH and industry can all work together to find cures, I point to cystic fibrosis. It's the very best example. Francis S Collins, MD, PhD, Director of the National Institutes of Health and a member of the international team that discovered the CF gene The authors of this paper, on behalf of the MyCyFAPP consortium, acknowledge the European Union and the Horizon 2020 Research and Innovation Framework Programme for funding the project (ref. 643806). They also acknowledge the support of the collaborators of Instituto de Investigación Sanitaria La Fe Sara Carrasco Llorens, Javier Ripoll Esteve, Victoria Fornes Ferrer, David Hervas Marin, Vanessa Garzón Rams, Sandra Martinez Barona and Etna Masip. Twitter: Follow Celia Martinez-Jimenez @celia4science Contributors: JC-L, CPM-J, AA, JPL-R and CR-K designed the research. ES, PC-E, CS, LP, JMH, LS, CC, CB and KdB contributed to the review and improvement of the project design. JC-L, CPM-J, AA, J-PL-R and CR-K drafted the first version of the manuscript and revised it critically for important intellectual content, and ES, PC-E, CS, LP, JMH, LS, CC, CB and KdB contributed to the revision of the manuscript ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All the authors approved the final version of the work. Funding: Horizon 2020 Research and Innovation Programme of the European Union. PHC-26-2014 call: self-management of health and disease: citizen engagement and mHealth, MyCyFAPP 643806. Competing interests: None declared. Ethics approval: Horizon 2020; Ethics Committee of Hospital Universitari La Fe. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: The project is currently in a pre-results stage. ==== Refs References 1 Colombo C , Littlewood J The implementation of standards of care in Europe: state of the art . J Cyst Fibros 2011 ;10 (Suppl 2 ):S7 –S15 . 10.1016/S1569-1993(11)60003-9 21658645 2 Gaskin KJ Nutritional care in children with cystic fibrosis: are our patients becoming better & quest . Eur J Clin Nutr 2013 ;67 :558 –64 . 10.1038/ejcn.2013.20 23462946 3 Stephenson AL , Mannik LA , Walsh S Longitudinal trends in nutritional status and the relation between lung function and BMI in cystic fibrosis: a population-based cohort study . Am J Clin Nutr 2013 ;97 :872 –7 . 10.3945/ajcn.112.051409 23388659 4 Kerem E , Viviani L , Zolin A Factors associated with FEV1 decline in cystic fibrosis: analysis of the ECFS Patient Registry . Eur Respir J 2014 ;43 :125 –33 . 10.1183/09031936.00166412 23598952 5 Fieker A , Philpott J , Armand M Enzyme replacement therapy for pancreatic insufficiency: present and future . Clin Experiment Gastroenterol 2011 :4 :55 –73 . 6 Sikkens EC , Cahen DL , Kuipers EJ Pancreatic enzyme replacement therapy in chronic pancreatitis . Best Pract Res Clin Gastroenterol 2010 ;24 :337 –47 . 10.1016/j.bpg.2010.03.006 20510833 7 Wilschanski M , Durie PR Patterns of GI disease in adulthood associated with mutations in the CFTR gene . Gut 2007 ;56 : 1153 –63 . 10.1136/gut.2004.062786 17446304 8 Woestenenk JW , van der Ent CK , Houwen RH Pancreatic enzyme replacement therapy and coefficient of fat absorption in children and adolescents with cystic fibrosis . J Pediatr Gastroenterol Nutr 2015 ;61 :355 –60 . 10.1097/MPG.0000000000000784 25782658 9 Haupt ME , Kwasny MJ , Schechter MS Pancreatic enzyme replacement therapy dosing and nutritional outcomes in children with cystic fibrosis . J Pediatr 2014 ;164 :1110 –15.e1 . 10.1016/j.jpeds.2014.01.022 24560182 10 Borowitz D , Durie PR , Clarke LL Gastrointestinal outcomes and confounders in cystic fibrosis . J Pediatr Gastroenterol Nutr 2005 ;41 :273 –85 . 10.1097/01.mpg.0000178439.64675.8d 16131979 11 Borowitz D , Gelfond D , Maguiness K Maximal daily dose of pancreatic enzyme replacement therapy in infants with cystic fibrosis: a reconsideration . J Cyst Fibros 2013 ;12 :784 –5 . 10.1016/j.jcf.2013.05.011 23809508 12 Turck D , Braegger CP , Colombo C ESPEN-ESPGHAN-ECFS guidelines on nutrition care for infants, children, and adults with cystic fibrosis . Clin Nutr 2016 ;35 :557 –77 . 10.1016/j.clnu.2016.03.004 27068495 13 Li Y , Hu M , McClements DJ Factors affecting lipase digestibility of emulsified lipids using an in vitro digestion model: proposal for a standardised pH-stat method . Food Chem 2011 ;126 : 498 –505 . 10.1016/j.foodchem.2010.11.027 14 Hunter JE Studies on effects of dietary fatty acids as related to their position on triglycerides . Lipids 2001 ;36 :655 –68 . 10.1007/s11745-001-0770-0 11521963 15 Stuknytė M , Cattaneo S , Masotti F Occurrence and fate of ACE-inhibitor peptides in cheeses and in their digestates following in vitro static gastrointestinal digestion . Food Chem 2015 ;168 : 27 –33 . 10.1016/j.foodchem.2014.07.045 25172679 16 Parada J , Aguilera JM Food microstructure affects the bioavailability of several nutrients . J Food Sci 2007 ;72 :R21 –32 . 10.1111/j.1750-3841.2007.00274.x 17995848 17 Singh H , Ye A , Horne D Structuring food emulsions in the gastrointestinal tract to modify lipid digestion . Prog Lipid Res 2009 ;48 :92 –100 . 10.1016/j.plipres.2008.12.001 19116160 18 Zhu X , Ye A , Verrier T Free fatty acid profiles of emulsified lipids during in vitro digestion with pancreatic lipase . Food Chem 2013 ;139 :398 –404 . 10.1016/j.foodchem.2012.12.060 23561123 19 Lesmes U , McClements DJ Controlling lipid digestibility: response of lipid droplets coated by β-lactoglobulin-dextran Maillard conjugates to simulated gastrointestinal conditions . Food Hydrocolloids 2012 ;26 :221 –30 . 10.1016/j.foodhyd.2011.05.011 20 Giang TM , Gaucel S , Brestaz P Dynamic modeling of in vitro lipid digestion: individual fatty acid release and bioaccessibility kinetics . Food Chem 2016 ;194 :1180 –8 . 10.1016/j.foodchem.2015.08.125 26471670 21 Nieva-Echevarría B , Goicoechea E , Manzanos MJ A study by 1H NMR on the influence of some factors affecting lipid in vitro digestion . Food Chem 2016 ;211 :17 –26 . 10.1016/j.foodchem.2016.05.021 27283602

PMC005xxxxxx/PMC5372032.txt

==== Front BMJ Open Diabetes Res CareBMJ Open Diabetes Res CarebmjdrcbmjdrcBMJ Open Diabetes Research & Care2052-4897BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjdrc-2016-00033710.1136/bmjdrc-2016-000337Clinical Care/Education/Nutrition/Psychosocial Research15061866Role of depression in diabetes management in an ethnic minority population: a case of Korean Americans with type 2 diabetes Kim Miyong To 1Kim Kim Byeng 2Ko Jisook 1Jang Yuri 3Levine David 4Lee Hochang Benjamin 5 1 School of Nursing, University of Texas at Austin, Austin, Texas, USA 2 Korean Resource Center, Ellicott City, Maryland, USA 3 School of Social Work, University of Texas at Austin, Austin, Texas, USA 4 School of Medicine, Johns Hopkins University, Baltimore, Maryland, USA 5 School of Medicine, Yale University, Boston, Connecticut, USACorrespondence to Dr Kim Byeng Kim; kbkim@ikorean.org2017 24 3 2017 5 1 e0003373 10 2016 12 1 2017 29 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Background Comorbid depression and diabetes mellitus (DM) compound challenges to disease management such as low health literacy, insufficient access to care, and social or linguistic isolation. Korean Americans (KAs), predominantly first-generation immigrants, suffer from a high prevalence of type 2 DM and depression. Limited research on KAs has prevented the development of effective interventions. Objectives To compare the prevalence of depression in KAs with DM and all Americans with/without DM, and to explore correlates of comorbid DM and depression and strategies to address KAs' DM and depression. Methods KAs' data were from a clinical trial of a community-based self-help intervention to improve KAs' DM and mental health outcomes. National Health and Nutrition Examination Survey data sets enabled comparison. Clinical indicators included hemoglobin A1C, lipid panel, and body mass index. Psychobehavioral indicators included self-efficacy for DM management, quality of life, and depression (Patient Health Questionnaire-9 (PHQ-9)). Results More KAs with DM had depression (44.2%) than did all Americans with DM (28.7%) or without DM (20.1%). Significantly more KAs with DM had mild (29.3%) or clinical (14.9%) depression than did Americans with DM (mild, 17.2%; clinical, 11.5%) or without (mild, 13.8%; clinical, 6.3%). One of six KAs with DM (16.9%) thought of suicide or self-harm (Americans with/without =5.0%, 2.8%). The self-help intervention reduced the mean PHQ-9 from 5.4 at baseline to 4.1 at 12 months. Limitations External validity might be limited; KAs' data were from one study site. Conclusions The prevalence of depression and DM among KAs warrants the development of efficacious interventions. Trial registration number NCT01264796. DepressionType 2 DiabetesSelf-CareMinoritiesNational Institute of Diabetes and Digestive and Kidney Diseaseshttp://dx.doi.org/10.13039/100000062R18DK083936 ==== Body Significance of this study What is already known about this subject? Depression is a significant comorbid risk for diabetes. What are the new findings? Korean Americans with type 2 diabetes mellitus have a high level of depression prevalence (44.2%). How might these results change the focus of research or clinical practice? A community-based self-help programme is effective in mitigating depression. Introduction Depression and diabetes mellitus (DM) are the most common comorbid conditions across all sociodemographic spectra.1 Although their etiological relationship is unclear, together they present significant challenges to disease management in individuals with type 2 DM.2–4 In comparison with people who have DM but no depression, individuals with concurrent depression and DM report lower levels of self-efficacy, self-care, and quality of life and have poorer clinical outcomes, including higher incidences of complications and reduced life expectancy.5 Comorbid depression and DM increase treatment costs, and clinical outcomes have not yet demonstrated their treatments' effectiveness.6–9 Also, the difficulty of DM management is amplified in vulnerable subgroups: individuals with disabilities, low income, and low health literacy, as well as members of linguistically isolated ethnic minority immigrant groups. Korean Americans (KAs) represent such a group. KAs, predominantly first-generation immigrants, suffer from a disproportionally high incidence of depression10 11 and DM,12 and they are socially isolated by linguistic and cultural barriers. A recent epidemiological study of elderly KAs (age ≥60) has reported a prevalence of depression of over 30% on the Patient Health Questionnaire (PHQ-9; mild depression score, 5–9; clinical depression, ≥10);13 DM was significantly correlated with depression. As depression affects individuals' abilities at DM self-management as well as clinical outcomes, future DM treatment and interventions will require a holistic approach, integrating strategies to improve mental and physiological health. Such intervention requires a thorough understanding of the target population, including the prevalence of factors that contribute to their chronic conditions as well as unique contextual factors that may influence their disease management. Such critical information, however, is not yet available due to the lack of empirical studies of KAs with type 2 DM that characterize the role of depression in their disease management. For the present study, we have analyzed data from a recently completed clinical trial of a community-based DM self-management intervention by estimating the prevalence of depression in the ethnic minority population of KAs and examining both the role of depression in their DM self-management behaviors and their clinical outcomes. To make clear inferences, we have also used national data from the National Health and Nutrition Examination Survey (NHANES) for the years 2005–2006, 2007–2008, 2009–2010, and 2011–2012 as a reference group. In addition to explicating empirical relationships between the comorbid conditions of DM and depression and traditional factors that influence DM self-management and clinical outcomes, we have evaluated several unique contextual factors associated with the KA population. In undertaking this study, we hypothesized that KAs with DM would have a higher prevalence of depression than would the general population and that this would have a significant impact on their ability to manage DM and thus on their quality of life. Material and methods Design The primary data source for this study was an open-label randomized control trial to test the effectiveness of a community-based self-help intervention for KAs with uncontrolled DM as measured by hemoglobin A1C ≥7.0%. A total of 250 KAs residing in the Baltimore–Washington region enrolled in the study. After 12 months, 105 in the intervention group (n=120) and 104 in the control group (n=130) completed the program for a retention rate of 83.6%. Recruitment, enrollment, and retention of participants have been detailed elsewhere.14 15 The study protocol was approved by the Johns Hopkins Institutional Review Board. After providing their consent, participants enrolled in the study from September 2009 to June 2014. Intervention The intervention group received 2 hours of didactic training each week for 6 weeks, followed by monthly telephone counseling by a team of registered nurses and community health workers. The control group received a shortened version of the intervention after the study was completed. The intervention's details have been presented elsewhere.14 15 In short, the training covered participants' DM management and control. It included the etiology of DM, healthy diet, and exercise. One hour devoted to stress management included ways to improve communication and solve problems. Measurements Our clinical trial and the NHANES followed the same data collection protocol and included (a) demographics, (b) dietary data, (c) examination data, (d) laboratory data, and (e) questionnaire data. Our data sets, however, were on a smaller scale. Whereas the NHANES data are biannual and cross-sectional, ours were panel data, collected at baseline and at 3, 6, 9, and 12 months, except for the dietary data and the questionnaire data for depression, which were not collected at 9 months. Demographic information included age, sex, marital status, family size, work status, and annual household income. Family size was an indicator of living arrangement, recoded as 1 for living alone or 0 for otherwise. In our study, participants reported actual monthly income, which was converted to a percentage of the federal poverty line by taking the family size into account. It was further reclassified into four categories, according to the Patient Protection and Affordable Care Act: 0 for an income ≤138%; 1 for 139–250%; 2 for 251–400%, and 3 for >400%. Dietary data were collected using the 24-hour recall method. Examination data included weight and height for the calculation of body mass index (BMI), as well as blood pressure measurement. Laboratory data included hemoglobin A1C, lipid panel (total cholesterol, triglycerides), and fasting plasma glucose. The PHQ-9 measured depression. The PHQ's 9 items probe the frequency of depressive experience—for example, ‘Over the past 2 weeks, how often have you been bothered by any of the following problems?’ Responses are 0=‘Not at all’, 1=‘Several days’, 2=‘More than half the days’, and 3=‘Nearly every day’. Thus, the total score ranges from 0 to 27. In our trial, we used the instrument's Korean version, the PHQ-9K.16 17 The total score was reclassified into three ordered categories of depression: (a) ‘normal’ if the score was <5, (b) ‘mild depression’ if the score was 5–9, and (c) ‘clinical depression’ if the score was 10 or higher.16 In addition, for detailed analysis, each item score was recoded as 0 for ‘no depression episode’ or 1 for ‘episode’. Also included as two dichotomous indicators were DM medication and whether or not an individual was on insulin (‘yes’=1 and ‘no’=0). Self-efficacy measured the level of confidence in managing DM with regard to diet, exercise, and general self-management behaviors, using a DM-related self-efficacy scale18 with eight 10-point Likert-type items (1=‘None’, and 10=‘Full’), yielding a total sum from 8 to 80. Social support from family members and friends was measured with the Personal Resource Questionnaire,19 20 with 11 5-point Likert-type items (1=‘Receiving no support’, and 5=‘A lot of support’). We also scored the gap between expected and received support on the same questionnaire. Total scores for self-efficacy, social support, and support gap were reclassified using tertiles: 0=‘Low’, 1=‘Middle’, and 2=‘High’. The self-efficacy and the support scales were available only in our clinical study. DM control was measured by hemoglobin A1C, the proportion of glycated red cells (expressed in percentages). A1C is a gold standard for the diagnosis of DM and indication of DM control.21 A person has DM if A1C ≥6.5%, and pre-diabetes if A1C is 5.7–6.4%. Analysis Depression prevalence among KAs with DM is based on our data, excluding one incomplete observation. For comparison, we selected two additional groups from the NHANES 2005–2006, 2007–2008, 2009–2010, and 2011–2012 cycles. One, the comparison group, included Americans (whites, blacks, Mexican Americans, Hispanics, and others, n=1707) with uncontrolled DM (A1C ≥7.0%); the other, the standard group, consisted of Americans (n=6880) without DM, pre-diabetes, or type 1 DM (A1C <5.7%). To explicate correlates of depression, ordered logistic regression was administered on a pooled sample of KAs at baseline, 3, 6, and 12 months (n=759 with listwise deletion), on the comparison and standard groups. Further, to explore the self-help intervention's effects on depression among KAs with uncontrolled DM, a mixed effects model was used on our panel data. We used STATA (V.14) for analysis; statistical significance was set at p<0.05. Results Sample characteristics The KAs from the clinical trial (N=249) were similar to the comparison group (Americans with DM, n=1707) in mean (±SE) age (58.3±0.5 vs 58.4±0.4 years). More men were included in the KA sample (56.6%) than in the comparison group (51.0%). More KAs were working either full or part-time (59.5% vs 44.8%), with greater education (86.3% vs 71.5% for high school graduate or above). More KAs reported low income than did the comparison group (67.1% vs 53.9% making less 251% of the federal poverty line). The majority of KAs were married (89.6% vs 59.0%), and fewer KAs were living alone (6.0% vs 16.1%) or covered by health insurance (52.2% vs 84.3%). Regarding physiological indicators, the mean (SE) BMI of KAs was 25.4 (0.2), lower than that of the comparison group (mean=32.2, SE=0.2). Although KAs had the same level of total cholesterol (193 mg/dL) as did Americans with DM, they had lower triglycerides (191 vs 240 mg/dL) and fasting serum glucose (160 vs 188 mg/dL). Also, more KAs were taking DM medication (63.1% vs 47.4%, unweighted z=4.63, 95% CI 0.09 to 0.22, p<0.001), but significantly fewer KAs were receiving insulin injection therapy (4.2% vs 23.2%; unweighted z=−6.90, 95% CI −0.22 to −0.16, p<0.001) (see table 1). Table 1 Comparison of demographic and physiological indicators between the Korean Americans (≥35 years old) with uncontrolled type 2 diabetes (A1C ≥7.0%) at baseline in our clinical study, the comparison group (Americans with the same conditions as Korean Americans), and the standard group (Americans with A1C <5.7% and not being diagnosed of diabetes, pre-diabetes or type 1 diabetes) in the NHANES Indicators\groups Korean Americans* (n=249) Comparison group† Standard group‡ (n=6880) Total (n=1707) Whites (n=537) Blacks (n=574) Mexican Americans (n=314) Hispanic (n=161) Others (n=121) Male (%) 56.6 51.0 53.2 45.5 54.3 46.1 49.5 48.1 Age, mean (SE) (years) 58.3 (0.5) 58.4 (0.4) 60.5 (0.6) 56.7 (0.6) 54.5 (1.0) 52.7 (1.5) 57.2 (1.6) 51.6 (0.3) Married (%) 89.6 59.0 .62.6 .44.7 59.6 56.8 73.7 65.4 Living alone (%) 6.0 16.1 18.6 18.5  9.1 7.0 7.2 14.1 Working full/part-time (%) 59.5 44.8 43.1 44.9 50.1 39.0 55.4 68.0 Education  <High school graduate (%) 13.7 28.5 18.9 32.1 70.2 51.1 18.5 14.2  High school graduate (%) 36.1 25.8 30.1 25.8 13.1 13.2 19.6 22.0  ≥College graduate (%) 50.2 45.7 51.0 42.1 16.7 35.8 62.0 63.8 Income (federal poverty line)  ≤138% 38.0 29.8 23.0 31.0 55.1 51.6 29.2 16.4  ≤250% 29.1 24.1 21.7 28.0 28.8 24.0 26.4 17.1  ≤400% 17.4 18.8 21.3 22.5 7.1 12.6 9.8 21.3  >400% 15.5 27.3 34.0 18.5 9.0 11.8 34.6 45.2 Health insurance (%) 52.2 84.3 90.9 84.1 55.3 73.8 81.1 85.4 BMI, mean (SE) 25.4 (0.2) 32.2 (0.3) 32.8 (0.5) 32.9 (0.4) 31.5 (0.5) 31.7 (0.6) 28.6 (1.2) 27.8 (0.1) Normal (BMI <25) 47.4 16.1 14.6 14.6 12.7 11.8 39.5 32.3 Overweight (BMI<30) 45.8 24.2 21.4 25.8 32.6 30.9 23.6 38.1 Obese (BMI ≥30) 6.8 59.7 64.0 59.6 54.7 57.3 36.9 28.6 A1C (%), mean (SE) 8.8 (0.1) 8.6 (0.1) 8.3 (0.1) 8.9 (0.1) 9.1 (0.1) 9.1 (0.3) 8.7 (0.2) 5.3 (0.0) Total cholesterol, mean (SE) 193 (2.9) 193 (1.7) 187 (2.4) 196 (3.0) 206 (3.5) 206 (6.5) 193 (7.6) 205 (0.8) Normal (<200) 59.6 59.4 59.6 65.6 55.0 45.5 40.3 47.3 Borderline (200–239) 26.4 26.5 24.2 20.5 30.1 29.9 35.4 35.8 High (≥240) 14.0 14.1 16.2 13.9 14.9 24.6 24.3 16.9 Triglyceride status (mg/dL) 191 (12.8) 240 (8.5) 250 (11.4) 170 (7.2) 276 (22.3) 265 (26.0) 252 (29.7) 151 (2.0) Normal (<150) 50.6 37.8 35.2 56.1 31.7 33.7 27.0 64.3 Borderline (150–199) 22.5 16.9 17.6 17.1 18.1 11.9 14.5 15.5 High (≥200) 26.9 45.3 47.2 26.8 50.2 54.4 58.5 21.2 Blood pressure, SBP/DBP (mg/Hg) 133/78 130/71 131/70 134/73 129/71 127/73 126/72 121/72 <120/80 37.3 29.8 30.0 25.2 31.4 36.4 33.1 45.8 ≤139/89 39.4 46.1 44.5 48.1 49.4 47.1 48.8 44.2 ≥140/90 23.3 24.1 25.5 26.7 19.2 16.5 18.1 10.0 Blood glucose (mg/dL) 160 (3.8) 188 (2.9) 181 (3.9) 193 (6.5) 212 (6.9) 202 (11.5) 176 (7.5) 92 (0.2) Normal (<100) (%) 9.2 9.2 08.8 14.2 04.5 09.1 8.4 79.2 Pre-diabetes (≤125) (%) 23.3 11.7 12.0 09.5 10.2 18.5 10.4 18.5 Diabetes (>125) (%) 67.5 79.1 79.2 76.3 85.3 72.4 81.2 2.3 Years having diabetes 8.1 (0.5) 10.6 (0.4) 10.6 (0.6) 11.0 (0.5) 10.4 (0.7) 9.3 (0.8) 10.4 (1.5) – On oral medication, % 63.1 47.4 49.8 42.8 47.5 44.2 44.0 – On insulin, % 4.2 23.2 27.1 20.4 16.0 15.0 17.0 – *At baseline and unweighted statistics. †Weighted statistics for N=11 767 732. ‡Weighted statistics for N=80 651 898. BMI, body mass index; DBP, diastolic blood pressure; NHANES, National Health and Nutrition Examination Survey; SBP, Systolic blood pressure. Prevalence of depression Overall, more KAs reported depression than did those in the comparison and standard groups. Differences were statistically significant for all items except item 8, ‘Moving or speaking so slowly that other people could have noticed…’ (unweighted z=1.43, 95% CI −0.02 to 0.08, p=0.153). In particular, about twice as many KAs than those in the comparison group reported ‘Little interest or pleasure in doing things’ (49.8% vs 27.7%; unweighted z=7.1, 95% CI 0.16 to 0.29, p<0.001), ‘Poor appetite or overeating’ (55.4% vs 26.4%; z=9.3, 95% CI 0.22 to 0.36, p<0.001), and ‘Trouble concentrating on things, such as reading the newspaper or watching television’ (29.3% vs 15.9%; z=5.2, 95% CI 0.07 to 0.19, p<0.001). The mean (±SE) depression scores of KAs (5.1±0.3) and Hispanics (5.0±0.7) were highest among races; differences between KAs and the comparison (3.7±0.1) and standard (2.7±0.1) groups were statistically significant (unweighted t=4.12 and 7.12, respectively, p<0.001). Accordingly, about a third (29.6%) of KAs had mild depression, which was higher than in whites (16.9%), blacks (16.6%), Mexican Americans (18.0%), Hispanics (20.1%), and others (17.5%). Similarly, clinical depression among KAs (14.9%) was second highest after clinical depression in Hispanics (19.7%) followed by whites (11.5%), blacks (11.2%), and Mexican Americans (11.0%). As a result, depression (mild and clinical depression combined) was most prevalent among KAs (44.2%), who were followed closely by Hispanics (39.8%) and other races with margins of >15 percentage points. About one out of six KAs (16.9%) had thoughts of hurting or killing themselves within the past 2 weeks, which was >3 times the percentage for other ethnic groups (4.8% for whites, 5.4% for blacks, and 5.5% for Mexican Americans and others). The KAs' rate on this item was 5.8 times higher than that for the standard group (2.9%). Nevertheless, KAs underestimated the impact of depression on daily activities (eg, doing one's work, taking care of things at home, or getting along with other people). One of four KAs (24.4%) felt impacts of depression, which was significantly lower than in the comparison group (33.6%) (unweighted z=−2.90, 95% CI −0.15 to −0.03, p=0.004), but similar to impacts of depression in the standard group (26.6%) (see table 2). Table 2 Comparison of prevalence of depression between the Korean Americans (≥35 years old) with uncontrolled type 2 diabetes (A1C ≥7.0%) at baseline in our clinical study, the comparison group (Americans with the same conditions as Korean Americans), and the standard group (Americans with A1C <5.7% and not being diagnosed of diabetes, pre-diabetes or type 1 diabetes) in the NHANES Depression\groups Korean Americans† (n=249) Comparison group‡ Standard group§ (n=6880) Total (n=1707) Whites (n=537) Blacks (n=574) Mexican (n=314) Hispanic (n=161) Others (n=121) Experienced at least once a week in the past 2 weeks (%):¶ Little interest or pleasure doing things 49.8*** 27.7 27.4 28.9 26.8 38.4 18.2 20.1 2. Feeling down, depressed, or hopeless 39.4*** 25.5 23.9 24.7 32.5 39.0 19.4 20.6 3. Trouble falling or staying asleep, or sleeping too much 59.0*** 40.9 42.9 36.2 35.0 44.1 44.0 38.6 4. Feeling tired or having little energy 70.3*** 53.9 58.2 47.2 45.8 49.8 54.5 46.5 5. Poor appetite or overeating 55.4*** 26.4 27.1 25.9 21.9 37.0 19.9 19.7 6. Feeling bad about yourself or that you are a failure or have let yourself or your family down 28.5*** 18.4 18.0 17.5 18.1 29.1 15.8 16.0 7. Trouble concentrating on things, such as reading the newspaper or watching television 29.3*** 15.9 15.8 15.8 12.6 26.9 12.6 15.6 8. Moving or speaking so slowly that other people could have noticed. Or the opposite being so fidgety or restless than usual 17.3 13.9 12.0 14.2 12.7 29.9 15.7 8.2 9. Thoughts that you would be better off dead, or of hurting yourself 16.9*** 5.0 4.2 5.5 4.7 13.1 3.8 2.8 10. Have these problems made it difficult for you to do your work, take care of things at home, or get along with other people? 24.4 33.6 33.2 30.0 33.3 35.1 46.1 26.6 PHQ-9, mean (SE) 5.1 (0.3)*** 3.7 (0.1) 3.6 (0.2) 3.7 (0.3) 3.6 (0.3) 5.0 (0.7) 3.1 (0.5) 2.7 (0.1) Status, χ2 statistics 23.6*** – – – – – – – Normal (PHQ-9<5) (%) 55.8 71.3 71.9 71.9 71.0 60.2 74.3 79.9 Mild (PHQ-9: 5–9) (%) 29.3 17.2 16.9 16.6 18.0 20.1 17.5 13.8 Clinical (PHQ-9 ≥10) (%) 14.9 11.5 11.2 11.5 11.0 19.7 8.2 6.3 ***p<0.001 for unweighted test statistics compared with the comparison group total. †At baseline. ‡N=11 767 732. §N=80 651 898. ¶Recoded as 0 if the score was 0, and 1 otherwise. NHANES, National Health and Nutrition Examination Survey; PHQ-9, Patient Health Questionnaire-9. Correlates of depression Ordered logistic regression analysis was performed to identify correlates of depression on KAs from our clinical trial, the comparison group, and the standard group. First, several demographic and physiological characteristics were correlated with the incidence of depression for KAs as well as other Americans, although the statistical significance varied across groups. In general, men, older people, and people with greater income and education were less likely to experience depression. For example, the effect on depression of being man was significantly lower in all groups: OR=0.67, 95% CI 0.47 to 0.95 in KAs; OR=0.62, 95% CI 0.39 to 0.97 in the comparison group; and OR=0.55, 95% CI 0.48 to 0.63 in the standard group. Depression decreased with age, but it was statistically significant in the standard group (OR=0.82, 95% CI 0.77 to 0.88). Second, people living alone experienced greater depression, which was consistent for all groups and statistically significant for the KAs and the comparison group. KAs living alone were almost three times more likely to be depressed (OR=2.92, 95% CI 1.65 to 5.18) than KAs who were living with someone. The effect of living alone on depression was also statistically significant in the standard group (OR=1.45, 95% CI 1.20 to 1.76). Also, oral medication and insulin for DM increased the likelihood of depression in KAs, but only oral medication was statistically significant (OR=1.49, 95% CI 1.07 to 2.09). DM medication was, however, inversely related to depression in the comparison group, although the relationship was not statistically significant. Third, the A1C level was significantly related to depression in the standard group (OR=1.45, 95% CI 1.20 to 1.76), and so was the high triglyceride level (OR=1.45, 95% CI 1.15 to 1.83). Other physiological indicators (high cholesterol and high blood pressure) were not significantly related to depression. Finally, self-efficacy for DM was correlated with depression. KAs in the middle and upper tertiles for self-efficacy were less likely to have depression (OR=0.57, 95% CI 0.40 to 0.82, and OR=0.45, 95% CI 0.30 to 0.66, respectively) than were KAs in the first tertile. Social support from family members and friends was significant, but the effect disappeared when the social support gap was introduced into the model. Those with a greater gap between expected and received support were one and a half times more likely to have depression (OR=1.54, 95% CI 1.06 to 2.23) (see table 3). Table 3 Comparison of people 35 years or older between Korean Americans (≥35 years old) with uncontrolled type 2 diabetes (A1C ≥7.0% at baseline) in our clinical study, and the comparison group (Americans with the same conditions as Korean Americans) and the standard group (Americans with A1C <5.7% and not being diagnosed of diabetes, pre-diabetes or type 1 diabetes) in the NHANES, using ordered logistic regressions on depression (normal=0, mild depression=1, and clinical depression=2) on listwise deletion of missing cases Group Korean Americans† (n=759) Comparison group‡ (n=1000) Standard group§ (n=6185) Indicators\model summary LR χ2 (20)=93.18 p<0.001 Pseudo R2=6.8% F(19, 45)=3.25 p<0.001 F(17, 48)=19.36 p<0.001 Sex (female=0, male=1) 0.67 (0.47, 0.95)* 0.62 (0.39, 0.97)* 0.55 (0.48, 0.63)*** Age (in 10s) 0.91 (0.76, 1.10) 0.91 (0.78, 1.06) 0.82 (0.77, 0.88)*** Income level (vs FPL <=138%)  139–250% 0.54 (0.37, 0.80)** 0.69 (0.41, 1.15) 0.56 (0.45, 0.71)***  251–400% 0.80 (0.52, 1.25) 0.69 (0.40, 1.20) 0.41 (0.33, 0.52)***  >400% 0.45 (0.28, 0.73)** 0.38 (0.22, 0.65)*** 0.27 (0.21, 0.36)*** Education level (vs 0–11th grade)  High school graduate 0.97 (0.59, 1.61) 0.77 (0.45, 1.34) 0.97 (0.77, 1.21)  College graduate or above 1.04 (0.62, 1.75) 0.70 (0.43, 1.17) 0.85 (0.68, 1.06) Living alone (yes=1) 2.92 (1.65, 5.18)*** 1.64 (0.99, 2.70) 1.45 (1.20, 1.76)*** Hemoglobin A1C (%) 1.02 (0.93, 1.13) 1.06 (0.93, 1.20) 0.76 (0.58, 1.00)* Triglycerides (vs <200 mg/dL)  Borderline (200–239 mg/dL) 1.01 (0.68, 1.49) 0.84 (0.49, 1.43) 1.17 (0.92, 1.49)  High (≥240 mg/dL) 1.34 (0.92, 1.97) 1.34 (0.88, 2.05) 1.45 (1.15, 1.83)** Obesity (vs BMI <25)  Overweight (BMI: 25–30) 0.95 (0.69, 1.30) 0.63 (0.31, 1.29) 0.95 (0.78, 1.16)  Obese (BMI >30) 0.66 (0.33, 1.32) 1.11 (0.55, 2.26) 1.27 (0.99, 1.63) DM oral medication (yes=1) 1.49 (1.07, 2.09)* 0.92 (0.65, 1.30) – DM insulin (yes=1) 1.72 (0.77, 3.82) 0.80 (0.53, 1.21) – Self-efficacy (1st tertile <34%)  Middle tertile (34–66%) 0.57 (0.40, 0.82)** – –  High tertile (67–100%) 0.45 (0.30, 0.66)*** – – Social support gap (1st tertile <34%)  Middle tertile (34–66%) 1.01 (0.70, 1.47) – –  High tertile (67–100%) 1.54 (1.06, 2.23)* – – *p<0.05, **p<0.01, ***p<0.001. †Pooled sample at baseline, months 3, 6, and 12. ‡N=7 017 657. §N=74 386 033; health insurance coverage, total cholesterol status, blood pressure control status, and social support were not reported due to non-significance. BMI, body mass index; DM, diabetes mellitus; FPL, federal poverty line; NHANES, National Health and Nutrition Examination Survey. Improvement of depression The mean depression score for KAs who completed the self-help DM intervention in the clinical trial (n=209) decreased over time from 5.3 (SE=0.4) at baseline to 4.7 (SE=0.3) at month 3; 4.6 (SE=0.3) at month 6; and 4.4 (SE=0.3) at 12 months. The changes from baseline were all statistically significant at p<0.05. Mean changes for the intervention group from baseline to months 3, 6, and 12 were −1.2 (SE=0.3), −0.8 (SE=0.4), and −1.3 (SE=0.4), respectively; all changes in the intervention group were statistically significant, but the changes in the control groups, which were −0.1 (SE=0.4), −0.7 (SE=0.4), and −0.5 (SE=0.5) for the same period, were not significant (table 4). Similarly, of the KAs who completed the intervention, the proportion with clinical depression decreased from 16.3% at baseline to 10.7%, 13.6%, and 8.1% at months 3, 6, and 12, respectively. These changes were statistically significant (χ2(6)=13.81, p=0.03) (table 5). Table 4 Depression score changes over time in the intervention and control groups in the clinical study (completer only, n=209) Group Baseline Month 3 Month 6 Month 12 Intervention, mean (SE) 5.4 (0.5) 4.2 (0.4) 4.6 (0.5) 4.1 (0.3) Change from baseline – 22121.2 (0.3)*** −0.8 (0.4)* −1.3 (0.4)** Control 5.3 (0.5) 5.2 (0.5) 4.6 (0.4) 4.8 (0.5) Change from baseline −0.1 (0.4) −0.7 (0.4) −0.5 (0.5) Total 5.3 (0.4) 4.7 (0.3) 4.6 (0.3) 4.4 (0.3) Change from baseline −0.6 (0.3)** −0.8 (0.3)** −0.9 (0.3)** *p<0.05, **p<0.01, ***p<0.001. Table 5 Change in distribution of depression over time in the clinical study (completer only, n=209)* Depression level Baseline (%) Month 3 (%) Month 6 (%) Month 12 (%) Total (%) Intervention group  Normal 53.3 61.9 68.6 68.6 63.1  Mild depression 28.6 29.5 21.8 22.9 24.5  Clinical depression 18.1 8.6 13.6 8.6 12.4 Control group  Normal 53.9 51.0 60.4 52.9 54.5  Mild depression 31.7 36.0 26.7 39.4 33.5  Clinical depression 14.4 13.0 12.8 7.7 12.0 Total  Normal 53.6 56.6 64.6 60.8 58.9  Mild depression 30.1 32.7 21.8 31.1 30.0  Clinical depression 16.3 10.7 13.6 8.1 12.2 *Pearson χ2 (6)=13.81, p=0.03. Discussion The findings of this research confirm our hypothesis that KA adults with DM would report a much higher rate of depressive symptoms (44.2%) than would Americans with DM (28.7%) or without DM (20.1%). It was clear that the net effect of diabetes on depression was significant across all groups. For example, the prevalence of mild depression among Americans with DM was 3.4% higher than in their counterparts without DM, and the difference was statistically significant (z=3.57, 95% CI 0.014 to 0.054, p<0.001). The prevalence of clinical depression among Americans with DM (11.5%) was almost twice that of their counterparts without DM (6.3%), and the difference was also significant (z=7.38, 95% CI 0.036 to 0.068, p<0.001) (table 2). A similar pattern was observed among KAs, although it was not feasible to directly deduce the net effect of diabetes on depression (note: no causality assumed) among KAs. Nevertheless, it is reasonable to assume that the effect was much greater among KAs than among Americans with DM. Considering the similarities in diabetes-related biomarkers (eg, A1C and total cholesterol) between KAs and Americans with DM, the net effect of diabetes on depression would have been same for both groups. Then the differences between the two groups in mild (12.1%; z=4.59, 95% CI 0.062 to 0.180, p<0.001) and clinical (3.4%; z=1.55, 95% CI −0.013 to 0.081, p=0.121) depression can be attributed to personal, social, and environmental factors, some of which are modifiable. In fact, KAs' depression prevalence (measured by the PHQ-9) was higher than any other previous depression estimations among KAs including those in our own studies with KA elderly (≥60 years old, 30.3%)13 and South Korean elderly (27.8%)21 as well as in a comparative epidemiological study of Korean, Japanese, and Chinese American adults (33.3%, 15.7%, and 20.4%, respectively).4 The high prevalence of depression among KAs with DM is alarming. These comorbid conditions impose substantial management burdens for this vulnerable group, who lack culturally and linguistically relevant interventions necessitated by language barriers and low health literacy. To make matters worse, KAs demonstrated significantly lower access to care than did the general population; nearly half of the participants (48%) reported that they had no medical insurance, whereas the uninsured rates for the comparison and standard groups were ∼15%. Given that almost a half (44.2%) of participants with uncontrolled DM (A1C ≥7%) had comorbid depression, this access gap highlights the magnitude of barriers to obtaining an optimal care among KAs with DM. Inadequate usage of healthcare services, including a lack of, or insufficient access to, care among KAs, is well documented.22 23 One explanation for this is that a large proportion of KAs work in or own small businesses that cannot afford medical insurance. Also, some recent KA immigrants over age 65 cannot enroll in Medicare because they lack an adequate employment history. Several researchers have argued that KAs and other Asian American immigrants are susceptible to depression because of their social and linguistic isolation and because of the insufficient access to care suffered by first-generation immigrants.13 24 Our findings show that those who have expected more social support than they ultimately receive are more likely to report depressive symptoms. A unique contextual factor associated with the high prevalence of depression among new immigrants may stem from an incongruity between educational attainment and both income and occupational prestige. The higher educational achievement among the first-generation KAs does not translate into professional jobs or high incomes in the USA. The incongruity between high aspirations before immigration and actual attainment after immigration quantified using socioeconomic status has been identified as a risk factor for poor mental health.25 The dissonance between expected and attained socioeconomic status can further exacerbate KAs' mental health via insufficient English skills and low health literacy. Many KAs have described their struggles in communicating with medical providers as well as a lack of access to health information in their mother tongue.26 In some KA households, children act as translators or brokers between their immigrant parents and healthcare providers as well as the mainstream culture. The reversed roles between parents and children can pose mental health challenges for some KAs because they compromise parents' cultural positions or status as ‘heads of household’ or ‘village leaders’, roles that are traditionally assured and reserved for parents in Korean and other Asian cultures.13 Thus, contextual factors related to acculturation and subsequent acculturative stress may explain the gaps in the prevalence of depression between KAs and other groups. In addition, unlike in other population literature, ‘marriage’, ‘living with family member’, and ‘high education level’ are not always good proxies for social support or personal resources that mediate stress levels. Regardless of the level of social support received from family members, most KAs feel that they are linguistically isolated from the larger mainstream community. Despite the high education level that many have obtained in their motherland, most KA immigrants suffer from high levels of English insufficiency and low health literacy, which also lead to inadequate healthcare access and information. Our findings are also consistent with previous research showing significant correlates of depressive symptoms: being woman, younger, in a lower income level, less educated, and living alone emerged as statistically significant correlates in all three groups (KAs with DM, Americans with DM, Americans without DM). Unique to the KAs with DM, taking DM medication was a significant correlate of depression. However, we are not certain that there is an etiological link between taking DM medication and depressive symptoms; it is plausible that taking medication might be a proxy for the severity of DM, because many KAs do not seek DM treatment until after their disease is advanced, owing to their limited access to care. The improvement in depressive symptoms over time differed between the intervention group, which received a self-help behavioral intervention, and the control group, which participated in data collection only. Although both groups improved in alleviating depressive symptoms during the project period (3, 6, and 12 months after the initial behavioral intervention), the intervention group's improvement was significant (>20% reduction in depressive symptoms). The intervention group and the control group showed statistically and clinically significant differences in glucose control: the intervention group achieved >1.5% sustained hemoglobin A1C reduction over 12 months after the initial intervention. The premise of the self-help intervention was to empower these individuals by cultivating their self-efficacy and their capabilities to manage their DM condition.14 15 Although the intervention was not specifically designed to improve mental health, the participants' improved problem-solving skills and self-confidence may yet have had a therapeutic effect. Further, our study demonstrates that depressive symptoms can be mitigated by enhancing one's coping skills, including self-efficacy, one's level of health literacy, and glucose control. To the best of our knowledge, our clinical trial was the first to target this understudied population by aiming to improve physical and mental health among KAs with uncontrolled DM for 12 months. Our findings demonstrate that a culturally tailored, community-based self-help intervention was effective for managing DM and depression in an ethnic/linguistic minority community. Although the sample was limited to one ethnic minority group, the study provides unique insights into the relationship between the most common comorbid conditions, DM and depression, as well as potential strategies to tackle such clinical challenges for similarly vulnerable populations. Limitations This study was carried out at a single site centered in the Baltimore–Washington area, targeting KAs only. We overcame this limitation by comparative analysis with a national sample that used the same measurements. Another inherent limitation is that the study's data were obtained from a DM intervention trial and did not include data from KAs without DM. Future studies should include such a comparison group to validate our findings. The authors are grateful for substantial editorial support with manuscript development provided by Dr John Bellquist at the Cain Center for Nursing Research and the Center for Transdisciplinary Collaborative Research in Self-Management Science (P30, NR015335) at The University of Texas at Austin School of Nursing. Contributors: MTK and KBK had full access to the data and drafted the manuscript. JK researched data and YJ, HBL, and DL discussed/edited the manuscript. Funding: This study was supported by a grant from the National Institute of Diabetes and Digestive and Kidney Diseases (R18DK083936) and with material support from LifeScan, including devices (OneTouch glucometers, OneTouch UltraSoft test strips, and OneTouch UltraSoft Lancets) for study participants. Also, the Johns Hopkins Institute for Clinical and Translational Research supported the cost of blood serum laboratory tests. Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or other supporters. Competing interests: None declared. Patient consent: Obtained. Ethics approval: The study received ethics approval by Johns Hopkins Medicine IRB. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: We shall make data available to the scientific community as few restrictions as feasible. However, because of the small sample size from a single ethnic community and some qualitative data (in Korean), PIs will maintain the electronic data file by removing all identifiable information. Data include demographic information, diabetes control status, and other standard blood serum test results. Further information can be obtained from the corresponding author. ==== Refs References 1 Anderson RJ , Freedland KE , Clouse RE The prevalence of comorbid depression in adults with diabetes: a meta-analysis . Diabetes Care 2001 ;24 :1069 –78 . 10.2337/diacare.24.6.1069 11375373 2 Bogner HR , Morales KH , Post EP Diabetes, depression, and death: a randomized controlled trial of a depression treatment program for older adults based in primary care (PROSPECT) . Diabetes Care 2007 ;30 :3005 –10 . 10.2337/dc07-0974 17717284 3 Holt RIG , de Groot M , Golden SH Diabetes and depression . Curr Diab Rep 2014 ;14 :491 .24743941 4 Kim HJ , Park E , Storr CL Depression among Asian-American adults in the community: systematic review and meta-analysis . PLoS ONE 2015 ;10 :e0127760 10.1371/journal.pone.0127760 26029911 5 Holt RIG , Katon WJ Dialogue on diabetes and depression: dealing with the double burden of co-morbidity . J Affect Disord 2012 ;142 :S1 –3 . 10.1016/S0165-0327(12)00632-5 23062852 6 Ell K , Aranda MP , Xie B Collaborative depression treatment in older and younger adults with physical illness: pooled comparative analysis of three randomized clinical trials . Am J Geriatr Psychiatry 2010 ;18 :520 –30 . 10.1097/JGP.0b013e3181cc0350 20220588 7 Ell K , Katon W , Lee P Depressive symptom deterioration among predominantly Hispanic diabetes patients in safety net care . Psychosomatics 2012 ;53 :347 –55 . 10.1016/j.psym.2011.12.009 22458987 8 Fuentes D , Aranda MP Depression interventions among racial and ethnic minority older adults: a systematic review across 20 years . Am J Geriatr Psychiatry 2012 ;20 :915 –31 . 10.1097/JGP.0b013e31825d091a 22828202 9 Ockene IS , Tellez TL , Rosal MC Outcomes of a Latino community-based intervention for the prevention of diabetes: the Lawrence Latino Diabetes Prevention Project . Am J Public Health 2012 ;102 :336 –42 . 10.2105/AJPH.2011.300357 22390448 10 Bernstein KS , Park S , Shin J Acculturation, discrimination and depressive symptoms among Korean immigrants in New York City . Community Ment Health J 2011 ;47 :24 –34 . 10.1007/s10597-009-9261-0 19888652 11 Park SY , Cho S , Park Y Factors associated with mental health service utilization among Korean American immigrants . Community Ment Health J 2013 ;49 :765 –73 . 10.1007/s10597-013-9604-8 23417654 12 Pistulka GM , Winch PJ , Park H Maintaining an outward image: a Korean immigrant's life with type 2 diabetes mellitus and hypertension . Qual Health Res 2012 ;22 :825 –34 . 10.1177/1049732312438778 22378835 13 Kim MT , Kim KB , Han HR Prevalence and predictors of depression in Korean American elderly: findings from the memory and aging study of Koreans (MASK) . Am J Geriatr Psychiatry 2015 ;23 :671 –83 . 10.1016/j.jagp.2014.11.003 25554484 14 Kim KB , Kim MT , Lee HB Community health workers versus nurses as counselors or case managers in a self-help diabetes management program . Am J Public Health 2016 ;106 :1052 –8 . 10.2105/AJPH.2016.303054 26985607 15 Kim MT , Kim KB , Huh B The effect of a community-based self-help intervention: Korean Americans with type 2 diabetes . Am J Prev Med 2015 ;49 :726 –37 . 10.1016/j.amepre.2015.04.033 26184986 16 Han C , Jo SA , Kwak JH Validation of the Patient Health Questionnaire-9 Korean version in the elderly population: the Ansan Geriatric Study . Compr Psychiatry 2008 ;49 :218 –23 . 10.1016/j.comppsych.2007.08.006 18243897 17 Shin J , Park S , Cho S Validation of Patient Health Questionnaire-9 Korean Version (PHQ-9K) Scale for screening depression among Korean Americans in community settings . J Theory Constr Test 2010 ;14 :45 –51 . 18 Lorig KR , Sobel DS , Ritter PL Effect of a self-management program on patients with chronic disease . Eff Clin Pract 2001 ;4 :256 –62 .11769298 19 Han H , Kim MT , Weinert C The psychometric evaluation of Korean translation of the Personal Resource Questionnaire 85-Part 2 . Nurs Res 2002 ;51 :309 –16 . 10.1097/00006199-200209000-00007 12352779 20 Weinert C A social support measure: PRQ85 . Nurs Res 1987 ;36 :273 –7 . 10.1097/00006199-198709000-00007 3306610 21 Park JH , Kim KW , Kim MH A nationwide survey on the prevalence and risk factors of late life depression in South Korea . J Affect Disord 2012 ;138 :34 –40 . 10.1016/j.jad.2011.12.038 22284016 22 Han H , Kim J , Kim MT Measuring health literacy among immigrants with a phonetic primary language: a case of Korean American women . J Immig Minor Health 2011 ;13 :253 –9 . 10.1007/s10903-010-9366-0 23 Kandula NR , Kersey M , Lurie N Assuring the health of immigrants: what the leading health indicators tell us . Annu Rev Public Health 2004 ;25 :357 –76 . 10.1146/annurev.publhealth.25.101802.123107 15015925 24 Mitra M , Iezzoni LI , Zhang J Prevalence and risk factors for postpartum depression symptoms among women with disabilities . Matern Child Health J 2015 ;19 :362 –72 . 10.1007/s10995-014-1518-8 24889114 25 Sorensen MV , Snodgrass JJ , Leonard WR Lifestyle incongruity, stress and immune function in indigenous Siberians: the health impacts of rapid social and economic change . Am J Phys Anthropol 2009 ;138 :62 –9 . 10.1002/ajpa.20899 18711739 26 Han H , Kang J , Kim KB Barriers to and strategies for recruiting Korean Americans for community-partnered health promotion research . J Immigr Minor Health 2007 ;9 :137 –46 . 10.1007/s10903-006-9022-x 17186370

PMC005xxxxxx/PMC5372033.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01296410.1136/bmjopen-2016-012964SurgeryProtocol1506173716941722Constraint choice in revision knee arthroplasty: study protocol of a randomised controlled trial assessing the effect of level of constraint on postoperative outcome Hommel Hagen 12Wilke Kai 1Kunze Daniel 1Hommel Peggy 1http://orcid.org/0000-0002-3549-398XFennema Peter 3 1 Krankenhaus Märkisch Oderland GmbH BT Wriezen, Klinik für Orthopädie, Sportmedizin und Rehabilitation, Wriezen, Germany 2 Lehrkrankenhaus der Medizinischen Hochschule Brandenburg (MHB) Theodor Fontane, Neuruppin, Germany 3 AMR Advanced Medical Research GmbH, Männedorf, SwitzerlandCorrespondence to Priv.-Doz. Dr. med. Hagen Hommel; H.Hommel@khmol.de2017 27 3 2017 7 3 e0129647 6 2016 4 1 2017 14 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction The proper management of total knee arthroplasty (TKA) in patients with severe deformities regarding the preferable prosthetic design and the required amount of constraint is a controversial subject. In the absence of any high-level clinical evidence, we designed a randomised clinical trial to investigate if rotating hinged (RTH) and constrained condylar knee (CCK) designs yield similar outcomes. Methods and analysis This study is a multicentre, randomised clinical trial including two groups of 85 patients. Patients will be randomised to a CCK knee design group or an RTH knee design group. Patients will be followed for 2 years. The study will be designed as an equivalence trial. The primary study outcome will be the postoperative functional outcome as measured by the self-administered Knee Injury and Osteoarthritis Outcome Score. Secondary outcomes will be postoperative joint awareness during various activities of daily living as measured by the Forgotten Joint Score-12, the Knee Society Score, along with the incidence and location of radiolucent lines using the Knee Society TKA radiographic evaluation system. Ethics and dissemination This study is approved by the ethics committee of the Landesärztekammer Brandenburg ((S 10(a)/2013) from 27.08.2013, amended on 25.04.2016) and will be conducted according to the principles of the World Medical Association Declaration of Helsinki and the ISO14155:2011. Trial registration number DRKS00010539. Osteoarthritis, kneearthroplasty, knee replacementtrials, randomized clinicalclinical equivalency ==== Body Strengths and limitations of this study This is the first randomised clinical study to assess differences in outcome between rotating hinged and condylar constraint knee arthroplasty. More insight will be gained into the problem of implant constraint, with patient reported outcome measures focusing on patient's relevancy. Limitation is the heterogeneity of the study population in terms of indication for surgery. Introduction The global increase in the incidence of total knee arthroplasty (TKA) has led to a concurrent rise in the incidence of complex primary and revision arthroplasty (RTKA) procedures. One study estimated the overall incidence of primary TKA and RTKA to grow by 174% and 600%, respectively, between 2005 and 2030.1 One of the major challenges of TKA is the management of instability, a factor which is a prerequisite for postoperative function as well as implant survival.2 Overcoming ligamentous instability requires a challenging combination of managing soft tissue deficiency, balancing the flexion and extension gaps and managing extensor mechanism insufficiency.3 The choice of implant constraint depends on the state of the ligaments of the knee and the severity of bone loss.2 The best pairing between bony defects during RTKA and the level of implant constraint required is a controversial topic. Posterostabilised designs may be appropriate when the ligaments of the knee are intact and bone defects are minimal.2 4 Hinged knee prostheses with a fixed axis were introduced to restore knee function and correct limb alignment in the presence of severe malalignment and/or instability.5 Disappointing long-term outcomes with low survival rates prompted a further refinement of the design of these prostheses.6 The rotating hinged (RTH) total knee prosthesis was introduced in 1982,7 and is typically used in cases of ligament absence or ligament disruption in combination with moderate or severe bone loss.2 8 A potential advantage of this knee design is the reduction of shear stresses around the bone–cement interface due to the additional rotation around the tibial axis.9 10 Studies reporting outcome of different types of RTH implants have been published; some had controversial conclusions.8 11–15 Condylar constrained knee (CCK) prostheses are uncoupled, non-hinged and semiconstrained implants. In patients where gross joint instability is not an issue, CCK prostheses represent an alternative to the RTH prostheses. CCK systems have been introduced more recently, and their use has become increasingly popular in patients with intermediate—but not complete—insufficiency of the ligaments and moderate bone loss.2 The preferable prosthetic design and required amount of constraint is still a controversial area.16 17 A significant shortcoming of constraint in TKA stems from the major increase of force transfer to the bone–implant interface. CCK prostheses are generally believed to be less constrained than RTH designs. CCK designs provide mediolateral and rotational constraint but no anteroposterior constraint. It should be taken into account that RTH prostheses offer more freedom in the axial plane than CCK designs.3 Previous studies have reported that RTH prostheses may, in contrast to common beliefs, be less constrained than CCK designs. Finite element models have shown that RTH designs are prone to less shear stress at the bone–implant interface when compared with CCK designs.18 However, there is a paucity of studies focussing on the clinical relevance of this difference in shear stress. Observational studies have shown that when implant selection is guided by inherent stability of the knee, CCK and RTH knee designs have similar outcomes.3 Fuchs et al19 found no significant differences in terms of Hospital for Special Surgery Score, Knee Society Score (KSS), Pain and the Tegner Score in their retrospective cohort study. Standard condylar revision implants yielded better postoperative mobility but had a lower Short Form 36 Mental Component Score. The latter finding may imply that patients tolerate the RTH design better than the standard condylar revision implant. Walker et al20 also compared the RTH design with standard condylar revision in a cohort study and found a high correlation in performance between the operated and non-operated side in the RTH group, indicating the hinges were capable of matching the non-operated knee performance. In our clinic, we studied 74 consecutive patients receiving a CCK design during primary or revision TKA from 2007 to 2013.21 This patient group was compared with a historical cohort of patients (n=93) receiving an RTH prosthesis as a primary or revision implant from 2003 to 2007. Table 1 summarises the baseline characteristics of the patient populations. Table 1 Baseline characteristics of the study population21 CCK RTH p Value Females:males* 42:32 69:33 0.307† Age (years)‡ 69.8±9.0 72.1±4.3 0.042§ BMI (m/kg2)‡ 30.0±3.1 31.1±3.4 0.028§ ASA* 18/14/12 14/57/22 0.224† HKA (varus/valgus)* 55:19 66:27 0.630† Baseline scores  KS‡ 24±4 26±4 0.031§  FS‡ 20±5 22±6 0.001§  WOMAC‡ 65±6 65±6 0.741§ *Presented as number of observations. †Pearson's chi-squared test. ‡Presented as mean±SD. §Student's t-test. ASA, American Society of Anesthesiologists Score; BMI, body mass index; CCK, constrained condylar knee; FS, Function Score; KS, Knee Score; RTH, rotating hinged; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index. The differences between follow-up and baseline values between the cohorts were compared (difference in difference (DID)). DIDs between CCK and RTH arthroplasty did not deviate significantly for either of the subscores of the KSS (the Function Score (FS) and the Knee Score (KS)) or for the Western Ontario and McMaster Universities Arthritis Index (WOMAC). Subgroup analysis of patients with valgus and varus osteoarthritis at the primary intervention yielded clinically equivalent results for the varus group, but significantly better KSSs were obtained in patients with an initial valgus deformity who had received an RTH prosthesis (table 2). Table 2 ‘Difference in difference’ in clinical outcome between CCK and RTH21 CCK RTH p Value Follow-up time (months) 54±25 106±33 <0.001 KS 63±6 63±4 0.712  Varus 65±5 63±4 0.039  Valgus 57±4 63±5 <0.001 FS 64±8 64±7 0.995  Varus 66±8 64±7 0.082  Valgus 59±6 65±8 0.025 WOMAC −40±7 −39±8 0.510  Varus −41±7 −39±7 0.177  Valgus −37±7 −39±8 0.351 Presented as mean±SD. §, Student's t-test. CCK, constrained condylar knee; FS, Function Score; KS, Knee Score; RTH, rotating hinged; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index. None of the studies were randomised, and confounding-by-indication is likely to have biased the conclusions of the aforementioned trials. In the absence of any high-level clinical evidence, we therefore designed a randomised clinical trial to investigate if RTH and CCK designs yield similar outcomes. Methods and analysis Based on the findings from observational studies, we designed the present randomised trial to verify the impact of implant constraint on the outcome of revision TKA. With this study, we plan to assess if revision knee arthroplasty using an RTH knee design and a standard condylar knee design have equivalent functional outcomes. Secondary objectives are to demonstrate that, compared with a CCK design, RTH knee design leads to similar clinical and radiographic outcomes, and has a similar complication rate. In addition, the ability to forget the knee prosthesis during activities of daily life will be compared between the two populations. This study is a multicentre, randomised clinical trial. Patients will be randomised before the surgical procedure to receive a CCK knee design (Legion, Smith and Nephew GmbH, Hamburg, Germany) or an RTH knee design (RT-Plus Solution, Smith and Nephew GmbH). Study participants will be allocated to one of the two study arms in a 1:1 ratio. Patients will be followed for two years after inclusion. This study was designed as an investigator-initiated trial, and the institution of the principal investigator (HH) will take the role of the sponsor. The study will be performed at four community clinics in Germany. Only clinics that perform at least 50 total knee revisions per year will be eligible. The current study will be designed as an equivalence trial: the study was designed to show that the two interventions do not differ in either direction by more than a prespecified unimportant or insignificant amount.22 All medical devices are used in the routine manner as specified in each product's instructions for use. Only devices with prior market authorisation will be applied in the study. The study protocol was drafted following the SPIRIT statement23 and the CONSORT extension for reporting noninferiority and equivalence trials.24 Outcomes The following characteristics will be retrieved from patient questionnaires, physical examination, the hospital information system or medical records: patient characteristics (ie, sex, age, height and weight); indication for surgery; American Society of Anesthesiologists classification; previous joint procedures or injury; present degree of deformity and degree of deformity before the index procedure; comorbidities; smoking and alcohol consumption and pain medication consumption. We will also note arthroplasty-related characteristics such as anaesthesia method, type of arthroplasty, surgical approach, postoperative analgesic consumption and arthroplasty-related complications. The primary outcome of the study will be the postoperative functional outcome as measured by the self-administered Knee Injury and Osteoarthritis Outcome Score (KOOS).25 26 The KOOS was developed in the 1990s to assess patients' opinions on the state of their knees after injury or surgery.26 27 The score also has sufficient sensitivity in the younger and more active patients with knee problems. In addition to questions about pain and stiffness, it examines knee function in daily life, sport and recreational activities and the quality of life. Secondary outcomes of this study are: Joint awareness during various activities of daily living, as measured by the Forgotten Joint Score-12 (FJS-12).28 The FJS-12 is a recently published patient-reported outcome measure to assess joint awareness in hips and knees during various activities of daily living.28 It consists of 12 questions.29 The FJS-12 has a high internal consistency (Cronbach's α 0.95), a low ceiling effect and the instrument discriminates well between good, very good and excellent outcomes after total hip arthroplasty (THA) and TKA.29 Patients' knees and functional abilities such as walking and stair climbing before and after TKA as assessed via the KSS, a simple, objective, globally recognised scoring system30 to track and report outcomes after total and partial knee arthroplasty. Conventional anteroposterior and lateral radiographs will be taken at designated time points. Postoperative radiographs will assess component position, alignment and the location and incidence of radiolucent lines using the Knee Society TKA radiographic evaluation system.31 The incidence of adverse events (AEs) and adverse device effects. Study population For a patient to qualify for study enrolment, the CCK and the RTH design should be indicated. If one of the two designs prevails in the physician's reasonable medical judgement, the patient should receive the treatment deemed most suitable by the patient's physician. Study participants will be recruited from eligible patients at the orthopaedic department of the Krankenhaus Märkisch Oderland GmbH BT, Wriezen, Germany scheduled for knee arthroplasty, and from three other clinical sites that have yet to be recruited. Eligible patients must require primary or revision surgery and must have an axial malalignment >18°, with one of the collateral ligaments (ie, medial collateral and/or lateral collateral ligament) being absent or incompetent.15 Indications are:32 rheumatoid arthritis; post-traumatic arthritis, osteoarthritis or degenerative arthritis in older patients whose age, weight and activity levels are compatible with an adequate long-term result; failed osteotomies, unicompartmental replacement or total knee replacement. In addition, the patient has to fulfil the following criteria to be eligible for the study: The patient is scheduled for elective primary or revision knee arthroplasty. The patient is 18 years or older. There are no upper age limits for the study, as long as the patient is healthy enough to undergo surgery. The patient is willing to provide informed consent. The following exclusion criteria shall be used for the study: genu recurvatum; gross anteroposterior instability; the patient participates in another concurrent study; major psychiatric disease; patients who do not comprehend the German language (read and speak); pregnancy, lactation or childbearing potential without using adequate contraception; acute or chronic infections, local or systemic (or corresponding previous incidents); severe muscle, nerve or vascular diseases that endanger the extremity in question; lacking bone substance or possessing inadequate bone quality that endangers a stable seating of the prosthesis; extreme insufficiency of the knee extensor mechanism as this can lead to excessive joint distortion; adipositas permagna; local tumours; known hypersensitivity to the material. Devices The Legion Revision System (Smith and Nephew, Memphis, Tennessee, USA) was introduced in Europe in 2006, and good postoperative outcome has been reported with the system.33 The system is available with a variety of wedges and femoral and tibial stems with and without offset. It is indicated for use in total knee replacement to overcome severe ligamentous imbalance, laxity or in the presence of deformity in excess of 20°. The RTH RT-Plus (Smith and Nephew) was introduced in Europe in 1997, and we previously reported good mid-term to long-term outcome with this device.15 The system is available with a variety of wedges and femoral and tibial stems. For the study, the device will be used within its indications for use while excluding severe ligamentous imbalances, ligamentous laxities or gross deformities in excess of 20°. Study withdrawal In accordance with the Declaration of Helsinki and with local regulations, study subjects have the right to withdraw at any time from the study without providing a reason. The investigator or regulatory authority can discontinue a subject's participation in the study at any time if medically or otherwise necessary. Unless the subject wishes to withdraw from the study completely, all scheduled examinations will be performed as planned. Study hypothesis The current study will be designed as an equivalence trial, that is, a trial designed to show that the two interventions do not differ in either direction by more than a prespecified unimportant or insignificant amount.34 The amount of permissible difference is the margin that delineates the ‘zone of indifference’. Within this zone of indifference, the two groups are considered equivalent.35 The KOOS has a minimally clinically important difference (MCID) of 8 to 10 points, and the SD in a cohort of TKA patients is ∼15 points.26 The null hypothesis (of non-equivalence) of the study is that functional outcome at 2 years of follow-up in the RTH group is not equal to the functional outcome at 2 years of follow-up in the CCK group, or: versus the equivalence assumption that functional outcome at 2 years of follow-up in the RTH group is equivalent to the functional outcome at 2 years of follow-up in the CCK group, or: Sample size calculation With α=0.05 and β=0.10, a sample size between 49 (MCID=10) and 77 (MCID=8) will be required. Calculation is based on the following formula:36 with: Attrition and dropout rate at the 2-year follow-up is estimated to be 10%. To account for attrition and drop-out, a total of 2×85 patients will be enrolled. Randomisation Patients will be randomly allocated to the treatment or control group in a 1:1 ratio, using block randomisation with random block sizes. Stratification based on study site will be performed. Under all circumstances, subjects will not be randomised before they have provided informed consent. Randomisation in advance of surgery (by at least 24 hours) is required in order to allow proper preoperative planning of the surgery, alongside preparation of the instrument sets. The investigator will randomise the patient through the internet (http://www.randomisation.com) and document the treatment group in the patient's chart. Blinding Surgeon and patients will not be blinded for the assigned treatment. In order to preserve the blinding of the study, a minimum number of examiners will have access to the randomisation schedule before the study is complete. To reduce measurement bias, the preoperative and follow-up assessments will be performed by independent examiners blinded to group allocation (one examiner plus one back up at each centre, both trained and neither involved in surgery). Before each follow-up visit, subjects will be reminded not to disclose their allocation. One independent single examiner will perform the radiographic assessments. Blinding of this examiner will not be possible. One interim analysis is planned to investigate the safety and the primary outcome after half of the patients have been enrolled. The statistician will provide a summary of the AEs and an analysis of the primary outcome. Study groups will be blinded; the group assignments will not be revealed. These results will be presented to the Clinical Study Steering Committee (CSSC). Based on the interim analysis, the CSSC will then decide whether to continue the study without adjustment, continue the study with adjustment(s) or to stop the study due to safety or efficacy concerns. The CSSC may request to reveal study arm assignments before coming to a definitive decision. Early termination of the study for efficacy at the interim analysis will be considered if the benefit of either one of the study groups is shown ‘beyond reasonable doubt.’ A nominal p value<0.003 (O'Brien Fleming α spending37) provides guidance in this instance but is not binding for this decision. As the products used in both study arms are CE-marked and are present in the market for <10 years, no formal criteria for early stopping for safety reasons have been defined. The CSSC's decision will be based on the statistical analysis and the clinical expertise of the board members. The interim results will be confidential and strictly limited to the members of the CSSC. This confidentiality minimises the risk of interim-result bias on the remaining course of the trial. All analyses will be performed primarily as an intention-to-treat-analysis, assessing all patients with available data for the outcomes according to the randomisation. For baseline characteristics, descriptive statistics will be used as appropriate. The primary hypothesis will be tested using the Westlake version of an equivalence test with known and fixed delta.38 Further exploration of the primary and secondary outcomes will be based on linear mixed models.39 Inferences on the random effect structures will be based on the restricted maximum likelihood and inferences for the fixed effect structure will be based on the standard maximum likelihood. With the exception of the interim analysis of the primary outcome, two-sided significance tests with an α of 0.05 will be applied. Stata 12.1 (StataCorp, College Station, Texas, USA) will be used for the analysis. Study procedures Patients that are screened and deemed ineligible prior to study inclusion will be tracked on a screening log along with a detailed reason for exclusion. Patients who are deemed eligible but who are not included in the study will be tracked on a screening log along with the reason for exclusion. Patients that are enrolled and are deemed ineligible during surgery will be tracked on a dedicated Termination Case Report Form along with the reason for exclusion. Informed consent Patients will be informed verbally by the investigator about the study objectives, design, risks, the study procedures and patient rights. Only patients who have signed the informed consent form 24 hours before surgery and meet all inclusion criteria and no exclusion criteria will be included in the trial. Patients will receive a copy of the signed informed consent form. Study visit definitions The screening date is defined as the date the informed consent is signed. The enrolment date is defined as the date of randomisation. Day 1 is defined as the surgery day. All on-study visits will be calculated from the initial day-1 visit. However, if a subject's visit is delayed, the subsequent visit date will be shifted. Study visit windows will be ±7 days for all the visits of the treatment and observation period and ±14 days for follow-up visits at the 1-year and 2-year follow-ups. Study procedures for a specific visit may be completed on multiple days as long as all the procedures are completed within the visit window. The subject's medical history will be obtained prior to enrolment and recorded in an electronic data capturing system. Table 3 shows all the assessments and procedures performed during each visit. Table 3 Schematic timeline Study period Enrolment Allocation Postallocation Close-out Time point Baseline Peri-op 3 months 6 months 1 year 2 years Enrolment: Eligibility screen X Informed consent X Allocation X Surgery X Assessments: PT characteristics X KOOS X X X X X KSS X X X X X FJS X X X X Radiography X X X X Adverse events X X X X X FJS, Forgotten Joint Score; KOOS, Knee Injury and Osteoarthritis Outcome Score; KSS, Knee Society Score; PT, patient. Concomitant care It is imperative that all study participants receive the best available care. This may lead to differences in concomitant care between the study groups. All concomitant medications and treatments will therefore be reported on a concomitant care case report form. Dissemination This clinical study will be conducted in accordance with the latest version of the World Medical Association Declaration of Helsinki, ISO14155, and German regulations. The study is registered at the German Clinical Trials Register (trial registration number DRKS00010539). The ethics committee will be notified should future amendments (ie, modifications that are likely to affect the safety or the scientific value of the trial) arise. The results of the study will be published in international peer-reviewed scientific journals, independent from the outcome. Anonymous patient data will be made available on request of the journal. We will provide a copy of the protocol, including the statistical analysis plan, to the medical journal considering a submitted manuscript for publication, if requested by the journal. Authorship The International Committee of Medical Journal Editors (ICMJE) recommendations regarding authorship will be followed. The investigators will identify the individuals who accept direct responsibility for the publication. All authors must take responsibility for the manuscript. Publications from individual institutions participating in the study will not precede the primary manuscript. An external professional medical or scientific writer may be employed to assist in producing publications. Names, involvement and funding of the external writer will be acknowledged in the publication. We will not engage ghostwriters. Study organisation No data and safety monitoring board will be installed, as the safety and efficacy of the study devices have been documented. A CSSC has been installed consisting of the principal investigators of the four study sites and a statistician. The CSSC has the primary responsibility for designing the study, maintaining the quality of study conduct, the ongoing monitoring of safety and efficacy and writing the study publications. The committee will also assess whether the endpoints meet the criteria of equivalence. Except for the statistician, the committee will be masked to the assigned study arm when performing their assessments. Data collection and monitoring Patients will be hospitalised during the whole study (due to the RTKA procedure). Study participants will therefore be under steady monitoring by surgeons and regular hospital staff. Dedicated study physicians and study nurses will collect data from the patients' records. Study nurses will schedule patient follow-ups to the clinic and will oversee the physical and the radiographic assessments. Study nurses will also provide and collect the questionnaires from the patients and will provide assistance with completing the questionnaires if requested by the patient. The clinical data management system used for the study will be the 21 CFR Part 11 and GCP compliant, CDISC-certified electronic data capture system Marvin (XClinical GmbH, Munich, Germany). Patients will be coded and patient-identifiable information will not be stored in the clinical data management system. Data entry comprises extensive data edit checks. The software will dynamically check and query illegal ranges and inconsistent entries. Data entry is recorded in an audit trail, and backup to a remote server is performed several times per day. The sponsor has appointed an independent contract research organisation (CRO) to perform site/clinical monitoring of the study to assure high-quality trial conduct. A monitor appointed by the CRO will perform biannual ‘on site’ monitoring of individual case histories, assess adherence to the protocol, ensure the ongoing implementation of appropriate data entry and quality control procedures and, in general, assess adherence to good clinical practices.40 The study monitor will remain blinded to study arm assignment. AEs and adverse device effects All AEs and adverse device effects reported spontaneously by the study subject or observed by the investigators or staff will be recorded. In case of a serious adverse device effect (SADE), the principal investigator will report the SADE to the ethics committee and the authorities (Bundesinstitut für Arzneimittel und Medizinprodukte (BfArM)). Prolonged hospitalisation (>14 days) will be reported as an SAE or as an SADE. Rehospitalisation (for any reason) will also be reported and handled as an SAE or as an SADE. All AEs will be followed until the event has resolved, or until a stable situation has been achieved. Study limitations There are several potential limitations to this study. First, heterogeneity of the study population in terms of indication for surgery introduces randomness in the data. The study is likely to be underpowered to assess possible effect modifiers such as preoperative deformity. Second, having a blinded examiner renders study procedures considerably more complex: the blinded examiner and the study nurse must have no contact with the surgical team; subjects must be briefed about the importance of keeping the examiner blinded. Twitter: Follow Peter Fennema @AMR_CRO Contributors: HH and PF participated in the conception of the idea for the study. HH, KW, DK, PH and PF participated in the design of the study and research protocol. HH and PF participated in the writing of the manuscript. HH, KW, DK, PH and PF were involved in the review and editing of the manuscript. HH, KW, DK, PH and PF have critically read the final draft of the paper and approved the final text. Competing interests: HH is a consultant for Smith and Nephew GmbH, Hamburg, Germany. Ethics approval: Landesärztekammer Brandenburg. (S 10(a)/2013). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Anonymous patient data will be made available on request. ==== Refs References 1 Kurtz S , Ong K , Lau E Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030 . J Bone Joint Surg Am 2007 ;89 :780 –5 . 10.2106/JBJS.F.00222 17403800 2 Vasso M , Beaufils P , Schiavone Panni A Constraint choice in revision knee arthroplasty . Int Orthop 2013 ;37 :1279 –84 . 10.1007/s00264-013-1929-y 23700251 3 Malcolm TL , Bederman SS , Schwarzkopf R Outcomes of varus valgus constrained versus rotating-hinge implants in total knee arthroplasty . Orthopedics 2016 ;39 :e140 –8 . 10.3928/01477447-20151228-07 26730689 4 Meijer MF , Reininga IH , Boerboom AL Poorer survival after a primary implant during revision total knee arthroplasty . Int Orthop 2013 ;37 :415 –9 . 10.1007/s00264-012-1739-7 23263508 5 Lettin AW , Deliss LJ , Blackburne JS The Stanmore hinged knee arthroplasty . J Bone Joint Surg Br 1978 ;60-b :327 –32 .681408 6 Deburge A , Guepar Guepar hinge prosthesis: complications and results with two years’ follow-up . Clin Orthop Relat Res 1976 ;120 :47 –53 . 7 Walker PS , Emerson R , Potter T The kinematic rotating hinge: biomechanics and clinical application . Orthop Clin North Am 1982 ;13 :187 –99 .7063191 8 Hwang SC , Kong JY , Nam DC Revision total knee arthroplasty with a cemented posterior stabilized, condylar constrained or fully constrained prosthesis: a minimum 2-year follow-up analysis . Clin Orthop Surg 2010 ;2 :112 –20 . 10.4055/cios.2010.2.2.112 20514269 9 Barrack RL , Lyons TR , Ingraham RQ The use of a modular rotating hinge component in salvage revision total knee arthroplasty . J Arthroplasty 2000 ;15 :858 –66 . 10.1054/arth.2000.9056 11061445 10 Neumann DR , Hofstaedter T , Dorn U Follow-up of a modular rotating hinge knee system in salvage revision total knee arthroplasty . J Arthroplasty 2012 ;27 :814 –19 . 10.1016/j.arth.2011.08.015 22000574 11 Gehrke T , Kendoff D , Haasper C The role of hinges in primary total knee replacement . Bone Joint J 2014 ;96-B :93 –5 . 10.1302/0301-620X.96B11.34143 25381417 12 Gudnason A , Milbrink J , Hailer NP Implant survival and outcome after rotating-hinge total knee revision arthroplasty: a minimum 6-year follow-up . Arch Orthop Trauma Surg 2011 ;131 :1601 –7 . 10.1007/s00402-011-1330-5 21656196 13 Joshi N , Navarro-Quilis A Is there a place for rotating-hinge arthroplasty in knee revision surgery for aseptic loosening? J Arthroplasty 2008 ;23 :1204 –11 . 10.1016/j.arth.2007.10.016 18534467 14 Hossain F , Patel S , Haddad FS Midterm assessment of causes and results of revision total knee arthroplasty . Clin Orthop Relat Res 2010 ;468 :1221 –8 . 10.1007/s11999-009-1204-0 20058112 15 Wilke K , Kunze D , Hommel P Rotating hinge total knee arthroplasty RT-PLUS Solution: a clinical and radiographic follow-up . J Orthopedics Rheumatol 2016 ;3 :4 . 16 Pour AE , Parvizi J , Slenker N Rotating hinged total knee replacement: use with caution . J Bone Joint Surg Am 2007 ;89 :1735 –41 . 10.2106/JBJS.F.00893 17671012 17 Lombardi AV Jr, Berend KR , Leith JR Posterior-stabilized constrained total knee arthroplasty for complex primary cases . J Bone Joint Surg Am 2007 ;89 (Suppl 3) :90 –102 . 10.2106/JBJS.G.00586 17908875 18 Samiezadeh S , Abolghasemian M , D'Lima DD Rotating hinge versus constrained condylar knee replacement: which one is more constrained? A finite element study. American Academy of Orthopaedic Surgeons Annual Meeting . New Orleans, LA , 2014 . 19 Fuchs S , Sandmann C , Gerdemann G Quality of life and clinical outcome in salvage revision total knee replacement: hinged vs total condylar design . Knee Surg Sports Traumatol Arthrosc 2004 ;12 :140 –3 . 10.1007/s00167-003-0401-8 12937894 20 Walker PS , Manktelow AR Comparison between a constrained condylar and a rotating hinge in revision knee surgery . knee 2001 ;8 :269 –79 .11706689 21 Hommel H Clinical outcome in total knee replacement: hinged vs total condylar design . Wriezen, Germany : Krankenhaus Märkisch Oderland , 2016 . 22 Greene CJ , Morland LA , Durkalski VL Noninferiority and equivalence designs: issues and implications for mental health research . J Trauma Stress 2008 ;21 :433 –9 . 10.1002/jts.20367 18956449 23 Agha RA , Altman DG , Rosin D The SPIRIT 2013 statement–defining standard protocol items for trials . Int J Surg 2015 ;13 :288 –91 . 10.1016/j.ijsu.2014.12.007 25498499 24 Piaggio G , Elbourne DR , Pocock SJ Reporting of noninferiority and equivalence randomized trials: extension of the CONSORT 2010 statement . JAMA 2012 ;308 :2594 –604 . 10.1001/jama.2012.87802 23268518 25 Roos EM , Roos HP , Ekdahl C , et al Knee injury and Osteoarthritis Outcome Score (KOOS)--validation of a Swedish version . Scand J Med Sci Sports 1998 ;8 :439 –48 .9863983 26 Roos EM , Toksvig-Larsen S Knee injury and Osteoarthritis Outcome Score (KOOS)—validation and comparison to the WOMAC in total knee replacement . Health Qual Life Outcomes 2003 ;1 :17 10.1186/1477-7525-1-17 12801417 27 Roos EM , Lohmander LS The Knee injury and Osteoarthritis Outcome Score (KOOS): from joint injury to osteoarthritis . Health Qual Life Outcomes 2003 ;1 :64 10.1186/1477-7525-1-64 14613558 28 Behrend H , Giesinger K , Giesinger JM The “forgotten joint” as the ultimate goal in joint arthroplasty: validation of a new patient-reported outcome measure . J Arthroplasty 2012 ;27 :430 –6.e1 . 10.1016/j.arth.2011.06.035 22000572 29 Giesinger JM , Kuster MS , Behrend H Association of psychological status and patient-reported physical outcome measures in joint arthroplasty: a lack of divergent validity . Health Qual Life Outcomes 2013 ;11 :64 10.1186/1477-7525-11-64 23601140 30 Insall JN , Dorr LD , Scott RD Rationale of the Knee Society clinical rating system . Clin Orthop Relat Res 1989 ;248 :13 –14 . 31 Ewald FC The Knee Society total knee arthroplasty roentgenographic evaluation and scoring system . Clin Orthop Relat Res 1989 ;248 :9 –12 . 32 Sells J Summary of safety and effictiveness . Smith & Nephew, Inc. LEGION cobalt chrome revision knee system , 2006 http://www.accessdata.fda.gov/cdrh_docs/pdf6/K060742.pdf 33 van Kempen RW , Schimmel JJP , van Hellemondt GG Reason for revision TKA predicts clinical outcome: prospective evaluation of 150 consecutive patients with 2-years followup . Clin Orthop Relat Res 2013 ;471 :2296 –302 . 10.1007/s11999-013-2940-8 23543418 34 No authors listed ICH Harmonised Tripartite Guideline. Statistical principles for clinical trials. International Conference on Harmonisation E9 Expert Working Group . Stat Med 1999 ;18 :1905 –42 .10532877 35 Blackwelder WC “Proving the null hypothesis” in clinical trials . Control Clin Trials 1982 ;3 :345 –53 .7160191 36 Julious SA Sample sizes for clinical trials with normal data . Stat Med 2004 ;23 :1921 –86 . 10.1002/sim.1783 15195324 37 Lan KKG , DeMets DL Discrete sequential boundaries for clinical trials . Biometrika. Biometrika 1093 ;70 :659 –63 . 38 Westlake WJ Symmetrical confidence intervals for bioequivalence trials . Biometrics 1976 ;32 :741 –4 .1009222 39 Verbeke G , Molenberghs G Linear mixed models for longitudinal data . New York : Springer , 2000 . 40 No authors listed. The establishment and operation of clinical trial data monitoring committees for clinical trial sponsors 2016. http://www.fda.gov/RegulatoryInformation/Guidances/ucm127069.htm

PMC005xxxxxx/PMC5372035.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01361610.1136/bmjopen-2016-013616OphthalmologyResearch1506171817371701Financial modelling of femtosecond laser-assisted cataract surgery within the National Health Service using a ‘hub and spoke’ model for the delivery of high-volume cataract surgery Roberts H W 12Ni M Z 3O'Brart D P S 12 1 Department of Ophthalmology, Guy's and St Thomas’ NHS Foundation Trust, London, UK 2 King's College London, London, UK 3 Division of Surgery, Department of Surgery and Cancer, Imperial College London, St Mary's Hospital, London, UKCorrespondence to Professor D P S O'Brart; davidobrart@aol.com2017 16 3 2017 7 3 e01361625 7 2016 31 10 2016 23 12 2016 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Aims To develop financial models which offset additional costs associated with femtosecond laser (FL)-assisted cataract surgery (FLACS) against improvements in productivity and to determine important factors relating to its implementation into the National Health Service (NHS). Methods FL platforms are expensive, in initial purchase and running costs. The additional costs associated with FL technology might be offset by an increase in surgical efficiency. Using a ‘hub and spoke’ model to provide high-volume cataract surgery, we designed a financial model, comparing FLACS against conventional phacoemulsification surgery (CPS). The model was populated with averaged financial data from 4 NHS foundation trusts and 4 commercial organisations manufacturing FL platforms. We tested our model with sensitivity and threshold analyses to allow for variations or uncertainties. Results The averaged weekly workload for cataract surgery using our hub and spoke model required either 8 or 5.4 theatre sessions with CPS or FLACS, respectively. Despite reduced theatre utilisation, CPS (average £433/case) was still found to be 8.7% cheaper than FLACS (average £502/case). The greatest associated cost of FLACS was the patient interface (PI) (average £135/case). Sensitivity analyses demonstrated that FLACS could be less expensive than CPS, but only if increased efficiency, in terms of cataract procedures per theatre list, increased by over 100%, or if the cost of the PI was reduced by almost 70%. Conclusions The financial viability of FLACS within the NHS is currently precluded by the cost of the PI and the lack of knowledge regarding any gains in operational efficiency. ==== Body Strengths and limitations of this study Data were collected and collated from four NHS foundation trusts of various sizes, locations and demographics to ensure the conclusions could be more representative. This is the only study investigating the financial implications of femtosecond laser-assisted cataract surgery (FLACS) which highlights the significance of the cost of the disposable patient interfaces over the capital cost of the laser machine itself. This is the only study investigating the financial implications of FLACS which has developed a working model incorporating a laser to improve, rather than impede productivity. This study falls short of providing an incremental cost-effectiveness ratio for FLACS. For the purposes of this model, clinical outcomes are assumed to be equivalent. This is currently supported by the latest evidence. Introduction In 2014–2015, over 370 000 cataract operations were performed on the National Health Service (NHS).1 This was 3.7 times the number performed in 1989.2 The need for cataract surgery is expected to rise further with increasing life expectancy, rising population size, growing patient expectations and an increase in age-related chronic diseases associated with cataracts, such as diabetes.3 With current financial constraints, this increased future demand for cataract surgery within the NHS is liable to be problematic. Femtosecond laser (FL) technology has been recently introduced into cataract surgery in an attempt to automate and improve the efficacy and safety of some of the surgical steps within this procedure.4 Within the scientific literature, there are now numerous prospective case series supporting its usage and continued development and more surgeons are adopting this new technology.4–9 However, while FL technology undoubtedly offers great surgical precision, a recent meta-analysis shows no significant advantages in terms of safety and efficacy of FL-assisted cataract surgery (FLACS) over conventional phacoemulsification cataract surgery (CPS).10 Two large multicentre randomised controlled trials (RCTs) are currently underway in France and the UK and may provide further evidence as to whether there is a difference in the clinical outcomes from FLACS.11 12 Until evidence exists of improved surgical outcomes, it is difficult at present to support the widespread implementation of FLACS. This is particularly pertinent as the introduction of FLACS has significant associated financial costs. These include initial purchase costs of the FL system itself, servicing, depreciation and the individual patient interfaces (PI), which call into question its financial viability, especially in a state-funded healthcare system. The majority of existing literature on the economics of FLACS originates from healthcare systems within countries such as the USA or Australia, where additional costs from procedures perceived as having a premium status may be passed onto the patient in the form of a copayment system.13–15 In these healthcare systems, the existing literature suggests that FLACS is not, at this time, a cost-effective solution. It is not surprising, therefore, that adoption of this technology within the NHS so far has been minimal and largely directed at research rather than service provision. Despite associated costs, by its very nature, the FL offers the potential to remove several steps of cataract extraction from needing to be performed by a fully trained surgeon in a fully equipped ophthalmic operating theatre. FL technology can automate several surgical steps of the cataract procedure, such as corneal incisions, arcuate keratotomies, capsulotomy and nuclear lens division, all of which can be potentially undertaken with this technology by a doctor in training or suitably trained nurse or technician in a clean room. By reducing the actual amount of time each patient spends within the operating theatre under the care of a trained surgeon, the volume of surgical cases undertaken in a given period of time might potentially be increased. This may be especially true if a ‘hub and spoke’ model is utilised, with the FL performing these initial automated steps and then allowing the completion of the surgical procedure to be undertaken in more than one operating theatre at a time. If the number of cases per theatre session can be increased sufficiently then the initial expenditure and additional costs associated with FL technology might be offset. For FLACS to see increased adoption by a state-funded healthcare system such as the NHS, it would need to be shown to be cost-effective based on an acceptable incremental cost-effectiveness ratio (ICER). The ICER is defined by the difference in the cost between two possible interventions divided by the difference in their clinical effectiveness. This study aims to investigate, in the absence of clinical outcomes from large RCTs showing any surgical benefit, the cost of incorporating FLACS into the NHS system in order to determine whether the increased costs of equipment may be offset by an increase in the volume of surgery performed. Methods Financial model A financial model was designed to compare FLACS against CPS for the provision of cataract surgery within the NHS. The inputs for this model can be seen in table 1. The model was based on data from four separate NHS Foundation Trust Ophthalmology Departments (Guy's and St Thomas' NHS Foundation Trust, Norfolk and Norwich NHS Foundation Trust, Peterborough and Stamford NHS Foundation Trust and West Suffolk NHS Foundation Trust) and four manufacturers of commercially available FL devices (Abbott Medical Optics, Santa Ana, California, USA; Ziemer Ophthalmic Systems AG, Switzerland; Alcon Laboratories, Fort Worth, Texas, USA and Bausch & Lomb, Rochester, New York, USA). The data were collated and averaged to ensure the results were more representative than had just one ophthalmology department or one FL been used. Table 1 Inputs for the model and nominal values Source Input Value (£) Range (£) A  Income NHS tariff for cataract surgery 789 729–917  Expenses Staffing (per session) Consultant surgeon 246 Band 5 nurse 79 Registrar/laser technician 101 Band 6 nurse/laser technician 102 Ward clerk 53 Overheads (per year) Ophthalmic day-case unit 525 620 30 112–1 061 481 2× operating theatres 585 676 353 245–962 287 Laser Initial cost 262 500 175k–350k Maintenance/year 28 333 20k–35k Cost of patient interface 134.75 99–170 Other costs Disposables and IOL (per case) 103 Cost of administration, management and pharmacy (per case) 50 B  Other variables   Number of cataract operations required per week 55 operations 27–96 Number of cases on CPS list 7 operations Number of cases on FLACS list* 10 operations Lifetime of FL 10 years *Based on the hub and spoke FLACS delivery model. Values for each input were derived from the following sources. Income for each procedure is reimbursed at the NHS national tariffs for 2014–2015 plus an additional market forces factor.16 17 Costs were divided into direct labour costs, equipment costs and overheads. Direct labour costs per theatre session were derived from NHS pay scales and midpoint values were chosen. This was then proportioned to the estimated duration of each theatre session. Costs relating to the FL were averaged from those provided by four manufacturers of commercially available FL devices. Costs such as estate, equipment and supplies were averaged from four NHS Foundation Trusts' departmental budgets (2014–2015). Pharmacy and administrative costs were obtained by reviewing the departmental budget at our institution. Baseline values for the number of cases achievable per 4-hour theatre session were given nominal values of 7 cases for CPS and 10 cases for FLACS. These initial values were then tested using sensitivity and threshold analyses. The model was tested based on two scenarios: FLACS versus CPS based on an average number of seven cases currently performed on a CPS cataract lists and a FLACS delivery model based on a ‘hub and spoke’ method with one FL in a clean room and operated by a doctor in training preparing patients for two operating theatres running in parallel with their associated surgeons, nursing and technical support staff. ‘Hub and spoke’ FLACS model Our theoretical ‘hub and spoke’ model for FLACS is based on a single FL platform in a clean room and operated by an ophthalmology registrar or suitably trained allied health professional and supported by a theatre nurse (figure 1). The laser would be programmed to perform capsulotomy, nuclear lens division and arcuate keratotomies (when indicated) for each individual patient. Patients would be prepared for two operating theatres running in parallel with their associated surgeons, nursing and technical support staff. The assumed FL treatment time is a maximum of 10 min per patient allowing for the preparation of up to 20 cataract surgery cases, 10 per theatre per 4-hour operating theatre session. The assumed theatre time is a maximum of 24 min per case. These values are based on our own experience with the FL. Figure 1 A proposed ‘hub and spoke’ model for femtosecond laser-assisted cataract surgery. Sensitivity analysis The model was constructed using Microsoft Excel (Microsoft Corp, Redmond, Washington, USA) based on the range of the above inputs (table 1). Univariate and bivariate sensitivity analyses were conducted by varying the inputs into the model to simulate the impact on the final service costs. The inputs chosen for the sensitivity analysis were as follows: capital cost of the FL, cost of the PI, number of cases possible on a FLACS theatre list, number of cases performed on a CPS list, number of cataract operations required per week. Threshold analyses were performed on the same variables as the sensitivity analyses to determine threshold values at which FLACS may break even with CPS. The results are reported as weekly costs. Results The first model tested FLACS versus CPS based on an average number of seven cases currently performed on CPS cataract lists. Our model estimated that the current CPS service at its existing productivity was costing £433 per case. Using a model that incorporates one FL into one theatre list, and therefore assuming no increase in productivity, the laser increases the cost per case by £167 to £600. Based on these values, the CPS service would be 72% of the cost of a FLACS service. Using the averaged and nominal values for our theoretical ‘hub and spoke’ model for FLACS, the use of the FL reduced the weekly theatre requirements from 8 CPS theatre sessions to 2.7 FLACS sessions with both theatres in the FL model running in parallel (total theatre sessions 5.4). This reduced the anticipated running costs of theatres, the ophthalmic day-case unit and staffing costs. However, the laser introduced additional costs into the model (FL equipment, supplies, maintenance and additional staff). Based on the nominal values, even with our hub and spoke model running optimally, the CPS service (average of £433/case) was found to be 86.3% of the cost of the FLACS service (average of £502/case). The capital cost of the FL when amortised over its lifetime of 10 years was £505/week. Maintenance of the laser was £545/week. The cost of 1 week's worth of PI (n=55) at £135 each was £7356 (figure 2). Figure 2 Comparison of the costs per week of conventional phacoemulsification surgery compared with femtosecond laser-assisted cataract surgery. The model was not affected when we changed the salary of the laser operator from a midpoint registrar to a band 6 nurse as the hourly rates were of negligible difference (table 1). Univariate sensitivity analyses were conducted by varying one variable at a time. Minimum and maximum values were obtained from the original data (table 2). Only when the number of operations on a CPS list was reduced or the number of operations on a FLACS list was increased, could the model give an output in favour of FLACS. Best and worst-case scenarios were constructed for CPS and FLACS, by aligning the most important variables all in favour of one or other modality, with costs of £371 and £515 for CPS and £381 and £545 for FLACS, respectively (table 2B). Table 2 (A) Univariate sensitivity analysis of the hub and spoke model based on range of values from data collected and (B) best-case scenarios for conventional phacoemulsification surgery and femtosecond laser-assisted cataract surgery Input Range of values Values inputted into hub and spoke model Cost of CPS service compared with FLACS (%) A Cataract workload/week   Minimum 27 82.7 Average 55 86.3 Maximum 96 87.8 Number of cataracts on CPS list*   Minimum 5 108.5 Nominal 7 86.3 Maximum 9 73.8 Number of cataracts on FL list†   Minimum 8 78.6 Nominal 10 86.3 Maximum 16 100.8 Initial cost of FL   Minimum £175 000 86.7 Average £262 500 86.3 Maximum £350 000 85.7 Cost of PI   Minimum £99 92.8 Average £135 86.3 Maximum £170 80.5 Cataract workload/week Number of cataracts on CPS list Number of cataracts on FL list Cost of PI Cost of CPS/case Cost of FLACS/case Cost of CPS service compared with FLACS (%) B Best-case scenario for CPS 27 9 9 135 £371  £515 72.1 Best-case scenario for FLACS 96 5 10 50 £545 £381 143.2 Bold indicates where FLACS is less expensive than CPS option. *Assuming 10 cases on FLACS list. †Assuming seven cases on CPS list. Univariate threshold analyses were performed to demonstrate the ‘break-even’ values of each input. Keeping all other inputs at their original values, the model could not find solutions by which the FL broke even when the capital cost of the FL or the number of operations performed per week was chosen. The costs of the services were equivalent if the true number of cases on a CPS list was 6, or if the FL could increase productivity to 16 cases/each theatre, or if the cost of the laser consumables was reduced to £66. It was thereby ascertained that these three parameters are the most important in this model for determining a cost-neutral scenario for FLACS. Bivariate sensitivity analyses were performed using combinations of the above inputs. For example, table 3 shows the outcomes of the model when the capacity for the number of cases on CPS and FLACS is simultaneously tested. It shows that the FLACS service would be required to approximately double the number of operations possible during a theatre list for FLACS to break even. Table 4 tests the outcome of the model based on an assumption that the NHS can negotiate lower PI costs based on the provision of a large number of operations per year. It shows that FLACS cannot break even unless the cost of the PI is significantly reduced (to approximately £50 per case). Table 5 compares the cost of the PI against the number of cases on a FLACS list. Table 3 Cost of femtosecond laser-assisted cataract surgery versus conventional phacoemulsification surgery Number of operations on CPS list (%) Number of operations on FLACS list 5 6 7 8 9 8 99.0 87.1 78.6 72.3 67.3 10 108.5 95.5 86.2 79.2 73.8 12 115.9 102.0 92.1 84.7 78.9 14 121.9 107.3 96.9 89.0 82.9 16 126.8 111.6 100.8 92.6 86.3 Bold indicates where FLACS is less expensive than CPS option. Bivariate sensitivity analysis: demonstrating relative costs of CPS service compared with FLACS when total number of cases on each theatre list are tested. Table 4 Cost of femtosecond laser-assisted cataract surgery versus conventional phacoemulsification surgery Number of cataract operations per year (%) Cost of PI (£) 2000 3000 4000 5000 6000 50 101.5 104.0 105.2 106.0 106.5 75 95.9 98.1 99.2 99.9 100.4 100 90.9 92.8 93.8 94.4% 94.9 125 86.4 88.1 89.0 89.6 89.9 150 82.3 83.8 84.7 85.2 85.5 Bold indicates where FLACS is less expensive than CPS option. Bivariate sensitivity analysis: demonstrating relative costs of CPS service compared with FLACS when cost of PI and total number of cases per year are tested. Table 5 Cost of femtosecond laser-assisted cataract surgery versus conventional phacoemulsification surgery Number of operations on FLACS list (%) Cost of PI 8 9 10 12 14 16 50 92.9 98.6 103.7 112.4 119.5 125.5 65 90.0 95.4 100.1 108.2 114.8 120.3 80 87.3 92.3 96.7 104.3 110.4 115.5 100 83.9 88.5 92.6 99.5 105.0 109.6 120 80.8 85.1 88.8 95.1 100.2 104.3 135 78.6 82.6 86.2 92.1 96.8 100.7 Bold indicates where FLACS is less expensive than CPS option. Bivariate sensitivity analysis: demonstrating relative costs of CPS service compared with FLACS when cost of PI and number of operations on FLACS list are tested. Discussion We have designed a hypothetical treatment delivery model based on a ‘hub and spoke’ service and utilising FLACS to improve the efficiency of cataract surgery in terms of number of cases undertaken per operating list. We then tested our model with sensitivity and threshold analyses to allow for variations or uncertainties. Even with our optimised delivery model, FLACS is still more expensive than CPS based on current estimates of costs. To break even, the incorporation of FLACS would have to approximately double the number of cataract operations performed per theatre list and indeed could not offer a cost-neutral solution if the number of cases on a CPS theatre list was 8 or more. Our model indicates that the greatest cost impediment to a FLACS service is the price of the PI (average cost £135/case) (figure 2), which represents almost 27% of the total cost per case. Unlike other service costs, the cost of the PI is not mitigated by potential increased productivity. It is therefore a major financial impediment to FLACS ever becoming cost-effective within the NHS, where the total tariff for each operation is fixed between £718 and £932.16 17 Potentially, this problem may be overcome by the manufacturer considerably discounting this cost to the NHS. In contrast to the PI, our financial model indicates that the costs of the laser itself, staffing and maintenance it were much less important (4.8% of total costs). There are three important unknowns with regard to our model. First, we are awaiting clinical results from large RCTs comparing FLACS with CPS.11 12 The latest meta-analysis shows no significant advantages in terms of safety of FLACS over CPS.10 However, there are advantages in terms of endothelial cell loss, effective phacoemulsification time and unaided visual acuity, albeit no difference in long-term best-corrected visual acuity and an increased risk of anterior capsular tear.18 We assumed in our financial modelling that there are no differences in outcomes and complication rates between the two procedures. If, however, FLACS were to show significant advantages in terms of patient safety and outcomes then such improvements then this may have additional positive financial implications. Second, potential gains in productivity from the FL are as yet unpublished and unrealised. Several studies investigating FLACS actually report decreased patient turnover with FLACS.13 19 20 This is because at present typically the operating surgeon is performing the FL treatment as well as the subsequent lens extraction. There are as yet no publications on the most effective way to design a FL-centric cataract service. We chose a ‘hub and spoke’ model based on one FL in a clean room, operated by an ophthalmic surgeon in training or ophthalmic technician/nurse. The FL then fed patients into two independent operating theatres, each with its own surgeon and support staff. This model is theoretical. It needs to be tested in the NHS setting to see if it is viable, and further work may need to be performed to determine a ‘best-practice’ and optimised efficiency model for FLACS. Third, it is likely that the costs of the PIs would be reduced below the values quoted to us by the manufacturers, as a large public sector ophthalmology department performing several thousand operations per year could negotiate on costs and capitalise on market competition. As discussed above, this would considerably improve the financial burdens associated with implementing FLACS. Abell and Vote13 have previously designed a hypothetical model to derive cost-effectiveness of FLACS. In the absence of better evidence, conservative estimates were used for complication rates with FLACS. Their use of the FL resulted in reducing their theatre efficiency by two cases per list, and subsequently, they estimated the additional cost of FL to be AUS$1065 per case, AUS$750 of which were the direct costs from the FL and AUS$315 from lost productivity. Our model was based on using the laser to improve, rather than impede, productivity. We estimated the cost per case to be £158, of which £135 is the PI. We chose to amortise the costs of the laser over 10 years rather than only 3, but reducing the lifetime of the laser to 3 years increased the cost of each operation to only £180. This demonstrates yet again the greatest cost of FLACS is the cost of the PI rather than the laser itself. In addition to the above, there are important limitations to mention regarding this hypothetical model. The model assumes that all patients are suitable for a high-volume FLACS theatre list. However, some patients may not be suited to FLACS or to a high-volume service, although the number of contraindications for FLACS is decreasing as experience with the technology improves.21–23 Departmental costs used in this model were obtained from a retrospective review of the financial records at four NHS foundation trusts. In order to ensure that the results were applicable to more than just one hospital with its associated population, we selected two teaching hospitals and two district general hospitals of varying sizes, with annual numbers of between ∼1400 and 5000 cataract operations. These hospitals serve urban and rural populations (range ∼275 000–823 000 served by each hospital) with a mixture of demographics (and include hospitals with one of the highest and one of the lowest cataract tariffs).17 The costs of consumables were assumed to be equal for FLACS and CPS. In reality, as the FL performs many stages of the procedure, the cost of some consumables may be reduced (vision blue, cystotome, etc) and some cataracts may no longer require any phacoemulsification.24 Our model incorporates the salary of a registrar to operate the laser,25 26 yet if FLACS becomes widely adopted within the UK, then technicians may be trained to perform this duty, perhaps at a reduced cost, but no money was saved when we modelled for the salary of a band 6 nurse to operate the laser. Overall, this model demonstrates that FLACS could only be financially viable if its implementation into the NHS allowed significant improvements in efficiency in the number of cases treated per theatre list and/or if the cost of the PI was considerably reduced. Further research is required on the clinical outcomes of FLACS compared with CPS as well as real-world evidence of the effect to surgical efficiency afforded by this technology. Guy's and St Thomas' NHS Foundation Trust, Norfolk and Norwich NHS Foundation Trust, Peterborough and Stamford NHS Foundation Trust and West Suffolk NHS Foundation Trust; Abbott Medical Optics, Santa Ana, California, USA; Ziemer Ophthalmic Systems AG, Switzerland; Alcon Laboratories, Fort Worth, Texas, USA and Bausch & Lomb, Rochester, New York, USA; Karen Bateman. Contributors: HWR, MZN and DPSOB contributed to acquisition of data, analysis and interpretation of data, drafting the article and revising it critically for important intellectual content. Funding: This research has been supported by a non-commercial research grant from Alcon Incorporated. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Further information is available by emailing HWR (harry.roberts@nhs.net). ==== Refs References 1 Department of Health . HES online Health Episode Statistics. 2015 https://www.google.co.uk/url?sa=t&rct=j&q=&esrc=s&source=web& cd=1&ved=0ahUKEwjg2-211cfSAhXJDywKHQSxCpUQFggaMAA &url=http%3A%2F%2Fcontent.digital.nhs.uk%2Fcatalogue%2FPU B19124%2Fhosp-epis-stat-admi-proc-2014-15-tab.xlsx&usg=AFQj CNFyjQSbEQpRdDNHEpJrqBrClJNaJg&sig2=NPzSIl2T6rr-QODx Hk3BPQ&bvm=bv.149093890,d.bGg. (Last accessed 8th March 2017) 2 Black N , Browne J , van der Meulen J Is there overutilisation of cataract surgery in England? Br J Ophthalmol 2008 ;93 :13 –17 . 10.1136/bjo.2007.136150 3 Minassian DC , Reidy A Future sight loss UK (2): an epidemiological and economic model for sight loss in the decade 2010–2020 2009 :1 –130 . http://www.rnib.org.uk/sites/default/files/FSUK_2.pdf (accessed 25 Oct 2016 ). 4 Nagy Z , Takacs A , Filkorn T Initial clinical evaluation of an intraocular femtosecond laser in cataract surgery . J Refract Surg 2009 ;25 :1053 –60 . 10.3928/1081597X-20091117-04 20000286 5 Ewe SY , Abell RG , Oakley CL A comparative cohort study of visual outcomes in femtosecond laser-assisted versus phacoemulsification cataract surgery . Ophthalmology 2016 ;123 :178 –82 . 10.1016/j.ophtha.2015.09.026 .26526634 6 Chee SP , Yang Y , Ti SE Clinical outcomes in the first two years of femtosecond laser-assisted cataract surgery . Am J Ophthalmol 2015 ;159 :714 –19.e2 . 10.1016/j.ajo.2015.01.016 25634531 7 Abell RG , Kerr NM , Vote BJ Femtosecond laser-assisted cataract surgery compared with conventional cataract surgery . Clin Exp Ophthalmol 2013 ;41 :455 –62 . 10.1111/ceo.12025 23078347 8 Abell RG , Darian-Smith E , Kan JB Femtosecond laser-assisted cataract surgery versus standard phacoemulsification cataract surgery: outcomes and safety in more than 4000 cases at a single center . J Cataract Refract Surg 2015 ;2015 :47 –52 . 10.1016/j.jcrs.2014.06.025 9 Roberts TV , Lawless M , Bali SJ Surgical outcomes and safety of femtosecond laser cataract surgery: a prospective study of 1500 consecutive cases . Ophthalmology 2013 ;120 :227 –33 . 10.1016/j.ophtha.2012.10.026 23218822 10 Chen X , Xiao W , Ye S Efficacy and safety of femtosecond laser-assisted cataract surgery versus conventional phacoemulsification for cataract: a meta-analysis of randomized controlled trials . Sci Rep 2015 ;5 :13123 10.1038/srep13123 26269445 11 Schweitzer C , Hayes N , Brezin A Re: Abell et al.: Cost-effectiveness of femtosecond laser-assisted cataract surgery versus phacoemulsification cataract surgery (Ophthalmology 2014;121:10–6) . Ophthalmology 2014 ;121 :e53 –4 . 10.1016/j.ophtha.2014.05.025 12 Day AC , Burr JM , Bunce C Randomised, single-masked non-inferiority trial of femtosecond laser-assisted versus manual phacoemulsification cataract surgery for adults with visually significant cataract: the FACT trial protocol . BMJ Open 2015 ;5 :e010381 10.1136/bmjopen-2015-010381 13 Abell RG , Vote BJ Cost-effectiveness of femtosecond laser-assisted cataract surgery versus phacoemulsification cataract surgery . Ophthalmology 2014 ;121 :10 –16 . 10.1016/j.ophtha.2013.07.056 24120324 14 Hansen MS , Hardten DR Financially efficient cataract surgery in today's healthcare environment . Curr Opin Ophthalmol 2015 ;26 :61 –5 . 10.1097/ICU.0000000000000120 25415299 15 Bartlett JD , Miller KM The economics of femtosecond laser-assisted cataract surgery . Curr Opin Ophthalmol 2016 ;27 :76 –81 . 10.1097/ICU.0000000000000219 26632918 16 Department of Health . 2015-16-eto-spreadsheet-3 . 2016 https://www.google.co.uk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0ahUKEwiYsoPOi9rNAhWMLsAKHcKfCaUQFggeMAA&url=https%3A%2F%2Fwww.england.nhs.uk%2Fwp-content%2Fuploads%2F2015%2F03%2F2015-16-eto-spreadsheet.xlsx&usg=AFQjCNHr4y_QdTX6KyU96-knpRLOAHcyLQ&sig2=-6Z6D9BLQpmdZcpuXAadog (accessed 25 Oct 2016 ). 17 Department of Health . 2015-16-eto-guid-market-forces-payment-4 2016 https://www.google.co.uk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0ahUKEwj3k-GBjdrNAhWYHsAKHRp2ARgQFggeMAA&url=https%3A%2F%2Fwww.england.nhs.uk%2Fwp-content%2Fuploads%2F2015%2F03%2F2015-16-eto-guid-market-forces-payment.xlsx&usg=AFQjCNFV4OKenNMsJOJuXIPSoFgkKaSeIw&sig2=SlcpHBWBquHKVoxxBaIk7g (accessed 25 Oct 2016 ). 18 Kohnen T , Mathys L , Petermann K Metaanalysis and systematic review of femtosecond laser lens surgery and conventional lens surgery . Copenhagen : European Society of Cataract and Refractive Surgeons , 2016 . 19 Bali SJ , Hodge C , Lawless M Early experience with the femtosecond laser for cataract surgery . Ophthalmology 2012 ;119 :891 –9 . 10.1016/j.ophtha.2011.12.025 22361311 20 Lubahn JG , Donaldson KE , Culbertson WW Operating times of experienced cataract surgeons beginning femtosecond laser-assisted cataract surgery . J Cataract Refract Surg 2014 ;40 :1773 –6 . 10.1016/j.jcrs.2014.03.024 25217069 21 Conrad-Hengerer I , Hengerer FH , Joachim SC Femtosecond laser-assisted cataract surgery in intumescent white cataracts . J Cataract Refract Surg 2014 ;40 :44 –50 . 10.1016/j.jcrs.2013.08.044 24269086 22 Dick HB , Schultz T Femtosecond laser-assisted capsulotomy rescue for capsulorhexis enlargement . J Cataract Refract Surg 2014 ;40 :1588 –90 . 10.1016/j.jcrs.2014.08.021 25263037 23 Hatch KM , Schultz T , Talamo JH Femtosecond laser-assisted compared with standard cataract surgery for removal of advanced cataracts . J Cataract Refract Surg 2015 ;41 :1833 –8 . 10.1016/j.jcrs.2015.10.040 26603391 24 Abell RG , Kerr NM , Vote BJ Toward zero effective phacoemulsification time using femtosecond laser pretreatment . Ophthalmology 2013 ;120 :942 –8 . 10.1016/j.ophtha.2012.11.045 23465860 25 Cohen MN , Intili A , Ni N Femtosecond laser-assisted cataract surgery in residency training . Curr Opin Ophthalmol 2015 ;26 :56 –60 . 10.1097/ICU.0000000000000114 25390862 26 Hou JH , Prickett AL , Cortina MS Safety of femtosecond laser-assisted cataract surgery performed by surgeons in training . J Refract Surg 2015 ;31 :69 –70 . 10.3928/1081597X-20141218-07 25599546

PMC005xxxxxx/PMC5372036.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01462010.1136/bmjopen-2016-014620PaediatricsProtocol150617191695Development of the Brussels Infant and Toddler Stool Scale (‘BITSS’): protocol of the study Vandenplas Yvan 1Szajewska Hania 2Benninga Marc 3Di Lorenzo Carlo 4Dupont Christophe 5Faure Christophe 6Miqdadi Mohamed 7Osatakul Seksit 8Ribes-Konickx Carmen 9Saps Miguel 4Shamir Raanan 10Staiano Annamaria 11on behalf of the BITSS Study GroupFranckx Johan Green Robin Hegar Badriul Lemmens Roel Salvatore Silvia Vieira Mario Verghote Marc Xinias Ioannis 1 Department of Pediatrics, UZ Brussel, Vrije Universiteit Brussel, Brussels, Belgium 2 Department of Pediatrics, The Medical University of Warsaw, Warsaw, Poland 3 Department of Pediatrics, Emma Children's Hospital/AMC, Amsterdam, The Netherlands 4 Division of Pediatric Gastroenterology, Hepatology, and Nutrition, Nationwide Children's Hospital, Columbus, Ohio, USA 5 Department of Pediatric Gastroenterology Hepatology and Nutrition, Hôpital Necker Enfants Malades, Paris, France 6 Pediatric Gastroenterology, Sainte-Justine Hospital, Montreal, Quebec, Canada 7 Sheikh Khalifa Medical City, Abu Dhabi, UAE 8 Department of Pediatrics, Faculty of Medicine, Prince of Songkla University, Hat Yai, Songkhla, Thailand 9 Pediatric Gastroenterology and Hepatology Unit, La Fe University Hospital, Valencia, Spain 10 Schneider Children's Medical Centre of Israel, Sackler Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel 11 Department of Translational Medical Science, Section of Pediatrics, University of Naples Federico II, Naples, ItalyCorrespondence to Dr Yvan Vandenplas; yvan.vandenplas@uzbrussel.beYV and HS are joint lead authors. 2017 29 3 2017 7 3 e01462011 10 2016 13 2 2017 28 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction The Bristol Stool Form Scale (BSS) which consists of 7 photographs of different stool forms allows assessment of stool consistency (scale 1 for hard lumps to scale 7 for watery stools), in an objective manner in adults. The BSS is also sometimes used to characterise the stools of infants and young children. Despite its use, there is general agreement among paediatric gastroenterologists that the BSS is not adequate to be used in infants and young children who wear diapers; thus, a new scale specifically designed for this population is needed. Our aim is to develop a paediatric stool scale, the Brussels Infant and Toddler Stool Scale (‘BITSS’), and to evaluate the interobserver agreement of stool assessment with the BITSS between the patient's parent and healthcare providers (physicians and nurses). Methods and analysis This study has two phases. In the first phase, 11 key-opinion leaders in the field of paediatric gastroenterology representing different areas of the world selected seven coloured photographs of infants and/or young children wearing diapers to match the original descriptors of the BSS. The selected photographs were used to create a new scale in which the drawings of stools of the BSS were replaced by infant/toddlers stool photographs. In phase II, we aim at demonstrating that parents, nurses and primary healthcare physicians interpret the stool-pictures of the BITSS with a high degree of consensus and that the agreement is independent of whether it is a parent or a healthcare provider. Interobserver variability of stool assessment with the BITSS between the patient's parent and healthcare providers will be assessed. Ethics and dissemination The study will be approved by the Ethics Committee of the participating centres. The findings of this study will be submitted to a peer-reviewed journal. Abstracts will be submitted to national and international conferences. Trial registration number NCT02913950. Bristol stool scalestooldefecation ==== Body Strengths and limitations of this study Every attempt to classify the form of stools has its limitations. The current lack of a ‘gold standard’ tool to assess stool consistency makes it prone to subjective interpretation. The Brussels Infant and Toddler Stool Scale will be a new tool to describe in an objective way stool consistency in non-toilet-trained infants and young children. Worldwide consensus between experts, primary healthcare physicians, nurses and parents will be obtained. Introduction It has been long thought in the medical community that the patients' or caregivers' description of stool consistency was highly subjective and poorly reliable. Thus, in 1992, the Bristol Stool Form Scale (hereafter, this scale is referred to as the BSS) was developed to provide an objective way of describing and reporting stool consistency in adults.1 2 In 1999, the Bristol Stool Chart was added.1 The BSS soon became the gold standard to classify stool consistency. In its present form, the 7-point BSS consists of 7 drawings of different stool forms representing a wide range of stool consistencies (scale 1 for hard lumps to scale 7 for watery stools) (see figure 1). The BSS has excellent reliability and agreement when used to rate individual stool type by different individuals.2 However, the BSS reliability and agreement decreases when applied to Rome III stool form categories.3–5 A modified BSS was shown to be reliable when used also in toilet-trained children 6 years and older.4 Figure 1 Illustration of the Modified Bristol Stool Scale.2 In 2009, the Amsterdam Stool Scale (hereafter, the Amsterdam Scale) was developed and validated, providing additional information on colour and volume of the defecations of non-toilet-trained children and infants.6 In brief, it consists of three visual descriptive scales assessing stool consistency (4-point scale: watery, soft, formed, and hard), amount (4-point scale) and colour (six categories). Compared with the BSS, the Amsterdam Scale was shown to be more suitable for use in infants. Still, the Amsterdam Scale is not universally used; it is too complex and too complicated for routine use by parents, nurses and primary healthcare physicians.7 A recent study showed that among parents of infants and toddlers only fair agreement existed between the BSS and parental report of stool consistency.8 Moreover, it was documented that different methods of assessment of stool consistency did not result in a difference in the prevalence of functional constipation.8 Since there is no better alternative, worldwide, the BSS is used to describe the stools of infants and young children. However, there is consensus among paediatric gastroenterologists that the BSS is not ideal for use in infants and young children who use diapers; thus, a new scale is needed. Objective To develop a paediatric stool form scale and to evaluate the interobserver reliability of stool assessment with the paediatric stool form scale, hereafter referred to as the Brussels Infants and Toddlers Stool Scale (BITSS), between the patient's parent and healthcare providers. Methods and analysis This study has two phases. In the first phase, which was completed between February and April 2016, the drawings of adult stools used in the BSS were replaced by colour photographs of the stools of infants and/or young children, not toilet trained, thus wearing diapers. For this, between February and April 2016, the principal investigator (YV) asked nurses to take photographs of the stools of hospitalised infants and young children. Three cameras were provided to the nurses. after parental consent was obtained, nurses took photographs of diapers containing fresh stools. The type of feeding was not considered as an exclusion or inclusion criterion. Diapers with evidence of melena or fresh blood or mucus were excluded. Then, 28 clear photographs best matching the 7 original descriptions and illustrations used in the original adult BSS were selected during 4 voting sessions with 12 nurses, 2 staff members of the ward and the principal investigator. For each of the 7 original BSS descriptions, the 4 best matching photographs were selected. One hundred per cent agreement was reached in the fourth anonymous voting round. When the photos were grouped, photos 1 and 2 indicated hard stools; photos 3 and 4 indicated normal formed stool; photo 5 indicated loose stools and photos 6 and 7 indicated watery stools. These 28 pictures were sent out to a first group of investigators (HS, AS, CR-K, CD and RS) who selected the 7 ‘pictures’ best representing the original descriptions of the BSS. After two voting rounds, 100% full consensus was reached. The principal investigator then mixed the order of the 7 selected photos and sent them, in combination with the original descriptions and illustrations of the BSS, to 11 key-opinion leaders in the field on paediatric gastroenterology representing different parts of the world. Each colleague was asked to identify the ‘best picture’ matching each of the BSS descriptions. The first voting round revealed a small interobserver variability (Fleiss κ value=0.86). In phase II, we plan to demonstrate that parents, nurses and primary healthcare physicians interpret the stool-photographs with a high degree of consensus and that a high degree of consensus can be achieved regardless of whether a parent or a healthcare provider rates the stools. In order to achieve this goal, we plan to evaluate interobserver variability of stool assessment with the BITSS between the patient's parents and healthcare providers. In most countries, the BSS has been translated into local languages. Otherwise, the local investigator will provide an appropriate translation. The protocol for this study is registered at https://register.clinicaltrials.gov. Participants All authors and members of the BITSS study group agreed to enrol a minimum of 50 parents, 25 nurses and 25 primary healthcare physicians, up to a maximum of 100 individuals (participating centres: see list of coauthors and members of BITSS study group). As a result, data will be obtained from at least 500 parents, 250 nurses and 250 physicians, distributed worldwide. The number of parents, nurses and physicians per centre was arbitrarily chosen. Outcome The primary outcome of this project is to develop a paediatric version of the BSS, based on a broad consensus between healthcare physicians, nurses and parents. Stool composition in non-toilet-trained children is classified according to seven preselected photographs. Stools will be described as follows (figure 1): type 1, ‘separate hard lumps’; type 2 ‘lumpy and sausage like’ (types 1 and 2 will be grouped together and considered as ‘hard stools’); type 3, ‘sausage shaped with cracks on the surface’; type 4, stool in the form of a ‘smooth, soft sausage’ (types 3 and 4 are grouped together and considered to be ‘normal, formed’ composition); type 5, ‘soft blobs with clear-cut edges’, or ‘loose’ stools, but they are also considered normal, for example in breastfed infants; type 6, ‘mushy’ consistency, and type 7, entirely liquid consistency with no solid pieces (types 6 and 7 are considered as ‘watery stools’). Statistical analysis The interobserver variability of stool assessment between the patient's parents and healthcare providers will be evaluated by calculating the proportion of exact agreement and the Fleiss' κ statistics. Correlations, based on the value of κ, will be categorised as poor (κ≤0.2), fair (0.21≤κ≤0.40), moderate (0.41≤κ≤0.60), good (0.61≤κ≤0.80) or excellent (0.81≤κ≤1.00).9 Fleiss’ κ is a statistical measure for assessing the reliability of agreement between a fixed number of raters when assigning categorical ratings to a number of items or classifying items. As the Fleiss' κ value decreases with an increasing number of raters, a ‘moderate’ κ value (>0.41 to 0.60) is accepted as indicative of a correlation that is high enough to be accepted. The Fleiss' κ contrasts with other κs such as Cohen's κ that can only be used to assess the agreement between no more than two raters or the intrarater reliability for one appraiser versus himself. The measure calculates the degree of agreement over that would be expected by chance. The following Fleiss' κ values will be calculated: overall, split-up per photograph, split-up per stool group (hard, normal, loose and watery stools), split-up per rater group (physicians, nurses, and parents) and split-up per country. In cases in which the Fleiss' κ value shows a poor or fair correlation, the reason for this failure will be evaluated and the photographs and/or descriptions will be adapted. Dissemination The findings of this study will be submitted to a peer-reviewed journal. Abstracts will be submitted to relevant national and international meetings. Conclusions The optimal way of classifying stool consistency in infants and young children not yet toilet trained is still debated. Our multicentre study, carried out at research centres with experience in conducting clinical studies in children with gastrointestinal disorders, intends to address a gap in describing stool consistency, providing a much needed tool for the objective assessment of stool consistency for infants and toddlers wearing diapers. Collaborators: The Paediatric Bristol Stool Scale Study Group: Johan Franckx (Belgium), Robin Green (South Africa), Badriul Hegar (Indonesia), Roel Lemmens (Belgium), Silvia Salvatore (Italy), Mario Vieira (Brazil), Marc Verghote (Belgium) and Ioannis Xinias (Greece). Contributors: YV conceptualised the study. YV and HS developed the first draft of the manuscript and contributed equally. All authors contributed to the development of the study protocol and approved the final draft of the manuscript. Funding: The project runs independently from industry, without any funding or financial compensation. All work was and will be carried out ‘freelance’. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. ==== Refs References 1 Heaton KW , Radvan J , Cripps H Defecation frequency and timing, and stool form in the general population: a prospective study . Gut 1992 ;33 :818 –24 . 10.1136/gut.33.6.818 1624166 2 Lewis SJ , Heaton KW Stool form scale as a useful guide to intestinal transit time . Scand J Gastroenterol 1997 ;32 :920 –4 . 10.3109/00365529709011203 9299672 3 Chumpitazi BP , Lane MM , Czyzewski DI Creation and initial evaluation of a Stool Form Scale for children . J Pediatr 2010 ;157 :594 –7 . 10.1016/j.jpeds.2010.04.040 20826285 4 Lane MM , Czyzewski DI , Chumpitazi BP Reliability and validity of a modified Bristol Stool Form Scale for children . J Pediatr 2011 ;159 :437 –41.e1 . 10.1016/j.jpeds.2011.03.002 21489557 5 Chumpitazi BP , Self MM , Czyzewski DI Bristol Stool Form Scale reliability and agreement decreases when determining Rome III stool form designations . Neurogastroenterol Motil. 2016 ;28 :443 –8 . 10.1111/nmo.12738 26690980 6 Bekkali N , Hamers SL , Reitsma JB Infant stool form scale: development and results . J Pediatr. 2009 ;154 :521 –26.e1 . 10.1016/j.jpeds.2008.10.010 19054528 7 Ghanma A , Puttemans K , Deneyer M Amsterdam infant stool scale is more useful for assessing children who have not been toilet trained than Bristol stool scale . Acta Paediatr 2014 ;103 :e91 –2 . 10.1111/apa.12422 24107091 8 Koppen IJ , Velasco-Benitez CA , Benninga MA Using the Bristol Stool Scale and parental report of stool consistency as part of the Rome III Criteria for functional constipation in infants and toddlers . J Pediatr 2016 ;177 :44 –8.e1 . 10.1016/j.jpeds.2016.06.055 27453373 9 Altman DG Practical statistics for medical research . London : Chapman & Hall , 1991 .

PMC005xxxxxx/PMC5372037.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01476910.1136/bmjopen-2016-014769Health Services ResearchResearch150617041727169416981703Scoping review of potential quality indicators for hip fracture patient care Pitzul Kristen B 1Munce Sarah E P 2Perrier Laure 3Beaupre Lauren 45Morin Suzanne N 6McGlasson Rhona 7Jaglal Susan B 128 1 Institute of Health Policy, Management, and Evaluation, University of Toronto, Toronto, Ontario, Canada 2 Toronto Rehabilitation Institute, University Health Network, Toronto, Ontario, Canada 3 Gerstein Science Information Centre, University of Toronto, Toronto, Ontario, Canada 4 Department of Physical Therapy, University of Alberta, Edmonton, Alberta, Canada 5 Department of Surgery, University of Alberta, Edmonton, Alberta, Canada 6 Department of Medicine, McGill University, Montreal, Quebec, Canada 7 Bone and Joint Canada, Toronto, Ontario, Canada 8 Department of Physical Therapy, University of Toronto, Toronto, Ontario, CanadaCorrespondence to Kristen B Pitzul; kristen.pitzul@mail.utoronto.ca2017 20 3 2017 7 3 e01476917 10 2016 17 1 2017 21 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective The purpose of this study is to identify existing or potential quality of care indicators (ie, current indicators as well as process and outcome measures) in the acute or postacute period, or across the continuum of care for older adults with hip fracture. Design Scoping review. Setting All care settings. Search strategy English peer-reviewed studies published from January 2000 to January 2016 were included. Literature search strategies were developed, and the search was peer-reviewed. Two reviewers independently piloted all forms, and all articles were screened in duplicate. Results The search yielded 2729 unique articles, of which 302 articles were included (11.1%). When indicators (eg, in-hospital mortality, acute care length of stay) and potential indicators (eg, comorbidities developed in hospital, walking ability) were grouped by the outcome or process construct they were trying to measure, the most common constructs were measures of mortality (outcome), length of stay (process) and time-sensitive measures (process). There was heterogeneity in definitions within constructs between studies. There was also a paucity of indicators and potential indicators in the postacute period. Conclusions To improve quality of care for patients with hip fracture and create a more efficient healthcare system, mechanisms for the measurement of quality of care across the entire continuum, not just during the acute period, are required. Future research should focus on decreasing the heterogeneity in definitions of quality indicators and the development and implementation of quality indicators for the postacute period. HEALTH SERVICES ADMINISTRATION & MANAGEMENTREHABILITATION MEDICINE ==== Body Strengths and limitations of this study This study includes potential indicators and indicators for hip fracture quality of care throughout the entire continuum of care and not just within the acute care period. The search strategy was performed by an experienced information scientist and peer-reviewed by another information scientist outside the study team. The screening and extraction were performed completely in duplicate. Non-English studies were not included, and there may therefore be a bias towards inclusion of studies performed in English-speaking countries. Background In 2000, ∼1.6 million people worldwide suffered from a fragility hip fracture (herein referred to as ‘hip fracture’), with this number projected to increase to 21 million by the year 2050.1 2 In the UK alone, there is predicted to be over 100 000 hip fractures by the year 2020.3 Half of the persons who suffer from hip fractures never return to premorbid function, even two years postfracture, and the direct attributable one-year mortality rates for hip fractures are between 20 and 30%.4–7 Patients with hip fractures have significantly higher acute care costs than matched controls as well as high postacute costs due to rehabilitation required after surgery.7–12 To help mitigate this extensive morbidity, mortality and healthcare use, it is imperative that quality care is delivered to patients with hip fracture. The Institute of Medicine defines quality care as “the degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge”.13 In essence, delivering quality care means delivering evidence-based care that has a good chance of improving a patient's health outcome(s). To ensure that quality care is delivered, a number of countries that have traditionally funded their institutions with global budgets (eg, Canada,14 the UK15) have begun implementing performance-based funding (ie, linking quality of care delivery to funding policies). One of the goals of performance-based funding, such as the Best Practice Tariff implemented in 2010 in the UK, is to assign increased accountability to institutions for the care that they deliver.16 Although the measurement of quality of care has always been imperative to healthcare delivery, the importance of choosing which aspects of quality of care to measure is highlighted in the context of performance-based funding models.17 Quality of care indicators are typically embedded within a performance framework (eg, a balanced scorecard) and are used to measure the structure, process and/or outcomes of care.18–20 Quality of care indicators measure how much deviation (if any) there is between the healthcare being delivered and best practice.13 Examples of the use of quality of care indicators to improve care delivery include the Surgical Safety Checklist (implemented in eight countries), and the National Hip Fracture Audit in the UK.15 21 There are a variety of approaches for developing quality of care indicators, including deductive (from concept to data) and inductive (from data to concept).17 22 23 Using either approach, when evidence is weak or non-existent, expert consensus is used to develop the indicator.23 Current hip fracture quality of care indicators target the acute care period, most likely due to the focus on performance-based funding for acute care institutions (ie, an incentive to measure quality). In the UK, The National Institute for Health and Care Excellence (NICE), along with the British Orthopaedics Association (BOA), has developed several hip fracture quality indicators (eg, time to surgery, assessment by ortho-geriatrician within 72 hours of admission) as part of their performance-based funding for acute care.3 15 24 25 These indicators have subsequently been used in other health systems worldwide (eg, Australia, Canada and other European countries).26–33 However, these indicators are focused solely in the acute care period and therefore do not measure the quality of care in the postacute care period. Although some studies have used process and outcome measures to evaluate care delivered in the postacute period (eg, a home-based rehabilitation programme), there is a lack of identified, evidence-based quality of care indicators in the postacute care period for patients with hip fracture.34–38 Without measures of quality of care in the postacute period for patients with hip fracture, frontline staff, administrators and policymakers are left without required information to assess the delivery of care during the postsurgical rehabilitation period.39 40 The objective of this study is to synthesise the evidence on existing or potential quality of care indicators for the acute period, the postacute period and across the entire continuum of care for patients following a hip fracture. Using a scoping review methodology, the specific research question to be addressed was: “What patient, institutional, and system-level indicators are currently in use or could potentially be used for measuring quality of care in the acute period, post-acute period, and across the continuum for older individuals following a hip fracture?”. Methods Study design and literature search strategies A scoping review methodology was employed, of which details are published elsewhere (see online supplementary file 1).41 Briefly, Arksey and O'Malley42 as well as Levac et al43 frameworks were used to guide the scoping review. Measures targeted at patients, institutions or health systems were included and encompassed care processes and outcomes in the acute and postacute period. For the purposes of this review, quality indicators were defined as validated process or outcome measures with a descriptive statement that were used to describe quality of care delivered.22 A potential quality indicator was defined as a process or outcome measure of care that was not specifically identified or referenced as an indicator of quality of care by the authors. This synthesis focused on quality of care indicators for older adults (aged 50 years and over) with non-pathological hip fracture caused by low trauma (eg, a fall from standing height or less). All study designs were included and only studies or abstracts published from the year 2000-January 2016, or in English were included to ensure relevance to the current healthcare context and feasibility. 10.1136/bmjopen-2016-014769.supp1supplementary file 1 Literature search strategies were developed using medical subject headings (MeSH) and text words related to hip fracture quality indicators. MEDLINE, EMBASE, CINAHL, Ageline, PEDRO (physiotherapy evidence database) and the Cochrane Central Register of Controlled Trials (CENTRAL) databases were searched on 18 January 2016, and the MEDLINE search was peer-reviewed.44 Searches were performed with no language restrictions and limited from 1 January 2000 (see online supplementary file 2). The search used combinations of the following terms: hip fracture, femoral fracture, process indicator, process measure and quality indicator. Appropriate wildcards were used in the search to account for plurals and variations in spelling. 10.1136/bmjopen-2016-014769.supp2supplementary file 2 Study selection and data abstraction Two reviewers (KBP and SEPM) piloted level 1 (titles and abstracts) and level 2 (full article texts) screening forms, as well as the extraction form (see online supplementary file 3). All screening and extraction were completed in duplicate. Disagreements were discussed between the two reviewers and a third party reviewer (LB, SNM or SBJ) was contacted if disagreements could not be resolved. 10.1136/bmjopen-2016-014769.supp3supplementary file 3 Abstracted data included study characteristics (eg, year of publication, country of study), indicator definitions (eg, length of stay defined as the number of total days stayed at institution without interruption) and numerator and denominator definitions when applicable (eg, per 1000 hip fractures). We examined the purpose and components of the indicators as well as the reported measurement properties, if applicable. Study setting was abstracted and defined as follows: acute care (any acute care institution or department within an acute care institution); postacute care (any institution or community setting used after discharge from acute care) or across the continuum of care (studies that include acute and postacute settings). Study quality was not assessed during the scoping review as the objective of a scoping review is to identify gaps in the literature and highlight future areas for systematic review.42 43 Studies were then summarised using numerical counts, and definitions of indicators or potential indicators were summarised. Results The literature searches yielded a total of 3828 articles (figure 1). After duplicates were removed, 2729 articles were included in level 1 screening. After level 1 screening was complete, 638 articles (23%) were included in full-text screening (ie, level 2). After level 2 screening was complete, data were extracted from 302 articles (11% of initial yield). Agreement between the two reviewers (KBP and SEPM) ranged from 75 to 85% for both searches. Reasons for article exclusion varied, but were primarily due to incorrect study population (ie, study included older adults with pathological hip fractures). Figure 1 The total number of articles yielded from the literature search in 2016, and the final number of articles included in the study. CENTRAL, the Cochrane Central Register of Controlled Trials; PEDRO, physiotherapy evidence database. Synthesis Owing to the volume of studies included in data extraction (N=302), indicators or potential indicators were grouped into process and outcome constructs that the authors were trying to measure. The creation of these constructs was therefore data-driven and included measures of mortality, time-to (eg, time to surgery, time spent in the emergency department), length of stay, functional ability, comorbidities and complications, discharge destinations, balance and mobility, quality of life, pain, cognitive, readmissions, the UK's Best Practice Tariff indicators (BPTs), prophylaxis (eg, antibiotic prophylaxis) and blood (eg, blood loss, blood transfusion), osteoporosis testing and medications, falls, healthcare usage, nutrition (eg, vitamin D levels), biometrics (eg, muscle strength tests), catheters, patient satisfaction, caregivers (eg, burden and stress), self-efficacy (eg, self-care ability) and other (eg, patient safety strategies). Developing these constructs greatly improved the feasibility of data synthesis, particularly for comparing variations in indicator or potential indicator definitions between studies. If the same indicator or potential indicator (eg, Berg Balance Scale) was used by different studies for measuring different constructs (eg, functional ability as well as balance and mobility), it was placed into all relevant constructs. Owing to the nature of the topic reviewed (ie, indicators or potential indicators), after screening, there were no qualitative studies for data extraction. General characteristics of included studies When individual countries were compared, most studies were conducted in the UK (27%), the USA (22%) or Australia and New Zealand (13%) (table 1). Most study settings were within the acute care period (78%), with a paucity of studies conducted in the postacute care period (8%) (table 1). The most common study design was retrospective cohort (28%), followed by prospective cohort (18%) and reviews (17%) (table 1). Within the included studies, the majority of indicators or potential indicators were at the patient level. Table 1 Country of study, study setting and study design, number of included studies (n, % total n) Number (% total*) of Studies Country of study  Other Europe 84 (28%)  UK 81 (27%)  USA 67 (22%)  Australia and New Zealand 39 (13%)  Nordic Countries (Norway, Denmark, Finland, Sweden) 48 (16%)  Canada 28 (9%)  Asia 13 (4%)  Middle East (Israel, India) 8 (3%) Study setting  Acute 237 (78%)  Postacute (any) 24 (8%)  Across the continuum of care 41 (14%) Study design  Retrospective cohort 85 (28%)  Prospective cohort 53 (18%)  Review (scoping, systematic, etc) 50 (17%)  Randomised controlled trial 34 (11%)  Clinical Audit 31 (10%)  Experimental (eg, pre-post) 19 (6%)  Population-based cohort 9 (3%)  Descriptive 7 (2%)  Cross-sectional 6 (2%)  Focus groups/interviews/consensus meetings 4 (1%)  Pilot study 2 (1%)  Survey 2 (1%) *The total percentage does not add up to 100% for country of study as some studies took place in multiple countries. Most common indicators or potential indicators When indicators or potential indicators were grouped into process and outcome constructs, mortality and time-to (eg, time to surgery) constructs were most commonly reported (42% and 35%, respectively, of the included studies) (table 2). Length of stay, functional ability, comorbidities and complications, discharge destinations and balance and mobility indicators or potential indicators were present in over 20% of the included studies. Indicators or potential indicators of self-efficacy, caregivers and patient satisfaction were the least commonly reported (2%, 2% and 2%, respectively). Table 2 Number of studies (n, % total N studies), by process or outcome constructs containing indicators or potential indicators Construct Indicators or potential indicators n (%N) Mortality In-hospital mortality; postdischarge mortality (eg, 30 days, 90 days) 125 (41%) Time-to Time from presentation to admission; time from admission to medical clearance; time from admission to surgery/surgical delay; operative time; time to rehabilitation 106 (35%) Length of stay During acute care; during intensive care; during rehabilitation; during a readmission 93 (31%) Comorbidities and complications Developed in-hospital; developed postoperatively; number present at admission; classification (major vs minor); adverse events in-hospital; pressure ulcers; urinary tract infections; venous thromboembolism; reoperation; infections 86 (29%) Functional ability Activities of daily living (ADLs) or instrumental ADLs (IADLs); short physical performance battery (SPPB); functional independence measure (FIM); Barthel or Modified Barthel Index; Katz or Modified Katz; Timed up and go; Harris Hip Score; SF36; Tinetti's Fall Efficacy Scale; Berg Balance Scale; sit to stand test; Frenchay's Activity Index; Activity Measure for Post-acute care (AM-PAC); Other measures 84 (28%) Discharge destinations Novel institutionalisation; change in premorbid level of care; discharge destination (eg, home, long-term care); successful community discharge 64 (21%) Balance and mobility Mobile yes/no; ability (eg, walking distance); ambulation decline; balance (eg, postural sway); weight bearing 62 (21%) Quality of life (QOL) EQ5D, EuroQOL (includes EQ5D and EQ-Visual Analog Scale); Health-related QOL (HRQOL); Dementia assessment for QOL (DEMQOL); Swedish QOL (SWED-QUAL); WHO's Brief QOL (WHOQOL-BREF); Short Form 12 (SF12), 36 (SF36) and 6D (SF6D); Western Ontario and McMaster Short Form (WOMAC-SF); Other (eg, Health Utilities Index) 39 (13.0%) Other Organisation's performance evaluation system (130 simple indicators and 50 composite measures); day of admission; maintainability (ie, unexpected event, including deaths, readmission or change in level of care); weight; composite poor outcome (eg, death or readmission); patient safety strategies 30 (10%) Pain Presence of pain (acute, chronic); pain score in EQ5D; assessment of pain (yes/no); use of analgesia (yes/no and type) 29 (10%) Readmissions 15, 28 or 30 days; 2 months, 4 months, monthly; 1 year 26 (9%) Cognitive Score or status (eg, mini mental status examination score); depression (yes/no); delirium (yes/no) 25 (8%) UK's Best Practice Tariff Indicators (BPT) Admission under consultant-led joint orthogeriatric care; admission using a multidisciplinary assessment protocol; Geriatric-directed multidisciplinary rehabilitation; perioperative assessment by geriatrician or ortho-geriatrician within 72 hours of admission to emergency department; Admission to ward from emergency department within 4 hours; Assessment for falls and bone protection 24 (8%) Prophylaxis and blood Antibiotic prophylaxis and anticoagulation (yes/no and type); blood loss (amount); blood transfusion (yes/no) 23 (8%) Osteoporosis testing and medication Bone mineral density testing; medication postoperatively or at acute discharge 21 (7%) Falls Crude count; prevention (eg, falls prevention programme); assessment for falls risk (in-hospital and postdischarge); Tinetti's fall efficacy scale; self-report falls at various time points postacute discharge 19 (6%) Healthcare usage Costs; community services; physical therapy visits (acute and postacute); composite measures (readmission, emergency department visit) 16 (5%) Nutrition Compliance with diet/nutrition interventions; vitamin D (amount); assessment (includes time to assessment) 13 (4%) Biometrics Neuromuscular assessment or status; muscle strength; muscle contraction; knee specific measures 11 (4%) Catheters Catheter yes/no; time to removal 8 (3%) Patient satisfaction Questionnaire/interview with various questions (eg, questions about satisfaction with information provided about hospital care) 5 (2%) Caregivers Support provided (eg, Social Support Scale); burden and stress (eg, Caregiver Strain Index) 4 (1%) Self-efficacy Self-care ability; self-efficacy for exercise 4 (1%) Indicators or potential indicators by study setting When the constructs are stratified by study setting, the paucity of potential indicators or indicators in the postacute period or across the continuum of care is highlighted (table 3). All of the studies with indicators or potential indicators in the Best Practice Tariff construct are set in the acute care period, as are the vast majority of studies with indicators or potential indicators in the ‘other’ construct (88%) or within the ‘time-to’ construct (87%) (table 3). Table 3 Construct of indicators or potential indicators, stratified by study setting (n and % total N of each construct) Construct Acute (n, %N) Postacute (n, % N) Across the continuum (n, %N) Mortality 95 (76%) 6 (5%) 24 (19%) Time-to 93 (88%) 0 (0%) 13 (12%) Length of stay 65 (70%) 8 (9%) 20 (22%) Comorbidities and complications 60 (69%) 2 (2%) 24 (28%) Functional ability 51 (50%) 22 (26%) 25 (30%) Discharge destination 38 (59%) 13 (20%) 13 (20%) Balance and mobility 33 (53%) 12 (19%) 17 (27%) Quality of life 15 (39%) 9 (22%) 15 (37%) Other 26 (87%) 0 (0%) 4 (13%) Pain 17 (59%) 4 (14%) 8 (28%) Readmissions 15 (58%) 3 (12%) 8 (31%) Cognitive 17 (68%) 4 (16%) 4 (16%) Best practice tariff 24 (100%) 0 (0%) 0 (0%) Prophylaxis and blood 17 (74%) 0 (0%) 6 (26%) Osteoporosis testing and medication 16 (76%) 2 (10%) 3 (14%) Falls 14 (74%) 3 (16%) 2 (11%) Healthcare utilization 8 (50%) 5 (31%) 3 (19%) Nutrition 10 (77%) 0 (0%) 3 (23%) Biometrics 2 (18%) 9 (82%) 0 (0%) Catheters 6 (69%) 0 (0%) 2 (25%) Patient satisfaction 2 (40%) 0 (0%) 3 (60%) Caregivers 1 (25%) 1 (25%) 2 (50%) Self-efficacy 2 (50%) 2 (50%) 0 (0%) The proportion of studies with indicators or potential indicators classified as functional ability or quality of life constructs was distributed between the acute care and the postacute periods as well as across the continuum of care (table 3). Since the goal of rehabilitation is to restore prefracture functional ability and quality of life, this broader distribution is not surprising. Finally, certain indicators or potential indicators are unlikely to occur in the postacute period because they are less relevant (eg, those within the catheter or prophylaxis and blood constructs) (table 3). When indicators or potential indicators were examined within each construct, there was substantial heterogeneity in definitions, including variations in when the indicator or potential indicator was measured, as well as the categorisation of categorical measures (eg, different cut-points on a scale). For example, length of acute care stay was measured from time to ward admission to discharge as well as from time to emergency department presentation to discharge.45 46 Definitions of time to surgery also varied, as some studies defined time to surgery as the time from medical stability to surgery;47 48 some studies defined time to surgery as time from admission to surgery,48–55 and others created a binary variable for time to surgery (eg, had surgery within 24 hours).56–59 The one exception was the UK's Best Practice Tariff indicators, which are clearly defined across studies. Compared to potential indicators or indicators implemented in the acute care period, even greater variability was seen for potential indicators or indicators implemented in the postacute period. Potential indicators and indicators within the functional ability and quality of life constructs are discussed below as exemplifiers of this extensive variability, as both constructs are established goals of rehabilitation and were prevalent in studies set in the postacute period and/or across the continuum of care (see online supplementary files 4 and 5). 10.1136/bmjopen-2016-014769.supp4supplementary file 4 10.1136/bmjopen-2016-014769.supp5supplementary file 5 Indicators or potential indicators in postacute: functional ability and quality of life Most measures of functional ability were validated scores or scales, such as the Functional Independence Measure (FIM),60–72 the Barthel Index (BI) (or Modified Barthel Index (MBI)),38 73–91 and the Activity Measure for Post-acute Care (AM-PAC)34 92 (see online supplementary file 4). Furthermore, change in functional ability (ie, difference in functional ability between two time points)36 69 70 78 93–97 was only used in 13.0% of studies measuring functional ability (see online supplementary file 4). Quality of life, similar to functional ability, was measured primarily using validated scores or scales, such as the SF3638 66 76 87 97–102 (see online supplementary file 5). Some studies used modified validated scales or scores, such as the EuroQOL (European Quality of Life measure which includes EQ5D and a visual analogue scale for pain).103 104 When quality of life was measured, it varied substantially from 3 to 4 days postoperatively to 1 year after acute care discharge.76 102 105 Changes in quality of life between two time points (compared to measurement of quality of life at one time point) were not measured in any of the included studies (see online supplementary file 5). Discussion The purpose of this study was to synthesise the literature for quality indicators or potential quality indicators for patients with hip fracture within the acute period, postacute period and across the continuum of care. Most studies were from the UK and contained patient-level indicators implemented within the acute care period. There was substantial variability in terms of indicator or potential indicator definitions among studies, particularly in the postacute period. This is particularly evident in the functional ability and quality of life constructs, two outcome constructs that are important to rehabilitation of patients with hip fracture. The most common process or outcome constructs were those that measured mortality, ‘time-to’ (eg, time spent in the emergency department on presentation) and length of stay. They were most often implemented in the acute care period. This prevalence may be, in part, due to the fact that these constructs contain measures currently in use as quality indicators in numerous health systems: in-hospital and 30-day mortality, time from emergency department to acute admission and time to surgery.3 14 24 However, even with respect to these known metrics, there were differences in time and type of measurement between studies. Although these differences may sometimes appear to be nuanced or negligible (eg, whether or not length of acute care stay includes time spent in the emergency department), they can be impactful if the indicator or potential indicators play a role in institutional funding (ie, performance-based funding). Differences in definitions may also be due to changes in best practice that occurred during the study time frame (eg, recommendation of time to surgery within 48 hours compared to time to surgery within 36 hours).3 24 This study highlights the lack of indicators or potential indicators implemented within the postacute care period for patients with hip fracture in the literature. This finding supports the conclusions of Duncan and Velozo39 and Leland and colleagues.40 Almost 10 years ago, Duncan and Velozo concluded that although validated outcome measures exist in the postacute rehabilitation period, there is a lack of quality indicators to actually assess the care delivered in the USA.39 Similar conclusions were made more recently by Leland and colleagues, who stated that owing to the limited number of quality of care measures in the postacute period, stakeholders (ie, patients, families, payers and providers) are left without required information to make important decisions for hip fracture rehabilitation in the USA.106 Two important potential indictors for quality of rehabilitative care in the postacute period that were found in the literature were functional ability and quality of life. Functional ability and quality of life constructs were very heterogeneous, in terms of potential indicators and indicator definitions, with no dominant measure reported, making comparisons among studies difficult. This also limits the utility of evidence in the development of quality indicators that can be applied to entire health systems and tied to financial models. The heterogeneity between performance measures for hip fracture was also discussed in a review performed by Giusti and colleagues, which concluded that measures for functional ability varied so substantially that results between studies were not comparable.107 The results of this study are supported by current literature in other rehabilitation populations. Mont et al108 found that few rating scales assessed all aspects of outcomes (including quality of life, rehabilitative and patient satisfaction) following total knee arthroplasty. A systematic review by Ritchie et al109 on measures of community integration for persons with traumatic brain injury found that more research is needed to inform best practice guidelines. Sleat et al110 reviewed current practice of trauma registries and found that most registries failed to measure outcomes such as morbidity and quality of life, which are needed to drive service improvement in the long term. Rinere O'Brien systematically reviewed the evidence to determine the impact of a new payment system implemented in the USA on quality of care indicators for inpatient stroke rehabilitation and found that lack of data with respect to the quality of care indicators made it difficult to ascertain conclusions.111 Recently, however, cardiac rehabilitation (which includes stroke rehabilitation) has made progress in terms of quality indicator development and implementation compared to other rehabilitation populations.112 Grace et al112 described the creation of quality indicators for cardiac rehabilitation (eg, ‘percentage of eligible in-patients referred to a cardiac rehabilitation program’ and ‘number of days between receipt of referral to a cardiac rehabilitation program and patient enrollment for eligible patients’) and secondary prevention through a literature review and consensus process led by the Canadian Cardiovascular Society. These advancements in quality indicators for cardiac rehabilitation can help inform future research and protocols on the development of indicators to assess quality of care delivered to patients with hip fracture and other rehabilitation populations in the postacute period. This study was not without limitations. First, scoping reviews do not assess study quality and, as such, information extracted from weak and strong studies is considered. Second, non-English studies were not included and there may therefore be a bias towards inclusion of studies performed in English-speaking countries. Third, owing to the considerable amount of time required to conduct scoping reviews, the search was completed 11 months ago and therefore more recent and relevant studies may be excluded. Fourth, inclusion of original research and review articles may have resulted in duplication of some results. Despite these limitations, this study has several strengths. First, it includes potential indicators and indicators for hip fracture quality of care throughout the entire continuum of care and not just within the acute care period. Second, the literature search was performed by an experienced information scientist, and the screening and extraction were performed completely in duplicate. Third, the search itself was peer-reviewed. To improve quality of care for patients and create a more efficient healthcare system, mechanisms for the measurement of quality of care are required. The implementation of quality of care indicators enables stakeholders to target areas for improvement in service delivery. Although acute care quality indicators for patients with hip fracture have been implemented in many health systems, there is a paucity of indicators and heterogeneity in potential indicators in the postacute care period. Owing to the requirement for rehabilitation after surgery for patients with hip fracture, the inability to measure quality of care in the postacute period is concerning. Future research should focus on collaborative efforts to decrease indicator heterogeneity as well as to develop a framework for indicators that could be shared globally. This would increase accountability and help ensure that quality care is delivered to patients with hip fracture worldwide. The authors thank Ms Saima Hossain, Ms Atiya Hemraj and Ms Julie Zhang for their help in retrieving articles. Contributors: KBP involved in writing protocol, screening for all levels, extraction, synthesis and writing of manuscripts and was the project coordinator. SEPM involved in editing of protocol, screening for all levels, extraction, synthesis and editing manuscripts. LP involved in editing of protocol, search strategy and editing manuscripts. LB contributed to editing of protocol, content expert input (rehabilitation), partial screening level 1 and editing manuscripts. SNM involved in editing of protocol, content expert input (clinical), partial screening level 1 and editing manuscripts. RM contributed to editing of protocol, stakeholder input (BJC), partial screening level 1 and editing manuscripts. SBJ involved in project conception, editing of protocol, partial screening level 1 and editing manuscripts and was the senior responsible investigator. Funding: This work was supported by a Technology Evaluation in the Elderly Network Knowledge Synthesis Grant # 2013–07, which is funded federally through the National Centers of Excellence. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Further details on studies included in this scoping review can be retrieved by contacting the corresponding author at kristen.pitzul@mail.utoronto.ca. ==== Refs References 1 Parker M , Johansen A Hip Fracture . BMJ 2006 ;333 :27 –30 . 10.1136/bmj.333.7557.27 16809710 2 Johnell O , Kanis JA An estimate of the worldwide prevalence and disability associated with osteoporotic fractures . Osteoporos Int 2006 ;17 :1726 –33 . 10.1007/s00198-006-0172-4 16983459 3 British-Orthopaedic-Association . The care of patients with fragility fractures . British Orthopaedic Association , 2007 http://www.bgs.org.uk/pdf_cms/pubs/Blue%20Book%20on%20fragility%20fracture%20care.pdf . 4 Bertram M , Norman R , Kemp L Review of the long-term disability associated with hip fractures . Inj Prev 2011 ;17 :365 –70 . 10.1136/ip.2010.029579 21486987 5 Magaziner J , Fredman L , Hawkes W Changes in functional status attributable to hip fracture: a comparison of hip fracture patients to community-dwelling aged . Am J Epidemiol 2003 ;157 :1023 –31 . 10.1093/aje/kwg081 12777366 6 Norton R , Butler M , Robinson E Declines in physical functioning attributable to hip fracture among older people: a follow-up study of case–control participants . Disabil Rehabil 2000 ;22 :345 –51 . 10.1080/096382800296584 10896094 7 Nikitovic M , Wodchis WP , Krahn MD Direct health-care costs attributed to hip fractures among seniors: a matched cohort study . Osteoporos Int 2013 ;24 :659 –69 . 10.1007/s00198-012-2034-6 22736067 8 Stukel TA , Fisher ES , Alter DA Association of hospital spending intensity with mortality and readmission rates in Ontario hospitals . J Am Med Assoc 2012 ;10 :1037 –45 . 10.1001/jama.2012.265 9 Wilkins K Health care consequences of falls for seniors . Health Rep 1999 ;10 :47 –55 .10607412 10 Haentjens P , Autier P , Barette M The economic cost of hip fractures among elderly women. A one-year, prospective, observational cohort study with matched-pair analysis. Belgian Hip Fracture Study Group . J Bone Joint Surg Am 2001 ;4 :493 –500 . 10.2106/00004623-200104000-00003 11 Dai K , Zhang W , Fan T Estimation of resource utilization associated with osteoporotic hip fracture and level of post-acute care in China . Curr Med Res Opinion 2007 ;23 :2937 –43 . 10.1185/030079907X233061 12 Pasco JA , Sanders KM , Hoekstra FM The human cost of fracture . Osteoporos Int 2005 ;16 :2046 –52 . 10.1007/s00198-005-1997-y 16228106 13 IOM . Crossing the Quality Chasm: A new health system for the 21st century . Washington (DC ): Institute of Medicine , 2001 . 14 HQO, MOHLTC . Quality-based procedures: clinical handbook for hip fracture . Toronto (ON ): Health Quality Ontario , 2013 . 15 Royal-College-of-Physicians . National Hip Fracture Database (NHFD) annual report 2014 . London : Royal-College-of-Physicians , 2014 . 16 Department-of-Health . Prevention package for older people. http://webarchive.nationalarchives.gov.uk/20130107105354/http://dh.gov.uk/en/index.htm 2009 . 17 Shekelle PG Quality indicators and performance measures: methods for development need more standardization . J Clin Epidemiol 2013 ;66 :1338 –9 . 10.1016/j.jclinepi.2013.06.012 24018346 18 Donabedian A The quality of care: how can it be assessed? J Am Med Assoc 1988 ;260 :1743 –8 . 10.1001/jama.1988.03410120089033 19 Arah OA , Klazinga NS , Delnoij DMJ Conceptual frameworks for health systems performance: a quest for effectiveness, quality, and improvement . Int J Qual Health Care 2003 ;15 :377 –98 . 10.1093/intqhc/mzg049 14527982 20 Kaplan RS , Norton DP The balanced scorecard-measures that drive performance . Harv Bus Rev 1992 ;70 :71 –9 .10119714 21 Haynes AB , Weiser TG , Berry WR A surgical safety checklist to reduce morbidity and mortality in a global population . N Engl J Med 2009 ;360 :491 –9 . 10.1056/NEJMsa0810119 19144931 22 Stelfox HT , Bobranska-Artiuch B , Nathens A Quality indicators for evaluating trauma care: a scoping review . Arch Surg 2010 ;145 :286 –95 . 10.1001/archsurg.2009.289 20231631 23 Stelfox HT , Straus SE Measuring quality of care: considering conceptual approaches to quality indicator development and evaluation . J Clin Epidemiol 2013 ;66 :1328 –37 . 10.1016/j.jclinepi.2013.05.017 24018342 24 NICE publishes advice on management of hip fracture. Guidelines in Practice 2011 ;14 :8 . 25 NHS . NHS performance indicators: a consultation . London : NHS , 2001 . 26 Sivakumar BS , McDermott LM , Bell JJ Dedicated hip fracture service: implementing a novel model of care . ANZ J Surg 2013 ;83 :559 –63 . 10.1111/j.1445-2197.2012.06201.x 22943562 27 Khan SK , Weusten A , Bonczek S The Best Practice Tariff helps improve management of neck of femur fractures: a completed audit loop . Br J Hosp Med (Lond) 2013 ;74 :644 –7 . 10.12968/hmed.2013.74.11.644 24220529 28 Currie CT , Hutchison JD Audit, guidelines and standards: clinical governance for hip fracture care in Scotland . Disabil Rehabil 2005 ;27 :1099 –105 . 10.1080/09638280500056329 16278178 29 Hughson J , Newman J , Pendleton RC Hip fracture management for the hospital-based clinician: a review of the evidence and best practices . Hosp Pract (Minneap) 2011 ;39 :52 –61 . 10.3810/hp.2011.02.374 30 Teo SP , Mador J Orthogeriatrics service for hip fracture patients in Dunedin hospital: achieving standards of hip fracture care . J Clin Gerontol Geriatr 2012 ;3 :62 –7 . 10.1016/j.jcgg.2012.04.007 31 Shaw CD , Groene O , Botje D The effect of certification and accreditation on quality management in 4 clinical services in 73 European hospitals . Int J Qual Health Care 2014 ;26 (Suppl 1):100 –7 . 32 Sunol R , Wagner C , Arah OA Evidence-based organization and patient safety strategies in European hospitals . Int J Qual Health Care 2014 ;26 (Suppl 1 ):47 –55 . 10.1093/intqhc/mzu016 24578501 33 Ellanti P , Cushen B , Galbraith A Improving hip fracture care in Ireland: a preliminary report of the Irish hip fracture database . J Osteoporos 2014 ;2014 :656357 .25548713 34 Chang FH , Latham NK , Ni P Does self-efficacy mediate functional change in older adults participating in an exercise program after hip fracture? A randomized controlled trial . Arch Phys Med Rehabil 2015 ;96 :1014 –1020.e1 . 10.1016/j.apmr.2015.02.009 25701101 35 Resnick B , D'Adamo C , Shardell M Adherence to an exercise intervention among older women post hip fracture . J Clin Sport Psychol 2008 ;2 :41 –56 . 10.1123/jcsp.2.1.41 20428489 36 Jung HY , Meucci M , Unruh MA Antipsychotic use in nursing home residents admitted with hip fracture . J Am Geriatr Soc 2013 ;61 :101 –6 . 10.1111/jgs.12043 23252409 37 Anand S , Buch K Post-discharge symptomatic thromboembolic events in hip fracture patients . Ann R Coll Surg Engl 2007 ;89 :517 –20 . 10.1308/003588407X202065 17694599 38 Smith TO , Hameed YA , Cross JL Enhanced rehabilitation and care models for adults with dementia following hip fracture surgery . Cochrane Database Syst Rev 2015 ;(6) :CD010569 10.1002/14651858.CD010569.pub2 .26074478 39 Duncan PW , Velozo CA State-of-the-Science on postacute rehabilitation: measurement and methodologies for assessing quality and establishing policy for postacute care . Arch Phys Med Rehabil 2007 ;88 :1482 –7 . 10.1016/j.apmr.2007.08.118 17964893 40 Leland NE , Gozalo P , Bynum J What happens to patients when they fracture their hip during a skilled nursing facility stay? J Am Med Dir Assoc 2015 ;16 :767 –74 . 10.1016/j.jamda.2015.03.026 25944177 41 Pitzul KB , Munce SEP , Perrier L Quality indicators for hip fracture patients: a scoping review protocol . BMJ Open 2014 ;4 :e006543 10.1136/bmjopen-2014-006543 42 Arksey H , O'Malley L Scoping studies: towards a methodological framework . Int J Soc Res Methodol 2005 ;8 :19 –32 . 10.1080/1364557032000119616 43 Levac D , Colquhoun H , O'Brien KK Scoping studies: advancing the methodology . Implement Sci 2010 ;5: 69 10.1186/1748-5908-5-69 20854677 44 Sampson M , McGowan J , Cogo E An evidence-based practice guideline for the peer review of electronic search strategies . J Clin Epidemiol 2009 ;62 :944 –52 . 10.1016/j.jclinepi.2008.10.012 19230612 45 Menzies IB , Mendelson DA , Kates SL The impact of comorbidity on perioperative outcomes of hip fractures in a geriatric fracture model . Geriatr Orthop Surg Rehabil 2012 ;3 :129 –34 . 10.1177/2151458512463392 23569706 46 Day GA , Swanson C , Yelland C Surgical outcomes of a randomised prospective trial involving patients with a proximal femoral fracture . ANZ J Surg 2001 ;71 (1):11 –14 . http://onlinelibrary.wiley.com/o/cochrane/clcentral/articles/663/CN-00400663/frame.html http://onlinelibrary.wiley.com/o/cochrane/clcentral/articles/663/CN-00400663/frame.html10.1046/j.1440-1622.2001.02019.x http://dx.doi.org/10.1046/j.1440-1622.2001.02019.x 11167590 47 Beaupre LA , Jones CA , Saunders LD Best practices for elderly hip fracture patients. A systematic overview of the evidence . J Gen Intern Med 2005 ;20 :1019 –25 . 10.1111/j.1525-1497.2005.00219.x 16307627 48 Collinge CA , McWilliam-Ross K , Beltran MJ , et al Measures of clinical outcome before, during, and after implementation of a comprehensive geriatric hip fracture program: is there a learning curve . J Orthop Trauma 2013 ;27 :672 –6 . 10.1097/BOT.0b013e318291f0e5 23515124 49 Baker K , Brierley S , Mitchell G Effecting change using careplans: experience from two fractured neck of femur pathways . Aust Health Rev 2012 ;36 :308 –12 . 10.1071/AH11021 22935123 50 Matityahu A , Elliott I , Marmor M Time intervals in the treatment of fractured femurs as indicators of the quality of trauma systems . Bull World Health Organ 2014 ;92 :40 –50 . 10.2471/BLT.13.120436 24391299 51 Merle V , Moret L , Pidhorz L Does comparison of performance lead to better care? A pilot observational study in patients admitted for hip fracture in three French public hospitals . Int J Qual Health Care 2009 ;21 :321 –9 . 10.1093/intqhc/mzp029 19692425 52 Hip Fracture Accelerated Surgical TCT Investigators . Accelerated care versus standard care among patients with hip fracture: the HIP ATTACK pilot trial . CMAJ 2014 ;186 :E52 –60 . 10.1503/cmaj.130901 24246589 53 Kristensen MT , Henriksen S Relative and absolute intertester reliability of the timed up and go test to quantify functional mobility in patients with hip fracture . J Am Geriatr Soc 2011 ;59 :565 –7 . 10.1111/j.1532-5415.2010.03293.x 21391955 54 Suhm N , Kaelin R , Studer P Orthogeriatric care pathway: a prospective survey of impact on length of stay, mortality and institutionalisation . Arch Orthop Trauma Surg 2014 ;134 :1261 –9 . 10.1007/s00402-014-2057-x 25060921 55 Al-Rashid M , Parker MJ Anticoagulation management in hip fracture patients on warfarin . Injury 2005 ;36 :1311 –15 . 10.1016/j.injury.2005.05.004 16214475 56 Della Rocca GJ , Crist BD Hip fracture protocols: what have we changed? Orthop Clin North Am 2013 ;44 :163 –82 . 10.1016/j.ocl.2013.01.009 23544822 57 Smektala R , Endres HG , Dasch B The effect of time-to-surgery on outcome in elderly patients with proximal femoral fractures . BMC Musculoskelet Disord 2008 ;9 :171 10.1186/1471-2474-9-171 19114019 58 Pickles SM , Coventry LL , Glennon DA ‘Making a difference’ a clinical pathway for hip fractures and the advance practice role in managing patients with minimal trauma hip fractures . Int J Orthop Trauma Nurs 2014 ;18 :205 –13 . 209p 10.1016/j.ijotn.2013.11.011 59 Simunovic ND , Devereaux PJ , Sprague S Effect of early surgery after hip fracture on mortality and complications: systematic review and meta-analysis . CMAJ 2010 ;182 :1609 –16 . 10.1503/cmaj.092220 20837683 60 Granger CV , Deutsch A , Graham JE The Uniform Data System for Medical Rehabilitation fourth in the series of impairment-specific longitudinal reports from the Uniform Data System for medical Rehabilitation (UDSMR) database . Am J Phys Med Rehabil 2011 ;90 :177 –89 . 10.1097/PHM.0b013e31820b18d7 21297397 61 Simpson P Clinical outcomes in transition program for older adults with hip fracture . Outcomes Manag 2002 ;6 :86 –92 .11949519 62 Adunsky A , Lerner-Geva L , Blumstein T Improved survival of hip fracture patients treated within a comprehensive geriatric hip fracture unit, compared with standard of care treatment . J Am Med Dir Assoc 2011 ;12 :439 –44 . 10.1016/j.jamda.2010.09.003 21450210 63 Eisler J , Cornwall R , Strauss E Outcomes of elderly patients with nondisplaced femoral neck fractures . J Orthop Trauma 2003 ;17 :S31 –37 . 10.1097/00005131-200309001-00008 14696776 64 Lai FH-Y , Soo AK-W , Wong SK-M Admission cognitive performance and functional gain following inpatient rehabilitation in geriatric patients with hip fractures . Hong Kong J Occup Ther 2003 ;13 :25 –33 . 65 Hershkovitz A , Kalandariov Z , Hermush V Factors affecting short-term rehabilitation outcomes of disabled elderly patients with proximal hip fracture . Arch Phys Med Rehabil 2007 ;88 :916 –21 . 10.1016/j.apmr.2007.03.029 17601474 66 Zidén L , Kreuter M , Frandin K Long-term effects of home rehabilitation after hip fracture—1-year follow-up of functioning, balance confidence, and health-related quality of life in elderly people . Disabil Rehabil 2010 ;32 :18 –32 . 10.3109/09638280902980910 19925273 67 Zidén L , Frändin K , Kreuter M Home rehabilitation after hip fracture a randomized controlled study on balance confidence, physical function and everyday activities [with consumer summary] . Clin Rehabil 2008 ;22 :1019 –33 . 10.1177/0269215508096183 19052241 68 Allen J , Koziak A , Buddingh S Rehabilitation in patients with dementia following hip fracture: a systematic review . Physiother Can 2012 ;64 :190 –201 . 10.3138/ptc.2011-06BH 23449813 69 Shiri-Sharvit O , Arad M , Mizrahi EH The association between psychotropic medication use and functional outcome of elderly hip-fracture patients undergoing rehabilitation . Arch Phys Med Rehabil 2005 ;86 :1389 –93 . 10.1016/j.apmr.2004.12.034 16003669 70 Cary MP , Baernholdt M , Anderson RA Performance-based outcomes of inpatient rehabilitation facilities treating hip fracture patients in the United States . Arch Phys Med Rehabil 2015 ;96 :790 –8 . 799p 10.1016/j.apmr.2015.01.003 25596000 71 Mizrahi EH , Arad M , Fleissig Y Gender differences in functional outcome of elderly hip fracture patients . Geriatr Gerontol Int 2014 ;4 :845 –50 . 10.1111/ggi.12178http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=emed12&NEWS=N&AN=24279758 72 Wang CY , Graham JE , Karmarkar AM FIM motor scores for classifying community discharge after inpatient rehabilitation for hip fracture . PM R 2014 ;6 :493 –7 . 10.1016/j.pmrj.2013.12.008 24389348 73 Lin PC , Chang SY Functional recovery among elderly people one year after hip fracture surgery . J Nurs Res 2004 ;12 :72 –82 . 10.1097/01.JNR.0000387490.71062.4a 15136965 74 Kammerlander C , Roth T , Friedman SM Ortho-geriatric service—a literature review comparing different models . Osteoporos Int 2010 ;21 (Suppl 4 ):S637 –646 . 10.1007/s00198-010-1396-x 21058004 75 Leigheb F , Vanhaecht K , Sermeus W The effect of care pathways for hip fractures: a systematic review . Calcif Tissue Int 2012 ;91 :1 –14 . 10.1007/s00223-012-9589-2 22476267 76 Crotty M , Whitehead C , Miller M Patient and caregiver outcomes 12 months after home-based therapy for hip fracture: a randomized controlled trial . Arch Phys Med Rehabil 2003 ;84 :1237 –9 . 10.1016/S0003-9993(03)00141-2 12917867 77 Beaupre LA , Cinats JG , Senthilselvan A . Does standardized rehabilitation and discharge planning improve functional recovery in elderly patients with hip fracture? Arch Phys Med Rehabil 2005 ; 86 :2231 –9 . 10.1016/j.apmr.2005.06.019 16344017 78 Tha HS , Armstrong D , Broad J Hip fracture in Auckland: contrasting models of care in two major hospitals . Intern Med J 2009 ;39 :89 –94 . 10.1111/j.1445-5994.2008.01721.x 18771434 79 Beringer TR , Clarke J , Elliott JR Outcome following proximal femoral fracture in Northern Ireland . Ulster Med J 2006 ;75 :200 –6 .16964812 80 Turner J , Gemmell LW , Slater A Seven year follow-up study of patients receiving surgical treatment for fractured neck of femur . Care Critic Ill 2006 ;22 :112 –14 . 81 Gregersen M , Borris LC , Damsgaard EM Blood transfusion and overall quality of life after hip fracture in frail elderly patients—the transfusion requirements in Frail Elderly Randomized Controlled Trial . J Am Med Dir Assoc 2015 ;9 :762 –6 10.1016/j.jamda.2015.03.022http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=emed13&NEWS=N&AN=2015976024 . 82 Doshi HK , Ramason R , Azellarasi J Orthogeriatric model for hip fracture patients in Singapore: our early experience and initial outcomes . Arch Orthop Trauma Surg 2014 ;134 :351 –7 . 10.1007/s00402-013-1900-9 24297214 83 Haywood KL , Griffin XL , Achten J Developing a core outcome set for hip fracture trials . Bone Joint J 2014 ;96-B :1016 –23 . 10.1302/0301-620X.96B8.33766 25086115 84 Bliemel C , Lechler P , Oberkircher L Effect of preexisting cognitive impairment on in-patient treatment and discharge management among elderly patients with hip fractures . Dement Geriatr Cogn Disord 2015 ;40 :33 –43 . 11p 10.1159/000381334 25896170 85 Bliemel C , Oberkircher L , Eschbach DA Impact of Parkinson's disease on the acute care treatment and medium-term functional outcome in geriatric hip fracture patients . Arch Orthop Trauma Surg 2015 ;11 :1519 –26 . 10.1007/s00402-015-2298-3http://link.springer.de/link/service/journals/00402/index.htm 86 Orive M , Aguirre U , García-Gutiérrez S Changes in health-related quality of life and activities of daily living after hip fracture because of a fall in elderly patients: a prospective cohort study . Int J Clin Pract 2015 ;69 :491 –500 . 410p 10.1111/ijcp.12527 25721490 87 Hall SE , Williams JA , Criddle RA A prospective study of falls following hip fracture in community dwelling older adults . Australas J Ageing 2001 ;20 :73 –8 . 10.1111/j.1741-6612.2001.tb00357.x 88 Street P , Thompson J , Bailey M Management of urinary catheters following hip fracture . Australas J Ageing 2013 ;19 :107 –8 . 10.1111/ajag.12166 89 Liem IS , Kammerlander C , Suhm N Identifying a standard set of outcome parameters for the evaluation of orthogeriatric co-management for hip fractures . Injury 2013 ;44 :1403 –12 . 10.1016/j.injury.2013.06.018 23880377 90 Muir SW , Yohannes AM The impact of cognitive impairment on rehabilitation outcomes in elderly patients admitted with a femoral neck fracture: a systematic review . J Geriatr Phys Ther 2009 ;32 :24 –32 . 10.1519/00139143-200932010-00006 19856633 91 Moseley AM , Sherrington C , Lord SR Mobility training after hip fracture: a randomised controlled trial . Age Ageing 2009 ;38 :74 –80 . 10.1093/ageing/afn217 19017676 92 Latham NK , Harris BA , Bean JF Effect of a home-based exercise program on functional recovery following rehabilitation after hip fracture: a randomized clinical trial . JAMA 2014 ;311 :700 –8 . 10.1001/jama.2014.469 24549550 93 Handoll HH , Parker MJ Conservative versus operative treatment for hip fractures in adults . Cochrane Database Syst Rev 2008 ;(3) :CD000337 10.1002/14651858.CD000337.pub2 18646065 94 Coleman EA , Kramer AM , Kowalsky JC A comparison of functional outcomes after hip fracture in group/staff HMOs and fee-for-service systems . Eff Clin Pract 2000 ;3 :229 –39 .11185328 95 Vidán M , Serra JA , Moreno C Efficacy of a comprehensive geriatric intervention in older patients hospitalized for hip fracture: a randomized, controlled trial . J Am Geriatr Soc 2005 ;53 :1476 –82 . 10.1111/j.1532-5415.2005.53466.x 16137275 96 Johansen I , Lindbaek M , Stanghelle JK Structured community-based inpatient rehabilitation of older patients is better than standard primary health care rehabilitation: an open comparative study . Disabil Rehabil 2012 ;34 :2039 –46 . 10.3109/09638288.2012.667193 22452632 97 Binder EF , Brown M , Sinacore DR Effects of extended outpatient rehabilitation after hip fracture: a randomized controlled trial . JAMA 2004 ;292 :837 –46 . 10.1001/jama.292.7.837 15315998 98 Chudyk AM , Jutai JW , Petrella RJ Systematic review of hip fracture rehabilitation practices in the elderly . Arch Phys Med Rehabil 2009 ;90 :246 –62 . 10.1016/j.apmr.2008.06.036 19236978 99 Bryant DM , Sanders DW , Coles CP Selection of outcome measures for patients with hip fracture . J Orthop Trauma 2009 ;23 :434 –41 . 10.1097/BOT.0b013e318162aaf9 19550231 100 Cranney AB , Coyle D , Hopman WM Prospective evaluation of preferences and quality of life in women with hip fractures . J Rheumatol 2005 ;32 :2393 –9 .16331770 101 Shyu YI , Liang J , Wu CC An interdisciplinary intervention for older Taiwanese patients after surgery for hip fracture improves health-related quality of life . BMC Musculoskelet Disord 2010 ;11 :225 10.1186/1471-2474-11-225 20920220 102 Shyu YI , Liang J , Tseng MY Comprehensive and subacute care interventions improve health-related quality of life for older patients after surgery for hip fracture: a randomised controlled trial . Int J Nurs Stud 2013 ;50 :1013 –24 . 10.1016/j.ijnurstu.2012.11.020 23245454 103 Gjertsen JE , Engesaeter LB , Furnes O The Norwegian Hip Fracture Register: experiences after the first 2 years and 15,576 reported operations . Acta Orthop 2008 ;79 :583 –93 . 10.1080/17453670810016588 18839363 104 Parsons N , Griffin XL , Achten J Outcome assessment after hip fracture: is EQ-5D the answer? Bone Joint Res 2014 ;3 :69 –75 . 10.1302/2046-3758.33.2000250 24648420 105 Hagsten B , Svensson O , Gardulf A Health-related quality of life and self-reported ability concerning ADL and IADL after hip fracture: a randomized trial . Acta Orthop Scand 2006 ;77 :114 –19 . 10.1080/17453670610045786 106 Leland NE , Gozalo P , Bynum J What happens to patients when they fracture their hip during a skilled nursing facility stay? J Am Med Dir Assoc 2015 ;9 :767 –74 . http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=emed13&NEWS=N&AN=2015993996 . 107 Giusti A , Barone A , Razzano M Optimal setting and care organization in the management of older adults with hip fracture . Eur J Phys Rehabil Med 2011 ;47 :281 –96 .21555985 108 Mont MA , Banerjee A , Jaurequi JJ What outcome metrics do the various knee rating systems for assessment of outcomes following total knee arthroplasty measure? A systematic review of literature . Surg Technol Int 2015 ;26 :269 –74 .26055019 109 Ritchie L , Wright-St Clair VA , Keogh J Community integration after traumatic brain injury: a systematic review of the clinical implications of measurement and service provision for older adults . Arch Phys Med Rehabil 2014 ;95 :163 –74 . 10.1016/j.apmr.2013.08.237 24016401 110 Sleat GK , Ardolino AM , Willett KM Outcome measures in major trauma care: a review of current international trauma registry practice . J Emerg Med 2011 ;28 :1008 –12 . 10.1136/emermed-2011-200326 111 Rinere-O'Brien S Trends in inpatient rehabilitation stroke outcomes before and after advent of the prospective payment system: a systematic review . J Neurol Phys Ther 2010 ;34 :17 –23 . 10.1097/NPT.0b013e3181cfd3ac 20212363 112 Grace SL , Poirier P , Norris CM Pan-Canadian development of cardiac rehabilitation and secondary prevention quality indicators . Can J Cardiol 2014 ;30 :945 –8 . 10.1016/j.cjca.2014.04.003 25064585

PMC005xxxxxx/PMC5372038.txt

==== Front RMD OpenRMD OpenrmdopenrmdopenRMD Open2056-5933BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR 28405469rmdopen-2016-00032710.1136/rmdopen-2016-000327Rheumatoid Arthritis1506Original articleWhat causes a small increase in radiographic progression in rheumatoid arthritis patients tapering TNF inhibitors? Bouman Chantal A M 1den Broeder Alfons A 12van der Maas Aatke 1van den Hoogen Frank H J 12Landewé Robert B M 3van Herwaarden Noortje 1 1 Department of Rheumatology, Sint Maartenskliniek Nijmegen, The Netherlands 2 Department of Rheumatology, Radboud University Medical Center, The Netherlands 3 Department of Rheumatology, Academic Medical Center, Amsterdam, The NetherlandsCorrespondence to Chantal A M Bouman; c.bouman@maartenskliniek.nl2017 20 3 2017 3 1 e0003275 7 2016 10 1 2017 15 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective In a randomised controlled trial investigating tapering of TNF inhibitors (TNFi) compared with usual care (UC) in rheumatoid arthritis patients, minimal radiographic progression was more frequent in patients who attempted tapering. Possible explanations include higher incidence of flaring, higher mean disease activity or lower TNFi use. Methods 18 months data from the DRESS study were used. Change in Sharp-van der Heijde (ΔSvdH) score (linear regression) and proportion of patients with >0.5 ΔSvdH (logistic regression) were used as outcomes. The cumulative incidence and number of short-lived and major flares per patient, mean time-weighted disease activity (MTW-DAS28-CRP) and TNFi use were used as independent variables. Regression models were performed stratified per study group and corrected for possible confounders. Results 175 of 180 patients had 18-month data available. The mean ΔSvdH were 0.75 and 0.15 units with 37 of 116 (32%) and 9 of 59 (15%) patients exceeding 0.5 points in the tapering and UC group, respectively (both p<0.05). MTW-DAS28-CRP, but not incidence or number of short-lived or major flares, or TNFi use, was independently associated with the mean progression score, but only in the tapering group. Additional analyses on DAS28-CRP subcomponents showed that this was mainly caused by MTW swollen joint count. No confounders were identified. Conclusions Radiographic progression was associated with higher MTW-DAS28-CRP (and especially swollen joint count), but only in patients who tapered TNFi. This finding stresses the importance of maintaining disease activity as low as possible in patients in whom TNFi is tapered and to check for radiographic progression regularly. Trial registration number NTR 3216; Post-results. Rheumatoid ArthritisTreatmentDMARDs (biologic)Disease ActivityAnti-TNF ==== Body Key messages What is already known about this subject? The TNF inhibitors (TNFi) tapering strategy used in the DRESS study resulted in an increase in radiographic progression for patients who attempted tapering compared with patients who continued TNFi dosing. Although this increase was minimal over 18 months, it may become significant in subsequent years and lead to disability. What does this study add? We investigated possible causes and found that higher disease activity (especially swollen joints) in combination with lower TNF inhibitors exposition was associated with the mean radiographic progression. How might this impact on clinical practice? This finding stresses the importance of maintaining a state of low disease activity or remission in patients in whom TNF inhibitors is tapered and to check for radiographic progression regularly. However, further progression in subsequent years is not to be expected, as higher disease activity is a temporary effect of a trial-and-error tapering strategy. Introduction Disease activity-guided tapering of TNF inhibitors (TNFi) in rheumatoid arthritis (RA) results in a significant reduction in TNFi use and subsequent cost, without compromising on important clinical outcomes.1 However, in the DRESS (Dose REduction Strategy of Subcutaneous TNF inhibitors) study, a minimal increase in radiographic progression was observed for patients who attempted tapering compared with patients who continued TNFi dosing. We propose three hypotheses that could explain this: first, in DRESS, short-lived flares were more frequent in patients tapering than in patients not tapering, which is a temporary effect of the trial-and-error type of tapering strategy. It could be hypothesised that the tapering strategy leads to a higher incidence of flares, thus causing radiographic progression (in both groups or tapering group alone).2 Second, a significantly higher mean time-weighted (MTW) disease activity was observed in the tapering group, again induced by the tapering and higher MTW disease activity could result in radiographic progression (in both groups or tapering group alone). Third, tapering causes lower TNFi exposition. Previous studies have suggested that TNFi use itself may prevent radiographic progression. Therefore, lower TNFi exposition could lead to progression, independent of increased disease activity.3–5 These hypotheses have different clinical implications. In the first two hypotheses, the effect is temporarily: progression is caused by a (sometimes unsuccessful) tapering attempt, not by lower TNFi use itself—so in subsequent years, damage would not progress further. Tight control should be optimised, and if flares could be predicted, progression would be reduced. The third hypothesis would mean an ongoing process of radiographic progression in following years (figure 1), and although the increase in progression that we found is minimal, it may become significant in subsequent years with consequent loss of function or pain symptoms. It would not be preventable by tight control alone and would require frequent radiographic monitoring and adaptation of TNFi use. Figure 1 Hypotheses for causes of progression and development over time. Therefore, we investigated the effects of the occurrence of short-lived or major flare, MTW disease activity and TNFi exposition on radiographic progression in patients tapering TNFi compared with patients not tapering. Patients and methods Patients and definitions Clinical and radiographic data from the DRESS study were used: an 18-month, open randomised clinical trial, investigating non-inferiority of a disease activity-guided tapering strategy of adalimumab or etanercept compared with usual care (UC).1 6 Radiographs from baseline and 18 months were scored pairwise and in chronological order using the Sharp-van der Heijde (SvdH) score by two researchers, blinded for clinical outcome and study group.7 The absolute SvdH score with subcomponents and change (Δ) in SvdH score between baseline and 18 months were calculated. The proportion of patients with minimal progression, defined as ΔSvdH>0.5 points, was calculated. Additionally, proportions of patients exceeding the minimal clinically important change (MCIC) (8 points per 18 months, based on previous values of 4 points per year)8 9 and smallest detectable change (SDC) (4.1 points)1 were calculated. Disease activity was defined using a 28 joint-based disease activity score (DAS28) with C reactive protein (CRP) and MTW-DAS28-CRP was calculated over 18 months. For (short-lived) flare, a validated flare criterion was used: DAS28 increase of >1.2 compared with baseline, or DAS28 increase of >0.6 and current DAS28 ≥3.2.10 A major flare was defined as a flare lasting >3 months. The cumulative incidence of patients with short-lived or major flare and number of short-lived or major flares per patient were calculated. TNFi use was calculated in the dose reduction and UC group, as the normalised proportion of the defined daily dose (DDD) of TNFi, with 1.0 as full-dose equivalent. DDD: 40 mg/14 days for adalimumab and 50 mg/7 days for etanercept. Statistical analyses STATA/IC V.13.1 was used. Descriptive statistics were performed, (non) parametrically when appropriate. Univariate and multivariate analyses were performed with cumulative incidence and number of short-lived and major flares per patient, MTW-DAS28-CRP and TNFi use as independent variables. The radiographic progression yes/no (ΔSvdH >0.5; logistic regression) and mean ΔSvdH (linear regression) were used as dependent variables. Possible confounders that were checked were: age, sex, body mass index, smoking, baseline SvdH score, DAS28-CRP, CRP, rheumatoid factor, anticitrullinated protein antibody status, oral glucocorticoid use and intramuscular or intra-articular glucocorticoid injections, number of glucocorticoid injections per patient and synthetic disease-modifying antirheumatic drug use. To check for effect modification, all analyses were performed stratified by allocation group (tapering or UC). Results Radiographic progression One hundred and seventy-five (116 taper group/59 UC) of 180 patients had clinical and radiographic data available. Baseline characteristics were comparable between patients with missing and non-missing data. The mean SvdH scores were 38.3 (SD 49.3) and 42.1 (58.7) at baseline (p=0.65), and 39.0 (49.6) and 42.2 (58.7) at 18 months (p=0.71) for the taper and UC groups, respectively (table 1). The mean ΔSvdH over 18 months were 0.75 (1.5) and 0.15 (1.1) in the taper and UC groups, respectively (p<0.05). The difference in ΔSvdH between groups was mainly caused by joint space narrowing; change in erosion score was similar (table 1). No patients exceeded the MCIC. The SDC was exceeded by five (4%) patients in the taper group and no patients in the UC group. Minimal progression was found in 37 of 116 (32%) and 9 of 59 (15%) patients in the taper and UC groups, respectively (p<0.05). Table 1 Radiographic outcomes Taper group (n=116) Usual care group (n=59) Difference (95% CI) Total (n=175) SvdH baseline* 38.3 (49.3) 42.1 (58.7) −3.79 (−20.4 to 12.8) 39.6 (52.5) SvdH 18 months* 39.0 (49.6) 42.2 (58.7) −3.19 (−19.9 to 13.5) 40.1 (52.7) Progression SvdH score* 0.75 (1.5) 0.15 (1.1) 0.60 (0.16 to 1.0) 0.55 (1.4) Progression erosion score* 0.29 (0.8) 0.12 (0.7) 0.17 (−0.07 to 0.42) 0.23 (0.8) Progression joint space narrowing* 0.46 (1.2) 0.03 (0.9) 0.43 (0.07 to 0.78) 0.32 (1.1) Progression >MCIC† 0 (0) 0 (0) 0 (0) 0 (0) Progression >SDC† 5 (4) 0 (0) 5 (4) 5 (3) Progression >0.5† 37 (32) 9 (15) 28 (17) 46 (26) *Mean with SD. †Number (%) of patients. MCIC, minimal clinical important change (8 units); Progression SvdH, Sharp-van der Heijde progression between baseline and 18 months; SDC, smallest detectable change (4.1 units); SvdH, Sharp-van der Heijde score. Disease activity and (major) flare MTW-DAS28-CRP was 2.3 (0.5) and 2.1 (0.6) in the taper and UC groups, respectively (p<0.01). For patients with minimal progression, the median MTW-DAS28-CRP was 2.3 (IQR 2.0–2.8) in the taper group and 2.0 (1.9–2.4) in the UC group. Additional data on the mean DAS28-CRP at certain time points are provided in online supplementary table S1. Short-lived flares occurred in 84 of 116 (72%) in the taper group and 16 of 59 (27%) in the UC group (p<0.001). The cumulative incidence of major flare was 14 of 116 (12%) and 6 of 59 (10%) in the taper and UC groups, respectively. 10.1136/rmdopen-2016-000327.supp1supplementary table mean DAS28-CRP values at baseline, 9 and 18 months and at flare visits TNFi exposition The median proportions of DDD were 0.47 (IQR 0.27–0.68) and 1.00 (IQR 0.95–1.00) in the taper and UC groups, respectively (p<0.0001). The lower bound of the IQR of the median proportion of DDD was slightly below 1.00 in the UC arm due to patients: discontinuing because of adverse events (n=6) or inefficacy (n=2); tapering because of low disease activity (n=5); being on lower than DDD dose at inclusion (n=2). Regression modelling Logistic regression with ΔSvdH >0.5 yes/no as dependent variable did not yield any association with short-lived or major flares, MTW-DAS28-CRP or TNFi use. In univariate linear regression with the mean ΔSvdH as dependent variable, only MTW-DAS28-CRP, not occurrence of short-lived or major flares or TNFi use, was independently associated with progression (β=0.51 (p=0.005)). In multivariate analyses, only MTW-DAS28-CRP remained significantly associated with the mean ΔSvdH. Effect modification was present by allocation group (table 2), with a significant association between MTW-DAS28-CRP and progression in the taper group, but not in the UC group. Stratified corrected analyses for the taper and UC groups showed non-significant associations, except for MTW-DAS28-CRP. Additional exploratory analyses on subcomponents of DAS28-CRP showed that MTW tender and swollen joint count (MTW-TJ and MTW-SJ) were significantly associated with the mean progression in the taper group. Patient global visual analogue scale (PG-VAS) and CRP were not significantly associated with the mean progression (table 2). Collinearity between MTW-TJ and MTW-DAS28-CRP was high (>0.7) but lower for MTW-SJ and MTW-DAS28-CRP, thus, MTW-SJ was added to the model. Afterwards, only MTW-SJ remained significantly associated with the mean progression in the taper group with β=0.52 (95% CI 0.05 to 0.99) (table 3). No significant confounding was identified. Table 2 Univariate linear regression models stratified by allocation group Tapering group Usual care group β 95% CI β 95% CI MTW-DAS28-CRP 0.64 0.14 to 1.14 0.17 −0.29 to 0.62 Constant −0.73 −0.20 MTW-TJ 0.24 0.07 to 0.10 0.05 −0.13 to 0.24 MTW-SJ 0.65 0.25 to 1.04 0.21 −0.19 to 0.62 MTW-PG-VAS 0.02 −0.0001 to 0.38 0.004 −0.02 to 0.03 MTW-CRP −0.001 −0.05 to 0.05 0.006 −0.03 to 0.05 Occurrence of flare 0.24 −0.38 to 0.87 0.091 −0.57–0.75 Constant 0.58 0.13 Number of flare per patient −0.025 −0.34 to 0.29 0.041 −0.36 to 0.44 Constant 0.78 0.14 Occurrence of major flare 0.69 −0.16 to 1.53 0.82 −0.86 to 1.08 Constant 0.67 0.14 Number of major flare per patient 0.71 −0.06 to 1.47 0.11 −0.86 to 1.08 Constant 0.66 0.14 TNFi use (% ddd) 0.43 −0.65 to 1.51 −0.068 −1.38 to 1.25 Constant 0.54 0.21 %ddd, percentage of the defined daily dose; CRP, C reactive protein; MTW-DAS28-CRP, mean time-weighted DAS28-CRP; PG-VAS, patient global visual analogue scale; SJ, swollen joint count; TJ, tender joint count. Table 3 Final linear regression model stratified by allocation group Tapering group Usual care group β 95% CI β 95% CI MTW-DAS28-CRP 0.28 −0.30 to 0.87 −0.02 −0.70 to 0.65 MTW-SJ 0.52 0.05 to 0.99 0.23 −0.37 to 0.84 Constant −0.24 −1.5 to 1.01 0.07 −1.15 to 1.29 MTW-DAS28-CRP, mean time-weighted DAS28-CRP; SJ, swollen joint count. Discussion In this study, we investigated possible causes of the minimal difference in radiographic progression in RA patients tapering TNFi compared with UC that was observed in the DRESS study. Analyses yielded an association between MTW disease activity and radiographic progression, but only in the tapering group. No association was found between the occurrence or number of flares or lower TNFi exposition and radiographic progression. Additional analyses on subcomponents of DAS28-CRP showed that radiographic progression was mainly caused by swollen joint count. Thus, it is the small overall increase in disease activity over time, and more specifically swollen joints, caused by the tapering strategy, and not the intermittent episodes of high disease activity (flares) that appear to cause progression. This suggests that radiographic progression occurs when two necessary causes (higher disease activity and tapering) are present. Therefore, tight control—although also important in non-tapering patients—is even more important when tapering TNFi, to prevent additional progression. However, further progression in subsequent years is not to be expected, as higher disease activity is a temporary effect of a trial-and-error tapering strategy, and disease activity was similar between dose reduction and the UC group at 18 months. Our study has some limitations. First, the follow-up time of 18 months was limited. Furthermore, the level of radiographic progression that is of clinical relevance is somewhat debatable. In 2006, Welsing et al8 established a level of five Sharp-Van der Heijde points per year as the minimal clinically important change. This level may be different for the current RA population treated with more strict tight control. Therefore, we also analysed progression with different cut-off levels (SDC and minimal progression <0.5 SvdH points), as well as with continuous SvdH score to be as sensitive as possible. Finally, the observed SDC is relatively high and some misclassification of patients with progression that is actually due to measurement error could be present. However, this would cause bias towards a null result, whereas we did find differences in radiographic progression and in associations between disease activity and progression. Our findings are in line with three studies that have shown some effect of discontinuation but no effect of tapering of TNFi tapering on radiographic progression in RA.11–13 In the STRASS study, patients were randomised to disease activity-guided TNFi tapering or continuation of treatment.11 Multiple tapering attempts were allowed. A difference in disease activity and relapse rate was observed, but no difference in radiographic progression. Follow-up time and SDC were comparable to our study, but sample size was smaller, which may explain the null result. In PRESERVE, patients were treated with etanercept and methotrexate for 36 weeks after which they were randomised to etanercept fixed dose halving, discontinuation or full-dose continuation.12 A significantly greater proportion of patients in the discontinuation group exceeded the SDC compared with patients continuing etanercept. This was explained by the fact that patients had moderate disease activity and were refractory to methotrexate monotherapy at study start. Furthermore, disease activity was not steered on, leading to a significant rise in DAS28 after etanercept discontinuation. Finally, Raffeiner et al13 showed that randomisation of RA patients in remission under etanercept, to either receive fixed halve dose etanercept or continuation of full-dose etanercept, did not lead to differences in radiographic progression after 2 years. However, this study did not include discontinuation of etanercept. In conclusion, disease activity-guided TNFi tapering may result in a small increase in radiographic progression. This is possibly due to the disappearance of the direct inhibitory effect of TNFi on radiographic progression (‘disconnect’), so that inflammation resumes driving this progression. These findings stress the increased importance of maintaining a state of low disease activity or remission—especially low swollen joint count—in patients in whom TNFi is tapered and to check for radiographic progression regularly. Long-term studies on TNFi tapering need to confirm that radiographic damage does not continue to progress over the years. Also, future studies should focus on predictors of successful tapering or discontinuation to further prevent the rise in disease activity that is the consequence of tapering. The authors would like to thank all rheumatologists and specialist nurses from The Sint Maartenskliniek Nijmegen and Woerden for their participation in data collection. The authors would also like to thank Kimberly Bouman for her support in data handling. Contributors: CAMB, AAdB, AvdM, FHJvdH and NvH were involved in the study design. CB, AAdB, AvdM and NvH were involved in the data collection. CAMB, AAdB, AvdM, RBML and NvH performed the data analyses. All authors were involved in writing and revision of the manuscript. Competing interests: AAdB reports that he received a congress invitation from ABBVIE, BIOGEN, CELLTRION and ROCHE and received an expert witness fee from AMGEN and BI, all outside the submitted work. The other authors have no competing interests to report. Patient consent: Obtained. Ethics approval: Ethical approval was given by the local ethics committee (CMO region Arnhem-Nijmegen; NL37704.091.11). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: The authors commit to making the relevant anonymised patient-level data available on reasonable request. ==== Refs References 1 van Herwaarden N , van der Maas A , Minten MJ Disease activity guided dose reduction and withdrawal of adalimumab or etanercept compared with usual care in rheumatoid arthritis: open label, randomised controlled, non-inferiority trial . BMJ 2015 ;350 :h1389 10.1136/bmj.h1389 25858265 2 Welsing PM , Landewé RB , van Riel PL The relationship between disease activity and radiologic progression in patients with rheumatoid arthritis: a longitudinal analysis . Arthritis Rheum 2004 ;50 :2082 –93 . 10.1002/art.20350 15248205 3 Smolen JS , Han C , Bala M Evidence of radiographic benefit of treatment with infliximab plus methotrexate in rheumatoid arthritis patients who had no clinical improvement: a detailed subanalysis of data from the anti-tumor necrosis factor trial in rheumatoid arthritis with concomitant therapy study . Arthritis Rheum 2005 ;52 :1020 –30 . 10.1002/art.20982 15818697 4 Landewé R , van der Heijde D , Klareskog L Disconnect between inflammation and joint destruction after treatment with etanercept plus methotrexate: results from the trial of etanercept and methotrexate with radiographic and patient outcomes . Arthritis Rheum 2006 ;54 :3119 –25 . 10.1002/art.22143 17009230 5 Smolen JS , Han C , van der Heijde DM Radiographic changes in rheumatoid arthritis patients attaining different disease activity states with methotrexate monotherapy and infliximab plus methotrexate: the impacts of remission and tumour necrosis factor blockade . Ann Rheum Dis 2009 ;68 :823 –7 . 10.1136/ard.2008.090019 18593759 6 den Broeder AA , van Herwaarden N , van der Maas A Dose REduction strategy of subcutaneous TNF inhibitors in rheumatoid arthritis: design of a pragmatic randomised non inferiority trial, the DRESS study . BMC Musculoskelet Disord 2013 ;14 :299 10.1186/1471-2474-14-299 24152421 7 van der Heijde D How to read radiographs according to the Sharp/van der Heijde method . J Rheumatol 1999 ;26 :743 –5 .10090194 8 Welsing PM , Borm GF , van Riel P Minimal clinically important difference in radiological progression of joint damage. A definition based on patient perspective . J Rheumatol 2006 ;33 :501 –7 .16511919 9 Bruynesteyn K , Boers M , Kostense P Deciding on progression of joint damage in paired films of individual patients: smallest detectable difference or change . Ann Rheum Dis 2005 ;64 :179 –82 . 10.1136/ard.2003.018457 15286006 10 van der Maas A , Lie E , Christensen R Construct and criterion validity of several proposed DAS28-based rheumatoid arthritis flare criteria: an OMERACT cohort validation study . Ann Rheum Dis 2013 ;72 :1800 –5 . 10.1136/annrheumdis-2012-202281 23178206 11 Fautrel B , Pham T , Alfaiate T Step-down strategy of spacing TNF-blocker injections for established rheumatoid arthritis in remission: results of the multicentre non-inferiority randomised open-label controlled trial (STRASS: Spacing of TNF-blocker injections in Rheumatoid ArthritiS Study) . Ann Rheum Dis 2016 ;75 :59 –67 . 10.1136/annrheumdis-2014-206696 26103979 12 Smolen JS , Nash P , Durez P Maintenance, reduction, or withdrawal of etanercept after treatment with etanercept and methotrexate in patients with moderate rheumatoid arthritis (PRESERVE): a randomised controlled trial . Lancet 2013 ;381 :918 –29 . 10.1016/S0140-6736(12)61811-X 23332236 13 Raffeiner B , Botsios C , Ometto F Effects of half dose etanercept (25 mg once a week) on clinical remission and radiographic progression in patients with rheumatoid arthritis in clinical remission achieved with standard dose . Clin Exp Rheumatol 2015 ;33 :63 –8 .25535985

PMC005xxxxxx/PMC5372039.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01172010.1136/bmjopen-2016-011720EpidemiologyResearch15061692169816831692Total volume and composition of fluid intake and mortality in older women: a cohort study Lim Wai H 12Wong Germaine 34Lewis Joshua R 13Lok Charmaine E 5Polkinghorne Kevan R 678Hodgson Jonathan 19Lim Ee M 10Prince Richard L 111 1 Sir Charles Gairdner Hospital Unit, University of Western Australia School of Medicine and Pharmacology, Perth, Western Australia, Australia 2 Department of Renal Medicine, Sir Charles Gairdner Hospital, Perth, Western Australia, Australia 3 Centre for Kidney Research, Children's Hospital at Westmead, Sydney, New South Wales, Australia 4 School of Public Health, Sydney Medical School, The University of Sydney, Sydney, New South Wales, Australia 5 Department of Medicine, University of Toronto; Division of Nephrology, Toronto General Hospital, Toronto, Ontario, Canada 6 Department of Nephrology, Monash Medical Centre, Clayton, Melbourne, Australia 7 Department of Medicine, Monash University, Clayton, Melbourne, Australia 8 Department of Epidemiology and Preventive Medicine, Monash University, Prahan, Melbourne, Australia 9 School of Medical and Health Sciences, Edith Cowan University, Perth, Western Australia, Australia 10 PathWest, Sir Charles Gairdner Hospital, Perth, Western Australia, Australia 11 Department of Endocrinology and Diabetes, Sir Charles Gairdner Hospital, Perth, Western Australia, AustraliaCorrespondence to Dr Wai H Lim; wai.lim@health.wa.gov.auWHL, GW and JRL contributed equally. 2017 24 3 2017 7 3 e01172029 2 2016 17 1 2017 19 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives The health benefits of ‘drinking at least 8 glasses of water a day” in healthy individuals are largely unproven. We aimed to examine the relationship between total fluid and the sources of fluid consumption, risk of rapid renal decline, cardiovascular disease (CVD) mortality and all-cause mortality in elderly women. Design, setting and participants We conducted a longitudinal analysis of a population-based cohort study of 1055 women aged ≥70 years residing in Australia. Main outcome measures The associations between total daily fluid intake (defined as total volume of beverage excluding alcohol and milk) and the types of fluid (water, black tea, coffee, milk and other fluids) measured as cups per day and rapid renal decline, CVD and all-cause mortality were assessed using adjusted logistic and Cox regression analyses. Results Over a follow-up period of 10 years, 70 (6.6%) experienced rapid renal decline and 362 (34.4%) died, of which 142 (13.5%) deaths were attributed to CVD. The median (IQR) intake of total fluid was 10.4 (8.5–12.5) cups per day, with water (median (IQR) 4 (2–6) cups per day) and black tea (median (IQR) 3 (1–4) cups per day) being the most frequent type of fluid consumed. Every cup per day higher intake of black tea was associated with adjusted HRs of 0.90 (95% CI 0.81 to 0.99) and 0.92 (95% CI 0.86 to 0.98) for CVD mortality and all-cause mortality, respectively. There were no associations between black tea intake and rapid renal decline, or between the quantity or type of other fluids, including water intake, and any clinical outcomes. Conclusions Habitual higher intake of black tea may potentially improve long-term health outcomes, independent of treating traditional CVD risk factors, but validation of our study findings is essential. fluid intakeEPIDEMIOLOGYelderlymortalitytea ==== Body Strengths and limitations of this study Completeness of the data set with long-term outcomes. Evaluation of multiple clinically important outcomes. Evaluation of habitual total fluid intake, with complete estimation of all ‘liquid’ volumes. Accuracy in the measurement of exposure factors including total fluid, water and tea intake. Potential for recall bias. Assessment of only baseline fluid consumption but not change in fluid consumption over time. Introduction It is a widely held belief that adequate water intake improves health outcomes, including reducing the risk of chronic kidney disease (CKD) and cardiovascular disease (CVD).1 2 The notion that high water intake could potentially prevent the development of CKD may, in part, be attributed to the inhibition of arginine vasopressin (AVP), a hormone known to be associated with glomerular hyperfiltration and injury.3 In addition to the effects on the glomerulus, high water intake had been shown to ameliorate tubulointerstitial injury in animal models, possibly via inhibition of the profibrotic cytokine, tumour necrosis factor-β.4 The association between water or total fluid intake and the risk of CVD is less clear; however, it is well established that CVD risk is highly associated with increasing severity of CKD.5–7 The potential beneficial effects of water and other fluid ingestion may, in part, be attributed to its effect on preventing pre-renal states, thus preservation of kidney function. Nevertheless, the exact association between fluid intake and CKD is often difficult to discern as dietary modifications, including fluid restriction, could potentially follow a diagnosis of CKD rather than precede it and therefore evaluating CVD mortality and all-cause mortality may be more appropriate clinical outcomes. Epidemiological studies of fluid intake have reported conflicting findings. The Adventist Health population cohort study, comprising 20 000 participants of both genders, suggested that a higher intake of water (but excluding other types of fluid) may be protective with a 40% reduction in CVD mortality among those drinking over five or more glasses of water per day compared with two or fewer glasses per day. On the contrary, the Netherlands cohort study of over 120 000 participants aged 55–69 years of both genders did not find any association between higher amounts of fluid consumption and CVD or stroke mortality.8–10 With the ongoing uncertainty of the beneficial health effects of water and other fluid consumption, plus the known differential effect of gender on CVD mortality, it is therefore crucial to examine the effects of the quantity of daily total fluid consumption and the types of fluid consumption on long-term health outcomes in gender-specific populations. We hypothesise that a higher intake of fluid consumption, particularly water intake, is associated with improved health outcomes in elderly women. The aims of this analysis are twofold. First, to determine whether a higher total daily fluid consumption is associated with a slower renal function decline and reduction in 10-year risk of CVD mortality and all-cause mortality; and second, to evaluate the relative kidney-related CVD mortality and all-cause mortality benefits of the different types of fluid compositions, including water and other beverages. Methods Study population One thousand five hundred women aged over 70 years, with a majority from higher socioeconomic groups, were recruited in 1998 in Perth, Western Australia to a 5-year prospective randomised controlled trial of oral calcium supplements (1.2 g of elemental calcium daily or matching placebo from 1998 to 2003) to prevent osteoporotic fractures. The details of the recruitment into this Calcium Intake Fracture Outcome Study (CAIFOS; Australian Clinical Trials Registry Registration Number: ACTRN012607000055404) have been previously published.11 At the completion of CAIFOS, participants were followed in an observational study for a further 10 years (2003–2013). Our study data and analyses are derived from these two associated studies. The Human Research Ethics Committee of the Western Australian Department of Health (DOHWA HREC) approved the data linkage study (approval number #2009/24) between CAIFOS and the follow-up observational study. Baseline data Details of baseline data collection have been previously published.11 In brief, baseline medical history (including diabetes, hypertension, smoking history) and medications were obtained from all participants, the latter verified with participants' general practitioners where required. Body mass index (BMI) measurements were obtained at baseline and at 5 years postrandomisation (1998 and 2003). At baseline, an average of three blood pressure readings were measured and recorded after the participants had been rested and seated for 5 min. Socioeconomic status (SES) was assessed using relative social advantage related to residential postcodes according to the Australian Bureau of Statistics method.12 This variable was divided into six categories: one being the most disadvantaged and six being the least disadvantaged. Fluid assessment A validated beverage intake questionnaire developed by the Cancer Council Victoria (Melbourne, Australia),13 which quantified habitual beverage consumption during the 12 months prior to study enrolment, was completed by each participant in 2003. Total fluid consumption, comprising water, black tea, coffee and other beverages (eg, chocolate beverages, herbal tea, juices and sugar beverages), excluding alcoholic beverages (including beer, wine and spirits), was calculated. The exposure study variable is cups (1 cup=250 mL) of total and types of fluids (specifically water, black tea, coffee, milk and other fluids) per day. Biochemistry Fasting blood samples were collected in 1998, 2003 and 2008 with sera stored in a −70°C freezer until analysis. Creatinine measurements were performed using stored sera with measurements of creatinine in 2003 and 2008 samples being available for 689 women (45.9%). Serum creatinine was analysed using an isotope dilution mass spectrometry traceable Jaffe kinetic assay for creatinine on a Hitachi 917 analyser (Roche Diagnostics GmbH, Mannheim, Germany). Estimated glomerular filtration rate (eGFR) was derived using the creatinine-based Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation.14 Assessment of clinical outcomes Participants' general practitioners verified their medical histories and medications where possible, and were coded using the International Classification of Primary Care—Plus (ICPC-Plus) method.15 Prevalent CVD was determined from hospital discharge data between 1980 and 1998 and were defined using diagnosis codes from the International Classification of Diseases, Injuries and Causes of Death Clinical Modification (ICD-9-CM, 390-459).16 Prevalent renal disease was collected between 1980 and 1998 using ICD-9-CM 17. These codes included glomerular diseases (ICD-9-CM codes 580–583); renal tubulointerstitial diseases (ICD-9-CM codes 593.3–593.5, 593.7); renal failure (ICD-9-CM codes 584–586) and hypertensive renal disease (ICD-9-CM code 403). The search for renal disease hospitalisations included any diagnosis code. The primary outcomes of the study were CVD mortality and all-cause mortality retrieved from the Western Australian Data Linkage System (WADLS) for each of the study participants from 2003 to 2013. CVD mortality was defined using primary death codes from ICD—the International Statistical Classification of Diseases and Related Health Problems, 10th Revision, Australian Modification (ICD-10-AM), I00-I99.17 Text fields from the death certificate were used to ascertain the cause(s) of deaths where coded death data were not yet available from the WADLS. Statistical analysis Baseline characteristics were expressed as mean and SD or median and IQR for continuous variables or as number and proportion for categorical variables. Correlation between total fluid intake and alcohol (gram/day (g/d)) was examined using Spearman's coefficient. Associations between fluid consumption and change in eGFR (absolute change in eGFR between 2003 and 2008) and risk of rapid renal decline (defined as reduction in eGFR of ≥3 mL/min/1.73 m2 annually between 2003 and 2008) were assessed using unadjusted and adjusted linear and logistic regression analyses, respectively. Restricted cubic splines were modelled to determine the linearity of the association between cups of fluid consumption per day and mortality. Associations between the total and types of fluids, CVD mortality and all-cause mortality were examined using the Cox proportional hazard regression models. The Cox models included covariates recorded at baseline in 1998 (ie, smoking history, SES, diabetes status, hypertension, systolic blood pressure, prevalent CVD, medications and treatment code (calcium supplementation vs no calcium supplementation)) and those recorded in 2003 (ie, fluid status, age, BMI and eGFR). Selection of covariates for inclusion in the multivariable-adjusted models were undertaken in two analytical approaches: (1) covariates with p<0.10 in the univariate models were selected (for each outcome including change in eGFR, rapid renal decline, CVD mortality and all-cause mortality), and (2) the β coefficient of each of the independent covariate and its association with each outcome in the univariate analyses was evaluated. If the coefficient of the independent covariate in the regression analyses changed by more than 10%, it was considered to be a confounder, and thus included in the final multivariable model. The covariates that were included in the multivariable models were selected from both analytical approaches. Fluid status and treatment code were included in all multivariable models irrespective of their associations in the univariate models. p Values of <0.05 in two-tailed testing were considered statistically significant. The proportional hazard assumptions of all Cox models were checked graphically by plotting the Schoenfield residuals, but there was no evidence of departures from proportional hazards for total and types of fluid intake. The data were analysed using SPSS (V.15; SPSS, Chicago, Illinois, USA) and STATA (V.11 StataCorp LP, College Station, Texas, USA). Results Baseline characteristics The baseline characteristics of the 1055 participants are shown in table 1. The mean age and burden of comorbidities of the 445 participants who were excluded (died between 1998 and 2003 or had missing data) were similar to the included cohort (data not shown). Over a follow-up time of 10 years between 2003 and 2013, 70 (6.6%) experienced rapid renal decline, and 362 (34.4%) died with 142 (13.5%) deaths attributed to CVD (table 2). The annual CVD mortality rate in our study cohort was 1346 deaths per 100 000 participants, as compared with an annual CVD mortality rate of 979 deaths per 100 000 population for women aged 75–84 years reported to the Australian Institute of Health and Welfare (AIHW; http://www.aihw.gov.au/cardiovascular-disease/deaths/#dt). Table 1 Baseline characteristics of participants Cohort (n=1055) Patient characteristics  Age (years in 2003, mean±SD) 80.0±2.6  Body mass index at entry (kg/m², mean±SD) 27.2±4.5  Body mass index at 60 m (kg/m², mean±SD) 27.2±4.7  Systolic blood pressure (mm Hg, mean±SD) 137.8±18.1  Diastolic blood pressure (mm Hg, mean±SD) 73.1±10.9  Previous and current smokers (n, %) 375 (35.6)  Diabetes (n, %) 53 (5.0)  Prevalent acute/chronic kidney disease (n, %) 0 (0.0)  Prevalent cardiovascular disease (n, %) 227 (21.5)  Socioeconomic status (n, %)*   Top 10% most highly disadvantaged 42 (4.0)   Highly disadvantaged 126 (11.9)   High–medium disadvantaged 172 (16.3)   Medium–low disadvantaged 162 (15.4)   Low disadvantaged 213 (20.2)   Top 10% least disadvantaged 333 (31.6) Medication use (n, %)  Aspirin 194 (18.4)  Statin 198 (18.8) Treatment with calcium supplements (n, %) 536 (50.8) Total and type of fluid composition in cups/day (median, IQR)  Total fluid 10.4 (8.5–12.5)  Water 4 (2–6)  Black tea 3 (1–4)  Coffee 1 (0–2)  Milk 1.5 (0.8–1.5)  Other fluids 0.4 (0–1) Data expressed as number (percentage), mean±SD or as median (IQR). *Missing socioeconomic status (n=7). Table 2 Clinical outcomes of participants Cohort (n=1055) Outcomes  eGFR 2003 (n=1055; mL/min/1.73 m², mean±SD) 57.2±14.7  eGFR 2008 (n=689; mL/min/1.73 m², mean±SD) 55.4±16.1  Change eGFR 2003–2008 (mL/min/1.73 m², mean±SD) −2.3±10.5  Rapid renal decline 2003–2008 (n, %) 70 (6.6)  CVD deaths 2003–2013 (n, %) 142 (13.5)  Deaths 2003–2013 (n, %) 362 (34.3) Data expressed as number (percentage) or as mean±SD. CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate. The median (IQR) intake of total fluid was 10.4 (8.5–12.5) cups per day, with water (median (IQR) 4 (2–6) cups per day) and black tea (median (IQR) 3 (1–4) cups per day) being the most frequent types of fluid consumed. Almost 22% of participants had prevalent CVD but no participants had prevalent CKD. Almost 70% of participants were from low-medium or low disadvantaged groups. There was no significant correlation between total fluid intake and alcohol intake (Spearman's correlation −0.024, p=0.45). Assessment of the relationship between total and types of fluid intake and outcomes The restricted cubic spline curves demonstrated that the relationship between total fluid intake and both CVD mortality and all-cause mortality was predominantly linear. Similarly, the relationships between water, black tea, coffee and other fluid intake and CVD mortality and all-cause mortality were also linear (see online supplementary figures S1 and S2). 10.1136/bmjopen-2016-011720.supp1supplementary figures Association between total fluid intake, 10-year CVD mortality and all-cause mortality There was no significant association between total fluid intake and CVD or all-cause mortality in the unadjusted and adjusted models. The unadjusted HRs for every cup per day higher intake of total fluid for CVD mortality and all-cause mortality were 0.96 (95% CI 0.91 to 1.05) and 0.97 (95% CI 0.93 to 1.02), respectively, whereas the adjusted HRs were 0.98 (95% CI 0.93 to 1.03) and 0.98 (95% CI 0.95 to 1.01), respectively. Other covariates associated with increased hazards of CVD mortality and all-cause mortality are shown in figure 1. Figure 1 Forest plots showing the association between total fluid consumption (every one cup higher intake per day), other covariates and risk of CVD and all-cause mortality in the adjusted Cox regression models. CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate. Association between fluid type, 10-year CVD mortality and all-cause mortality There was no association between water intake and risk of CVD mortality or all-cause mortality. The unadjusted HRs for every cup per day higher intake of water for CVD mortality and all-cause mortality were 1.03 (95% CI 0.96 to 1.10) and 1.01 (95% CI 0.96 to 1.05), respectively, whereas the adjusted HRs were 1.02 (95% CI 0.95 to 1.10) and 0.99 (95% CI 0.95 to 1.04), respectively. For every cup per day higher intake of black tea, the unadjusted HRs of CVD mortality and all-cause mortality were 0.89 (95% CI 0.82 to 0.98) and 0.91 (95% CI 0.86 to 0.96), respectively. The adjusted HRs for every cup per day higher intake of black tea for CVD mortality and all-cause mortality were 0.90 (95% CI 0.81 to 0.99) and 0.92 (95% CI 0.86 to 0.98), respectively (figure 2). There were no associations between coffee, milk and other fluid types (excluding water, black tea and coffee) intake and CVD mortality and all-cause mortality in the unadjusted and adjusted models. Figure 2 Forest plots showing the association between types of fluid consumption (every one cup higher intake per day), including water, black tea, coffee, milk and other fluids and risk of CVD and all-cause mortality, adjusted for age, blood pressure, prevalent CVD, diabetes and eGFR. CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate. Association between total and type of fluid intake and renal function decline There was no association between total fluid intake and odds of rapid renal decline, with unadjusted and adjusted ORs of 1.02 (95% CI 0.95 to 1.10) and 1.04 (95% CI 0.96 to 1.12), respectively, for every cup per day higher intake of total fluid. There was no association between fluid type and rapid renal decline. The adjusted OR for every cup per day higher intake of water and black tea were 1.04 (95% CI 0.94 to 1.16) and 1.03 (95% CI 0.90 to 1.18), respectively, adjusted for age, BMI, hypertension and diabetes. Only diabetes was associated with rapid renal decline with an adjusted OR of 3.38 (95% CI 1.47 to 7.75). There was no association between total fluid intake and change in eGFR. For every cup per day higher intake of total fluid, there was a −0.12 mL/min/1.73 m2 (95% CI −0.38 to 0.13) and −0.18 mL/min/1.73 m2 (95% CI −0.42 to 0.07) reduction in 5-year eGFR, respectively, in the unadjusted and adjusted models. There was no association between fluid types and change in eGFR. Only the presence of diabetes was associated with a decline in eGFR over 5 years of −6.28 mL/min/1.73 m2 (95% CI −10.38 to −2.19). Discussion The long-standing advice that drinking adequate amounts of fluid to avoid dehydration may lead to an improvement in health outcomes is currently not supported by consistent evidence. In this analysis of a prospective contemporary population-based cohort of older women over the age of 75 years, we have found no association between total fluid intake and 10-year hazards of CVD mortality and all-cause mortality. However, higher consumption of black tea was associated with a significant reduction in CVD mortality and all-cause mortality, independent of traditional CVD risk factors. Similar to our study, another Australian prospective, population-based cohort study of 3858 men and women aged over 48 years found no association between total fluid intake (from food and beverages excluding water) and all-cause and CVD mortality.8 Similar findings (water and other fluids excluding food were evaluated) were corroborated in another large population-based cohort of 120 852 men and women aged between 55 and 69 years recruited in the Netherlands.9 In contrast, an inverse association between pure water intake (excluding water from food) and risk of fatal coronary artery disease (CAD) was shown in a large cohort of 20 297 men and women aged at least 38 years without prevalent vascular disease or diabetes in the USA. Compared with individuals who consumed two or less glasses of water per day, those who consumed five or more glasses of water per day had experienced over 30% reduction in the risk of fatal CAD.10 The beneficial effects of higher fluid intake in reducing the rates of vascular events including recurrent stroke, myocardial infarction or all-cause mortality were further shown in a small prospective study of 465 patients with stroke.18 Our study could not find an association between total fluid intake and CVD mortality and all-cause mortality, even though we established a best-fit linear relationship between the continuous exposure factor, total fluid intake and outcomes. The challenges in comparing and interpreting the results from these epidemiological studies include the vastly different populations with varying characteristics, different eras, variations in the definition of fluid consumption and dissimilar follow-up periods. The basis of the suggestion that fluid intake is associated with beneficial health outcomes has been extrapolated from studies involving surrogate markers of health outcomes and as such the evidence supporting this assertion is weak. For example, mechanistic theories suggest that adequate fluid intake to avoid dehydration can lead to suppression of AVP and avoidance of adverse physiological processes such as elevated plasma viscosity and fibrinogen (putative surrogate markers of CVD),19 20 which potentially may improve renal blood flow and perfusion resulting in improved kidney function and health outcomes.3 To shed light on the inconsistent negative association between quantity of total fluid consumption and important health outcomes, the next logical step is to evaluate the association between composition of beverage consumption and health outcomes. In our study, we have shown an independent inverse association between consumption of black tea and CVD mortality and all-cause mortality. Higher intake of black tea was associated with a reduction in the relative hazards of CVD mortality and all-cause mortality. Nevertheless, our findings are in contrast to those of other epidemiological studies where no consistent patterns between types of beverages and clinical outcomes could be found. In the US study, a higher intake of fluid other than water (not examined by specific fluid types) was associated with an increased risk of fatal CAD in women but not in men, whereas the Netherlands study suggests that the inverse relationship between tea intake and CVD mortality was only apparent in men. A gender differential in the risk of CVD mortality and stroke mortality has been shown and although epidemiological studies have suggested that traditional vascular risk factors remain important in predicting mortality in men and women, it does raise the question of whether differences in lifestyle, dietary or behavioural factors may contribute to the gender disparity in CVD mortality.21 While the health benefits of black tea have been attributed to the antioxidant and anti-inflammatory properties of flavonoids (a major composition in tea—100–300 mg per serving of tea),22–24 there may be other beneficial components in black tea that remain unexplored. In a survey of US adults, women were more likely to consume a greater amount of flavonoids in their diet compared with men and the greatest source of intake of flavonoids was from tea.25 Of note, the flavonoid content in non-herbal tea depends on the strength, type of tea (particularly black tea) and the process of oxidisation. Furthermore, there are other food and beverage sources of flavonoids such as chocolate, nuts and red wine that may have modified the association between black tea and outcomes. The finding of a significant inverse association between tea consumption and CVD and all-cause mortality in women with CKD is intriguing but requires validation in other cohorts. Finally, we have previously shown that increased consumption of proanthocyanidins, a class of flavonoids, was associated with improved kidney function and reduced risk of kidney disease-related clinical events but did not address the association of renal function decline.26 In this study, there was no association between total fluid, water or black tea consumption and eGFR decline or risk of rapid renal decline, findings that have been contradicted in other large cohort studies. A cross-sectional analysis of 3427 adults with a mean age of 46 years from the National Health and Nutrition Examination Survey (NHANES) showed that CKD, defined as eGFR between 30 and 60 mL/min/1.73 m2, was associated with a lower intake of plain water of <2 L/day with adjusted OR of 2.36 (95% CI 1.10 to 5.06), compared with those with a higher water intake of >4.3 L/day.27 A similar protective effect between high fluid intake and development of CKD has also been shown in a cohort of older men and women with mean ages of 65 years.28 However, these studies defined CKD using a single time-point eGFR in the absence of other markers of CKD such as albuminuria. This association needs to be examined in other population cohorts using other markers of CKD and change in eGFR before this assertion can be substantiated. Our study has several strengths and limitations. The prospective nature and completeness of the data set suggest that selection and ascertainment biases between the exposure factor and outcomes are minimised. The analytical data on fluid intake were calculated before examination of the relation to clinical outcome data and were subjected to meticulous covariate analysis in an attempt to identify important collinearity that may have accounted for the observed associations. Identification of causality is limited by the potential for recall bias, the complexity of estimating total fluid intake and the variability of the composition of the various beverages. Our analysis only focused on baseline fluid consumption but not change in fluid consumption over time, which may have modified our result findings. In addition, fluid derived from dietary consumption, which is likely to account for a small percentage of overall daily fluid intake, amount/intensity of exercise and energy expenditure were not included in this study because of the uncertainty over the accuracy of this information but could have modified the association between fluid intake and outcomes. In addition, we had excluded alcohol in the estimation of total fluid intake because epidemiological studies have shown an independent association between alcohol intake and health outcomes (including CVD mortality).29 However, despite these limitations, the association between volume of black tea consumption and mortality remains significant even after adjustment for other factors. Future studies will determine the generalisability of our results to a wider population, including men and younger individuals. Conclusion We found no association between the amount or type of daily fluid intake and renal function decline. However, an inverse linear relationship between the daily amount of black tea ingested and long-term hazards of CVD mortality and all-cause mortality was found in a cohort of elderly women. External validation of our findings in men and women of varying ages is required to establish the true clinical significance of black tea intake. The authors wish to thank the staff at the Data Linkage Branch, Hospital Morbidity Data Collection and Registry of Births, Deaths and Marriages for their work in providing the data for this study. The authors would also like to thank the participants who were involved in the initial and follow-up studies and have consented to data collection linkage. Contributors: All authors participated in the design of the work, interpretation of the data (WHL, GW and JRL participated in the conception and analysis), drafting of the work, given final approval and are accountable for all aspects of the work. The lead author (WHL) affirms that the manuscript is an honest, accurate and transparent account of the study being reported; and that any discrepancies from the study as planned have been explained. Funding: The study was supported by Healthway Health Promotion Foundation of Western Australia, Sir Charles Gairdner Hospital Research Advisory Committee Grant and by project grants 254627, 303169 and 572604 from the National Health and Medical Research Council of Australia. The salary of JRL is supported by a National Health and Medical Research Council of Australia Career Development Fellowship. Competing interests: None declared. Ethics approval: The Human Ethics Committee of the University of Western Australia approved the study protocol and consent forms (approval number 05/06/004/H50). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Data regarding this cohort (plus other available variables for this cohort) can be requested from http://www.lsaw.com.au/pages/view/about. ==== Refs References 1 Valtin H “Drink at least eight glasses of water a day.” Really? Is there scientific evidence for “8×8”? Am J Physiol Regul Integr Comp Physiol 2002 ;283 :R993 –1004 . 10.1152/ajpregu.00365.2002 12376390 2 McCartney M Waterlogged? BMJ 2011 ;343 :d4280 10.1136/bmj.d4280 21750074 3 Edwards RM , Trizna W , Kinter LB Renal microvascular effects of vasopressin and vasopressin antagonists . Am J Physiol 1989 ;256 :F274 –8 .2916660 4 Sugiura T , Yamauchi A , Kitamura H High water intake ameliorates tubulointerstitial injury in rats with subtotal nephrectomy: possible role of TGF-beta . Kidney Int 1999 ;55 :1800 –10 . 10.1046/j.1523-1755.1999.00443.x 10231442 5 Go AS , Chertow GM , Fan D Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization . N Engl J Med 2004 ;351 :1296 –305 . 10.1056/NEJMoa041031 15385656 6 Sarnak MJ Cardiovascular complications in chronic kidney disease . Am J Kidney Dis 2003 ;41 :11 –17 . 10.1016/S0272-6386(03)00372-X 12776309 7 Sarnak MJ , Levey AS , Schoolwerth AC Kidney disease as a risk factor for development of cardiovascular disease: a statement from the American Heart Association Councils on Kidney in Cardiovascular Disease, High Blood Pressure Research, Clinical Cardiology, and Epidemiology and Prevention . Circulation 2003 ;108 :2154 –69 .14581387 8 Palmer SC , Wong G , Iff S Fluid intake and all-cause mortality, cardiovascular mortality and kidney function: a population-based longitudinal cohort study . Nephrol Dial Transplant 2014 ;29 :1377 –84 .24398890 9 Leurs LJ , Schouten LJ , Goldbohm RA Total fluid and specific beverage intake and mortality due to IHD and stroke in the Netherlands Cohort Study . Br J Nutr 2010 ;104 :1212 –21 . 10.1017/S0007114510001923 20456812 10 Chan J , Knutsen SF , Blix GG Water, other fluids, and fatal coronary heart disease: the Adventist Health Study . Am J Epidemiol 2002 ;155 :827 –33 . 10.1093/aje/155.9.827 11978586 11 Prince RL , Devine A , Dhaliwal SS Effects of calcium supplementation on clinical fracture and bone structure: results of a 5-year, double-blind, placebo-controlled trial in elderly women . Arch Intern Med 2006 ;166 :869 –75 . 10.1001/archinte.166.8.869 16636212 12 Australian Bureau of Statistics . Socio-economic indexes for areas . Canberra : CGPS , 1991 . 13 Hodge A , Patterson AJ , Brown WJ The Anti Cancer Council of Victoria FFQ: relative validity of nutrient intakes compared with weighed food records in young to middle-aged women in a study of iron supplementation . Aust N Z J Public Health 2000 ;24 :576 –83 . 10.1111/j.1467-842X.2000.tb00520.x 11215004 14 Levey AS , Stevens LA , Schmid CH A new equation to estimate glomerular filtration rate . Ann Intern Med 2009 ;150 :604 –12 . 10.7326/0003-4819-150-9-200905050-00006 19414839 15 Britt H A new coding tool for computerised clinical systems in primary care—ICPC plus . Aust Fam Physician 1997 ;26 (Suppl 2 ):S79 –82 .9254947 16 World Health Organization . Manual of the international statistical classification of diseases, injuries, and causes of death: based on the recommendations of the ninth revision conference, 1975, and adopted by the Twenty-ninth World Health Assembly. 1975 revision . Geneva : World Health Organization , 1977 . 17 World Health Organization . ICD-10: International Statistical Classification of Diseases and related health problems: tenth revision . 2nd edn Geneva : World Health Organization , 2004 . 18 Mucke S , Grotemeyer KH , Stahlhut L The influence of fluid intake on stroke recurrence—a prospective study . J Neurol Sci 2012 ;315 :82 –5 . 10.1016/j.jns.2011.11.024 22169398 19 Ernst E , Resch KL Fibrinogen as a cardiovascular risk factor: a meta-analysis and review of the literature . Ann Intern Med 1993 ;118 :956 –63 . 10.7326/0003-4819-118-12-199306150-00008 8489110 20 Lowe GD , Lee AJ , Rumley A Blood viscosity and risk of cardiovascular events: the Edinburgh Artery Study . Br J Haematol 1997 ;96 :168 –73 . 10.1046/j.1365-2141.1997.8532481.x 9012704 21 Janghorbani M , Hedley AJ , Jones RB Gender differential in all-cause and cardiovascular disease mortality . Int J Epidemiol 1993 ;22 :1056 –63 . 10.1093/ije/22.6.1056 8144286 22 Serafini M , Peluso I , Raguzzini A Flavonoids as anti-inflammatory agents . Proc Nutr Soc 2010 ;69 :273 –8 . 10.1017/S002966511000162X 20569521 23 Ivey KL , Hodgson JM , Croft KD Flavonoid intake and all-cause mortality . Am J Clin Nutr 2015 ;101 :1012 –20 . 10.3945/ajcn.113.073106 25832340 24 Ivey KL , Lewis JR , Prince RL Tea and non-tea flavonol intakes in relation to atherosclerotic vascular disease mortality in older women . Br J Nutr 2013 ;110 :1648 –55 . 10.1017/S0007114513000780 23628082 25 Chun OK , Chung SJ , Song WO Estimated dietary flavonoid intake and major food sources of U.S. adults . J Nutr 2007 ;137 :1244 –52 .17449588 26 Ivey KL , Lewis JR , Lim WH Associations of proanthocyanidin intake with renal function and clinical outcomes in elderly women . PLoS ONE 2013 ;8 :e71166 10.1371/journal.pone.0071166 23940710 27 Sontrop JM , Dixon SN , Garg AX Association between water intake, chronic kidney disease, and cardiovascular disease: a cross-sectional analysis of NHANES data . Am J Nephrol 2013 ;37 :434 –42 . 10.1159/000350377 23594828 28 Strippoli GF , Craig JC , Rochtchina E Fluid and nutrient intake and risk of chronic kidney disease . Nephrology (Carlton) 2011 ;16 :326 –34 . 10.1111/j.1440-1797.2010.01415.x 21342326 29 Ronksley PE , Brien SE , Turner BJ Association of alcohol consumption with selected cardiovascular disease outcomes: a systematic review and meta-analysis . BMJ 2011 ;342 :d671 10.1136/bmj.d671 21343207

PMC005xxxxxx/PMC5372041.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01249210.1136/bmjopen-2016-012492Health Services ResearchResearch150617041703Exploring similarities and differences in hospital adverse event rates between Norway and Sweden using Global Trigger Tool http://orcid.org/0000-0001-6527-5636Deilkås Ellen Tveter 12Risberg Madeleine Borgstedt 3Haugen Marion 4Lindstrøm Jonas Christoffer 2Nylén Urban 5Rutberg Hans 67Michael Soop 5 1 National Patient Safety Program, Norwegian Directorate of Health, Oslo, Norway 2 Health Services Research Center, Akershus University Hospital, Lørenskog, Norway 3 Center for Healthcare Development, County Council of Östergötland, Sweden 4 Norwegian Computing Center, Oslo, Norway 5 National Board of Health and Welfare, Stockholm, Sweden 6 Division of Health Care Analysis, Department of Medical and Health Sciences, Linköping University, Linköping, Sweden 7 Swedish Association of Local Authorities and Regions, Stockholm, SwedenCorrespondence to Dr Ellen Tveter Deilkås; elde@ahus.no2017 20 3 2017 7 3 e0124923 5 2016 18 1 2017 23 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives In this paper, we explore similarities and differences in hospital adverse event (AE) rates between Norway and Sweden by reviewing medical records with the Global Trigger Tool (GTT). Design All acute care hospitals in both countries performed medical record reviews, except one in Norway. Records were randomly selected from all eligible admissions in 2013. Eligible admissions were patients 18 years of age or older, undergoing care with an in-hospital stay of at least 24 hours, excluding psychiatric and care and rehabilitation. Reviews were done according to GTT methodology. Setting Similar contexts for healthcare and similar socioeconomic and demographic characteristics have inspired the Nordic countries to exchange experiences from measuring and monitoring quality and patient safety in healthcare. The co-operation has promoted the use of GTT to monitor national and local rates of AEs in hospital care. Participants 10 986 medical records were reviewed in Norway and 19 141 medical records in Sweden. Results No significant difference between overall AE rates was found between the two countries. The rate was 13.0% (95% CI 11.7% to 14.3%) in Norway and 14.4% (95% CI 12.6% to 16.3%) in Sweden. There were significantly higher AE rates of surgical complications in Norwegian hospitals compared with Swedish hospitals. Swedish hospitals had significantly higher rates of pressure ulcers, falls and ‘other’ AEs. Among more severe AEs, Norwegian hospitals had significantly higher rates of surgical complications than Swedish hospitals. Swedish hospitals had significantly higher rates of postpartum AEs. Conclusions The level of patient safety in acute care hospitals, as assessed by GTT, was essentially the same in both countries. The differences between the countries in the rates of several types of AEs provide new incentives for Norwegian and Swedish governing bodies to address patient safety issues. HEALTH SERVICES ADMINISTRATION & MANAGEMENTPUBLIC HEALTHEPIDEMIOLOGY ==== Body Strengths and limitations of this study The samples are drawn from all eligible hospital admissions in both countries in 2013. The samples represent 1.9% of all eligible hospital admissions in Norway and 1.4% in Sweden. This is the first explorative cross-country comparison of adverse event rates and types based on Global Trigger Tool. The study does not include demographic data or other patient characteristics. The reviewing process differed slightly between the countries and inter-rater reliability between review teams across countries was not assessed. Introduction Interpreting the extent to which patient safety indicators accurately reflect international differences in patient safety must be done with caution. Differences may reflect the way that countries report, code and calculate rates of adverse events (AEs) as seen by the Organisation for Economic Co-operation and Development (OECD) indicators.1–3 In some cases, higher AE rates may signal more developed patient safety monitoring systems rather than substandard care. In 2007, the Nordic Council of Ministers initiated a project to develop and strengthen Nordic efforts to measure and monitor quality and safety in healthcare. Results of the Nordic project were reported in 2010.4 As part of the project, Nordic expert groups were constituted in various fields of patient safety like patient safety culture surveys and Global Trigger Tool (GTT).5 6 Nordic experiences from using GTT have been reported previously.7 The Nordic co-operation and exchange of experiences promoted the use of GTT in Norway and Sweden which, to the best of our knowledge, are among the few countries that have required use of GTT in all hospitals as part of a national government policy. A research scan from the Health Foundation has considered GTT to compare well to other approaches, to be relatively sensitive and to identify significantly more AEs compared with self-reporting or other chart audit methods.8 GTT produces substantial inter-rater reliability within and between independent internal teams.9 10 An evidence scan from the New Zealand Health Quality and Safety Commission considers GTT to be the most accurate and efficient method to identify AEs, although further work is needed to confirm its reliability.11 12 Most of the studies describing its use are based on large samples from multiple hospitals.12 13 Although GTT is considered relevant for measuring AEs at the national level,8 we are only aware of one publication describing the process of doing so.14 This publication presents Norwegian results from 2010 to 2013, while in this study we compare results from 2013 between Norway and Sweden. Norway and Sweden have similar structural conditions and contexts for healthcare as tax-based funding, similar socioeconomic status, demographic characteristics, publically funded education of healthcare employees and democratic policies pursuing equal access.15 Both countries are politically stable and have over decades pursued co-operation and mutual learning.4 Sweden is a step ahead regarding quality registers and did many hospital mergers in the mid-1990s, while these happened 10 years later in Norway.16 In 2010 when the Norwegian and Swedish national patient safety campaigns were prepared, Sweden had already made a national AE study based on the Harvard Medical Practice Study (HMPS) protocol.17 The national patient safety campaign initiatives in the two countries were not co-ordinated, but target areas like, for example, hospital-acquired infections, pressure ulcers and falls were similar. In Norway, GTT was considered a cost-effective method to monitor levels of AEs at hospital level, in relation to the patient safety campaign efforts. Sufficient knowledge and experience was at hand for the purpose.4 18 Evidence of AE rates from other countries created an additional momentum in Norway to facilitate, so that the local GTT data could be used to estimate national AE rates.14 In the Swedish campaign, it was decided to use GTT rather than the HMPS method as GTT is not just a tool for measurement but also a tool for learning and patient safety improvement work locally. During planning and implementation of the GTT tool, experiences were exchanged in expert meetings and correspondence within the network derived from the Nordic Council of Ministers. In this paper, we explore similarities and differences in hospital AE rates between Norway and Sweden based on GTT in 2013. Norwegian patient safety campaign In 2011, the Norwegian government launched a national patient safety campaign.18 The aims were to reduce patient harm, increase knowledge and competence in patient safety and improve patient safety culture. GTT was chosen as the way to do continuous medical record reviews in all acute care hospitals. The patient safety campaign also set requirements for the hospitals to aim at reducing preventable AEs by 20% during the campaign period, which ended in 2013. To monitor AEs, annual rates were estimated at national level and data at hospital level were plotted in run charts. Swedish patient safety campaign A national initiative to increase patient safety was launched by the government and The Swedish Association of Local Authorities and Regions (SALAR) for the period 2011–2014.19 The initiative involved financial incentives and the agreement focused on patient safety culture, hospital-acquired infections, pressure ulcers, prescription of antibiotics and medication errors. As a part of this initiative, all Swedish acute care hospitals have performed medical record reviews according to the GTT method since 2012. Education of review teams and creation of a national database for registration of data from the reviews were included in this initiative. Methods Medical record review by use of the GTT GTT is an internationally recognised and standardised procedure for medical record review to identify and estimate rates and severity of AEs among adult patients in non-psychiatric hospital admissions.6 It involves an educated team of two primary reviewers, often nurses, and a physician. A list of criteria (triggers) that indicate a higher probability of AEs is used to identify details in the record that might indicate possible AEs. Readmission to hospital within 30 days after discharge and infection during the hospital stay are examples of triggers. The primary reviewers examine records on their own before they compare results with each other and ultimately validate them with the physician. The identified AEs are then classified regarding type and rated on a 5-point severity scale (table 1). Table 1 Categories used to assess the severity of adverse events (AEs) E AE contributed to temporary harm to the patient which required intervention F AE contributed to temporary harm to the patient which required initial or prolonged hospitalisation G AE contributed to permanent patient harm H Intervention was required to sustain life I AE contributed to patient death Sampling The data presented in this study were collected separately by independent government-initiated processes in each country. The study was designed after data had been collected. Data from reviews of medical records from hospital admissions in 2013 were used. The reviews were based on the GTT methodology on randomly selected medical records. Selection criteria were completed records from admissions of at least 18-year-old patients with in-hospital stays that lasted at least 24 hours. Admissions in psychiatric or rehabilitation departments were excluded. The reviews included the whole period of hospitalisation even if the patients had been treated in different departments during the hospital stay. The admissions that were completed 1 month earlier were eligible for the random selection as specified by the GTT method. The whole population of eligible admissions in each hospital had to be covered by one or more randomly selected samples that could not overlap. For identification of records to review, lists of all eligible admissions were retrieved locally by the electronic administrative system in each hospital, and a random selection of the required number of records was done. Settings In Norway, 21 acute care hospitals (all except one) and two small non-acute care hospitals, with a total of 45 GTT teams, participated in the review. The teams selected at least 10 medical records bimonthly, which two trained registered nurses (RNs) first reviewed separately. They then compared their results and validated them with a physician. All 63 Swedish acute care hospitals participated. The minimum monthly number of randomly selected admissions reviewed was 40 for university hospitals, 30 for the central county council hospitals and 20 for the small hospitals. Each hospital had its own review team. In some of the university hospitals, more than one team performed the reviews but the results were then pooled into one list for the hospital. An RN searched the records for triggers. After that a team, with one or more RN and one or more physicians, assessed records with positive triggers, to identify possible AEs and classify them. Reviewers in both countries were employed in the hospitals where they worked clinically. They were all clinically experienced reviewers although some worked in the quality department, where they sometimes did not do clinical work at the time of the review. Our strategy leans on a study where internal GTT teams found more AEs than external teams.9 Translation and validation In Norway, a professional translator translated the original GTT white paper to Norwegian in 2010.20 It was then slightly modified regarding a few triggers.14 A protocol with rules for the national co-ordination of the GTT review was provided. The list of frequently asked questions (FAQ) was expanded with experience during the first 6 months of the review, which started in January 2011. The translated original GTT white paper, together with the FAQ list and the protocol, constituted the Norwegian manual. In 2008, a Swedish translation and adaptation of the GTT method was introduced in some hospitals. The Swedish manual was revised in 2012.21 Accordingly, triggers that were seldom found or rarely indicated an AE were omitted and four new triggers were introduced to increase sensitivity for AEs in non-surgical care. The descriptions of triggers were reformulated and markedly expanded in order to facilitate assessment of the severity of the AEs.22 Training and standardisation In Norway, training was organised and delivered by the secretariat of the national patient safety campaign, starting in January 2011. The teams were since then trained by a physician specialised in internal medicine, who also provided support by phone and email. The standard 1-day course, according to the GTT manual, included review of five training records. Forty additional medical records were required to be reviewed before participating in the ordinary reviews. Teams were gathered annually, by the secretariat of the national patient safety campaign, at national meetings to present results, cases, and to review and calibrate routines and definitions. In Sweden, training in the GTT methodology for review teams was initiated in the beginning of 2012 by educational teams appointed by SALAR. Each educational session was led by a physician and a nurse, both experts in GTT, and comprised a combination of lectures and training for 1 day. Support from the central team of educators was continuously available by phone. Meetings for review teams were arranged regionally for calibration and support. Definitions, categorising and reporting of data In both countries, an AE was defined according to the GTT definition of harm as an ‘unintended physical injury resulting from, or contributed to, by medical care that requires additional monitoring, treatment or hospitalisation or that results in death’.6 In both countries, AEs were categorised according to severity and type. Severity was categorised according to ‘National Coordinating Council for Medication Error Reporting and Prevention (NCC MERP) index’.6 An AE could only be categorised to one grade of severity (table 1). Two compounded severity categories were constructed in the Norwegian monitoring system in order to make more robust measures; E-I which include all AEs and F-I which include the four most serious categories (table 2).14 The category F-I had previously been used in the American study of AEs among Medicare beneficiaries.23 Both compounded categories were applied in the comparison of Norwegian and Swedish AE rates. Table 2 Comparison of adverse event rates according to severity category Severity category Norway—mean (95% CI) Sweden—mean (95% CI) Sweden minus Norway difference (95% CI) E-I 12.96% (11.68% to 14.25%) 14.41% (12.59% to 16.34%) 1.44 (−0.82 to 3.78) F-I 7.59% (6.62% to 8.68%) 7.70% (6.71% to 8.75%) 0.11 (−1.38 to 1.56) E 6.42% (5.67% to 7.17%) 8.10% (6.72% to 9.54%) 1.68 (0.06 to 3.34) F 6.68% (5.70% to 7.78%) 7.01% (6.10% to 7.97%) 0.32 (−1.09 to 1.76) G 0.72% (0.53% to 0.92%) 0.42% (0.31% to 0.56%) −0.29 (−0.53 to −0.06) H 0.20% (0.08% to 0.33%) 0.13% (0.08% to 0.18%) −0.07 (−0.22 to 0.06) I 0.30% (0.21% to 0.40%) 0.28% (0.20% to 0.38%) −0.02 (−0.15 to 0.11) Rates (%) of hospital admissions with at least one adverse event in compounded severity categories E-I and F-I, and individual severity categories E, F, G, H and I, in Norway (N=45) and Sweden (N=63) in 2013. In Norway, 23 types of AEs were specified (table 3) based on categories adopted from Sweden. AEs were reported in an Excel template, together with supplementary information regarding type, severity and numbering according to admission, so that AEs related to the same admission could be considered and counted together in the compounded severity categories E-I and F-I. When AEs were reported, the types were not mutually exclusive. If relevant, one AE could be categorised as more than one type, for example, both a postoperative bleeding and a reoperation. The template was annually sent to the secretariat of the national patient safety campaign. AE rates were also plotted by the Norwegian hospitals in run charts, which facilitated monitoring. Table 3 Comparison of adverse event (AE) rates in severity category E-I according to type Type of adverse event Norway (%) Sweden (%) Sweden minus Norway difference (95% CI) Urinary tract infection 2.119 1.549 −0.570 (−0.966 to −0.064) Surgical complication 1.888 0.885 −1.003 (−1.425 to −0.4209) Adverse drug event 1.854 1.625 −0.229 (−0.885 to 0.473) Postoperative wound infection 1.624 1.588 −0.037 (−0.385 to 0.550) Other infection 1.336 1.558 0.222 (−0.215 to 0.765) Pneumonia all types 1.279 0.885 −0.394 (−0.770 to −0.045) Postoperative haematoma 0.834 0.562 −0.272 (−0.521 to 0.026) Reoperation 0.809 0.596 −0.213 (−0.463 to 0.071) Pressure ulcer 0.600 1.285 0.684 (0.334 to 1.004) Other 0.435 1.036 0.601 (0.231 to 1.106) Organ injury 0.414 0.289 −0.125 (−0.293 to 0.061) Thrombosis 0.400 0.325 −0.075 (−0.239 to 0.089) Falls 0.372 0.756 0.384 (0.139 to 0.579) Allergic reaction 0.306 0.259 −0.047 (−0.162 to 0.093) Postpartum AE 0.249 0.396 0.147 (−0.075 to 0.395) Medical equipment 0.028* 0.033* 0.005 (−0.039 to 0.045) Ventilator-related pneumonia 0.084 0.137 0.053 (−0.027 to 0.178) Central line infection 0.073* 0.161 0.088 (0.026 to 0.189) Wrong-side surgery 0.000* 0.019* 0.019 (0.004 to 0.042) Types specified only in Sweden  Distended urinary bladder 1.667  Skin or vessel 1.131  Sepsis 0.623  Vital signs 0.438  Haemorrhage—not surgery 0.431  Anaesthesia-related AE 0.130  Neurological 0.108 Types specified only in Norway  Haemorrhage all types 0.990  Postoperative respiratory AE 0.215  Deterioration of chronic condition 0.183  Fractures 0.069 Rates (%) of hospital admissions with at least one AE according to type, severity category E-I combined, in Norway and Sweden in 2013. *Types with five or less observations greater than zero in Norway or Sweden; the corresponding CI must be interpreted with caution. In Sweden, 26 types of AEs were specified (table 3). Nineteen AE types were specified in both countries giving opportunity for comparisons. The results from each hospital in Sweden were entered into a national database where local results could be compared with national average results. The findings of the record review on a national level were presented in a report produced by SALAR.24 In addition, the teams in both countries reported the total number of admissions that the investigated records had been randomly selected from. This was used for weighting the team results, when making national estimates. Ethics Approval from research ethics committee was not required or applied for in either country for this data collection and analysis. The Ministry of Health and Care Services in Norway concluded that GTT could be applied as a means of quality assurance at the hospitals within the framework of national regulation, provided that data reported from the hospitals to the national campaign were anonymous. Data reported from hospitals in Norway did not identify individual patients and were evaluated by the Norwegian Data Protection Authority to be anonymous. The study was thus not affected by the Norwegian Health Research Act. In Sweden, the record review was conducted in accordance to national regulation regarding the use of medical records and seen as a part of quality improvement initiatives in the hospitals. Personal identification numbers were not collected or entered into the national database. Statistical analysis Cross-sectional analysis was performed on data from the GTT teams to calculate national AE rates with associated 95% CIs, divided into types and severities. The national AE rates were calculated as a weighted average of individual means for the GTT teams. The weight of one GTT team was equal to the number of admissions that investigated records had been randomly selected from for the team, divided by the total number of admissions that investigated records had been randomly selected from for all teams. Non-parametric CIs for the national AE rates were calculated using 10 000 bootstrap simulations25 since AEs according to type and severity were not symmetrically distributed. The bootstrap was performed by drawing randomly with replacement from the individual means for all 45 GTT teams in Norway and all 63 hospitals in Sweden (each with its own review team). An observation was drawn from the data with probability equal to the weight of the GTT team. For all analyses, the significance level is 5%. The calculations were performed with Microsoft Excel V.2010 and the R statistical software (R: A Language and Environment for Statistical Computing [program]. Vienna, Austria: R Foundation for Statistical Computing, 2015.) Results A total of 10 986 records, randomly selected from 569 714 admissions in 23 hospitals (with a total of 45 GTT teams) were reviewed in Norway in 2013. This is equivalent to 1.9% of all eligible admissions. A total of 19 141 medical records randomly selected from 1 345 506 admissions in 63 hospitals (each with its own GTT team) were reviewed in Sweden. This is equivalent to 1.4% of all eligible admissions. In the samples, a total of 1672 AEs across all severity categories were identified in Norway and 3217 in Sweden. Rates of admissions with one or more AEs in all severity categories are presented in table 2. There was no significant difference between Norway and Sweden in the rates of hospital admissions with AEs in severity categories E-I, that also include the less severe AEs. Nor was there a significant difference in categories F-I which only include AEs of higher severity, or in category I; AEs that contributed to patient's death. The variation in AE rates between GTT teams in Norway, and between hospitals in Sweden was large; for E-I: minimum 2.1% and maximum 21.7% in Norway and minimum 3.6% and maximum 31.1% in Sweden; for F-I: minimum 0% and maximum 15.0% in Norway and minimum 0.8% and maximum 19.3% in Sweden. In table 3, rates of different AE types in severity category E-I are compared. There were significantly higher rates of surgical complications in Norwegian hospitals compared with Swedish hospitals. Swedish hospitals had significantly higher rates of pressure ulcers, falls and ‘other’ AEs. In table 4, rates of different types of AE in severity category F-I are compared. Norwegian hospitals had significantly higher rates of surgical complications than Swedish hospitals. Swedish hospitals had significantly higher rates of postpartum AEs. Eighty-seven per cent of Norwegian AEs of higher severity (F-I) were in the F category, while 90% of Swedish AEs of higher severity were in the F category. Table 4 Comparison of adverse event (AE) rates in severity category F-I according to type Type of adverse event Norway (%) Sweden (%) Sweden minus Norway difference (95% CI) Surgical complication 1.429 0.569 −0.860 (−1.187 to −0.420) Postoperative wound infection 1.204 1.403 0.199 (−0.125 to 0.737) Adverse drug event 1.133 0.860 −0.273 (−0.804 to 0.259) Other infection 0.893 0.814 −0.079 (−0.334 to 0.227) Reoperation 0.762 0.550 −0.213 (−0.460 to 0.054) Pneumonia all types 0.728 0.605 −0.122 (−0.424 to 0.104) Postoperative haematoma 0.611 0.325 −0.286 (−0.453 to −0.076) Urinary tract infection 0.518 0.530 0.012 (−0.211 to 0.264) Other 0.343 0.597 0.254 (0.009 to 0.562) Organ injury 0.342 0.206 −0.135 (−0.311 to 0.035) Thrombosis 0.333 0.226 −0.107 (−0.237 to 0.046) Falls 0.148 0.182 0.034 (−0.124 to 0.132) Allergic reaction 0.127 0.070 −0.057 (−0.114 to 0.011) Pressure ulcer 0.067* 0.243 0.176 (0.037 to 0.249) Central line infection 0.059* 0.078 0.019 (−0.028 to 0.084) Ventilator-related pneumonia 0.024* 0.086 0.062 (0.016 to 0.119) Postpartum AE 0.019 0.217 0.198 (0.101 to 0.307) Medical equipment 0.005* 0.018* 0.013 (−0.015 to 0.031) Wrong-side surgery 0.000* 0.017* 0.017 (0.004 to 0.039) Rates (%) of hospital admissions with at least one AE according to type, severity category F-I combined, in Norway and Sweden in 2013. *Types with five or less observations greater than zero in Norway or Sweden; the corresponding CI must be interpreted with caution. Discussion We have described findings of an explorative study where medical record review with GTT has been used to compare national rates of AEs in acute care hospitals. There are no statistically significant differences between the overall estimates of AEs between the two countries. There are, however, differences in types of AEs between Norway and Sweden. Similar overall results between Norway and Sweden were expected considering the similar conditions and contexts of the two countries' healthcare systems.15 The finding that estimated AE rates in Norwegian hospitals correlate with patients’ surveyed perceptions of patient safety supports the validity of the results.26 Combined, this may indicate that AE rates produced by GTT may be useful for comparing hospital patient safety between countries. More research on issues like inter-rater reliability and validity across countries is, however, needed before further cross-country comparison of GTT results is advised.10 In Norwegian hospitals, there were significantly higher rates of surgical complications in the combined severity category E-I, compared with Swedish hospitals. Swedish hospitals had significantly higher rates of pressure ulcers, falls and ‘other’ AEs. Norwegian hospitals had significantly higher rates of surgical complications also of higher severity (F-I) than Swedish hospitals. Swedish hospitals had significantly higher rates of postpartum AEs of higher severity. It is well known that an increased volume of a specific surgical procedure performed by the same department and team improves medical outcome. In both countries, there are geographical aspects and a question of availability of hospital care that influence the organisation of national health services and the proportion of smaller hospitals as compared with large hospitals and centralised care. Such aspects might contribute to the differences in the amount of AEs related to surgical procedures with higher numbers seen in Norway. The higher rates in Sweden of pressure ulcers, falls and ‘other’ AEs also indicate that there are differences in the outcome of nursing care between the countries. Differences in the organisation of departments, workload, nursing competence and other factors may have contributed.27 Results presented in ‘Health at a glance 2013’2 show that obstetric trauma with lacerations of perineum after deliveries with instrument as well as without instrument are more frequent in Sweden than in Norway. This supports the finding of the present study that postpartum AEs (corresponding to lacerations of the perineum) were more common in Swedish hospitals. In Sweden, distended urinary bladder was used as a specific trigger and also did show to be a common type of AE. Late detection of distended urinary bladder has earlier been identified as a frequent AE in a Swedish orthopaedic clinic and it has also been reported that patients that have suffered distended urinary bladder sometimes have urinary, psychosocial and emotional problems for long afterwards.28 29 In Norway, distended urinary bladder was not used as a trigger and this type of AE was classified as ‘other’ and therefore not possible to identify separately. Prevention of urinary tract infection by restrictive use of urinary bladder catheters were part of both the Norwegian and Swedish patient safety campaigns. The tendency to lower rate of urinary tract infections in Swedish hospitals may reflect a more restrictive use of catheters, which in turn may have entailed an increased rate of distended urinary bladder. We wish to emphasise that the rates of hospitalisations with AEs that contributed to death (severity category I); 0.3% in both countries, do not reflect deaths that could have been avoided as we did not assess the extent to which the AEs contributed to death nor their preventability. When interpreting the results in this study, it should be born in mind that the GTT methods used in the two countries were similar but not identical. Both countries adhered to the original definition and method, paying attention to context so that the translated version should be comprehensible. However, in Sweden the initial review in search for triggers in most teams was done by only one instead of two reviewers. This modification could potentially have influenced the results. However, studies report κ values from 0.53 to 0.73 on triggers and 0.40 to 0.60 on AEs on agreement between primary nurse reviewers which is fair.12 One could also presume that using only one primary reviewer would lead to lower AE rates of all types. That is not the case in our study; some AE types were more common in Norway and some were more common in Sweden. We therefore consider that modifying the reviewing process in Sweden probably has had minor influence on the results. Still we recommend avoiding any deviation from the original reviewing method in future cross-country comparisons using the GTT. Cross-country comparisons based on medical record review require that the samples reviewed are representative for the type of hospital care in each country. In this study, the numbers of records reviewed are probably the largest random samples of admissions drawn from all hospitals in two countries that have been reviewed with GTT. That should with reason be sufficient for comparing overall results as well as specific AE type rates. Unfortunately, data protection regulations did not allow us to collect individual demographic or other patient data, which would have allowed us to identify and correct for possible differences in characteristics between the populations of the two countries which could have influenced the results. However, the risk for such differences is probably minor as the general demographics and other population characteristics of the two countries are very similar. The variation in AE rates was large between GTT teams in Norway, and between hospitals in Sweden. That is expected since rates are based on cross-sectional analyses of small samples allowing large random variation. In addition, there are differences between characteristics of hospitals' patient populations and activities. For this reason, we do not use GTT results for comparison between hospitals. No previous studies have to the best of our knowledge compared national rates of hospital AEs between countries based on the GTT. Other studies have presented results for regions and healthcare systems,14 30 31 as well as subgroups at national level.23 International comparisons of levels of patient safety are of great interest as a means to study organisational development in healthcare.2 3 32 Comparison based on nationwide data should, however, be interpreted with caution and in order to improve measures and data quality for international comparisons, several methods have been developed.1 4 Presently, there is no generally accepted gold standard for such comparisons. Agency for Healthcare Research and Quality (AHRQ) and OECD indicators based on hospital discharge data are used but some indicators remain unreliable for international comparisons due to varying data quality, undercoding and lack of precision in documentation. Furthermore, there are differences in International Classification of Diseases (ICD) coding guidelines and practices, and in many countries coding is influenced by financial incentives.3 In a recent publication, it is argued that the current version of the ICD, the 10th revision, is not optimal for the capture of healthcare-related harm and injury events which has resulted in the development of a new framework for the coding and capture of healthcare-related harm.33 ICD codes that better identify harm could be a good supplement to medical record review with GTT, since ICD codes also can be used as triggers. Conclusions AE rates as assessed by GTT may be useful for comparing hospital patient safety between countries. In our explorative study, the overall level of patient safety in acute care hospitals was essentially the same in both countries. This is expected considering similarities between the two countries' healthcare systems. The difference between the countries in the rates of several types of AEs merit further investigation and provide new incentives for Norwegian and Swedish governing bodies to address patient safety issues. We thank GTT teams in all the hospitals for their contributions to the study. We thank the referees for valuable comments, which have led to great improvements to the quality of the paper. Contributors: ETD was responsible for designing the study in Norway, training of reviewer teams, data collection, analysis and interpretation of data and writing the manuscript. MBR contributed with Swedish data collection and in their interpretation. MH participated in data collection in Norway, in the statistical analysis and interpretation, in addition to critically revising the manuscript. JCL participated in the statistical analysis and interpretation, in addition to revising the manuscript. UN participated in designing the study in Sweden, training of reviewer teams, data collection, analysis and interpretation of data, and revising the manuscript critically. HR participated in designing the study in Sweden, training of reviewer teams, data collection, analysis and interpretation of data and revising the manuscript critically. MS participated in designing the study in Sweden, analysis and interpretation of data and revising the manuscript critically. All authors read and approved the final manuscript. Funding: No grants were specifically awarded for this research. Reviewer teams were funded by the hospitals. The research was funded by Akershus University Hospital's Health Services Research Unit, the Norwegian Computing Center and by the Swedish Association of Local Authorities and Regions (SALAR). Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: The data are administered by the Norwegian Patient Safety Program, at the Norwegian Directorate of Health and by the Swedish Association of Local Authorities and Regions (SALAR). No additional data are available. ==== Refs References 1 Mainz J , Bartels P , Rutberg H International benchmarking. Option or illusion? Int J Qual Health Care 2009 ;21 :151 –2 . 10.1093/intqhc/mzp001 19188141 2 OECD . Health at a Glance 2013: OECD Indicators. Health at a glance , 2013 . 3 Drösler SE , Klazinga NS , Romano PS Application of patient safety indicators internationally: a pilot study among seven countries . Int J Qual Health Care 2009 ;21 :272 –8 . 10.1093/intqhc/mzp018 19395469 4 Nordisk Ministerråd. Nordisk kvalitetsmåling i sundhedsvæsenet. København, Denmark , 2010 :119 –23 . 5 Working group on Patient Safety under Nordic Ministry Council . A report on patient safety Copenhagen : Styrelsen for Patientsikkerhed , 2016 . 6 Griffin F , Resar R IHI global trigger tool for measuring adverse events IHI innovation series . 2nd edn Cambridge, MA: Institute for Healthcare Improvement ; 2009 . 7 Doupi P , Svaar H , Bjørn B Use of the Global Trigger Tool in patient safety improvement efforts: Nordic experiences . Cogn Tech Work 2015 :17 –45 . 10.1007/s10111-014-0302-2 8 Health Foundation . Evidence scan: global trigger tools . London, UK : The Health Foundation , 2010 . 9 Sharek PJ , Parry G , Goldmann D Performance characteristics of a methodology to quantify adverse events over time in hospitalized patients . Health Serv Res 2010 ;46 :654 –78 . 10.1111/j.1475-6773.2010.01156.x 20722749 10 Hanskamp-Sebregts M , Zegers M , Vincent C Measurement of patient safety: a systematic review of the reliability and validity of adverse event detection with record review . BMJ Open 2016 ;6 :e011078 10.1136/bmjopen-2016-011078 11 Health Quality & Safety Commission . The global trigger tool: a review of the evidence—report for the health quality & safety commission New Zealand . Wellington : Health Quality & Safety Commission , 2013 . 12 Good VS , Saldaña M , Gilder R Large-scale deployment of the Global Trigger Tool across a large hospital system: refinements for the characterisation of adverse events to support patient safety learning opportunities . BMJ Qual Saf 2011 ;20 :25 –30 . 10.1136/bmjqs.2008.029181 13 Rutberg H , Borgstedt Risberg M , Sjødahl R Characterisations of adverse events detected in a university hospital: a 4-year study using the Global Trigger Tool method . BMJ Open 2014 ;4 :e004879 10.1136/bmjopen-2014-004879 14 Deilkås E , Bukholm G , Lindstrøm JC Monitoring adverse events in Norwegian hospitals from 2010 to 2013. BMJ Open 2015 ;5 :e008576 10.1136/bmjopen-2015-008576 15 Veggeland Noralv The current Nordic welfare state model . New York : Nova Science Pub Inc , 2016 . 16 Magnussen J , Vrangbaek K , Saltman RB Nordic health care systems—recent reforms and current policy challenges . Berkshire, UK : Mc Graw Hill Open University Press , 2009 . 17 Soop M , Fryksmark U , Köster M The incidence of adverse events in Swedish hospitals: a retrospective medical record review study . Int J Qual Health Care 2009 ;21 :285 –91 . 10.1093/intqhc/mzp025 19556405 18 The Norwegian patient safety campaign secretariat . Final report. Secondary Final report 2014. http://www.pasientsikkerhetsprogrammet.no/no/I+trygge+hender/L%C3%A6r+om+programmet/_attachment/2925?_ts=146d6eb36d2 19 Patientsäkerhetssatsning 2011 överenskommelse mellan staten och Sveriges Kommuner och Landsting om förbättrad patientsäkerhet. In: Government TS, ed. Stockholm , 2011 . 20 Den nasjonale pasientsikkerhetskampanjen. Strukturert journalundersøkelse, ved bruk av Global Trigger Tool for å identifisere og måle forekomst av skader i helsetjenesten , 2010 . 21 Sveriges Kommuner och Landsting . Markörbaserad journalgranskning—för att identifiera och mäta skador i vården. LTAB: Sveriges Kommuner och Landsting , 2012 . 22 Sveriges Kommuner och Landsting . Markörer med definitioner Stockholm: Sveriges Kommune og Landsting , 2012 . 23 Levinson DR Adverse events in hospitals. National incidence among Medicare beneficiaries . Department of Health and Human Services, Office of Inspector General , 2010 . 24 Sveriges Kommuner och Landsting . Vårdskador VAD TRODDE VI DÅ—VAD VET VI NU? Stockholm : Sveriges Kommuner och Landsting , 2016 . 25 Efron B , Tibshirani RJ An introduction to the bootstrap London : Chapman & Hall , 1993 . 26 Bjertnæs Ø , Skudal KE ; Iversen HH The Patient-Reported Incident in Hospital Instrument (PRIH-I): assessments of data quality, test-retest reliability and hospital level reliability . BMJ Qual Saf 2013 ;22 , 743 –51 . 10.1136/bmjqs-2012-001756 27 Aiken LH , Sermeus W , Van den Heede K Patient safety, satisfaction, and quality of hospital care: cross sectional surveys of nurses and patients in 12 countries in Europe and the United States . BMJ 2012 ;344 ;e1717 10.1136/bmj.e1717 22434089 28 Unbeck M , Dalen N , Muren O Healthcare processes must be improved to reduce the occurrence of orthopaedic adverse events . Scand J Caring Sci 2010 ;24 :671 –7 . 10.1111/j.1471-6712.2009.00760.x 20409063 29 Joelsson-Alm E , Nyman CR , Svensé et al Micturition problems after bladder distension during hospitalization in Sweden: “I'm not ill, just damaged for the rest of my life” . Nurs Res 2014 ;63 :418 –25 . 10.1097/NNR.0000000000000057 25350541 30 Landrigan CP , Parry GJ , Bones CB Temporal trends in rates of patient harm resulting from medical care . N Engl J Med 2010 ;363 :2124 –34 . 10.1056/NEJMsa1004404 21105794 31 Garrett PR Jr., Sammer C , Nelson A Developing and implementing a standardized process for global trigger tool application across a large health system . J Comm J Qual Patient Saf 2013 ;39 :292 –7 . 10.1016/S1553-7250(13)39041-2 32 Braithwaite J , Matsuyama Y , Mannion R Healthcare Reform, Quality and Safety: perspectives, participants, partnerships, and prospects in 30 countries . Dorchester : Ashgate , 2015 . 33 Southern DA , Pincus HA , Romano PS Enhanced capture of healthcare-related harms and injuries in the 11th revision of the International Classification of Diseases (ICD-11) . Int J Qual Health Care 2015 10.1093/intqhc/mzv099

PMC005xxxxxx/PMC5372043.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01319810.1136/bmjopen-2016-013198Public HealthResearch1506172418451738Prevalence of undiagnosed asymptomatic bacteriuria and associated risk factors during pregnancy: a cross-sectional study at two tertiary centres in Cairo, Egypt Abdel-Aziz Elzayat Mohamed 12Barnett-Vanes Ashton 23Dabour Mohamed Farag Elmorsy 1Cheng Feng 2 1 Faculty of Medicine, Al-Azhar University Egypt, Cairo, Egypt 2 Research Center for Public Health and Center for Global Health & Infectious Diseases, Tsinghua University, Beijing, China 3 Faculty of Medicine, Imperial College London, London, UKCorrespondence to Professor Feng Cheng; fcheng@tsinghua.edu.cnMA-AE is the first author. 2017 21 3 2017 7 3 e01319829 6 2016 13 1 2017 2 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Background The prevalence of asymptomatic bacteriuria (ASB) during pregnancy is poorly understood in Egypt—a country with a high birth rate. Objectives To determine the prevalence of ASB among pregnant women booking at El Hussein and Sayed Galal Hospitals in Al-Azhar University in Egypt; and to observe the relationship between ASB prevalence and risk factors such as socioeconomic level and personal hygiene. Setting Obstetrics and gynaecology clinics of 2 university hospitals in the capital of Egypt. Both hospitals are teaching and referral hospitals receiving referrals from across over the country. They operate specialist antenatal clinics 6 days per week. Participants A cross-sectional study combining the use of questionnaires and laboratory analysis was conducted in 171 pregnant women with no signs or symptoms of urinary tract infection (1 case was excluded). Samples of clean catch midstream urine were collected and cultured using quantitative urine culture and antibiotic sensitivity tests were performed. Results Of 171 pregnant women, 1 case was excluded; 17 cases (10%, 95% CI 5.93% to 15.53%) were positive for ASB. There was a statistically significant relation between the direction of washing genitals and sexual activity per week—and ASB. Escherichia coli was the most commonly isolated bacteria followed by Klebsiella. Nitrofurantoin showed 100% sensitivity, while 88% of the isolates were resistant to cephalexin. Conclusions The prevalence of ASB seen in pregnant women in 2 tertiary hospitals in Egypt was 10%. E. coli and Klebsiella are the common organisms isolated. The direction of washing genitals and sexual activity significantly influences the risk of ASB. Pregnant women should be screened early for ASB during pregnancy; appropriate treatment should be given for positive cases according to antibiotic sensitivity screening. Cephalexin is likely to be of limited use in this management. Asymptomatic BacteriuriaUrinary Tract InfectionUrine CultureEgypt ==== Body Strengths and limitations of this study This study holds implications for clinical providers and policymakers in Egypt regarding screening and prevention of asymptomatic bacteriuria (ASB). This study provides the first insights into the prevalence of ASB among pregnant women in Egypt; and outlines causative organisms, risk factors and appropriate antimicrobial therapy. Negatives of this study include: Positive cases with ASB were not followed-up to determine their adverse outcomes. We were unable to track patients through follow-up urine specimen testing to determine efficacy of antimicrobial treatment. With greater study duration, more patients would be enrolled strengthening the power of the study. Introduction Urinary tract infection (UTI) is one of the most common infections during pregnancy, affecting up to 20% of expectant mothers.1 2 It is defined as microbial contamination of the urine as well as tissue invasion of any part of the urinary tract.3 UTI does not always cause signs and symptoms; if asymptomatic but the urine still contains a significant number of ≥105 colony-forming units (CFU)/mL of bacteria, this condition is termed asymptomatic bacteriuria (ASB).4 ASB during pregnancy is influenced by a range of physiological and anatomical factors, including mechanical compression and changes in the immune and renal systems.5 In addition, there are a range of risk factors that predispose expectant mothers to developing ASB including age, gestational stage, parity, sexual activities and other factors as summarised in the online supplementary appendix table S1.6–8 10.1136/bmjopen-2016-013198.supp1supplementary appendix The prevalence of ASB ranges from 2% to 11% during pregnancy;5 9–11 Escherichia coli is found in 70–90% of isolates that cause ASB.12 13 Other bacteria involved include Klebsiella, Proteus, Pseudomonas and Staphylococcus Saprophyticus.13 14 Most of these pathogens exist naturally in the periurethral area and in the perianal area—and their ascension through the urethral orifice can lead to UTI.12 15 Quantitative urine culture is the gold standard for diagnosis of ASB—the optimal time for screening is the 16th gestational week.16 If ASB is left undiagnosed, there is a risk of developing acute pyelonephritis, seen in up to 40% of pregnant women.6 13 17 Pyelonephritis is associated with preterm labour,12 which is one of the main contributors to neonatal mortality and morbidity worldwide. Early diagnosis and treatment of ASB can drastically reduce the incidence of pyelonephritis,18 and prevent preterm labour by up to 20%.4 However, in developing countries, including Egypt, screening for ASB in pregnancy is not viewed as an essential component of antenatal care; and as a result, there is little understanding of the prevalence of ASB. This is particularly important given Egypt's high birth rate of 23.35 births/1000 population—nearly double that seen in Western Europe or the USA.19 Accordingly, this study—conducted in two teaching and referral hospitals in Egypt—sought to determine the prevalence of ASB during pregnancy, identify the causative organisms and antibiotic sensitivity, and establish the relationship between ASB and common risk factors; with the aim of making recommendations to improve obstetric practice in Egypt and other middle-income countries. Methodology Study design The study was a cross-sectional study combining the use of questionnaires and laboratory analysis of samples obtained from participants (questionnaire survey used is included in the online supplementary appendix figure S1) between January and February 2016 at the obstetrics and gynaecology clinics of El Hussein and Sayed Galal Hospitals of Al-Azhar University in Cairo Governorate which is the capital of Egypt. Both hospitals are teaching and referral hospitals receiving patients from across the country. Study procedures Pregnant females were interviewed using precoded, pretested, interviewer-administered questionnaires to collect and record maternal social demographic characteristics. Laboratory forms were used to record data and results after sample analysis. Selection criteria The full study inclusion criteria are included in the online supplementary appendix figure S2. Briefly, pregnant women aged 18–41 years attending the antenatal clinic sites of this study were invited to enrol. Exclusion criteria included a history of UTI or recent use of antibiotics. Participants were asked to provide blood and urine sample for further testing as described below. We had excess of 10% participant recruitment to meet the expected non-response or loss of questionnaires, giving a minimum sample size of 121 cases. The sample size was increased to 171 cases to maximise the validity of the study and improve the data quality measures.20 Further details are included in the online supplementary appendix figure S3. Data management/analysis Data were entered into a secured personal computer using Microsoft Excel software and analysed using Epi Info V.7.2 computer software. Frequency distribution of selected variables was performed first. Means were compared using the t-test and χ2 test was used to assess the difference between proportions. A p value <0.05 was considered statistically significant. Ethical consideration Agreement for this study was obtained from the hospital's ethical committee, and informed consent was obtained from pregnant women after adequate provision of information regarding the study requirements, purpose and risks. Further details are included in the online supplementary appendix figure S4. Laboratory investigations Blood samples From each participant, 5 mL of blood sample was collected; 2 mL in EDTA-containing tube and tested for complete blood count (CBC) using an automated CBC analyser (Sysmex KX-21N) and the remaining 3 mL of blood was collected in a plain tube, left to coagulate and then centrifuged. The serum was kept in an Eppendorf tube at 0°C for further tests; blood glucose levels were measured using a Hitachi modular analyser (Roche cobas 8000) and rapid HIV test was performed using ELISA (IMMULITE 2000). Urine samples Urine collection and macroscopy Participants were taught how to collect midstream urine in a sterile universal bottle. The sample processing was carried out within 4 hours of specimen collection. Urine samples were examined macroscopically by observing the colour, aspect, deposit and blood clots or debris. Each sample was divided into three portions: microscopic analysis, culture and chemical analysis to avoid contamination of the samples. About 5 mL of each well-mixed urine sample was centrifuged at 3000 rpm for 10 min. A drop of properly mixed sediment was placed on a glass slide and examined under light microscope to detect pus cells (indicating ingested bacteria), Trichomonas vaginalis, Schistosoma ova, white cell count, red blood cells, casts, crystals and yeast-like cells. The presence of 10 pus cells/mm3 or more was regarded as pyuria.21 Drops of the urine were applied to microscope slides, allowed to air dry, stained with Gram stain, and examined microscopically (primary Gram staining). Quality control was performed.22 The supernatant of the centrifuged urine was tested using Combi screen 10 urinalysis strips, with the existence of nitrite and leucocyte esterase in the urine being suggestive of infection.23 24 Culturing of bacteria from urine samples A sterile disposable calibrated loop delivering 0.01 mL of urine was used for streaking cystine lactose electrolyte deficient (CLED) agar plates following standard procedure.25 Specimens were also streaked on the blood agar plate and MacConkey agar plate and then incubated at 37°C for 24 hours. After 24 hours, the CLED agar plates were observed for confluent growth, which shows significant bacteriuria, and if not confluent, the colonies were counted then multiplied by the size of the inoculums of the calibrated loop, which is 1/100. Significant ASB was considered when the bacterial value was ≥105. For cultures with no or insignificant bacterial growths, incubation was continued for a further 24 hours. After a description of colonies, Gram staining was performed from pure colonies. Biochemical tests were performed from the pure colonies for identification. The antibiogram determination was performed using pure colonies from the CLED agar plates. Sensitivity tests Organisms showing significant bacteriuria were inoculated into peptone water before plating on Mueller-Hinton agar. Commercially organised antimicrobial discs of known minimum inhibitory concentrations (MICs) were placed over the surface of the sensitivity agar and pressed down with sterile forceps to make enough contact with the agar. The plates were incubated at 37°C for 24 hours and the zones of growth inhibition were estimated.26 The antimicrobial sensitivity discs used were: amoxycillin-clavulanate, imipenem, ceftazidime, ceftriaxone, cefotaxime, cefuroxime, cefaclor, norfloxacin, ciprofloxacin, nitrofurantoin, amikacin and sulfamethoxazole-trimethoprim. Results A total of 171 pregnant women were examined for ASB; 1 case was excluded (microscopic urine analysis reported pus cells more than 10 cells/high-power field (HPF)). Hence, 170 pregnant women were included in this study. Table 1 describes the demographic characteristics of the participants and their ASB results. The mean age of patients was 28.52±5.36 years ranging from 18 to 41 years. Among the participants, 75% were in their third trimester, 70% were multiparous; regarding their educational status—47% had completed high school; 61% were in a ‘low’ socioeconomic level based on (Kuppuswamy's Socio-economic Status (SES) Scale for 2016) online tool.27 Table 1 Demographic characteristics of pregnant women included in this study Characteristics Frequency Positive culture (N) Percentage p Value Age (years)  <20 2 0 0  20–30 114 12 11 0.29  >30 54 5 9 Gestational age  First trimester 14 0 0  Second trimester 27 5 19 0.86  Third trimester 129 12 9 Parity  Grand multipara 12 0 0  Multiparous 119 13 11  Primigravida 39 4 10 0.11 Educational level  College 19 1 6  Elementary 15 1 7  Graduate 32 3 9 0.69  High school 80 10 13  Junior school 24 2 8 Socioeconomic level  High 14 0 0  Intermediate 52 3 6 0.08  Low 104 14 13 Direction of wash genitals  Back to front 102 15 15  Front to back 68 2 3 0.03 Number of bathing and changing underwear (week)  1–3 times 119 12 10  >3 times 51 5 10 0.69 Number of sexual intercourse (week)  1–2 times 92 6 7  >2 times 78 11 14 0.01 Of the 170 pregnant women tested, 17 cases were positive for significant bacteriuria (CFU≥105/mL), giving an overall prevalence of 10% (95% CI 5.93% to 15.53%; figure 1A). E. coli was the most predominant organism followed by Klebsiella; no other isolated organisms showed significant growth (figure 1B). On microscopic examination of positive cases, 10 (59%) had pus cells (<10)/HPF, 5 cases (29%) had red blood cells, 1 case (6%) had epithelial cells and 1 case (6%) had crystals (figure 1C). Figure 1 Urine culture and microscopic urinalysis. Proportion (%) of pregnant women with ASB in the study (A); proportions of causative uropathogens isolated from positive cases (B); microscopic analysis of bacterially positive urine cases (C). ASB, asymptomatic bacteriuria; E. coli, Escherichia coli; RBC, red blood cell. We then examined the sensitivity of these to antibiotics. Overall, nitrofurantoin, imipenem and amikacin demonstrated 100% sensitivity (figure 2). A range of other antibiotics showed good sensitivity including norfloxacin and ceftazidime; however, 88% of the urinary isolates were resistant to cephalexin (figure 2). Investigating whether there were isolate-specific differences in antimicrobial susceptibility, we found that only E. coli demonstrated resistance across the range of antibiotics tested (table 2). However, of note, cephalexin showed poor efficacy across both bacteria. Table 2 Susceptibility of isolated uropathogens to different antibiotics using discs' diffusion method Organism sensitivity N (%) Antibiotic E. coli sensitive N (%) Klebsiella sensitive N (%) AUG 3 (25%) 5 (100%) CAZ 9 (75%) 5 (100%) CRO 9 (75%) 5 (100%) CTX 5 (41.7%) 5 (100%) CXM 5 (41.7%) 3 (60%) F 12 (100%) 5 (100%) NOR 9 (75%) 5 (100%) CIP 9 (75%) 5 (100%) AK 12 (100%) 5 (100%) SXT 9 (75%) 5 (100%) IPM 12 (100%) 5 (100%) CL 2 (16.7%) 0 (0%) Positive cases=17; E. coli=12 cases; Klebsiella=5 cases. AUG, amoxycillin-clavulanate; AK, amikacin; CAZ, ceftazidime; CIP, ciprofloxacin; CL, cefaclor; CRO, ceftriaxone; CTX, cefotaxime; CXM, cefuroxime; E. coli, Escherichia coli; F, nitrofurantoin; IPM, imipenem; NOR, norfloxacin; SXT, sulfamethoxazole-trimethoprim. Figure 2 The proportion (%) of sensitivity/resistance susceptibility of isolated bacteria to different antibiotics using discs' diffusion method; commercially purchased antimicrobial discs of known MICs were placed aseptically over the surface of the sensitivity agar. The plates were incubated for 24 hours, and the zones of growth inhibition were estimated. MIC, minimum inhibitory concentration. Regarding the relationship between ASB and the range of demographic and personal hygiene risk factors examined in this study, ASB was predominant in participants with higher sexual activity: 78 (65%) participants reported their sexual activity as greater than twice per week, and 11 of the 17 ASB cases were seen in this cohort (p=0.01). ASB was also significantly higher among participants who reported washing their genitals from back to front after defaecation (88%, p=0.03; table 1). There were no statistically significant differences between ASB and age, gestational age, parity, educational level, socioeconomic level or haemoglobin concentration (table 1). No HIV+ cases were identified in this study. Discussion In the present study, the prevalence of ASB during pregnancy was found to be 10% (95% CI 5.93% to 15.53%). The prevalence in this study is comparable to that reported in Nigeria,14 but lower than studies from Ethiopia.28–31 These discrepancies between and within countries may be due to differences in the study participants' socioeconomic levels, and cultural7 and religious behaviours related to personal hygiene and sexual contact. ASB had a significant relationship with sexual activity as seen in other studies.32 33 Sexual intercourse may increase the probability of transfer of uropathogens into the urethra; and as reported elsewhere, ASB had a significant relationship with the direction of washing genitals after urination or defaecation.34 Washing of genitals from back to front is more likely to lead to the spread of anal or vaginal flora into the urethra. Education on the direction of washing and advice to micturate shortly after sexual activity can reduce the prevalence of UTI.35 However, there was no statistically significant association between parity, maternal age, socioeconomic class, educational level or gestational age and ASB (p>0.05). This is most probably because of the small sample size. Multiparous women had the highest frequency of ASB, similar to findings in another study.36 This is believed to be because high parity leads to the descent of pelvic organs, and a widening of the urethral orifice, which influences the ascent of microbes.37–39 ASB appears predominant in women aged between 20 and 30 years, which is similar to findings from other studies.40 41 The vulnerability of these age groups could be explained by early and intensive sexual intercourse which may cause minor urethral trauma and transfer bacteria from the perineum into the bladder.42 Accurate diagnosis of causative organisms is critical to the appropriate selection and completion of an antibiotic course. In this study, E. coli and Klebsiella were causative, with E. coli dominant in most cases, as reported previously.40 41 43 Choice of antibiotics must also consider potential side effects; while all isolates were sensitive to nitrofurantoin, there have been concerns over its potential impacts on the fetus.12 Of concern for clinicians, 88% of E. coli and Klebsiella isolates in this study were resistant to cephalexin. The antimicrobial sensitivity and resistance patterns vary between communities and hospitals. This is likely because of the emergence of resistant strains, caused in part by inappropriate antibiotic prescription. Today, antimicrobial resistance is recognised as a looming international health crisis;44 and as such is now a global health priority. Certain regions are already experiencing high levels of bacterial resistance rates to common frontline antibiotics such as amoxicillin or ampicillin.13 Accordingly, a range of guidelines has been established for the screening and diagnosis of ASB, including from the UK's National Institute for Health and Care Excellence (NICE) and the Center for Disease Control and Prevention (CDC) in the USA.45 46 Early and regular check-ups by medical providers are vital in assessing the physical status and early recognition of complications during pregnancy. Yet, the provision of regular antenatal care is still low in Egypt, especially in rural areas. Antenatal care coverage for at least one visit is 74% and antenatal care coverage for at least four visits is 66%; 69% of pregnant women are examined by routine urine analysis only.47 48 These findings, combined with the prevalence of ASB found in this study betray an antenatal care system in need of improvement. This is all the more urgent given the high fertility rate in Egypt, on average 3.5 children per woman compared with 1.83 in the UK.49 The implications of this study for clinical providers and policymakers in Egypt are threefold. First, physicians must be educated on the importance of screening and prevention of ASB, and informed of the latest antimicrobial resistance data in their country or region. Second, pregnant women must be educated on personal hygiene and ASB to ensure they recognise the implications for their health and their children; and third, policymakers must recognise the cost-benefits of diagnosis of ASB early before it progresses to other more serious diseases such as pyelonephritis and preterm labour. Recommendations Screening for ASB must become an essential part of antenatal care. We recommend periodic screening at each trimester especially at 9–17 gestational weeks by quantitative urine culture. Selection of the appropriate antibiotic based on antibiotic sensitivity testing of uropathogens (control resistant strains in the future). It is important to remember that therapy must be safe for mother and fetus; the practice should be guided by bacterial sensitivity/resistance profiles. Nitrofurantoin is recommended to be used for patients in the first and second trimesters, as it is cheap, showed 100% sensitivity and is reported safe and efficacious in the treatment of ASB during pregnancy; however, concerns exist for its use in the third trimester. This antibiotic could replace cephalosporins (if isolates show sensitivity to it).50–52 Conclusion The prevalence of ASB seen in pregnant women in two tertiary hospitals in Egypt was 10%. E. coli was the dominant organism isolated. The direction of washing genitals and sexual activity significantly influences the risk of ASB. Quantitative urine culture is the ideal test for detection of ASB. Nitrofurantoin is the most efficient antimicrobial for the treatment of ASB. Early detection and treatment are essential to safeguard the health of mother and fetus. Further larger studies could provide cost-benefit data9 53 54 necessary to inform a national screening programme. Contributors: All authors have participated fully in the conception, writing and critical review of this manuscript. All have seen and agreed to the submission of the final manuscript. MA-AE is the first author, principle investigator, and was involved in design of the work, data collection, sample collection, data analysis and interpretation, writing and drafting the article. AB-V was involved in writing and critical review. MFED is the laboratory physician, and was involved in sample collection and processing, writing, and critical review. FC is the corresponding author, and was involved in writing and critical review. Funding: The project was financed by Chinese Government Scholarship ‘Youth of Excellence Scheme Of China (Yes China)’. Competing interests: None declared. Patient consent: Obtained. Ethics approval: El Hussein and Sayed Galal Hospital's Ethical Committee. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Technical appendices of survey questions and additional data are included in the online supplementary material. ==== Refs References 1 Parveen K , Momen A , Begum AA Prevalence of urinary tract infection during pregnancy . J Dhaka Natl Med Coll Hosp 2012 ;17 :8 –12 . 2 Ebidor UL , Tolulope A , Deborah O Urinary tract infection amongst pregnant women in Amassoma, Southern Nigeria . Afr J Microbiol Res 2015 ;9 :355 –9 . 10.5897/AJMR2014.7323 3 Najar MS , Saldanha CL , Banday KA Approach to urinary tract infections . Indian J Nephrol 2009 ;19 :129 –39 . 10.4103/0971-4065.59333 20535247 4 Kazemier BM , Schneeberger C , De Miranda E Costs and effects of screening and treating low risk women with a singleton pregnancy for asymptomatic bacteriuria, the ASB study . BMC Pregnancy Childbirth 2012 ;12 :52 10.1186/1471-2393-12-52 22892110 5 Ghafari M , Baigi V , Cheraghi Z The prevalence of asymptomatic bacteriuria in Iranian pregnant women: a systematic review and meta-analysis . PLoS ONE 2016 ;11 :e0158031 10.1371/journal.pone.0158031 27336476 6 Jain V , Das V , Agarwal A Asymptomatic bacteriuria & obstetric outcome following treatment in early versus late pregnancy in north Indian women . Indian J Med Res 2013 ;137 :753 –8 . http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3724257&tool=pmcentrez&rendertype=abstract (accessed 25 Nov 2015 ).23703344 7 Muharram SH , Ghazali SNB , Yaakub HR A preliminary assessment of asymptomatic bacteriuria of pregnancy in Brunei Darussalam . Malaysian J Med Sci 2014 ;21 :34 –9 . 8 Smaill F , Vazquez JC Antibiotics for asymptomatic bacteriuria in pregnancy . Cochrane Database Syst Rev 2007 ;(2) :CD000490 10.1002/14651858.CD000490.pub2 17443502 9 Hazhir S Asymptomatic bacteriuria in pregnant women . Urol J 2007 ;4 :24 –7 .17514607 10 Jassim ZM Asymptomatic bacteriuria and pyuria in pregnant women in Hilla city: causative agents and antibiotic sensitivity . J Babylon Univ 2013 ;21 :2755 . 11 Sau-yee F , Ny F , Hy H The prevalence of asymptomatic bacteriuria in pregnant Hong Kong women . Hong Kong J Gynaecol Obs Midwifery 2012 ;16 :39 –45 . 12 Schnarr J , Smaill F Asymptomatic bacteriuria and symptomatic urinary tract infections in pregnancy . Eur J Clin Invest 2008 ;38 (Suppl 2 ): 50 –7 . 10.1111/j.1365-2362.2008.02009.x 18826482 13 Le J , Briggs GG , McKeown A Urinary tract infections during pregnancy . Urin Tract Infect J 2004 ;3 :1 –28 . 14 Aminu KY , Aliyu UU Asymptomatic bacteriuria in pregnant women in the antenatal booking clinic at Aminu Kano Teaching Hospital, Kano, Nigeria . Open J Obstet Gynecol 2015 ;5 :286 –97 . 10.4236/ojog.2015.55042 15 Khazal N , Hindi K , Hasson SO Bacteriological study of urinary tract infections with antibiotics susceptibility to bacterial isolates among honeymoon women in Al Qassim Hospital, Babylon Province, Iraq . Br Biotechnol J 2013 ;3 :332 –40 . 10.9734/BBJ/2013/3573 16 Stenqvist K , Dahlén-Nilsson I , Lidin-Janson G Bacteriuria in pregnancy. Frequency and risk of acquisition . Am J Epidemiol 1989 ;129 :372 –9 . 10.1093/oxfordjournals.aje.a115140 2912046 17 Cheung KL , Lafayette RA Renal physiology of pregnancy . Adv Chronic Kidney Dis 2013 ;20 :209 –14 . 10.1053/j.ackd.2013.01.012 23928384 18 Matuszkiewicz-rowińska J , Małyszko J , Wieliczko M Urinary tract infections in pregnancy: old and new unresolved diagnostic and therapeutic problems . Arch Med Sci 2015 ;11 :67 –77 . 10.5114/aoms.2013.39202 25861291 19 Breu F , Guggenbichler S , Wollmann J World Health Statistics . 2013 Geneva: WHO Press . 20 Requirements SS, Difference M . 19: Sample size, precision, and power sample size requirements for estimating a mean or mean difference sample size requirements for testing a mean or mean difference . Power 2000 ;1 :3 –6 . 21 Alnaaimi AS , Sabri M Validity of pyuria and bacteriuria (detected by Gram-stain) in predicting positive urine culture in asymptomatic female children Rajah JT Al-Ma'amoory* . 2007 ;20 :349 –53 . 22 Ismail M , Assurance Q Quality assurance in . Indian J Med Microbiol 2011 ;22 :85 –91 . 23 Simerville JA , Maxted WC , Pahira JJ Urinalysis: a comprehensive review . Am Fam Physician 2005 ;71 :1153 –62 .15791892 24 Schmiemann G , Kniehl E , Gebhardt K The diagnosis of urinary tract infection: a systematic review . Dtsch Ärzteblatt Int 2010 ;107: 361 –7 . 10.3238/arztebl.2010.0361 25 Services M UK standards for microbiology investigations . Bacteriol Health Engl 2015 ;B 55 :1 –21 . 26 Shaifali I , Gupta U , Mahmood SE Antibiotic susceptibility patterns of urinary pathogens in female outpatients . N Am J Med Sci 2012 ;4 :163 –9 . 10.4103/1947-2714.94940 22536558 27 Sharma R Online interactive calculator for real-time update of the Kuppuswamy's socioeconomic status scale. http://www.scaleupdate.weebly.com (accessed 5 Mar 2016 ). 28 Tadesse E , Teshome M , Merid Y Asymptomatic urinary tract infection among pregnant women attending the antenatal clinic of Hawassa Referral Hospital, Southern Ethiopia . BMC Res Notes 2014 ;7 :155 10.1186/1756-0500-7-155 24636218 29 Oli AN , Okafor CI , Ibezim EC The prevalence and bacteriology of asymptomatic bacteriuria among antenatal patients in Nnamdi Azikiwe University Teaching Hospital Nnewi; South Eastern Nigeria . Niger J Clin Pract 2010 ;13 :409 –12 .21220855 30 Rajaratnam A , Baby NM , Kuruvilla TS Diagnosis of asymptomatic bacteriuria and associated risk factors among pregnant women in Mangalore, Karnataka state . J Clin Diagn Res 2014 ;8 :OC23 –5 .25386490 31 Kehinde AO , Adedapo KS , Aimaikhu CO Significant bacteriuria among asymptomatic antenatal clinic attendees in Ibadan, Nigeria . Trop Med Health 2011 ;39 :73 –6 . 10.2149/tmh.2011-02 32 Amala SE , Nwokah EG Prevalence of asymptomatic bacteriuria among pregnant women attending antenatal in Port Harcourt Township, Nigeria and antibiogram of isolated bacteria . Am J Biomed Sci 2015 ;7 :125 –33 . 10.5099/aj150200125 33 Labi A-K , Yawson AE , Ganyaglo GY Prevalence and associated risk factors of asymptomatic bacteriuria in ante-natal clients in a large teaching hospital in Ghana . Ghana Med J 2015 ;49 :154 –8 .26693190 34 Obiora CC , Dim CC , Ezegwui HU Asymptomatic bacteriuria among pregnant women with sickle cell trait in Enugu, South Eastern Nigeria . Niger J Clin Pract 2014 ;17 :95 –9 . 10.4103/1119-3077.122856 24326816 35 Moustafa MF , Makhlouf EM Association between the hygiene practices for genital organs and sexual activity on urinary tract infection in pregnant women at women's Health Center, at Assiut University Hospital . J Am Sci 2012 ;8 :515 –22 . 36 Nisha AK , Etana AE , Tesso H Prevalence of asymptomatic bacteriuria during pregnancy in Adama city, Ethiopia . Int J Microbiol Immunol Res 2015 ;3 :58 –63 . 37 Fong SY , Tung CW , Yu YNY The Prevalence of Asymptomatic Bacteriuria in Pregnant Hong Kong Women Hong Kong J Gynaecol Obstet Midwifery 2013 ;13 :40 –4 . 38 Shruthi A Asymptomatic bacteriuria in pregnancy: bacteriological profile and antibiotic sensitivity pattern in a tertiary care hospital, Bengaluru . Int J Health Sci Res 2015 ;5 :157 –62 . 39 Ojide CK , Wagbatsoma VA , Kalu EI Asymptomatic bacteriuria among antenatal care women in a tertiary hospital in Benin, Nigeria . Niger J Exp Clin Biosci 2014 ;2 :79–85 . 40 Sujatha R , Nawani M Prevalence of asymptomatic bacteriuria and its antibacterial susceptibility pattern among pregnant women attending the antenatal clinic at Kanpur, India . J Clin Diagn Res 2014 ;8 :2 –4 . 41 Khan S , Singh P , Siddiqui Z Pregnancy-associated asymptomatic bacteriuria and drug resistance . J Taibah Univ Med Sci 2015 ;10 :340 –5 . 42 Jalali M , Shamsi M , Roozbehani N Prevalence of urinary tract infection and some factors affected in pregnant women in Iran Karaj City 2013 . Middle East J Sci Res 2014 ;20 :781 –5 . 43 Olamijulo JA , Adewale CO , Olaleye O Asymptomatic bacteriuria among antenatal women in Lagos . J Obstet Gynaecol (Lahore) 2016 ;3615 :1 –4 . 44 Smaill F Asymptomatic bacteriuria in pregnancy . Best Pract Res Clin Obstet Gynaecol 2007 ;21 :439 –50 . 10.1016/j.bpobgyn.2007.01.004 17347050 45 Catherine M Screening for asymptomatic bacteriuria in pregnancy. External review against programme appraisal criteria for the UK National Screening Committee (UK NSC) . UK Natl Screen Com 2011 ;2 :1 –15 . 46 Horan TC , Andrus M , Dudeck MA CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting . Am J Infect Control 2008 ;36 :309 –32 . 10.1016/j.ajic.2008.03.002 18538699 47 El-Zanaty F , Way A Egypt Demographic and Health Survey—the DHS Program ICF International Rockville, Maryland, U.S.A . Minist Health Popul 2005 ;2 :21 –30 . 48 Care P Children in Egypt Chapter 2 Births and Perinatal Care. Published Online First: 2015. http://www.unicef.org/egypt/Children_in_Egypt-_data_digest_2014.pdf 49 Office for National Statistics . Statistical Bulletin Births in England and Wales, 2013 2014 ;2012 :1 –11 . 50 Grabe M , Bartoletti R , Johansen TEB Guidelines urological Infections. Published Online First: 2015. http://uroweb.org/wp-content/uploads/19-Urological-infections_LR2.pdf 51 Mittal P , Wing DA Urinary tract infections in pregnancy . Clin Perinatol 2005 ;32 :749 –64 . 10.1016/j.clp.2005.05.006 16085031 52 Simoes JA , Aroutcheva AA , Heimler I Antibiotic resistance patterns of group B streptococcal clinical isolates . Infect Dis Obstet Gynecol 2004 ;12 :1 –8 . 10.1080/10647440410001722269 15460188 53 Evans DB , Edejer TT , Adam T Methods to assess the costs and health effects of interventions for improving health in developing countries . BMJ 2005 ;331 :1137 –40 . 10.1136/bmj.331.7525.1137 16282411 54 Selimuzzaman A , Ullah M , Haque M Asymptomatic bacteriuria during pregnancy: causative agents and their sensitivity in Rajshahi City . J Teach Assoc RMC 2006 ;19 :66 –9 .

PMC005xxxxxx/PMC5372044.txt

==== Front BMJ Open GastroenterolBMJ Open GastroenterolbmjgastrobmjgastroBMJ Open Gastroenterology2054-4774BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjgast-2016-00012910.1136/bmjgast-2016-0001291506CancerIs there a standard for surgical therapy of hepatocellular carcinoma in healthy and cirrhotic liver? A comparison of eight guidelines Manzini Giulia 1Henne-Bruns Doris 1Porzsolt Franz 2Kremer Michael 11 Department of General and Visceral Surgery, University of Ulm, Ulm, Germany2 Health Care Research Group at the Hospital of General and Visceral Surgery, University Hospital of Ulm, Ulm, GermanyCorrespondence to Professor Doris Henne-Bruns; doris.henne-bruns@uniklinik-ulm.de2017 24 3 2017 4 1 e00012913 12 2016 8 2 2017 13 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017BMJ Open GastroenterologyThis is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Background and aims Liver resection (LR) and transplantation are the most reliable treatments for hepatocellular carcinoma (HCC). Aim was to compare different guidelines regarding indication for resection and transplantation because of HCC with and without underlying cirrhosis. Methods We compared the following guidelines published after 1 January 2010: American (American Association for the Study of Liver Diseases (AASLD)), Spanish (Sociedad Espanola de Oncologia Medica (SEOM)), European (European Association for the study of liver-European Organization for Research and Treatment of Cancer (EASL-EORTC) and European Society for Medical Oncology-European Society of Digestive Oncology (ESMO-ESDO)), Asian (Asian Pacific Association for the Study of Liver (APASL)), Japanese (Japan Society of Hepatology (JSH)), Italian (Associazione Italiana Oncologia Medica (AIOM)) and German (S3) guidelines. Results All guidelines recommend resection as therapy of choice in healthy liver. Guidelines based on the Barcelona Clinic Liver Cancer staging system recommend resection for single HCC<2 cm and Child-Pugh A cirrhosis and for HCC≤5 cm with normal bilirubin and portal pressure, whereas transplantation is recommended for multiple tumours between Milan criteria and for single tumours ≤5 cm and advanced liver dysfunction. Patients with HCC and Child-Pugh C cirrhosis are not candidates for transplantation. JSH guidelines recommend LR for patients with Child-Pugh A/B with HCC without tumour size restriction; APASL guidelines in general exclude patients with Child-Pugh A from transplantation. In patients with Child-Pugh B, transplantation is the second-line therapy, if resection is not possible for patients within Milan criteria. German and Italian guidelines recommend transplantation for all patients within Milan criteria. Conclusions Whereas resection is the standard therapy of HCC in healthy liver, a standard regarding the indication for LR and transplantation for HCC in cirrhotic liver does not exist, although nearly all guidelines claim to be evidence based. Surprisingly, despite European guidelines, Germany and Italy use their own national guidelines which partially differ from the European. Possible solutions of the problems are discussed. CIRRHOSISLIVER TRANSPLANTATIONHEPATOCELLULAR CARCINOMASURGICAL RESECTION ==== Body Article summary ▸ Surgery is the only curative option for hepatocellular carcinoma (HCC). Several guidelines exist that provide recommendations regarding indication for resection and transplantation. Although nearly all guidelines claim to be evidence-based, we only find consensus in regard to indication for liver resection and transplantation for HCC in healthy liver, but a standard for the treatment of HCC with underlying liver cirrhosis does not exist. Traditional guidelines are based on efficacy but not yet effectiveness data. Only when outcomes, conditions, patient characteristics and interventions are described transparently, it will be possible to discuss possible reasons for different guidelines in different countries. Traditional guidelines are based on efficacy but not yet effectiveness of data. Progress in the development of guidelines will be made when the reasons that explain the differences in the existing guidelines can be identified. Promising prognostic factors considering tumor biology as well as liver function tests should be included in future guidelines. Introduction Hepatocellular carcinoma (HCC) is the 5th most common cancer and the 3rd leading cause of cancer-related deaths worldwide.1 2 Surgical resection and transplantation are the most reliable treatments for local control and are up to date the only potential curative treatment. Surgical resection is the treatment of choice in patients without cirrhosis, who account for 5% of the cases in Western countries and for about 40% in Asia as these patients tolerate major resections with low morbidity.3 When HCC is diagnosed in cirrhotic liver, the indication for liver resection (LR) should be carefully given.4 5 The 5-year survival after resection can exceed 50%.3 Early diagnosis and accurate evaluation of the preoperative liver function allow the identification of those patients, in which a resection could lead to postoperative liver failure with higher probability.3 Next to the Child-Pugh classification, also the assessment of the presence of portal hypertension plays a central role in the identification of candidates for surgical resection. Studies have shown that normal bilirubin concentration and a hepatic vein pressure gradient <10 mm Hg are the best predictors of excellent outcomes after surgery, with almost no risk for postoperative liver failure.4 6 These selected patients may achieve a 5-year survival of more than 70%,3 7 whereas <50% 5-year survival is to be expected in patients with portal hypertension. The 5-year survival of patients with elevated bilirubin value and portal hypertension and/or multifocal disease is <30%, regardless of their Child-Pugh stage.4 8 Different guidelines exist for the same problem. By presenting similarities and differences between guidelines of different countries regarding the indication for LR, liver transplantation (LT) as well as the recommendations regarding expansion of the transplant criteria, bridging and downstaging therapies and living-donor LT (LDLT), aim of this work is to evaluate, interpret and present solutions for the accounted problems. Material and methods Systematic literature search To generate a standardised basis for the systematic literature research, uniform comparison criteria were established within the guideline group. Criteria for selection were: guidelines should be in English, German, Italian or Spanish and published after 1st January 2010 to ensure that outdated guidelines are excluded. The guidelines should be generated by expert groups of internationally recognised organisations and based on evidence-based publications. If evidence-based guidelines were not provided, we included consensus-based guidelines. Tumour classification can be based on the Barcelona Clinic Liver Cancer classification (BCLC) staging system,9 10 which links staging of HCC in cirrhosis with treatment modalities or not. We performed a systematic research with Ovid. We screened the database Medline, Cochrane and PubMed. Table 1 illustrates the results in English language. Our key words for the search were: “guidelines hepatocellular carcinoma” and “guidelines HCC”. On websites of medical institution we found additional results in their respective native language (table 2). We conducted our online search on 21 April 2016. Table 1 Results of the systematic literature search conducted on 21 April 2016 Database Key words Results (n) PubMed Guidelines hepatocellular carcinoma 1014 Guidelines HCC 566 Medline Guidelines hepatocellular carcinoma 123 Guidelines HCC 82 Cochrane Guidelines hepatocellular carcinoma 1 Guidelines HCC 19 HCC, hepatocellular carcinoma. Table 2 Websites of medical institution Institution Key words Results Sociedad Espanola de Oncologia Medica (http://www.seom.org) Hepatocarcinoma 1 Arbeitsgemeinschaft der Wissenschaftlichen Medizinischen Fachgesellschaften (http://www.AWMF.org) Hepatozelluläres Karzinom 2 Associazione italiana di oncologia medica (http://www.aiom.it) Epatocarcinoma 3 Japan Society of Hepatology (JSH) (http://www.jsh.or.jp/English/) Hepatocellular carcinoma 1 Selection of the guidelines Two authors (GM and MK) screened the results manually and independently, by looking at the title and the abstracts. If the inclusion criteria were met, the manuscript was analysed. The search on the database in English language retrieved following guidelines: American (American Association for the Study of Liver Diseases (AASLD)),3 Asian (Asian Pacific Association for the Study of Liver (APASL)),11 Hong Kong,12 Japanese (Japan Society of Hepatology (JSH)),13 14 European Association for the study of liver-European Organization for Research and Treatment of Cancer (EASL-EORTC)15 and ESMO-ESDO16 and Spanish (Sociedad Espanola de Oncologia Medica (SEOM)).17 The Spanish guidelines and the evidence-based Japanese guidelines are a synopsis. The entire version of the Spanish guidelines in original language is freely accessible only for members of the Spanish Society of Medical Oncology (http://www.seom.org). As we could not have access to the entire version, we included the synopsis in our analysis. The found article about the updated version of the evidence-based Japanese guidelines suggested in the introduction a link to the homepage of the Japanese Society of Hepatology, where the entire new guidelines version was freely accessible and in English (http://www.jsh.or.jp/English/). We excluded the consensus-based Japanese guidelines, as evidence-based guidelines were also found, as well as the Hong Kong consensus recommendations because of a rather small population and because of the inclusion of the APASL guidelines. The search with the medical institutions in table 2 found following regional guidelines: Italian18 and German19 in the original language and the full version of the evidence-based Japanese guidelines. These were all included. We finally included in our analysis a total of five international and three regional guidelines, which are listed in table 3. Table 3 List of the eight included guidelines Country Organisation Title Year America American Association for the Study of Liver Diseases (AASLD) Management of hepatocellular carcinoma: an update 2010 Asia Asian Pacific Association for the Study of Liver (APASL) APASL consensus recommendation on hepatocellular carcinoma 2010 Japan Japan Society of Hepatology (JSH) Evidence-based clinical practice guidelines for hepatocellular carcinoma: the Japan Society of Hepatology 2013 update (3rd JSH-HCC Guidelines) 2013 Europe European Association for the study of liver (EASL) European Society for Medical Oncology (ESMO) EASL-EORTC clinical practice guidelines: management of hepatocellular carcinoma Hepatocellular carcinoma: ESMO-ESDO clinical practice guidelines for diagnosis, treatment and follow-up 2012 2012 Spain Sociedad Espanola de Oncologia Medica (SEOM) Clinical guidelines SEOM: hepatocellular carcinoma 2015 Germany Deutsche Gesellschaft für Verdauungs- und Stoffwechselkrankheiten e. V.(DGVS) S3-Leitlinien “Hepatozellulärer Karzinom”—Diagnostik und Therapie des hepatozellulären Karzinoms 2013 Italy Associazione Italiana di Oncologia Medica (AIOM) Linee guida epatocarcinoma 2015 EORTC, European Organization for Research and Treatment of Cancer; ESDO, European Society of Digestive Oncology; HCC, hepatocellular carcinoma. Comparison of the guidelines We compared the guidelines included in table 3 regarding the indications for LR and LT in patients with and without cirrhosis. Additionally, we analysed the recommendations regarding the expansion of the transplantation criteria beyond Milan, regarding bridging therapies of patients still on waiting list for transplantation as well as downstaging of patients initially beyond Milan criteria and finally regarding LDLT. In order to assess the treatment recommendations of different guidelines in a comparable way, we generated uniform comparison criteria. To present this comparison clearly, we chose a colour coding: in the event that no or no clear treatment recommendation was given, a black circle was assigned. Once a treatment option was recommended as first-line therapy, a green circle was assigned. A yellow circle was assigned when therapy can be carried out or when is recommended as second-line therapy. Rejection of a therapy was symbolised by a red-circled white dot. By this, we could summarise the mentioned treatment options and their associated recommendations very clearly in a tabular form. Results Surgical treatment of HCC without cirrhosis In all analysed guidelines, surgical resection is the treatment of choice for resectable HCC in absence of cirrhosis. In several guidelines, a more precise indication is given. According to the Spanish guidelines,17 LR is preferred in patients with early stage HCC who have no cirrhosis and an anticipated liver remnant of at least 20%.4 According to the ESMO-ESDO guidelines,16 resection is the recommended treatment in patients without advanced fibrosis, as long as an R0 resection can be carried out without causing postoperative liver failure due to a too small liver remnant.20 The German S3-guidelines19 define the criteria of non-resectability as follows: non-resectable extrahepatic tumour manifestation, patients’ general comorbidities, tumour infiltration in all three liver veins and a too small liver remnant.3 21 Also, re-resection in case of a recurrence appears to be feasible, as 5-year survival rates of up to 80% can be achieved as long as no extrahepatic tumour manifestation is found.22 Adequate postoperative liver function and portal hypertension need to be taken into account on functional resectability. In healthy liver, a minimum of 25–30% of liver parenchyma is needed to prevent risk of postoperative liver failure.19 Only the Italian18 and German S3-guidelines19 express a clear recommendation regarding LT for HCC in absence of cirrhosis. According to the Italian guidelines, a LT can be performed if surgical resection is not possible and no vascular and nodal invasion detectable. Moreover, LT should be offered to those patients who develop a local recurrence not suitable for resection at least 1 year after the primary resection. According to the German S3-guidelines, LT without cirrhosis should only be considered in the specific case of local recurrence of fibrolamellar carcinoma in absence of lymphonodal metastases.23 Figure 1 illustrates the comparison of the guidelines using the assigned colour codes as described in the methods section. Figure 1 Comparison of the guidelines regarding liver resection (LR) and liver transplantation (LT) for HCC in absence of cirrhosis. Green circle: first-line therapy. Yellow circle: therapy can be carried out or therapy recommended as second line. Black circle: no or no clear treatment recommendations. HCC, hepatocellular carcinoma. Surgical treatment of HCC with cirrhosis The American,3 the SEOM15 and the European guidelines EASL-EORTC and ESMO-ESDO13 14 base the therapy of HCC on the BCLC. For this reason, indication for resection and transplantation are mostly similar. The Asian,11 Italian18 and German19 guidelines base the treatment on the Child-Pugh Score. The Japanese evidence-based guidelines base the treatment algorithm on three major factors: liver function (Child-Pugh score), number and size of tumours. Child-Pugh class A According to the AASLD guidelines,10 resection is the first-line therapy for patients who have a single lesion irrespective of the size and a still preserved liver function, normal bilirubin and hepatic vein pressure gradient <10 mm Hg. An increased bilirubin, a significant portal hypertension or minor fluid retention requiring diuretic therapy exclude resection also in case of Child A, and LT is indicated if the patient is within the Milan criteria (BCLC-A). In case of multinodular HCC within Milan criteria, LT is the first-line therapy. Resection is not indicated. For multinodular tumour outside the transplant criteria, neither resection nor transplantation is indicated. The SEOM guidelines17 recommend LR for patients with solitary or limited multifocal HCC (stage BCLC-0 and BCLC-A), with no major vascular invasion or extrahepatic spread, no portal hypertension (defined as hepatic venous pressure gradient <11 mm Hg or platelet count >100 000), adequate liver reserve and an anticipated liver remnant of at least 30–40%.4 Patients within Milan criteria could be considered for LT from either a dead or living donor.17 Similarly, according to the EASL-EORTC guidelines,15 resection is the first-line therapy option for patients with solitary tumours and well-preserved liver function, defined as normal bilirubin with either hepatic venous pressure gradient <10 mm Hg or platelet count ≥100 000. LT is the first treatment choice for patients with small multinodular tumours meeting Milan criteria (≤3 nodules ≤3 cm) or those with single tumours ≤5 cm and advanced liver dysfunction. In case of recurrence, the patient is reassessed with BCLC and treated accordingly. According to the ESMO-ESDO guidelines,16 in case of cirrhosis, resection is effective and safe (postoperative mortality <5%) in early BCLC stages (0 and A), provided that one is dealing with a single lesion, a good performance status and no clinically important portal hypertension.24 25 LR is the first-line curative treatment of solitary or multifocal HCC confined to the liver, anatomically resectable and with satisfactory liver function reserve according to the APASL guidelines.11 Definite contraindications for resection are distant metastasis, main portal vein thrombosis and inferior vena cava thrombosis. In case of non-resectable HCC within Milan criteria in Child A cirrhosis, local ablation is recommended. According to the evidence-based Japanese guidelines, LR is indicated for HCC if there are three or fewer tumours and all are limited to the liver. There is no restriction on tumour size.8 26 It is suggested that patients with tumour invasion to the portal vein be indicated for surgery if the tumour has not progressed beyond the first-order branches. In fact, portal vein invasion is consistently reported as the most powerful prognostic factor for HCC.27–29 No transplantation is indicated at this stage of cirrhosis. According to the Italian18 and German19 guidelines, LT is the treatment of choice for patients with Child-Pugh A cirrhosis within the Milan criteria.3 30 According to the Italian guidelines, a hepatic resection can be done for patients within the Milan criteria not eligible for transplantation (age, comorbidities) in Child-Pugh A patients. The best survival results are achieved for patients with good performance status, without comorbidities and with single tumours. For single tumour sized 2–3 cm, the 5-year survival of 60–70% and the perioperative mortality is about 2–3%.31–34 Portal hypertension (portal-hepatic gradient >12 mm Hg or platelets count <100 000/mL with splenomegaly or oesophageal varices) is associated with poor prognosis, but does not exclude resection in well-selected patients.8 In case of unifocal HCC beyond the Milan criteria regarding size (>5 cm), surgical resection is the main indication, if feasible, and if the liver remnant is large enough. According to the German S3-guidelines,19 patients not suitable for transplantation with Child A cirrhosis can be resected or treated with radiofrequency ablation (RFA) according to tumour size and number. Adequate postoperative liver function and portal hypertension need to be taken into account on functional resectability. In Child A cirrhosis, a minimum of 40% of liver parenchyma is needed to minimise risk of postoperative liver failure.19 Figure 2 shows the comparison of the indications to LR and LT using the assigned colour codes as described in the methods section. Figure 2 Treatment of HCC in Child-Pugh A cirrhosis. Green circle: first-line therapy. Yellow circle: therapy can be carried out or therapy recommended as second line. Black circle: no or no clear treatment recommendations. Red-circled white dot: rejection of the therapy. HCC, hepatocellular carcinoma. Child-Pugh class B According to the BCLC-based guidelines and JSH guidelines, the treatment recommendations for HCC on Child-Pugh B cirrhosis are identical to Child-Pugh A cirrhosis as described in the previous chapter. In particular, the SEOM guidelines17 recommend LR for patients with solitary or limited multifocal HCC (stage BCLC-0 and BCLC-A), with no major vascular invasion or extrahepatic spread, no portal hypertension (defined as hepatic venous pressure gradient <11 mm Hg or platelet count >100 000), adequate liver reserve and an anticipated liver remnant of at least 30–40%. Anatomical resections are recommended. Patients within Milan criteria could be considered for LT (from either a dead or living donor).4 LR is the first-line curative treatment of solitary or multifocal HCC confined to the liver, anatomically resectable and with satisfactory liver function reserve according to the APASL guidelines.11 LT can be offered to patients within the Milan criteria when resection is not possible. According to the Italian and German guidelines, LT is the treatment of choice for patients with Child-Pugh B cirrhosis within the Milan criteria.3 35 According to the Italian guidelines for patients with Child B cirrhosis non-eligible for transplantation, LR represents an option in case of a single tumour which can be removed with a limited resection in particular for those patients without clinically manifested portal hypertension. Patients not suitable for transplantation with Child B cirrhosis can be resected or treated with RFA according to tumour size and number according to the German S3-guidelines.19 Figure 3 shows the comparison of the indications for LR and LT. Figure 3 Treatment of HCC in Child-Pugh B cirrhosis. Green circle: first-line therapy. Yellow circle: therapy can be carried out or therapy recommended as second line. Black circle: no or no clear treatment recommendations. Red-circled white dot: rejection of the therapy. HCC, hepatocellular carcinoma. Child-Pugh class C According to AASLD,3 SEOM17 and EASL-EORTC15 guidelines, Child-Pugh C score defines an end-stage disease. Neither transplantation nor resection is recommended. According to the ESMO-ESDO guidelines,16 patients with poor liver synthetic function and tumour extent within the Milan criteria should not be denied the possibility of LT and are therefore not classified as terminal stage. According to APASL guidelines, LT provides the best curative treatment within Milan criteria associated with Child C cirrhosis and without radiological evidence of venous invasion or distant metastasis. In Japan, transplantation is recommended at this stage of cirrhosis for patients with HCC within Milan criteria and age ≤65, if disease control is not possible using other treatment methods. Tumour diameter, tumour number, tumour marker levels, extent of vascular invasion and degree of tumour differentiation are strong predictors of recurrence. According to the AIOM18 and German S3-guidelines,19 LT is the treatment of choice for patients with Child-Pugh C cirrhosis within the Milan criteria. For Child C cirrhosis, no LR is recommended according to the Italian guidelines. Figure 4 shows the comparison of the indications for LR and LT. Figure 4 Treatment of HCC in Child-Pugh C cirrhosis. Green circle: first-line therapy. Yellow circle: therapy can be carried out or therapy recommended as second line. Black circle: no or no clear treatment recommendations. Red-circled white dot: rejection of the therapy. HCC, hepatocellular carcinoma. Expansion of the criteria beyond Milan The AASLD guidelines,3 APASL,11 evidence-based Japanese guidelines and German19 do not recommend the expansion of the listing criteria beyond the standard Milan criteria. ESMO-ESDO guidelines16 give no statement in this regard. According to the SEOM guidelines,17 patients with tumour characteristics slightly beyond Milan criteria and without microvascular invasion may be considered for LT. However, this indication requires prospective validation. The EASL-EORTC guidelines15 state that the extension of tumour limit criteria for LT for HCC has not been established. Modest expansion of Milan criteria applying the ‘up-to-seven’ criteria (new Milan criteria: HCC with seven as the sum of the size of the largest tumour (in cm) and the number of tumours) proposed by Mazzaferro et al in 200936 in patients without microvascular invasion achieves competitive outcomes, and thus this indication requires prospective validation. In Italy, the expansion of the criteria was proposed, but the probability that a patient beyond Milan is transplanted is very low. LT for patients beyond Milan cannot be recommended according to the German S-3 guidelines. Bridging therapy for liver transplant candidates already on waiting list Generally, all guidelines recommend bridging therapy if the waiting list time exceeds 6 months. According to the ESMO-ESDO guidelines16 in case of a long-anticipated waiting time (>6 months), patients may be offered resection, local ablation or transarterial chemoembolisation in order to minimise the risk of tumour progression and to offer a ‘bridge’ to transplant. In Italy, bridging therapies are also allowed under progression while on waiting list. According to the German S3 guidelines,19 bridging is recommended when a long waiting time until transplantation is expected. According to the APASL guidelines,11 bridging therapy using local ablation or chemoembolisation may reduce the dropout rate with long waiting time of more than 6 months. According to the EASL-EORTC guidelines,15 patients already on the waiting list with tumour progression beyond Milan criteria and liver-only disease should be placed on hold until downstaging by local ablation or chemoembolisation is achieved and maintained for a period of at least 3 months. In the SEOM17 and JSH guidelines, no recommendation is given about bridging therapy for patients on waiting list. Downstaging of patients beyond Milan criteria According to the SEOM guidelines,17 downstaging cannot be recommended. According to the EASL-EORTC guidelines,15 downstaging policies for HCCs exceeding conventional criteria cannot be recommended and should be explored in the context of prospective studies aimed at survival and disease progression end points. According to the APASL guidelines,11 downstaging therapy using local ablation or chemoembolisation may reduce the dropout rate with long waiting time of more than 6 months, but there is no proven benefit in long-term survival or downstaging to allow expanded indication. The Japanese evidence-based guidelines state that there is insufficient scientific evidence to support tumour downstaging prior to LT to improve HCC prognosis. The role of transplantation after downstaging is not established in Italy because of lack of high-quality evidence. On the basis of the available data, it is reasonable that patients slightly beyond Milan and in good general conditions can receive a consultation for possible transplantation. According to the German S3 guidelines,19 downstaging can be considered in order to achieve Milan criteria. AASLD3 and ESMO-ESDO16 offer no recommendation regarding downstaging. Living donor LT According to the AASLD guidelines,3 LDLT is a reasonable approach if the waiting time exceeds 7 months by taking into account the risk of dropout while waiting (4% per month), the expected survival of the recipient (70% at 5 years) and the risk for the donor (0.3–0.5% mortality).37 This procedure should be only performed by expert surgeons. According to the SEOM guidelines,17 patients within Milan criteria could be considered for LT from either a dead or living donor, achieving a 5-year overall survival of more than 70% and a 5-year recurrence rate of <10%.30 According to the EORTC-EASL guidelines,15 LDLT is an alternative option in patients with a waiting list exceeding 6–7 months. It is not recommended for any extended indications, except in the context of research studies, and should be restricted to centres of excellence in hepatic surgery. According to the APASL guidelines,11 LDLT is theoretically a more preferred choice for HCC patients, because the waiting list time is significantly reduced. However, risk of donor hepatectomy (0.3–0.5%) and recipient complications (20–40%) need to be considered in offering such treatment. LDLT is the main type of transplantation performed in Japan and does not involve wait list. In Italy, it represents only 0.6% of all transplantations. According to the German S3-guidelines,19 LDLT is an option for patients in whom a tumour progress is likely while on the waiting list with the risk of drop out. By using LDLT, waiting time can be avoided and thus tumour progression can be prevented. Additionally, it relieves the limited pool of deceased donor organs. As the potential risk of complications for the donor in experienced centres is relatively low, this possibility in absence of an appropriate postmortem donor and therefore a long-anticipated waiting time should be evaluated. Morbidity and mortality after LDLT is comparable with the recipient of a postmortem LT. The guidelines ESMO-ESDO16 give no recommendation regarding live donor LT. Figure 5 shows the comparison of the recommendations regarding the expansion of the criteria beyond Milan, bridging therapy, downstaging and LDLT. Figure 5 Comparison of the recommendation regarding expansion of the criteria beyond Milan, bridging therapy, downstaging and LDLT. Green circle: first-line therapy. Yellow circle: therapy can be carried out or therapy recommended as second line. Black circle: no or no clear treatment recommendations. Red-circled white dot: rejection of the therapy. LDLT, living-donor liver transplantation. Discussion There is no worldwide consensus on the recommendation for surgical treatment of HCC, although the evidence is the same. Relative homogeneity in indications exists for the countries using the BCLC classification, with the exception of patients within Milan criteria and Child-Pugh C cirrhosis. These patients are classified as patients with end-stage disease according to the AASLD,3 SEOM17 and EASL-EORTC15 and therefore are consequently excluded from transplantation. Therapeutic-suggested option is best supportive care. The ESMO-ESDO guidelines16 allow transplantation for HCC within Milan on Child-Pugh C, and these patients are not classified as end stage. It is remarkable that the two European guidelines differ in a so important point, in one case excluding Child-Pugh C patients from transplantation (EASL-EORTC) and in the other allowing it (ESMO-ESDO). The Italian18 and German guidelines19 recommend transplantation for Child-Pugh C patients within Milan criteria. The question about the effective usefulness of European guidelines, when, for example, European countries like Germany and Italy use their national guidelines, remains open. Moreover, there is no homogeneity between European guidelines itself. Spain has also its own guidelines which are only accessible to members of the Spanish society of medical oncology. Another critical point where misunderstanding can arise is the treatment of single tumours between 2 and 5 cm in liver cirrhosis Child-Pugh A/B according to the EASL-EORTC and SEOM guidelines. In fact, both rely on the updated BCLC staging system (2011). As the original BCLC classification, on which the AASLD guidelines rely, clearly states that first-line treatment option is LR if no portal hypertension and elevated bilirubin are present and LT is indicated only in case of advanced liver dysfunction, the EASL-EORTC and SEOM guidelines are unclear in this point. While the text of the guidelines suggests treatment according to the original BCLC classification LR as first-line therapy, the graphical representation of the treatment algorithm in both guidelines suggests first-line therapy for such tumours is transplantation and not resection. Taking into consideration the graphical representation of patients with early stage disease (BCLC A) that they are not candidate for LR, we interpreted the guidelines according to the text and not according to the figure. However, this possible double interpretation needs to be mentioned and future guidelines should state without ambiguity the therapeutic strategy for these tumours. The major difference between the treatment algorithm used in Japan and the BCLC system is the indication for hepatectomy for HCC with ≤3 lesions and a diameter ≤3 cm on Child-Pugh A/B. The BCLC system recommends LT or RFA for HCC with two or three nodules and a diameter ≤3 cm. In contrast, the treatment algorithm in Japan recommends hepatectomy for HCC with ≤3 lesions if liver function is good, regardless of the tumour size. According to the Japanese guidelines, as well as for Italy, Germany and European ESMO-ESDO guidelines, first-line therapy for patients with Child-Pugh C cirrhosis and HCC within Milan criteria is transplantation. In Japan, the majority of transplantations are LDLT, whereas in Italy, only 0.6% of the patients treated with transplantation are LDLT. In general, cultural attitudes in Asia regarding life, death, ethics and religion have influenced their attitude towards organ transplantation from deceased donors greatly. In highly specialised centres, the survival after LDLT is comparable with the survival after postmortem transplantation (70% at 5 years). Although donor morbidity and mortality is low, a reported mortality between 0.3% and 0.5% does not appear to be acceptable.38 Donor safety is paramount and has been a topic of much discussion in the transplant community worldwide. The donor risk appears to be low overall, with a favourable long-term quality of life. The latest trend has been a gradual shift from right lobe grafts to left lobe grafts to reduce donor risk, provided that the left lobe can provide adequate liver volume for the recipient.39 Significant low morbidity and mortality rates of donor patients are reported by high-volume centres in Asia due to high case load and standardised perioperative and postoperative treatment. Also, already published in 2007 on US-data, it appears that LDLT at experienced centres results in best long-term survival compared with all other groups.37 Moreover, LDLT offers the advantage over the deceased donation of a clinically more stable recipient and an optimal time of transplantation avoiding long waiting time.19 As a result of the high dropout rate for patients with HCC, worldwide the priority of liver graft allocation has been reconsidered. First, waiting list priority was determined primarily by liver disease severity based on the Model for End-Stage Liver Disease (MELD) score in order to reduce the rate of death on the waiting list.40 Second, patients with HCC that fulfilled the Milan criteria were registered with an adjusted score and were subsequently assigned additional scores at regular intervals to reflect their risk for dropout as a result of tumour progression. With such priority listing, the access to timely transplant liver for patients with HCC has improved in the USA.41 However, introduction of the MELD score to reduce death on the waiting list did not only achieve positive results in all countries. Especially in Germany, because of lack of donor organs, only high MELD scores, partially high in the 30s, result in allocation of liver grafts. Although waiting list mortality was decreased, this basically means to transplant patients in Child C status with high risk of poor outcome, and increased morbidity and mortality of those extremely sick patients is common.42 The treatment algorithm of Japanese evidence-based guidelines includes grade of liver damage, tumour number and tumour diameter. Extrahepatic disease and vascular spread are not included in the algorithm in contrast with the AASLD and the APASL guidelines. This was explained by the need to keep the treatment algorithm simple, few evidences available to recommend a certain treatment option for HCC with vascular invasion and extrahepatic HCC at the time of initial diagnosis was considered rare in daily practice in Japan.13 Interestingly, extrahepatic spread and vascular invasion are included in the treatment algorithm of the consensus-based Japanese guidelines,14 whereas neoplastic invasion of bile ducts plays no role in all guidelines so far. Interestingly, only in the evidence-based Japanese guidelines an age limit of ≤65 is set to be transplantable. In several countries, as Germany or Italy, for example, no patients >65 yrs are routinely transplanted although no age limit is expressed in the respective guidelines. Quantitative liver function tests allow a more precise assessment of postoperative morbidity and mortality. The relevance of quantitative liver function tests has so far found consideration only in the guidelines of the JSH. In the JSH-HCC guidelines, the indocyanine green (ICG) test as an indicator of liver function is considered indispensable for surgical decision-making, but is not routinely performed before non-surgical treatments like RFA or Trans-Arterial Chemo-Embolization (TACE). As several publications demonstrate the usefulness of ICG clearance alone43 or in combination with other parameters44 or imaging-based liver function tests45 as a predictor of postoperative death, extended liver surgery has been made safer to avoid postoperative liver failure. Additionally to the ICG clearance, the LiMAx test has been found to be valuable to quantify liver function.46 Perioperative morbidity and mortality was reduced after implementing LiMAx algorithms in LRs,47 and after LT the LiMAx score was predictive for postoperative liver failure. However, so far LiMAx has not been recognised in any of the guidelines for treatment of HCC or LT. Also, new innovations in liver surgery like portal vein embolisation,48 two-stage hepatectomy,49 Associating Liver Partition and Portal Vein Ligation for Staged Hepatectomy (ALPPS)50 or partial ALPPS51 allow extensive LR with acceptable morbidity and mortality, even when transplantation because of tumour load is no option anymore. However, extended LR can only be performed in healthy liver, thus again leaving transplantation as only potential curative treatment. Another critical point which is not addressed in any of the guidelines is the tumour biology. Tumour biology, immunological and genetic tumour-specific treatments gain more and more impact on diagnosis, interdisciplinary treatment and outcome. One promising field regarding risk of acute rejection after LT is gene expression profiling. Thude et al52 demonstrated that genotyping of liver recipients for specific genetic polymorphisms might be useful to stratify liver transplant recipients according to the risk of acute liver transplant rejection. Mazzaferro,53 who introduced the Milan criteria in the field of LT, recently published an article about an adaptive approach for selection and allocation in LT for HCC. He proposes to maximise ‘all tumour and therapy heterogeneities in a model that utilizes variations in HCC presentation and response to treatment as adjusting factors to reconcile selection and allocation logistics, with the ultimate aim of increasing the benefit, effectiveness, and justice of transplantation for cancer’. Unfortunately, as stated before, not one actual guideline considers these important developments in individualised and specific treatment. In conclusion, whereas we find a consensus in HCC treatment in healthy livers, the analysed international recommendations about the treatment of HCC in cirrhotic livers show several variations, although nearly all guidelines claim to be evidence based. Moreover, promising prognostic factors considering tumour biology as well as liver function tests should be included in future guidelines. One possible explanation for the inhomogeneity among the guidelines included in our analysis might be cultural difference as well as variation in the healthcare system. Progress in the development of guidelines will be made when the reasons that explain the differences in the existing guidelines can be identified. These reasons can be identified when the burden and risks that have to be accepted and the outcomes, that is, the achieved survival and quality of life, can be assessed.54 Meaningful assessments require two essential conditions. First, the conditions under which ‘costs’ and ‘consequences’ are compared have to reflect the situation of day-to-day clinical practice and second, the conditions have to be standardised. The traditional method for comparative assessments of clinical outcomes is the randomised controlled trial. These trials measure effects under ideal study conditions, that is, efficacy, but not effects that can be detected under real-world conditions, which is effectiveness. Traditional guidelines are based on efficacy but not yet effectiveness data. Methods that compare effectiveness under real-world conditions have only recently been proposed.55 Some of these methods include risk stratification which means that only patients with similar risks (high, low or intermediate) can be compared and the baseline risks of each patient have to be related to each of the outcomes that will be assessed. These assessments under real-life conditions can be completed in any community hospital and will be important as basis of clinical guidelines. When outcomes, conditions, patient characteristics and interventions are described transparently, it will be possible to discuss possible reasons for different guidelines in different countries. Contributors: FP and DH-B contributed substantially to the conception and design of the study. GM and MK contributed in the analysis and interpretation of the data. GM and MK drafted the article. All authors gave the final approval of the version to be published. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Parkin DM , Bray F , Ferlay J , et al Estimating the world cancer burden: Globocan 2000 . Int J Cancer 2001 ;94 :153 –6 . doi:10.1002/ijc.144011668491 2 Bosch FX , Ribes J , Diaz M , et al Primary liver cancer: worldwide incidence and trends . Gastroenterology 2004 ;127 (Suppl 1 ):S5 –16 . doi:10.1053/j.gastro.2004.09.01115508102 3 Bruix J , Sherman M American Association for the Study of Liver Diseases: management of hepatocellular carcinoma: an update . Hepatology 2011 ;53 :1020 –2 . doi:10.1002/hep.2419921374666 4 Llovet JM , Fuster J , Bruix J Intention-to-treat analysis of surgical treatment for early hepatocellular carcinoma: resection versus transplantation . Hepatology 1999 ;30 :1434 –40 . doi:10.1002/hep.51030062910573522 5 Arii S , Yamoka Y , Futagawa S , et al Results of surgical and nonsurgical treatment for small-sized hepatocellular carcinomas: a retrospective and nationwide survey in Japan. The Liver Cancer Study Group of Japan . Hepatology 2000 ;32 :1224 –9 .11093728 6 Bruix J , Castells A , Bosch J , et al Surgical resection of hepatocellular carcinoma in cirrhotic patients: prognostic value of preoperative portal pressure . Gastroenterology 1996 ;111 :1018 –22 .8831597 7 Huo TI , Lin HC , Hsia CY , et al The model for end-stage liver disease based cancer staging system are better prognostic models for hepatocellular carcinoma: a prospective sequential survey . Am J Gastroenterol 2007 ;102 :1920 –30 . doi:10.1111/j.1572-0241.2007.01370.x17573792 8 Ishizawa T , Hasegawa K , Aoki T , et al Neither multiple tumors nor portal hypertension are surgical contraindications for hepatocellular carcinoma . Gastroenterology 2008 ;134 :1908 –16 . doi:10.1053/j.gastro.2008.02.09118549877 9 Llovet JM , Bru C , Bruix J Prognosis of hepatocellular carcinoma: the BCLC staging classification . Semin Liver Dis 1999 ;19 :329 –38 . doi:10.1055/s-2007-100712210518312 10 Bruix J , Sherman M Practice guidelines committee, American Association for the Study of Liver Diseases. Management of hepatocellular carcinoma . Hepatology 2005 ;42 :1208 –36 .16250051 11 Omato M , Lesmana LA , Tateishi R , et al Asian Pacific Association for the study of the liver consensus recommendations on hepatocellular carcinoma . Hepatol Int 2010 ;4 :439 –74 . doi:10.1007/s12072-010-9165-720827404 12 Poon RT , Cheung TT , Kwok PC , et al Hong Kong consensus recommendations on the management of hepatocellular carcioma . Liver Cancer 2015 ;4 :51 –69 . doi:10.1159/00036772826020029 13 Kokudo N , Hasegawa K , Akahane M , et al Evidence-based clinical practice guidelines for hepatocellular carcinoma: The Japan Society of Hepatology 2013 update (3th JSH-HCC Guidelines) . Hepatol Res 2015 ;45 :123 –7 . doi:10.1111/hepr.12464 14 Kudo M , Matsui O , Izumi N , et al JSH consensus-based clinical practice guidelines for the management of hepatocellular carcinoma: 2014 update by the Liver Cancer Study Group of Japan . Liver Cancer 2014 ;3 :458 –68 . doi:10.1159/00034387526280007 15 European Association for the Study of the Liver, European Organisation for Research and Treatment of Cancer . EASL-EORTC Clinical Practice Guidelines: management of hepatocellular carcinoma . J Hepatol 2012 ;56 :908 –43 . doi:10.1016/j.jhep.2011.12.00122424438 16 Verslype C , Rosmorduc O , Rougier P , On behalf of the ESMO guidelines working group . Hepatocellular carcinoma: ESMO-ESDO clinical practice guidelines for diagnosis, treatment and follow up . Ann Oncol 2012 ;23 (Suppl 7 ):vii41 –8 . doi:10.1093/annonc/mds22522997453 17 Sastre J , Diaz-Beveridge R , Garcia-Foncillas J , et al Clinical guidelines SEOM: hepatocellular carcinoma . Clin Transl Oncol 2015 ;17 :988 –95 .26607931 18 Associazione Italiana Oncologia Medica (AIOM) . Linee Guida epatocarcinoma . Edizione , 2015 . 19 Deutsche Gesellschaft für Verdauungs- und Stoffwechselkrankheiten . Diagnostik und Therapie des hepatozellulären Karzinoms. Version 1.0. Mai 2013 . 20 Lang H , Sotiropoulos GC , Dömland M , et al Liver resection for hepatocellular carcinoma in non-cirrhotic liver without underlsing viral hepatitis . Br J Surg 2005 ;92 :198 –202 . doi:10.1002/bjs.476315609381 21 Benson AB III, Abrams TA , Ben-Josef E , et al NCCN clinical practice guidelines in oncology: hepatobiliary cancers . J Natl Compr Canc Netw 2009 ;7 :350 –91 .19406039 22 Minagawa M , Makuuchi M , Takayama T , et al Selection criteria for repeat hepatectomy in patients with recurrent hepatocellular carcinoma . Ann Surg 2003 ;238 :703 –10 . doi:10.1097/01.sla.0000094549.11754.e614578733 23 Schlitt HJ , Neipp M , Weimann A , et al Recurrence patterns of hepatocellular and fibrolamellar carcinoma after liver transplantation . J Clin Oncol 1999 ;17 :324 –31 . doi:10.1200/JCO.1999.17.1.32410458250 24 Belghiti J , Hiramatsu K , Benoist S , et al Seven hundred forty-seven hepatectomies in the 1990s an update to evaluate the actual risk of liver resection . J Am Coll Surg 2000 ;191 :38 –46 . doi:10.1016/S1072-7515(00)00261-110898182 25 Llovet JM , Schwartz M , Mazzaferro V Resection and liver transplantation for hepatocellular carcinoma . Semin Liver Dis 2005 ;25 :181 –200 . doi:10.1055/s-2005-87119815918147 26 Ho MC , Huang GT , Tsang YM , et al Liver resection improves the survival of patients with multiple hepatocellular carcinomas . Ann Surg Oncol 2009 ;16 :848 –55 . doi:10.1245/s10434-008-0282-719159983 27 Wu CC , Hsieh SR , Chen JT , et al An appraisal of liver and portal vein resection for hepatocellular carcinoma with tumor thrombi extending to portal bifurcation . Arch Surg 2000 ;135 :1273 –9 . doi:10.1001/archsurg.135.11.127311074879 28 Shi J , Lai EC , Li N , et al Surgical treatment of hepatocellular carcinoma with portal vein tumor thrombus . Ann Surg Oncol 2010 ;17 :2073 –80 . doi:10.1245/s10434-010-0940-420131013 29 Inoue Y , Hasegawa K , Ishizawa T , et al Is there any difference in survival according to the portal tumor thrombectomy method in patients with hepatocellular carcinoma? Surgery 2009 ; 145 :9 –19 . doi:10.1016/j.surg.2008.09.00519081470 30 Mazzaferro V , Regalia E , Doci R , et al Liver transplantation for the treatment of small hepatocellular carcinomas in patients with cirrhosis . N Engl J Med 1996 ;334 :693 –9 . doi:10.1056/NEJM1996031433411048594428 31 Llovet JM , Bruix J Novel advancements in the management of hepatocellular carcinoma in 2008 . J Hepatol 2008 ;48 (Suppl 1 ):S20 –37 . doi:10.1016/j.jhep.2008.01.02218304676 32 Mazzaferro V , Romito R , Schiavo M , et al Prevention of hepatocellular carcinoma recurrence with alpha-interferon after liver resection in HCV cirrhosis . Hepatology 2006 ;44 :1543 –54 . doi:10.1002/hep.2141517133492 33 Poon RT , Fan ST , Lo CM , et al Extended hepatic resection for hepatocellular carcinoma in patients with cirrhosis: is it justified? Ann Surg 2002 ;236 :602 –11 . doi:10.1097/01.SLA.0000033038.38956.5E12409666 34 Roayaie S , Blume IN , Thung SN , et al A system of classifying microvascular invasion to predict outcome after resection in patients with hepatocellular carcinoma . Gastroenterology 2009 ;137 :850 –5 . doi:10.1053/j.gastro.2009.06.00319524573 35 Mazzaferro V , Bhoori S , Sposito C , et al Milan criteria in liver transplantation for hepatocellular carcinoma: an evidence-based analysis of 15 years of experience . Liver Transpl 2011 ;17 (Suppl 2 ):S44 –57 . doi:10.1002/lt.2236521695773 36 Mazzaferro V , Llovet JM , Miceli R , et al Predicting survival after liver transplantation in patients with hepatocellular carcinoma beyond the Milan criteria: a retrospective, exploratory analysis . Lancet Oncol 2009 ;10 :35 –43 . doi:10.1016/S1470-2045(08)70284-519058754 37 Berg CL , Gillespie BW , Merion RM , et al Improvement in survival associated with adult-to-adult living donor liver transplantation . Gastroenterology 2007 ;133 :1806 –13 . doi:10.1053/j.gastro.2007.09.00418054553 38 Sarasin FP , Majno PE , Llovet JM , et al Living donor liver transplantation for early hepatocellular carcinoma: a life-expectancy and cost-effectiveness prespective . Hepatology 2001 ;33 :1037 –79 . doi:10.1053/jhep.2001.23311 39 Kim PT , Testa G Living donor liver transplantation in the USA . Hepatobiliary Surg Nutr 2016 ;5 :133 –40 . doi:10.3978/j.issn.2304-3881.2015.06.0127115007 40 Wiesner R , Edwards E , Freeman R , et al Model for end-stage liver disease (MELD) and allocation of donor livers . Gastroenterology 2003 ;124 :91 –6 . doi:10.1053/gast.2003.5001612512033 41 Yao FY , Bass NM , Ascher NL , et al Liver transplantation for hepatocellular carcinoma: Lessons from the first year under the Model of End-Stage Liver Disease (MELD) organ allocation policy . Liver Transpl 2004 ;10 :621 –30 . doi:10.1002/lt.2015915108253 42 Kaltenborn A , Hartmann C , Salinas R , et al Risk factors for short- and long-term mortality in liver transplant recipients with MELD score ≥30 . Ann Transplant 2015 ;20 :59 –69 . doi:10.12659/AOT.89232225630462 43 Haegele S , Reiter S , Wanek D , et al Perioperative non-invasive indocyanine green-clearance testing to predict postoperative outcome after liver resection . PLoS ONE 2016 ;11 :e0165481 doi:10.1371/journal.pone.016548127812143 44 Lau H , Man K , Fan ST , et al Evaluation of preoperative hepatic function in patients with hepatocellular carcinoma undergoing hepatectomy . Br J Surg 1997 ;84 :1255 –9 . doi:10.1002/bjs.18008409179313707 45 Geisel D , Lüdemann L , Hamm B , et al Imaging-based liver function tests—past, present and future . Rofo 2015 ;187 :863 –71 . doi:10.1055/s-0035-155330626230140 46 Stockmann M , Lock JF , Riecke B , et al Prediction of postoperative outcome after hepatectomy with a new bedside test for maximal liver function capacity . Ann Surg 2009 ;250 :119 –25 . doi:10.1097/SLA.0b013e3181ad85b519561474 47 Jara M , Reese T , Malinowski M , et al Reductions in post-hepatectomy liver failure and related mortality after implementation of the LiMAx algorithm in preoperative work-up: a single-centre analysis of 1170 hepatectomies of one or more segments. . HPB (Oxford) 2015 ;17 :651 –8 . doi:10.1111/hpb.1242426058324 48 Glantzounis GK , Tokidis E , Basourakos SP , et al The role of portal vein embolization in the surgical management of primary hepatobiliary cancers. A systematic review . Eur J Surg Oncol 2017 ;43 :32 –41 . doi:10.1016/j.ejso.2016.05.02627283892 49 Adam R , Imai K , Castro Benitez C , et al Outcome after associating liver partition and portal vein ligation for staged hepatectomy and conventional two-stage hepatectomy for colorectal liver metastases . Br J Surg 2016 ;103 :1521 –9 . doi:10.1002/bjs.1025627517369 50 Kremer M , Manzini G , Hristov B , et al Impact of neoadjuvant chemotherapy on hypertrophy of the future liver remnant after associating liver partition and portal vein ligation for staged hepatectomy . J Am Coll Surg 2015 ;221 :717 –28.e1 . doi:10.1016/j.jamcollsurg.2015.05.01726232303 51 Linecker M , Kambakamba P , Reiner CS , et al How much liver needs to be transected in ALPPS? A translational study investigating the concept of less invasiveness . Surgery 2017 ;161 :453 –64 . doi:10.1016/j.surg.2016.08.00427814957 52 Thude H , Janssen M , Sterneck M , et al 14-bp ins/del polymorphism and +3142C>G SNP of the HLA-G gene have a significant impact on acute rejection after liver transplantation . Hum Immunol 2016 ;77 :1159 –65 . doi:10.1016/j.humimm.2016.09.00927664842 53 Mazzaferro V Squaring the circle of selection and allocation in liver transplantation for HCC: an adaptive approach . Hepatology 2016 ;63 :1707 –17 . doi:10.1002/hep.2842026703761 54 Porzsolt F , Kaplan RM , eds. Optimizing health—improving the value of healthcare delivery . New York : Springer , 2006 :313 ISBN 0-387-33920-5 . 55 Porzsolt F , Rocha NG , Toledo-Arruda AC , et al Efficacy and effectiveness trials have different goals, use different tools, and generate different messages . Pragmat Obs Res 2015 ;6 : 47 –54 . doi:10.2147/POR.S8994627774032

PMC005xxxxxx/PMC5372045.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01330710.1136/bmjopen-2016-013307Health Services ResearchProtocol1506170417041724‘right@home’: a randomised controlled trial of sustained nurse home visiting from pregnancy to child age 2 years, versus usual care, to improve parent care, parent responsivity and the home learning environment at 2 years Goldfeld Sharon 123Price Anna 12Bryson Hannah 12Bruce Tracey 4Mensah Fiona 35Orsini Francesca 5Gold Lisa 6Hiscock Harriet 123Smith Charlene 7Bishop Lara 2Jackson Dianne 7Kemp Lynn 4 1 Centre for Community Child Health, The Royal Children's Hospital, Parkville, Victoria, Australia 2 Population Health, Murdoch Childrens Research Institute, Parkville, Victoria, Australia 3 Department of Paediatrics, University of Melbourne, Parkville, Victoria, Australia 4 Ingham Institute, Western Sydney University, Sydney, New South Wales, Australia 5 Clinical Sciences and Biostatistics Unit, Murdoch Childrens Research Institute, The Royal Children's Hospital, Parkville, Victoria, Australia 6 School of Health and Social Development, Deakin University, Burwood, Victoria, Australia 7 Australian Research Alliance for Children and Youth, Canberra City, Australian Capital Territory, AustraliaCorrespondence to Professor Sharon Goldfeld; sharon.goldfeld@rch.org.au2017 20 3 2017 7 3 e0133073 7 2016 8 11 2016 4 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction By the time children start school, inequities in learning, development and health outcomes are already evident. Sustained nurse home visiting (SNHV) offers a potential platform for families experiencing adversity, who often have limited access to services. While SNHV programmes have been growing in popularity in Australia and internationally, it is not known whether they can improve children's learning and development when offered via the Australian service system. The right@home trial aims to investigate the effectiveness of an SNHV programme, offered to women from pregnancy to child age 2 years, in improving parent care of and responsivity to the child, and the home learning environment. Methods and analysis Pregnant Australian women (n=722) are identified after completing a screening survey of 10 factors known to predict children's learning and development (eg, young pregnancy, poor mental or physical health, lack of support). Consenting women—surveyed while attending clinics at 10 hospitals in Victoria and Tasmania—are enrolled if they report having 2 or more risk factors. The intervention comprises 25 home visits from pregnancy to 2 years, focusing on parent care of the child, responsivity to the child and providing a good quality home learning environment. The standard, universal, Australian child and family health service provides the comparator (control). Primary outcome measures include a combination of parent-reported and objective assessments of children's sleep, safety, nutrition, parenting styles and the home learning environment, including the Home Observation of the Environment Inventory and items adapted from the Longitudinal Study of Australian Children. Ethics and dissemination This study is approved by the Royal Children's Hospital Human Research Ethics Committees (HREC 32296) and site-specific HRECs. The investigators and sponsor will communicate the trial results to stakeholders, participants, healthcare professionals, the public and other relevant groups via presentations and publications. Trial registration number ISRCTN89962120, pre-results. home visitschild developmentmaternal healthrandomised controlled trialsocioeconomic factors ==== Body Strengths and limitations of this study First multisite, multijurisdictional randomised controlled trial to test the effectiveness of sustained nurse home visiting in Australia. Evaluation will inform the ongoing provision and delivery of the universal child and family health services in Victoria and Tasmania. This study is crucial for generating Australian evidence of an effective intervention to reduce the impact of social and environmental factors predisposing children to inequitable outcomes. The exclusion criteria mean the findings may not generalise to non-English-speaking women or women with severe intellectual disability. While we use a population-based sampling strategy for recruitment, women stop receiving the intervention if they move out of a study region. This could be avoided if the service is delivered across the participating states (ie, following the intention of the real-life design). Background By the time children start school, inequities in learning, development and health outcomes are already evident, due to the failure of health, education and welfare systems to adequately ameliorate the impacts of early adversity. The clear social gradient associated with children's vocabulary, emerging literacy, well-being and behaviour is evident from birth to school entry.1 These trajectories track into adolescence and correspond to poorer educational attainment, income and health across the life course.2–10 Neuroimaging research extends the evidence for these suboptimal trajectories, showing that children raised in poverty from infancy are more likely to have delayed brain growth with smaller volumetric size of the regions particularly responsible for executive functioning and language.11 This evidence supports the need for further effort to redress inequities that arise from the impact of adversity during the potential developmental window of opportunity in early childhood. The Australian Early Development Census is a population-level measure of early childhood development collected on every student by teachers at school entry (N>260 000) every 3 years.12 It measures five domains of early childhood development (physical health and well-being; social competence; emotional maturity; language and cognitive skills; and communication skills and general knowledge). Results show that 17.4% of children who live in areas subject to the greatest socioeconomic disadvantage are developmentally vulnerable on two or more of these domains. In other words, they are not equipped with the developmental abilities they need to flourish at school. This proportion is almost triple the 6.5% of children living in the most advantaged areas.12 Families experiencing the most adversity are often the least able to access health resources and support services.13 To address the need for better reach, sustained nurse home visiting (SNHV) has become increasingly popular as a model of service delivery to improve outcomes for these families.14 Internationally, the best known SNHV programme is the US Nurse-Family Partnership (NFP), also known as Family Nurse Partnership (FNP) in the UK. Designed and led by Professor David Olds, NFP/FNP has grown in popularity as multiple trials have concluded effectiveness for a variety of outcomes for young, first-time mothers and their children. This includes improved birth, health and child development outcomes, and reductions in child maltreatment.15 A number of SNHV programmes have also shown favourable effects on healthcare usage, including rates of well-child healthcare visits.15 In Australia, SNHV-type programmes are becoming more widespread, with a number of states offering an array of outreach and home visiting programmes to parents via universal (predominantly nursing) healthcare platforms.16 However, only the Maternal (formerly Miller) Early Childhood Sustained Home-visiting (MECSH)17 programme has been rigorously evaluated when delivered in this ‘real-life’ setting. Conducted by Kemp et al,18 MECSH recruited 208 participants from Miller, a Sydney suburb known for experiencing significant socioeconomic disadvantage. Pregnant women of any age or parity were eligible if they reported one or more risk factors on the antenatal psychosocial screening interview, which is a standard clinical tool collected in New South Wales' birthing hospitals. The SNHV programme aimed to improve family, maternal, and child health and developmental outcomes measured when children turned 2 years old. Compared with ‘usual care’ mothers who received the universally available and free programme of nurse-delivered well-child checks, MECSH ‘intervention’ mothers were more responsive to children's needs (effect size (d)=0.26, p=0.02). Effects at child age 2 years were most pronounced for women who were first-time mothers, had more than one antenatal risk factor or had poorer mental health.18 Intervention mothers who were born overseas (n=62) also breast fed for longer (d=0.87, p<0.001) and reported an improved experience of being a mother (d=0.54, p=0.003) than the equivalent usual care subgroup. There were no differences between groups in child development, immunisation rates, maternal health or smoking rates.18 Despite the international interest and considerable financial investment required to implement SNHV programmes, the literature shows variable results. Even the most successful SNHV programmes have moderate effects in the short term, and mixed benefits in the longer term.19 One reason for this may be that the measures assessed by these trials cover a broad range of child and parent outcomes that are not always explicitly targeted by the intervention. A review by Segal et al20 of SNHV programmes designed to reduce child maltreatment found that programme logic helped target explicit outcomes and was related to effectiveness. They noted the use of programme logic to be a key feature missing from many trials including those targeting developmental and behavioural outcomes. The context and design of the evaluation of programmes may also be a key factor to evidence of effectiveness. Benefits observed in one system (eg, USA) may not translate in different service systems with different populations and reach. For example, a recent evaluation of the effectiveness of FNP delivered in England's broadly based, publicly funded, healthcare setting, concluded no evidence of benefit for the primary outcomes versus usual care, that is, smoking in pregnancy, birth weight, emergency hospital attendance and admission for the child, and subsequent pregnancy.21 That said, some secondary outcomes (eg, language development concern, child safety) did show some benefit even within a trial where the control group received high levels of routine care. This suggests that outcomes for SNHV need to be carefully considered with thoughtful programme logic, rigorous study design and an understanding of the likely, specific healthcare system benefits.21 In the Australian context, the results from the MECSH trial suggest that SNHV holds promise for improving children's learning and development outcomes. However, with more SNHV programmes in Australia gaining traction without evaluation in randomised controlled trials (RCTs), and evidence for only moderate effects from international research, it is important to empirically determine whether an Australian SNHV programme can help families overcome the effects of adversity by improving the learning and development of their children. In response to this need, the state governments of Victoria and Tasmania and philanthropic organisations have funded a new partnership committed to the development and evaluation of the first multisite, multijurisdictional, randomised trial of SNHV delivered via the universal nurse-delivered child and family health (CFH) service. This is the largest SNHV trial in Australia. The programme is known as ‘right@home’. This paper reports the research protocol for the right@home RCT. The primary hypotheses are that at child age 2 years, when compared with those in usual care, the intervention mothers will demonstrate improved: (1) parental care of the child; (2) parent responsivity and (3) a supportive home learning environment. Secondary hypotheses are that (1) mothers will have improved pregnancy outcomes, quality of life, mental health, general health and well-being, parenting self-efficacy and health service use; (2) children will demonstrate improved general health and functioning; and (3) siblings will have improved mental health and behaviour. In developing this trial, we found a relative absence of detail in the international home visiting literature on process details for intervention and evaluation. This protocol therefore provides detailed description of the intervention, study design and research methods employed. Methods and design Study design The right@home trial is an RCT of SNHV from pregnancy to child age 2 years, compared with usual care. It is conducted as a superiority trial with two parallel groups and a primary end point at child age 2 years. Parental care of the child, parental responsivity to the child and the home learning environment are evaluated at 2 years by researchers who are blinded to intervention status. Setting This is a multisite trial conducted in the states of Victoria and Tasmania in Australia where each state is responsible for the delivery of their local healthcare system. Participants are recruited from the public maternity hospitals servicing four local government areas (councils) in Victoria (Ballarat, Dandenong, Frankston, Whittlesea) and three regions in Tasmania (South, North, North West). The trial regions are selected for their high prevalence of families experiencing socioeconomic and psychosocial adversity, a mix of metropolitan and regional areas, and interest from the universal CFH services in participating in the trial. Participants Eligible participants are pregnant women attending the antenatal clinics from May 2013 to August 2014: With expected due dates before 1 October 2014, Are <37 weeks gestation, Have sufficient English proficiency to verbally answer interview questions, Have 2 or more of 10 risk factors identified by risk factor screening (see Recruitment and eligibility section and table 2), and Have home addresses within the travel boundaries specified by the local councils/regions managing the intervention nurses. Women are excluded if they: Are enrolled in the Tasmanian Department of Health and Human Services CU@Home visiting programme (for first-time mothers aged 15–19 years), Do not comprehend the recruitment invitation (eg, have an intellectual disability such that they are unable to consent to entering the study, or have insufficient English to complete face-to-face assessments), or Have no mechanism for contact (landline or mobile telephone, or email address). Recruitment and eligibility Trained researchers are provided with scripts and approach all women in the waiting rooms of antenatal clinics, inviting them to complete the hardcopy risk factor survey (table 1). Women with insufficient English proficiency for participation are most frequently identified before completing the screening survey, although a proportion is identified afterwards. The majority of women who are ineligible due to other exclusion criteria are most often identified after completing the screening survey (eg, after reporting their local postcode). When the antenatal clinics are extremely busy, the researchers cannot invite all women to complete the survey. Researchers then verbally assess eligibility regarding gestation and address before offering the survey. To accommodate low literacy, researchers ask women if they would like the survey read to them or to complete it alone. Researchers check all surveys for completeness. Surveys were collected until the necessary sample size for the RCT was reached and the recruitment period ended. Eligibility based on risk is defined as 2 or more of 10 risk factors as identified by a pilot study of the recruitment processes conducted in February to March 2013. The risk factor measures are summarised in table 2. Table 1 Graphical depiction (Perera diagram) of the trial components shared and unique to the trial arms RCT, randomised controlled trial; SNHV, sustained nurse home visiting. Table 2 Description of measures Item Description Screening criteria to establish eligibility at waiting room survey*  Young pregnancy Calculated from year of birth and dichotomised into ‘<23’ vs ‘≥23 years’22  Living with another adult ‘Yes’ vs ‘no’  Support in pregnancy Anyone supporting participant through pregnancy, for example, financially, emotionally or practically? (This could be a partner/husband, relative or friend) (yes/no)  Smoking ‘Yes’ vs ‘no’  Global health Single 5-point item (‘poor’ to ‘excellent’) from the self-reported SF6,23 dichotomised into ‘poor/fair/good’ vs ‘very good/excellent’24  Long-term illness Health problem or disability that limits daily activities (yes/no), drawn from the UK 2001 Census25  Anxious mood Matthey 2-item Generic Mood Question, which has shown good correlation with longer, validated anxiety measures including the EPDS and the Hospital Anxiety Depression Scale26  Education Highest level of school completed, dichotomised into ‘<year 12’ vs ‘completed year 12’ reflecting completion of secondary level education in Australia  Income Person in household who currently has paid work/earns an income (yes/no)  Ever worked Participant has ever had a job before (yes/no) Primary outcomes collected at 2 years  Regular meal times Single 5-point item (‘never’ to ‘always)’. Study designed based on Sleep Well Be Well Regular Bedtime item27  Food choices 12-item measure of food choices over past 24 hours, rated on a 3-point scale (not at all/once/more than once), drawn from LSAC28  Regular bedtime Single 5-point item (‘never’ to ‘always’), adapted from the ‘Sleep Well Be Well’ study27  Regular bed routine Single 5-point item (‘never’ to ‘always’), drawn from the ‘Sleep Well Be Well’ study27  Safety of the environment Items assessing aspects of home safety, which are dichotomised into ‘safe’ vs ‘not safe’. Study designed based on Royal Children's Hospital Safety Centre and KidSafe checklists30 31  Warm parenting 6-item measure assessing parental warmth. Items rated on a 5-point scale (‘never/almost never’ to ‘always/almost always’), drawn from LSAC29  Hostile parenting 5-item measure assessing parental hostility. Items rated on a 10-point scale (‘not at all’ to ‘all of the time’), drawn from LSAC29  Parent responsivity and the home learning environment HOME:32 45-item measure comprised of observation and parent report, assessing the quality and quantity of stimulation and support available to a child in the home environment. Items dichotomised (‘not observed or reported’ vs ‘observed and/or reported’) and summed. Continuous total score ranging 0–45, with higher scores indicating a better home environment. 6 subscale scores: parental responsivity (11 items), acceptance of the child (8 items), organisation of the environment (6 items), learning materials (9 items), parental involvement (6 items), variety in experience (5 items) Secondary outcomes at 2 years  Child ever breast fed Single item ‘yes’ vs ‘no’  Duration of breast feeding Age in months at which breast feeding stopped  Age started solids Age in months  Drink choices apart from milk/formula Child regularly has drinks other than milk or formula ‘yes’ vs ‘no’; if yes, list of regular drink choices28  Feeding problem Single 4-point item report of child feeding problems, dichotomised into yes (moderate/severe) vs no (none/mild), study designed based on LSAC sleep problem item28  Child ate breakfast today Single item ‘yes’ vs ‘no’, drawn from LSAC28  Sleep problems Single 4-point item report of child sleep problems, dichotomised into yes (moderate/severe) vs no (none/mild), drawn from LSAC28  Child–parent relationship Single 5-point item (‘poor’ to ‘excellent’), study designed based on the single global health item drawn from the self-reported SF6  Parenting efficacy 4-item parenting efficacy scale. Items rated on a 10-point scale (‘not at all how I feel’ to ‘exactly how I feel’) drawn from LSAC, and a single 5-point parenting efficacy item assessing mother's feelings about herself as a parent (‘not very good’ to ‘very good’) drawn from LSAC29  Global health See description in Screening measures above  Maternal mental health DASS:33 21-item measure, rated on a 4-point scale (‘not at all’ to ‘most of the time’) assessing the negative emotional states of depression, anxiety and tension/stress. Three subscales (7 items each): depression, anxiety and stress scales  Life satisfaction Personal Well-being Index (International Well-being Group, 2013):34 single item assessing general life satisfaction, and 8 items assessing satisfaction with specific life domains, rated a 10-point scale (‘no satisfaction at all’ to ‘completely satisfied’)  Locus of control 3 items assessing mother's self-efficacy or locus of control, drawn from the UK Millennium Cohort35  Maternal quality of life AQoL-8D:36 37 35-item measure assessing health-related quality of life. Provides a single overall utility-based quality of life measure, and 8 separately scored scales: independent living, happiness, mental health, coping, relationships, self-worth, pain, senses, which can be totalled into 2 super dimension scales: physical and psychosocial  Smoking ‘yes’ vs ‘no’  Current education Mother currently undertaking study or training ‘yes’ vs ‘no’; if yes, type of qualification, drawn from LSAC.28  Current employment Mother currently employed ‘yes’ vs ‘no’; if yes, type of employment and age of child when mother returned to work.  Maternal stress Hair cortisol as a measure of maternal stress response over the past 3 months. The hair sample is a minimum length of 3 cm, with the total density of the sample equating to approximately half a pencil's width (30–50 mg). Cortisol concentrations to be reported as a continuous measure38  Child global health See description in Screening measures above, collected for child and parent  Child stress Hair cortisol, see description for maternal stress above  Communication and symbolic behaviour CSBS:39 6 items from the CSBS Developmental Profile Infant/Toddler Checklist (item numbers 7, 8, 11, 13, 18, 24) selected to assess child communication, which have not shown a ceiling effect when assessed at child age 2 years in a population cohort of child language.40 Items rated on a 3-point scale (not yet/sometimes/often)  Maternal–child interactions Drawn from fine analysis of maternal–child video, identifying maternal responsive behaviours (eg, expansions, imitations, questions, labels, directives) that are associated with child language outcomes, adapted from the work by Levickis et al40  Sibling mental health and behaviour 25-item Strengths and Difficulties Questionnaire (4–10 years old version):41 42 assesses emotional and conduct behaviour, and total difficulties scores (higher scores indicate greater problem)  Parental enablement Modified Parent Enablement Index:43 6-item measure assessing parent enablement as a result of services provided by the child health nurse, rated on a 3-point scale (much better/better/same or less)  Parent satisfaction Modified Parent Satisfaction Questionnaire:44 45 10-item measure assessing parent satisfaction with services provided by the child health nurse, rated on a 5-point scale (‘strongly agree’ to ‘strongly disagree’) *Eligible participants report 2 or more of the 10 risk factors identified by risk factor screening. AQoL-8D, Assessment of Quality of Life-8D; CSBS, Communication and Behaviour Scales; DASS, Depression Anxiety and Stress Scales; EPDS, Edinburgh Postnatal Depression Scale; HOME, Home Observation for Measurement of the Environment; LSAC, Longitudinal Study of Australian Children; SF6, Short Form-6. Eligible women are invited into the RCT and the initial baseline visit is booked immediately where possible. Interested women are visited in their homes by trained researchers who collect informed consent and conduct a comprehensive baseline questionnaire including questions about their mental and physical health, psychosocial circumstances and pregnancy (table 3). The questionnaire is conducted face-to-face to minimise participant burden and the potential impact of low literacy. Table 3 Data collection schedule Antenatal Postnatal Measures Screening Base 6-week 6-month 12-month 18-month 24-month Screening  Young pregnancy ●  Gestation ●  Postcode/zip code ● ● ● ● ● ● ●  First child ●  Living with another adult ● ● ● ● ● ● ●  Support in pregnancy ●  Global health ● ● ● ● ● ● ●  Smoking ● ● ●  Long-term illness ●  Anxious mood ● ● ● ● ● ●  Education ● ● ● ●  Income ● ● ● ● ● ● ●  Ever worked ● Primary outcomes  Care outcomes   Regular meal times ● ● ●   Food choices ● ●   Regular bedtime ● ● ● ●   Regular bed routine ● ●   Safety of the environment ● ●  Responsivity outcomes   Warm parenting ● ●   Hostile parenting ● ●   Parent responsivity, acceptance, and involvement ▴● ▴●  Home learning environment outcomes   Home organisation of the environment, learning materials and variety ▴● ▴● Secondary outcomes  Care outcomes   Child ever breast fed ● ●   Duration of breast feeding ● ●   Age started solids ● ●   Drink choices apart from milk/formula ●   Feeding problem ● ●   Child ate breakfast today   Sleep problem ● ● ● ●  Responsivity outcomes   Child–parent relationship ● ● ● ●  Maternal outcomes   Parenting efficacy ● ●   Global health ● ● ● ● ● ● ●   Anxious mood ● ● ● ● ● ●   Maternal mental health ● ● ● ● ● ●   Life satisfaction ● ● ●   Locus of control ● ● ●   Maternal quality of life ● ● ●   Maternal stress + +   Smoking ● ● ●   Current employment ● ●   Current education ●  Child outcomes   Child global health ● ● ● ●   Child stress +   Communication and symbolic behaviour ●   Maternal–child interactions ▴ ▴  Impact   Parental enablement ● ●   Parent satisfaction ● ●  Sibling outcomes   Sibling mental health and behaviour ● ● ● ●Collected by parent report; ▴collected by observation; +collected by hair sample. Randomisation Sequence generation: Participants are randomly assigned to either usual care (control) or programme (intervention) arm with a 1:1 allocation following a computer-generated randomisation schedule stratified by site (Ballarat Hospital, Dandenong Hospital, Frankston Hospital, Northern Health, Launceston General Hospital, Hobart Royal Hospital, Northwest Regional Hospital) and parity (first time parent vs those with children already) using permuted blocks of sizes 2, 4 or 6. Allocation concealment mechanism and implementation: Participants are randomised using a study-designed, online, central randomisation service. To ensure baseline allocation concealment, the project coordinator does not randomise until the participant is recruited into the trial; that is, after the participant provides informed consent, all the inclusion and exclusion criteria are addressed, the eligibility of the participant is confirmed and all baseline measurements are completed. At the end of each baseline assessment, the researcher checks in with project coordinator who attempts to contact each participant to conduct randomisation allocation almost immediately. Usual care Families in participating sites are provided with well-child checks which are delivered by community-based nurses. These checks are available to all families from birth until 5 years at no out-of-pocket cost. In the first 2 years, all families are offered six (Tasmania) or nine checks (Victoria). The first visit occurs in families' homes and successive visits occur at a local centre. Nurses also provide a needs-based ‘enhanced’ service, which involves additional home and/or centre visits. Eligibility for the enhanced model of care is decided by the local area's CFH service. This ‘usual service’ provides the comparator for the right@home intervention. Intervention The right@home SNHV programme is designed as an integral element of the universal CFH services in Victoria and Tasmania. This approach reflects the concept of proportionate universalism, where action to address social gradients in child health and development ‘must be universal, but with a scale and intensity that is proportionate to the level of disadvantage’46 (ie, higher dose and intensity for higher need). Development of the intervention: The intervention programme was developed based on a series of three literature reviews conducted to inform the trial,19 47 48 which addressed the overarching question: What features of an SNHV programme are likely to bring about improved learning and development outcomes for young children whose families could benefit from greater support? The original intention was not to undertake three literature reviews (only the first); however, the findings of each highlighted the need to undertake a subsequent review with a different focus. The first literature review concluded that there is limited knowledge about what makes SNHV programmes effective. Factors that did emerge as important for impact were programmes being delivered by a more professionally skilled workforce, visits starting in the antenatal period, being offered over a longer period of time and being offered to the families experiencing greatest adversity or complexity.49 The most effective SNHV programmes are those targeted to populations identified as most likely to benefit from additional support.49 The second literature review then investigated factors associated with successfully working with families experiencing adversity. From a range of disciplines the review determined that, to be successful, programmes must involve a partnership between the family and nurse, focus on goals that parents prioritise, build competencies, be non-stigmatising and maintain continuity of care.48 The third review went on to investigate specific evidence-based interventions that focused on this trial's primary outcome areas and had the potential to enhance the effectiveness of SNHV programmes. The outcome areas were derived from reviewing the early childhood evidence which highlighted the importance of the home learning environment,50 parent responsivity and language development40 and the differential effect of adversity on executive functioning and therefore self-regulation.51 All of these areas are considered necessary precursors for optimising children's learning and development trajectories. Intervention overview: The right@home intervention is structured around the core MECSH framework and programme.17 18 This core is bolstered by five evidence-based strategies for content and two for the process of delivery, termed ‘focus modules’ (see description of content below).47 Taking heed of Segal et al's20 review demonstrating the importance of programme logic for effective home visiting programs, the intervention content is selected to align with evidence of impact on the primary objectives. The intervention schedule includes a minimum of 25 home visits offered to the woman, primarily by the same specially trained right@home nurse. Three visits are scheduled antenatally, with the remainder during the first two years postbirth. The actual number of antenatal home visits that a woman receives is determined by gestation and may vary. For example, a woman recruited at 20 weeks gestation should receive the three visits. If a woman is recruited later, more frequent visits may be offered to catch up. After 36 weeks gestation, one antenatal visit is scheduled if possible, unless it is appropriate to delay until the very early postnatal period. Postnatal visits are scheduled to occur within 1 week of birth; at least weekly until 6 weeks; fortnightly until 12 weeks; 3-weekly to 6 months; 6-weekly to 12 months and bi-monthly until 2 years. In preparation for discharge from right@home, families are assisted to re-enter the usual care service, which is available until child age 5. Within the right@home intervention, the nurses incorporate the well-child checks that are delivered via usual care (described above) into the home visits, ensuring right@home builds on the universal platform. Intervention staffing: Nurses are recruited from the usual care service and trained to deliver the right@home programme. The right@home nurses must be qualified CFH nurses; that is, Baccalaureate-registered nurses (or equivalent) with postgraduate qualifications in CFH, who have also completed Family Partnership Model Training, online and face-to-face training in the core MECSH programme, and additional training in the right@home focus modules. Line managers provide nurses with clinical supervision. Each nurse is expected to receive a minimum of 1 hour per month of reflective practice supervision—this may be in a group or individually—and is ideally facilitated by someone other than the line manager. In addition to reflective practice supervision, there is case review, where each family is reviewed by the right@home clinical team, and additionally any other clinical professionals relevant to the cases being reviewed, at least once every 6 months. This occurs through scheduled monthly case conference meetings. Each site has a dedicated social care practitioner, who is a member of the programme team. There is one full-time social care practitioner per 100 families in the programme. The role of the social care practitioner is to provide support for the nursing team and psychosocial support for the families, such as brief counselling interventions, and instrumental support, including advocating for and assisting families with housing, service access and financial issues. Intervention content: The following evidence-based strategies contribute to the ‘focus modules’ that are aligned with the primary outcomes: Parental care of the child: keeping children safe within a less chaotic and more structured environment (eg, feeding and sleeping routines) promotes self-regulation, decreases rates of injury and is importantly related to executive functioning and school success.52 The nurse goes through the following with families at scheduled time points, and reinforces the content as necessary during the course of the intervention. Safety: nurse-led KidSafe audit30 of the internal and external safety of the child's home; Sleep: from 0 to 6 months: anticipatory guidance on normal infant sleep and positive bedtime routines; from 6 months onwards: a behavioural sleep intervention;53 54 Nutrition: ‘Get up and Grow’ healthy eating guidelines.55 Parent responsivity: the construct of responsivity incorporates both bonding with the child and promoting language (ie, is also related to the home learning environment (the third primary outcome)). Neuroscience suggests that infant brain development related to emotional attachment is most rapid in the first 12 months of life and predictive of infants' ongoing social and emotional development.51 Parental verbal responsivity is also strongly predictive of child's vocabulary and language.40 ‘Promoting First Relationships’56 programme of materials and activities for parents promoting secure and healthy relationships with their children. Home learning environment: research has shown that the home learning environment (including aspects like the number of books in the home, and activities like reading stories and recognising numbers and shapes) independently predicts school outcomes. Importantly this research demonstrated that the home environment promoted children's learning and development regardless of socioeconomic status.50 ‘Learning to Communicate’ programme,57 from 0 to 12 months and a modified version of the ‘smalltalk’ programme,58 from 13 to 24 months, to enhance the ability of parents to provide appropriate stimulation for their babies, which will facilitate their development. While some of the above content (eg, sleep, safety, nutrition) and supports are provided in the usual care system, they are not provided systematically as in right@home. In this programme, although the focus modules are designed to be implemented at specific developmental points, nurses still structure each home visit flexibly to best address each mother's needs, skills, strengths and capacity. They are guided by a strengths-based approach and joint goal setting, an integral part of the Family Partnership Training and aligned with our literature review findings.48 The nurse supports and enables the mother and the family to: ▸ Enhance their coping and problem-solving skills, and ability to mobilise resources; ▸ Foster positive parenting skills; ▸Support the family to establish supportive relationships in their community; ▸Mentor maternal–infant bonding and attachment; and ▸ Provide proactive primary healthcare and anticipatory health education, including but not limited to evidence-based information regarding immunisation, Sudden Infant Death Syndrome (SIDS) risk reduction, infant nutrition and child safety. Nurses use an additional two ‘process’ focus modules—video feedback and motivational interviewing strategies—to help parents instigate behavioural change.47 Nurses and the social care practitioners also help parents access early childhood health services, volunteer home visiting services and family support services; hold group activities specifically for intervention families; and link women into community activities, as needed. The key differences between the right@home intervention and the usual care are: Home visiting starting antenatally; Continuity of care by the same nurse throughout the 2½-year programme; Care by nurses with additional training in the programme model; Postnatal home visiting programme to the child's second birthday including: the MECSH structured programme; well-child checks; proactive (rather than needs-based) preventive and anticipatory primary healthcare and health education; and standardised focus modules aligned with primary outcomes; Dedicated social care practitioner in the team; Group activities specifically tailored for the right@home families. Intervention fidelity: For the purpose of this study, dose refers to the number of visits from a nurse for each intervention participant. For the delivery of scheduled programme content, nurses and/or social care practitioners enter data into the research database following each visit via ‘checklists’, indicating the occasion, duration and content delivered in the session. The quality of the intervention, including dose, client retention and delivery of programme content is systematically monitored by the MECSH Support Service at the Western Sydney University through quarterly review of programme delivery and feedback on performance to the participating sites. Blinding: The research managerial staff, the participants and nurse teams delivering the intervention are aware of the allocation to treatment arm. Control clients will be on the caseload of usual care nurses. Intervention clients may also, on occasion, access the usual care service. Great care is taken to prevent usual care nurses knowing which specific clients are in the study; however, they will be aware that the study is underway and some of their clients may be in the study. At each site a nominated ‘special contact’, usually the nurse unit manager is informed of all research participants and their intervention or control allocation. The special contact is the only person who knows all of the participants. Research staff that are responsible for conducting outcome assessments are blinded to treatment allocation. Families are asked not to disclose their randomisation status at assessments. If the research staff are unblinded at face-to-face assessments (see below), then the unblinding is recorded in the study database and attempts are made to organise the next annual assessment with another researcher. Researchers complete phone assessments with participants from different sites, that is, the ones they do not complete face-to-face assessments with, to minimise the opportunities for unblinding. Any data cleaning, coding and/or analysis undertaken by the data managers and statisticians excludes randomisation variables to maintain blinding until all 2-year data are collected. Emergency unblinding should not be necessary as intervention families, healthcare staff and senior study staff are aware of randomisation status. Assessments All assessments are conducted via participant–researcher interviews. At baseline and 1 and 2 years, interviews are conducted face-to-face in the participant's home. At 6 weeks, 6 and 18 months, interviews are conducted via telephone. All questionnaires except the initial screening survey are developed to be collected electronically on tablets. Women are able to voluntarily skip questions. Paper versions of assessments are provided in the case of electronic/technical malfunction, or if the woman cannot complete or declines a home visit but is happy to complete a hardcopy version. Described in detail in tables 1 and 2, the assessment at 2 years takes place in women's homes and includes measures of the primary outcomes and secondary outcomes. The procedure comprises: (1) standardised interview and observational assessment of the majority of outcome measures, (2) videoing of maternal–child interactions for later analysis, and (3) sampling of hair for cortisol testing. Methods for retention Researchers make every reasonable effort to follow each participant for the entire study period, recognising the importance of retention in maintaining the sample size, generalisability and comparability between the groups randomised to the intervention programme or usual care. The right@home families are considerably mobile, thus our sample size has been calculated for a retention rate of 60% of mothers until the assessment of the primary outcomes at age 2 years. Retention is promoted in the following ways: Maintaining regular contact with brief phone interviews at 6 weeks, 6 and 18 months in between the face-to-face assessments; Distribution of reminder postcards before each assessment; Reminder phone calls and text messages before face-to-face assessments; Distribution of end-of-year newsletters and seasons' greeting cards; Giving a $30 gift card for a national supermarket chain (excluding alcohol and tobacco purchases) as a token of appreciation for each of the three face-to-face assessments completed; Recording up to two alternate contacts for each participant, who the research team can contact in the case that they lose contact with the participant; and Consent from women to contact the Australian Department of Human Services for their updated contact details recorded by the Centrelink programme (an agency that provides a rebate for childcare costs, as well as means-tested social support and unemployment benefits). Data management All participants and nurses are given unique numerical identifiers (an ID code) for use throughout the study. A single, secure, purpose-built online electronic database (using Umbraco software) is used to record and store all participant and nurse details. Video data (collected by intervention nurses to conduct video feedback with families and by researchers at the face-to-face follow-up assessments) are uploaded as electronic files to external hard drives that are securely stored with written materials and hair samples in locked filing cabinets. Following Human Research Ethics Committee (HREC) storage requirements, all project materials are stored for the required period of time, that is, indefinitely if the participant consents to providing their data for data pooling or, otherwise, until the youngest participant is 25 years old. After that time, hardcopy materials will be destroyed by shredding, and any password-protected electronic archives are permanently deleted. After hair samples are tested at an external laboratory, they are destroyed according to the laboratory's protocol. Sample size calculation Existing SNHV trials show relatively modest effects (effect sizes of 0.2–0.4 SDs) for outcomes such as child mental health and behaviour, and cognitive and language development, from infancy to mid-childhood.19 While effect sizes of 0.25–0.3 SDs can be meaningful and impactful at the whole of population level,59 targeted public health interventions such as SNHV include a cost and intensity such that larger effects in the short-to-medium term might be necessary to justify implementation at a population level. We do note the longer term cost-benefits that have been achieved despite the more limited short-term benefits. Interestingly, previous home visiting RCTs have rarely published sample size calculations. Given the primary objectives of the trial and measures collected in existing RCTs,18 49 we chose to anchor our sample size calculation around detection of a minimum effect size of 0.3 for the responsivity subscale of the Home Observation Measurement of the Environment (HOME) Inventory (see table 2), to allow comparisons with the original MECSH trial and other international SNHV programmes.60 The sample size applies across all of the subscales of the HOME Inventory and other continuous outcomes as based on number of SDs rather than the actual outcome distributions. This is the first SNHV RCT to account for the potential effect of clustering in relation to the impact of each nurse on a group of women. The sample size was calculated twice: originally based on the expected nurse staffing for the intervention arm (n=14) and then based on the finalised staffing (n=18). The revised calculation in June 2014 considered the final staffing load of 18 nurses in the intervention arm and 18 pseudo-clusters in the usual care arm. To detect a minimum effect size of 0.3 with 80% power at the 0.05 significance level and assuming a modest average intraclass correlation of 0.02 within the clusters, the total sample size, allowing for attrition of 40% by 2 years, was N=714 (ie, n=357 in each arm). The anticipated attrition rate is based on results from other SNHV studies.49 Statistical analyses The baseline characteristics of the mothers will be presented for each treatment arm using descriptors such as the mean, SDs, median and IQR for continuous data and proportions for categorical data. So that the effects attrition may have on the study findings may be considered, comparability between mothers participating at baseline and those who completed follow-up to 2 years will be examined for each of the treatment groups. These analyses will be used to determine the selectivity and loss of representativeness resulting from sample attrition. Maternal and child outcome measures will be described by treatment arm. Comparisons will be made using regression models respecting the nature of the distribution of the outcomes, that is, linear regression for continuous or semicontinuous data, with presentation of mean differences and 95% CIs; and logistic regression for binary data with presentation of ORs and 95% CIs. Tobit regression will be used to confirm the sensitivity of linear regression to a non-normal distribution for outcomes with a censored normal distribution, and ordinal logistic regression for outcomes with up to five ordered categories. All regression analyses will be adjusted for study site and maternal parity in line with the stratification of the study randomisation. All regression analyses will also take account of any effects of the nurse (clustering), so that accurate effects of the intervention, regardless of child and family nurse delivering it, are estimated. Subsequent analyses will adjust for factors that may not be balanced by randomisation and that are associated with family and child outcomes. These analyses will take into account maternal baseline and child characteristics identified a priori, for example, child's gender and age (at assessment), and maternal age, parity, antenatal risk, self-efficacy, mental health, education and socioeconomic status. As noted in the UK Medical Research Council guidance described by Craig et al,61 it is recommended that multiple outcomes are considered in evaluating the effectiveness of interventions which are complex in nature and are likely to result in responses across a diversity of family and child domains. As such, each of the multiple outcomes will be analysed individually with interpretations made across the consensus of evidence provided. This will involve careful examination and consideration of the magnitude, direction and statistical significance of the responsiveness estimated for each outcome. In recognition of the increased potential for false-positive findings arising through analysis of multiple outcomes, findings will be interpreted cautiously and in context with one another rather than in isolation. Patterns and consistency in the responsiveness of outcomes, and the overall balance of the evidence, will be examined rather than isolated findings which may have arisen by chance. It is particularly important that sufficient data are presented to enable comparability across SHNV programmes because of the complexity of this type of intervention and likely influence across multiple domains, and the extent to which SNHV programmes vary in their content, setting and target population. Subgroup analyses: We will examine whether there is evidence that the intervention effect is modified for subgroups within the trial participants using tests of interaction between intervention and child and family factors as follows: parity (first-born vs other), antenatal risks (2 vs 3 or more risk factors at screening), maternal mental health at baseline (high vs low score)18 62 63 and self-efficacy at baseline (poor vs normal mastery)35 using the regression models described above with additional terms for interaction between subgroup and trial arm. Should any of these interaction terms reveal evidence that the intervention effect varies between these groups, specific subgroup estimates and CIs will be presented. As we have not powered the trial to consider subgroups, these analyses are considered exploratory. Per protocol analysis: In addition to the intention-to-treat analyses, we will conduct a per protocol analysis to examine how fidelity is related to effectiveness. In the intervention arm, fidelity is defined as having at least one antenatal visit and at least 19 visits in total with a right@home nurse during the course of the programme, that is, received 75% of the dose. In the usual care arm of the trial, fidelity is defined as having at least one visit with a CFH nurse and having fewer than 11 CFH nurse visits in total. This is to compare right@home full dose to usual care expected dose. Women will be excluded from the per protocol analysis if they either do not fulfil the definition of fidelity, or do not complete the 2-year follow-up researcher assessment, or the child is removed from the primary carer, or the family experiences a critical event (such as miscarriage, late termination of pregnancy, stillbirth or neonatal death, or own critical health event). Child removal will be defined as children who have spent <4 nights per week with a primary carer as reported at the 2-year survey. If parents have shared custody then they will be excluded if the child spends <4 nights per week in the participants' care. Missing data: The frequency and patterns of missing data will be examined and sensitivity analyses will be performed to compare the results of analyses restricted to families with complete data and analyses where those with missing data are considered using multiple imputation.64 65 Data monitoring: No data monitoring committee is needed for this study due to the known minimal risks. No interim analyses or stopping rules will be applied. Cost evaluation The economic analysis will use a cost consequences analysis from a government-as-payer perspective.66 It will compare any additional costs experienced over the first 2 years of children's lives in the intervention group compared with the usual care group, to the changes in the multiple outcome measures at 2 years described in table 2. Costs are based on the health resources used by women from recruitment to child age 2 years. Data on health resource use are available by provider (nurse and maternity hospital) administrative records and by women's recalled service use in 6-monthly interviews. Provider data include the number and type (home or clinic based) of visits attended in intervention and usual care, and referrals made. Women report their use of health and other services (referred or other) over the previous 6 months. Measured health resource use will be valued with standard unit costs (eg, award rates for nurse salaries, Medicare fee schedule for referred services) and presented in 2016 Australian dollars, with second-year costs discounted at 5%. The trial-based economic evaluation results will be expressed as the change in costs of the intervention compared with usual care, relative to the change in effects of the intervention over and above the usual care arm at 2 years. Study governance The study is governed by a tri-partite partnership between the Australian Research Alliance for Children and Youth, a national not-for-profit organisation (responsible for overall project management including nurse contracts), the Centre for Community Child Health at the Murdoch Childrens Research Institute, Royal Children's Hospital, Melbourne (responsible for research evaluation) and the Translational Research and Social Innovation team, Western Sydney University (responsible for implementation of intervention). Study partners meet face-to-face quarterly, have regular meetings with state government partners and provide regular reports to funders. Ethics, consent and permissions A condition of approval is that any proposed amendments to the project, including changes to the protocol, participant information and consent form/s and participant materials are submitted to the reviewing HRECs for approval before use. The managerial research staff make safety and progress reports to the HRECs at least annually and within 3 months of study termination or completion at each site. Consent: At the baseline home visit with the researcher, a signed consent form is obtained for each participant before any further survey administration. The consent form describes the purpose of the study, the procedures to be followed, and the risks and benefits of participation. The trained researcher conducts the informed consent discussion and checks that the participants comprehend the information provided and answer any questions about the study. Additional consents are collected throughout the study for aspects such as hair sampling and data linkage. To accommodate low literacy, women are offered the option of reading all consent invitations alone or reading it through with the researcher. Consent is voluntary and free from coercion. At all times it is made clear that non-participation in the study does not affect the usual routine clinical management offered by any health providers, for example, the care they receive from the hospital or as part of the standard CFH service. The researcher who conducts the consent discussion also co-signs the informed consent forms. A copy of the consent form is given to the participant. Participant consent to the study is documented in their record on the study's electronic database. Each participant can choose to stop participating in the nurse service (intervention or control depending on randomisation status) at any point. Participants who choose to stop participating in the nurse service will continue to be followed with the research assessments, unless they request to withdraw from the trial, in which case all research assessments will cease. Confidentiality: Participant confidentiality is strictly held in trust by the investigators, research staff, and the sponsoring institutions and their agents, and is extended to cover clinical information relating to participants. The study protocol, documentation, data and all other information generated are held in strict confidence and in ‘locked’ electronic files. No information concerning the study or the data is released to any unauthorised third party, without prior written approval of the sponsoring institution. Investigators and students have access to the final data set via permissions maintained by the data managers. Dissemination: The investigators and sponsor will communicate trial results to stakeholders, participants, healthcare professionals, the public and other relevant groups via presentations and publications. Discussion This is the first multisite, multijurisdictional, Australian RCT to examine the effectiveness of SNHV in improving parenting and the home learning environment, when delivered via the existing universal child and family healthcare platform. The trial has been established as a partnership between academia, government, non-government and philanthropic organisations. The intent is to achieve the best ‘real-life’ study, focusing on generalisability, and within the confines of practicalities and budget. Should the intervention prove effective and cost-effective, this approach provides the greatest opportunity for research translation and full-scale implementation. The trial has been designed with a number of methodological strengths. Recruitment is conducted via public hospitals, providing a study cohort that is representative of these target populations and thus generalisability of the findings to Australian families experiencing adversity. The intervention is built on those components for effective home visiting that were identified by a series of literature reviews (including the Australian MECSH study)18 47–49 and is being conducted based on programme logic linking the intervention to impact and outcomes.20 In 2014, we secured competitive funding from Australia's National Health and Medical Research Council, which will provide for follow-up to child age 5 years, allowing assessment of the effects and cost-effectiveness of the SNHV programme to school entry. At these older ages, assessments will incorporate more objective, face-to-face assessments of children's outcomes. Finally, the study governance arrangements allow for ‘arm's length’ evaluation of the intervention through a separate research organisation. There are some limitations. As the trial does not blind participants, outcome reporting may be influenced by maternal perception and feelings about being in the trial, health knowledge, and well-being. However, as the parent is most often the closest observer of the child, they are best placed to report on the child's immediate environment and behaviour67 and, at 2 years, parent report is the most feasible and powerful way to pick up any early signals that families and children are responding to the intervention. Direct observation measures like the HOME Inventory also help mitigate this limitation. The exclusion criteria mean the findings may not generalise to non-English-speaking women or women with severe intellectual disability. The former is a limitation of home visiting trials generally, as using interpreters and translators may alter the type of family–nurse partnership necessary for effective behavioural change. While we use a population-based sampling strategy for recruitment, women stop receiving the intervention if they move out of a study region. This could be avoided if the service is delivered across the participating states (ie, following the intention of the real-life design). This study is crucial for generating Australian evidence of an effective intervention to reduce the impact of social and environmental factors predisposing children to inequitable outcomes. The rigour and scope of this trial will make it possible to determine the effect of this comprehensive Australian SNHV programme. Despite the rhetoric regarding the benefit of SNHV, this is the first trial in Australia to test, at scale, the benefit and cost-benefit of an intervention programme that is delivered within the context of an existing (and therefore sustainable) universal health service system. In addition, the research and the intervention programme are being undertaken by two distinct organisations, with a third providing project management of the collaboration. This provides a more independent assessment of effectiveness than in many other SNHV trials where the research and implementation teams are the same. As such, this trial is a best practice implementation and evaluation model for professional home visiting in Australia. Addressing inequity in outcomes for children across health and education is an issue of timely and significant policy interest at a state and federal level.68 69 If right@home is effective and demonstrates benefit, the study design enables replicability at scale, with significant implications for the development of early childhood policy and strategy throughout Australia and internationally. The ‘right@home’ sustained nurse home visiting trial is a research collaboration between the Australian Research Alliance for Children and Youth (ARACY); the Translational Research and Social Innovation (TReSI) Group at Western Sydney University; and the Centre for Community Child Health (CCCH), which is a department of the Royal Children's Hospital and a research group of Murdoch Childrens Research Institute. The authors thank all families, the research assistants, and nurses and social care practitioners working on the right@home trial, the antenatal clinic staff at participating hospitals who helped facilitate the research, and the Expert Reference Group for their guidance in designing the trial. Contributors: The original study design was conceived by SG, AP, LK, LB, FM, LG and HH; SG, AP, HB, LK, CS, LB and DJ implemented the study design. CS and DJ, on behalf of the sponsor ARACY, is the funding holder. FM and FO provided statistical expertise in the trial design; HB, FM and FO are conducting the data cleaning; and FM and FO are conducting the statistical analysis. LG provided health economics expertise in the trial design and is conducting the cost evaluation analysis. All authors contributed to refinement of the study protocol and approved the final manuscript. Funding: This work is supported by the Victorian Department of Education and Training, the Tasmanian Department of Health and Human Services, the Ian Potter Foundation, Sabemo Trust, Sidney Myer Fund, the Vincent Fairfax Family Foundation, and the National Health and Medical Research Council (NHMRC, 1079418). The MCRI administered the research grant for the study and provided infrastructural support to its staff but played no role in the conduct or analysis of the trial. Research at the MCRI is supported by the Victorian Government's Operational Infrastructure Support Programme. SG was supported by an NHMRC Career Development Award (1082922). FM was supported by NHMRC Early Career and Career Development Fellowships (1037449 and 1111160). LG was supported by a NHMRC Early Career Fellowship (1035100). HH was supported by an NHMRC Career Development Award (607351). Competing interests: None declared. Ethics approval: This study is approved by the Human Research Ethics Committees of: the Royal Children’s Hospital (HREC 32296); Peninsula Health (HREC/13/PH/14); Ballarat Health Services (HREC/13/BHSSJOG/9); Southern Health (HREC 13084X); Northern Health (HREC P03/13) in Victoria, Australia; and the University of Tasmania (HREC H0013113), Tasmania, Australia. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Unpublished data will be available for sharing once core data are published and data sharing agreements are formalised with both parties. ==== Refs References 1 Nicholson JM , Lucas N , Berthelsen D Socioeconomic inequality profiles in physical and developmental health from 0–7 years: Australian National Study . J Epidemiol Community Health 2012 ;66 :81 –7 . 10.1136/jech.2009.103291 20961874 2 Béatrice N , Lise G , Victoria ZM Longitudinal patterns of poverty and health in early childhood: exploring the influence of concurrent, previous, and cumulative poverty on child health outcomes . BMC Pediatr 2012 ;12 :141 .22947499 3 Blackburn CM , Spencer NJ , Read JM Is the onset of disabling chronic conditions in later childhood associated with exposure to social disadvantage in earlier childhood? A prospective cohort study using the ONS Longitudinal Study for England and Wales . BMC Pediatr 2013 ;13 :101 10.1186/1471-2431-13-101 23802581 4 Centre for the Study of Social Policy W, DC. Centre for the Study of Social Policy . Results-based public policy strategies for reducing child poverty . Washington DC : Centre for the Study of Social Policy , 2014 http://www.cssp.org/policy/2014/Results-Based-Public-Policy-Strategies-for-Reducing-Child-Poverty.pdf (accessed 4 Nov 2015 ). 5 Duncan G , Magnuson K The importance of poverty early in childhood . Policy Q 2013 ;9 :12 –17 . 6 Gibb SJ , Fergusson DM , Horwood LJ Childhood family income and life outcomes in adulthood: findings from a 30-year longitudinal study in New Zealand . Soc Sci Med 2012 ;74 :1979 –86 . 10.1016/j.socscimed.2012.02.028 22486841 7 Hirsch D Estimating the costs of child poverty . York, UK : The Joseph Rowntree Foundation , 2008 https://www.jrf.org.uk/report/estimating-costs-child-poverty (accessed 4 Nov 2015 ). 8 Lamb ME Mothers, fathers, families, and circumstances: factors affecting children's adjustment . Appl Dev Sci 2012 ;16 :98 –111 . 9 Pavalko EK , Caputo J Social inequality and health across the life course . Am Behav Sci 2013 ;57 :1040 –56 . 10 Schoon I , Jones E , Cheng H Family hardship, family instability, and cognitive development . J Epidemiol Community Health 2012 ;66 :716 –22 . 10.1136/jech.2010.121228 21507894 11 Hanson JL , Hair N , Shen DG Family poverty affects the rate of human infant brain growth . PLoS ONE 2013 ;8 :e80954 10.1371/journal.pone.0080954 24349025 12 CCCH TICHR . A snapshot of Early Childhood Development in Australia: AEDI National Report 2012 . Canberra , 2012 . 13 Woolfenden S , Goldfeld S , Raman S Inequity in child health: the importance of early childhood development . J Paediatr Child Health 2013 ;49 :E365 –E9 . 10.1111/jpc.12171 23551940 14 United States of America, Department of Health and Human Services (DHHS) . Home visiting evidence of effectiveness (HOMVEE) , 2014 http://homvee.acf.hhs.gov (accessed 4 Nov 2015 ). 15 Avellar SA , Supplee LH Effectiveness of home visiting in improving child health and reducing child maltreatment . Pediatrics 2013 ;132 (Suppl 2) :S90 –9 . 10.1542/peds.2013-1021G 24187128 16 Schmied V , Donovan J , Kruske S Commonalities and challenges: a review of Australian State and Territory maternity and child health policies . Contemp Nurse 2011 ;40 :106 –17 . 10.5172/conu.2011.40.1.106 22545909 17 Kemp L , Harris E , McMahon C Miller Early Childhood Sustained Home-visiting (MECSH) trial: design, method and sample description . BMC Public Health 2008 ;8 :424 10.1186/1471-2458-8-424 19113994 18 Kemp L , Harris E , McMahon C Child and family outcomes of a long-term nurse home visitation programme: a randomised controlled trial . Arch Dis Child 2011 ;96 :533 –40 . 10.1136/adc.2010.196279 21429975 19 McDonald M , Moore T , Goldfeld S Sustained home visiting for vulnerable families and children: a literature overview . Melbourne : Centre for Community Child Health (CCCH) & Murdoch Children's Research institute (MCRI) , 2012 . 20 Segal L , Sara Opie R , Dalziel KIM Theory! The missing link in understanding the performance of neonate/infant home-visiting programs to prevent child maltreatment: a systematic review . Milbank Q 2012 ;90 :47 –106 . 10.1111/j.1468-0009.2011.00655.x 22428693 21 Robling M , Bekkers MJ , Bell K Effectiveness of a nurse-led intensive home-visitation programme for first-time teenage mothers (Building Blocks): a pragmatic randomised controlled trial . Lancet 2016 ;387 :146 –55 . 10.1016/S0140-6736(15)00392-X 26474809 22 Kiernan KE Becoming a young parent: a longitudinal study of associated factors . Br J Sociol 1997 ;48 :406 –28 .9372635 23 Sanson AV , Hawkins MT , Misson S The development and validation of Australian indices of child development—part II: validity support . Child Indicators Res 2010 ;3 :293 –312 . 24 Mensah FK , Hobcroft J Childhood deprivation, health and development: associations with adult health in the 1958 and 1970 British prospective birth cohort studies . J Epidemiol Community Health 2008 ;62 :599 –606 . 10.1136/jech.2007.065706 18559442 25 ONS. UK Office for National Statistics . Census 2001 Key Statistics 21: long term illness. UK Census . London : Office for National Statistics , 2001 http://data.london.gov.uk/dataset/census-2001-key-statistics-21-long-term-illness (accessed 18 Dec 2015 ). 26 Matthey S , Valenti B , Souter K Comparison of four self-report measures and a generic mood question to screen for anxiety during pregnancy in English-speaking women . J Affect Disord 2013 ;148 :347 –51 . 10.1016/j.jad.2012.12.022 23380518 27 Quach J , Gold L , Arnup S Sleep well—be well study: improving school transition by improving child sleep: a translational randomised trial . BMJ Open 2013 ;3 :e004009 10.1136/bmjopen-2013-004009 28 Australian Institute of Family Studies, Growing Up in Australia: The Longitudinal Study of Australian Children. http://growingupinaustralia.gov.au/ (accessed 11 May 2016). 29 Zubrick SR , Lucas N , Westrupp EM Parenting measures in the Longitudinal Study of Australian Children: construct validity and measurement quality, Waves 1 to 4 . Canberra : Department of Social Services , 2014 . 30 Kidsafe Child Accident Prevention Foundation of Australia . Home Safety Checklist. https://www.kidsafevic.com.au/images/stories/pdfs/Safety%20Checklist.pdf (accessed 11 May 2016) . 31 The Royal Children's Hospital, Melbourne. Safety Centre, Home Safety Checklist. http://www.rch.org.au/uploadedFiles/Main/Content/safetycentre/120544%20Home%20Safety%20Checklist%20A4.pdf (accessed 11 May 2016 ). 32 Caldwell BM , Bradley RH Home observation for measurement of the environment: administration manual . Tempe, AZ : Family & Human Dynamics Research Institute, Arizona State University , 2003 . 33 Lovibond PF , Lovibond SH The structure of negative emotional states: comparison of the Depression Anxiety Stress Scales (DASS) with the Beck Depression and Anxiety Inventories . Behav Res Ther 1995 ;33 :335 –43 .7726811 34 International Wellbeing Group . Personal Wellbeing Index . 5th edn Melbourne : Australian Centre on Quality of Life, Deakin University , 2013 http://www.acqol.com.au/iwbg/wellbeing-index/pwi-a-english.pdf (accessed 6 Jun 2016 ). 35 Kiernan KE , Mensah FK Maternal indicators in pregnancy and children's infancy that signal future outcomes for children's development, behaviour and health: evidence from the Millennium Cohort Study . York : University of York , 2009 . 36 Richardson J , Khan M , Iezzi A Preliminary results for the validation of the assessment of quality of life AQoL-8D instrument. Research paper 47 . Melbourne : Centre for Health Economics, Monash University , 2010 . 37 Richardson J , Iezzi A , Khan MA Validity and reliability of the Assessment of Quality of Life (AQoL)-8D multi-attribute utility instrument . Patient 2014 ;7 :85 –96 . 10.1007/s40271-013-0036-x 24271592 38 Stalder T , Kirschbaum C Analysis of cortisol in hair--state of the art and future directions . Brain Behav Immun 2012 ;26 :1019 –29 . 10.1016/j.bbi.2012.02.002 22366690 39 Wetherby AM , Prizant BM CSBS developmental profile infant/toddler checklist . Baltimore : Brookes Publishing , 2001 . 40 Levickis P , Reilly S , Girolametto L Maternal behaviors that promote early language acquisition in slow-to-talk toddlers: prospective community-based study . J Dev Behav Pediatr 2014 ;35 :274 –81 . 10.1097/DBP.0000000000000056 24748010 41 Goodman R The Strengths and Difficulties Questionnaire: a research note . J Child Psychol Psychiatry 1997 ;38 :581 –6 .9255702 42 Goodman R , Scott S Comparing the strengths and difficulties questionnaire and the child behavior checklist: is small beautiful? J Abnorm Child Psychol 1999 ;27 :17 –24 .10197403 43 Howie JG , Heaney DJ , Maxwell M A comparison of a Patient Enablement Instrument (PEI) against two established satisfaction scales as an outcome measure of primary care consultations . Fam Pract 1998 ;15 :165 –71 .9613486 44 Armstrong KL , Fraser JA , Dadda MR Promoting secure attachment, maternal mood and child health in a vulnerable population: a randomized controlled trial . J Paediatr Child Health 2000 ;36 :555 –62 .11115031 45 Marshall GN , Hays RD The patient satisfaction questionnaire short-form (PSQ-18) . CA : Rand Santa Monica , 1994 . 46 Marmot M , Allen J , Goldblatt P Fair society, healthy lives: the Marmot review; strategic review of health inequalities in England post-2010 . Marmot Review , 2011 . 47 Moore T , McDonald M , Sanjeevan S Evidence-based service modules for a sustained nurse home visiting program: a literature review . Melbourne : Centre for Community Child Health (CCCH) & Murdoch Children's Research institute (MCRI) , 2012 . 48 Moore TG , McDonald M , Sanjeevan S Sustained home visiting for families and children: a literature review of effective processes and strategies. Prepared for Australian Research Alliance for Children and Youth . Parkville, Victoria : The Royal Children's Hospital Centre for Community Child Health, Murdoch Childrens Research Institute , 2012 . 49 McDonald M , Moore T , Goldfeld S Sustained home visiting for families and children: a review of effective programs. Prepared for Australian Research Alliance for Children and Youth . Parkville, Victoria : The Royal Children's Hospital Centre for Community Child Health, Murdoch Childrens Research Institute , 2012 . 50 Sylva K , Melhuish E , Sammons P Technical paper 12: the final report: effective pre-school education 2004 . 51 Farah MJ , Betancourt L , Shera DM Environmental stimulation, parental nurturance and cognitive development in humans . Development 2008 ;11 :793 –801 . 52 Kiernan KE , Mensah FK Poverty, family resources and children's early educational attainment: the mediating role of parenting . Br Educ Res J 2011 ;37 :317 –36 . 53 Hiscock H , Bayer J , Gold L Improving infant sleep and maternal mental health: a cluster randomised trial . Arch Dis Child 2007 ;92 :952 –8 . 10.1136/adc.2006.099812 17158146 54 Hiscock H , Bayer JK , Hampton A Long-term mother and child mental health effects of a population-based infant sleep intervention: cluster-randomized, controlled trial . Pediatrics 2008 ;122 :e621 –7 . 10.1542/peds.2007-3783 18762495 55 Australian Government Department of Health . Get up & grow: healthy eating and physical activity for early childhood . Commonwealth of Australia , 2009 . 56 Kelly JF , Zuckerman TG , Sandoval D Promoting first relationships: a program for service providers to help parents and other caregivers nurture young children's social and emotional development . Seattle, Washington : NCAST-AVENUW Publications , 2008 . 57 Anderson T Learning to communicate: a guide to infant communication development . Liverpool : South West Sydney Area Health Service , 1997 . 58 Parenting Research Centre . The Early Home Learning Study (EHLS) and smalltalk program and materials. http://www.smalltalk.net.au/ (accessed 11 May 2016 ). 59 Rose GA , Khaw KT , Marmot M Rose's strategy of preventive medicine: the complete original text . Oxford University Press , USA , 2008 . 60 Totsika V , Sylva K The home observation for measurement of the environment revisited . Child Adolesc Ment Health 2004 ;9 :25 –35 . 61 Craig P , Dieppe P , Macintyre S Developing and evaluating complex interventions: the new Medical Research Council guidance . BMJ 2008 ;337 :a1655 .18824488 62 Olds DL , Robinson J , O'Brien R Home visiting by paraprofessionals and by nurses: a randomized, controlled trial . Pediatrics 2002 ;110 :486 –96 .12205249 63 DuMont K , Mitchell-Herzfeld S , Greene R Healthy Families New York (HFNY) randomized trial: impacts on parenting after the first two years . New York State Office of Children & Family Services , 2006 . 64 Lee KJ , Simpson JA Introduction to multiple imputation for dealing with missing data . Respirology 2014 ;19 :162 –7 . 10.1111/resp.12226 24372814 65 Sterne JA , White IR , Carlin JB Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls . BMJ 2009 ;338 :b2393 .19564179 66 Drummond MF Methods for the economic evaluation of health care programmes . Oxford University Press , 2005 . 67 Mensah FK , Kiernan KE Parents’ mental health and children's cognitive and social development . Soc Psychiatry Psychiatr Epidemiol 2010 ;45 :1023 –35 . 10.1007/s00127-009-0137-y 19823757 68 Productivity Commission, Childcare and Early Childhood Learning: Overview, Inquiry Report No. 73, 2014; Canberra. http://www.pc.gov.au/inquiries/completed/childcare/report/childcare-overview.pdf 69 Commonwealth of Australia. Investing in the early years –a National Early Childhood Development Strategy: an initiative of the Council of Australian Governments. 2009; Canberra. http://www.startingblocks.gov.au/media/1104/national_ecd_strategy.pdf

PMC005xxxxxx/PMC5372046.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01478110.1136/bmjopen-2016-014781Renal MedicineResearch150617281722172517281698To dialyse or delay: a qualitative study of older New Zealanders’ perceptions and experiences of decision-making, with stage 5 chronic kidney disease Lovell Sarah 1Walker Robert J 2Schollum John B W 2Marshall Mark R 345McNoe Bronwen M 6Derrett Sarah 6 1 School of Health Sciences, University of Canterbury, Christchurch, New Zealand 2 Department of Medicine, University of Otago, Dunedin, New Zealand 3 Department of Renal Medicine, Counties Manukau Health, Auckland, New Zealand 4 School of Medicine, University of Auckland, Auckland, New Zealand 5 Medical Affairs, Baxter Healthcare (Asia) Pte Ltd, Singapore 6 Department of Preventive and Social Medicine, University of Otago, Dunedin, New ZealandCorrespondence to Professor Robert J Walker; rob.walker@otago.ac.nz2017 29 3 2017 7 3 e01478119 10 2016 30 1 2017 16 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Background Issues related to renal replacement therapy in elderly people with end stage kidney disease (ESKD) are complex. There is inadequate empirical data related to: decision-making by older populations, treatment experiences, implications of dialysis treatment and treatment modality on quality of life, and how these link to expectations of ageing. Study population Participants for this study were selected from a larger quantitative study of dialysis and predialysis patients aged 65 years or older recruited from three nephrology services across New Zealand. All participants had reached chronic kidney disease (CKD) stage 5 and had undergone dialysis education but had not started dialysis or recently started dialysis within the past 6 months. Methodology Serial qualitative interviews were undertaken to explore the decision-making processes and subsequent treatment experiences of patients with ESKD. Analytical approach: A framework method guided the iterative process of analysis. Decision-making codes were generated within NVivo software and then compared with the body of the interviews. Results Interviews were undertaken with 17 participants. We observed that decision-making was often a fluid process, rather than occurring at a single point in time, and was heavily influenced by perceptions of oneself as becoming old, social circumstances, life events and health status. Limitations This study focuses on participants' experiences of decision-making about treatment and does not include perspectives of their nephrologists or other members of the nephrology team. Conclusions Older patients often delay dialysis as an act of self-efficacy. They often do not commit to a dialysis decision following predialysis education. Delaying decision-making and initiating dialysis were common. This was not seen by participants as a final decision about therapy. Predialysis care and education should be different for older patients, who will delay decision-making until the time of facing obvious uraemic symptoms, threatening blood tests or paternalistic guidance from their nephrologist. Trial registration number Australasian Clinical Trials Registry ACTRN 12611000024943; results. ElderlyQUALITATIVE RESEARCHdecision-makingHealth Research Council of New Zealandhttp://dx.doi.org/10.13039/50110000150510/354 ==== Body Strengths and limitations of this study Decision-making with respect to renal replacement therapy was a fluid process influenced by their perception of age, social circumstances, life events and health status. Delayed decisions were very common and frequently misunderstood by nephrologists as opting for conservative care. Predialysis care and education needs to be different for older patients. Limitation: small numbers interviewed. Patient focus only does not include nephrologists' perspective and may have limited applicability to low-middle-income countries or non-Western countries. Introduction Population ageing and an ‘epidemic’ of end stage kidney disease (ESKD) among older people are driving demand for dialysis.1 The needs of this group are usually more complex than those of younger patients, given their increased prevalence of multimorbidity, subsequent frailty and functional dependence, and relatively short longevity.2 This situation limits the applicability of traditional paradigms for managing dialysis in the elderly, necessitating a more individualised approach.3–5 Core questions, intersecting age and clinical need, influence decisions of dialysis in the elderly—pertaining to the question of ‘if’ and also ‘when’. It is axiomatic that delivery of dialysis to the elderly incorporates patient-centred outcomes. Relevant research has highlighted concerns among elderly patients that dialysis will impede their freedom,4 that they, and their partners, can feel overwhelmed by the impact of dialysis on their lives, and the extent to which they must adjust to the imposition of dialysis.6 7 In addition, the relatively short longevity of older patients means that the impact of time spent dialysing or travelling to dialysis is different in the elderly compared with those who are younger.2 8 9 In recent years, these factors have increasingly influenced the healthcare delivery of dialysis to the elderly. For instance, conservative management, that is, the active management of symptoms without resorting to dialysis, has been increasingly adopted for elderly patients who are concerned about the burden of dialysis,10–12 as have assisted and unassisted home-based (as opposed to clinic-based or hospital-based) dialysis therapies.2 13 14 Such models of care, however, are often developed out of cumulative clinical experience, and the development of systematic approaches has been limited by the inadequate amount of empirical data for: decision-making by older populations, their treatment experiences, the implications of dialysis treatment and treatment modality on quality of life or expectations of ageing. This study examined the experiences of older adults (aged ≥65 years) living with chronic kidney disease (CKD) as they chose whether or not to begin dialysis or continue with conservative management. We focus on factors that influence decision-making about dialysis and dialysis modality among older New Zealanders with ESKD, within a health delivery system that imposes no restriction on modality selection through reimbursement or policy2 15 16 This context allows insight into unfiltered patient's perspectives of the progression of CKD, the decision to begin dialysis, and how patients' perceptions of ageing and health status may influence that decision-making process. Such data are critical for developing an evidence-based and patient-centred model of care, and implementing optimal programmatic measures in an effective manner. Methods This reporting of this study is based on the Consolidated Criteria for Reporting Qualitative Health Research (COREQ)17 and the Standards for Reporting Qualitative Research (SRQR).18 This study is registered with the Australasian Clinical Trials Registry ACTRN 12611000024943. Participant selection Participants for this study were selected from a larger quantitative study (dialysis outcomes for people with CKD aged ≥65 years (DOS65)). DOS65 takes a census-based approach, and includes all older CKD5 patients from three nephrology services across New Zealand, each serving different patient populations and varying in organisational and clinical models of dialysis delivery.19 The three units were Middlemore Hospital, Hawkes Bay and Southern Region. The Middlemore unit is a large urban centre providing in-centre, satellite and to a lesser extent home haemodialysis as well as peritoneal dialysis in South Auckland, an area with a high proportion of Māori and Pacific Island patients. Hawkes Bay is a smaller regional unit with a more rural population providing in-centre haemodialysis and peritoneal dialysis. The Southern Region is a medium-sized teaching hospital with an exclusively home-based dialysis programme (haemodialysis and peritoneal dialysis). Participants for this qualitative study were purposely selected from DOS65 if they had undergone dialysis education and had reached CKD stage 5 without yet starting dialysis or recently started dialysis within the past 6 months. Participants were approached by their physician at the time of recruitment into the DOS65 and invited to take part in this additional qualitative study. Study sites were selected to maximise ethnic diversity. Initial convenience sampling of eligible patients was modified due to rapid recruitment of male-intended/current peritoneal dialysers. Subsequent purposive sampling20 was used to allow us to capture diversity and balance in experiences and opinions among participants and enhance representation of current/intended patients with haemodialysis and non-dialysers. Participants did not receive any reimbursement for participating in the study. Initially, 27 participants were identified for invitation. Of these, five declined and another five were unable to be contacted or meet for the scheduled interview. As a longitudinal qualitative study, serial interviews (serial qualitative interviews (SQIs)) were undertaken with 17 participants (11 from Middlemore Hospital, 1 from Hawkes Bay and 5 from the Southern Region) spaced over a 2–3-year period, to capture the participants' experiences on the decision-making processes and subsequent treatment experiences of patients with CKD stage 5 as they selected and pursued a treatment pathway.9 21 Data collection Qualitative interviews were undertaken to explore the temporal changes in the perceptions and needs of patients undergoing transitions in their health and healthcare, specifically related to decision-making processes and subsequent treatment experiences of patients with ESKD. Participants were not known to the interviewer (SL), although SL provided them with the goals for the project as part of the informed consent process. Participants were interviewed in English (face-to-face), at home (only two were interviewed in a clinic setting), close to the time of treatment decision-making to precede key transitions in health status and treatment. Interviews were semistructured drawing from a predetermined set of questions (informed by a previous pilot study13 and the wider literature) but with flexibility to pursue ideas raised by participants.22 Interviews ceased when theoretical saturation was reached, that is, when few or no new concepts or topics were raised. Serial interviews were spaced to allow for establishing a dialysis routine, stabilised health or decline to end of life care9 23 Analysis All interviews were audio-recorded and transcribed verbatim. The transcripts were entered into NVivo software and then compared with the body of the interviews (QSR International Pty : NVivo 10.0 edition). Two authors (SL and SD) coded the transcripts independently, and reconciled any differences by discussion. Coding and thematic analysis followed the principles of framework analysis.22 Coding of decision-making interviews saw the inductive generation of initial codes from the data. To address the large amount of data generated through SQI's, a purely deductive review of satisfaction with treatment path options demonstrated a wide range of care trajectories and outcomes during follow-up interviews. The participants' general satisfaction with their own decisions meant follow-up interviews provided few additional insights into their dialysis decision-making in the subsequent interviews. NVivo was used to generate a report of all codes with the corresponding text allowing similar concepts to be grouped into themes. Conceptual links and patterns among themes and subthemes were identified and mapped into a thematic schema. Researcher triangulation was conducted whereby SL and SD discussed the preliminary themes with other researchers (MRM, JBWS, RJW) who read the transcripts independently and confirmed that the themes captured the full range of participants' perspectives. The longitudinal data documented diverse care trajectories among participants but yielded few insights into participant's earlier decision-making behaviour. Pseudonyms have been used in place of participants' actual names. Results Interviews were undertaken with 17 participants. Participant characteristics are presented in table 1. The first set of interviews (17 participants) took place between March 2011 and April 2012; the second follow-up set (16 out of 17 participants) occurred 6–13 months later and the third and final interviews (11of the 16 participants) took place 22–27 months after the first interview. At the time of the first interview, 4 participants had just started dialysis (<3 months) and by the second interview, two additional participants had started dialysis. One participant was deceased by the second interview and three further individuals died between the second and third interviews. One individual declined to participate in a third interview. In three cases, the final interviews were cancelled (two participants) or substantially delayed (one participant) due to hospitalisations. The current manuscript is primarily informed by the first set of interviews, which generated the greatest amount of data on participants' dialysis decision-making (table 1). In several instances, participant's thoughts on dialysis continued to evolve and data from the second interview was incorporated into the decision-making analysis. Table 1 Demographics of participants Pseudonym Sex Age Cause of ESRD Interview analysed* Comorbidities (N)† eGFR‡ Jean Female 66 Mesangial proliferative (IgA+) 1 3 12 Dawn Female 80 Uncertain diagnosis 1 5 12 Mary Female 84 Renal vascular disease—type unspecified 1 and 2 4 14 Raymond Male 69 Obstructive nephropathy 1 4 4 Donald Male 69 Polycystic kidney disease 1 2 11 Paul Male 69 Presumed glomerular nephritis 1 2 13 Graham Male 69 Diabetes—type 2 1 3 12 Marcus Male 71 Uncertain diagnosis 1 2 7 Richard Male 71 Renal vascular disease—due to hypertension 1 3 8 John Male 70 Renal vascular disease—due to hypertension 1 2 12 Vincent Male 72 Diabetes—type 2 1 5 12 Fraser Male 75 Renal cell carcinoma—nephrectomy 1 1 7 Malcolm Male 75 Uncertain diagnosis 1 3 10 Daniel Male 79 Presumed glomerular nephritis 1 3 12 Douglas Male 83 Renal vascular disease—due to hypertension 1 1 12 Kevin Male 84 Renal vascular disease—due to hypertension 1 4 5 Neville Male 90 Obstructive uropathy 1 1 13 *Up to three interviews were carried out with all participants, however only interviews where dialysis decision-making was the focus of the interview, were analysed for the purposes of this paper. Among most participants, there was considerable overlap in content in the follow-up interview(s) with additional new data that informed decision-making only evident in one participant as noted in table. †Comorbidities were recorded as cardiovascular disease, cerebrovascular disease, peripheral vascular disease, diabetes, lung disease, cancer, musculoskeletal disease and other comorbidities. ‡eGFR (mL/min/1.73 m2) determined at the time of recruitment into the study. eGFR, estimated glomerular filtration rate; ESRD, end-stage renal disease. Declining dialysis: independence and dependents Maintaining one's independence was a primary concern in dialysis decision-making; those without the practical and emotional support of a spouse gave more serious consideration to declining dialysis. Participants spoke of support from a range of sources, including their children; however, the presence of a supportive spouse was a critical factor in the decision to start dialysis. Only four participants in our study reported giving serious thought to conservative management; all were aged over 80 and two lived alone while the other two were primary caregivers for their wives who lived with the effects of stroke or dementia. These participants anticipated regular trips to the nephrology unit and dialysis as a tie that would undermine their quality of life and caregiving responsibilities. However, the decision to refuse dialysis was rarely straightforward; only one woman (Dawn, 80) had settled on a decision of conservative management (no dialysis) at the time of our first interview; a further woman (Mary 84) made the same decision ahead of her second interview. Dawn and Mary, were single, living alone and experiencing comorbidities; both saw dialysis as a threat to the weekly social activities they enjoyed. With a strong sense of independence but restricted mobility and only her niece and nephew for support, Dawn did not consider home-based dialysis a viable option. Having cared for her mother for many years, Dawn did not wish to be a ‘burden’ on anyone. Dawn's fluctuating energy levels also led her to worry about the burden of travelling to the clinic and dialysing in centre:So [the doctor] said about going on to dialysis and I said that I didn't want to go on there because, I said, it will interfere with my life. I said three days a week, I said, a couple of those days are the days that I might go out …. And I said and I'm getting older and you don't live forever and I don't want to be a burden to anybody cause, you see, [I] got no—you know—well, I've only got nieces and nephews. (Dawn, 80) All participants discussed the negative impacts that they anticipated dialysis would have on their lifestyle yet, for most, this burden was outweighed by the benefits of a longer life that would see them enjoy more time with their spouse and grandchildren. Experiencing a transition in personal circumstances was identified by participants as a trigger for consideration of conservative management. For example, Neville, 90 years, had been diagnosed with CKD ∼20 years earlier. He intended to begin dialysis ‘when the time came’ as it would enable him to continue caring for his wife who was living with the effects of stroke. However, he explained that his wife's presence was critical to the choice to undertake dialysis:If [she] predeceases me then I've only got to think of myself. A lot would depend on my general health. If she doesn't predecease me well the question [of whether to choose dialysis] doesn't get raised [he would do it when it became necessary to allow him to continue in his role as carer]. (Neville, 90) Similarly, caregiver Kevin, 84 years, whose wife's advancing dementia was requiring increased care, had no immediate support from family members living nearby. Kevin had already chosen conservative management when a visit from his daughter led him to reconsider that decision:Last November I felt real crook and ah, I ah, [the doctor] got me in there and they said ‘Oh geez, you're right down to eight percent, kidney function’ so then they ah, I had to make up my mind. I turned things down for a start. I said ‘I'm not going on a machine’. Then this peritoneal dialysis, my daughter [was visiting] so she arranged to meet the specialist, second in charge, and they arranged a meeting in the dialysis unit so that I could see what was going on. It's quite involved … they had an old chap there, he was 76, and he was being trained and he said ‘Look, I haven't felt better’; he said ‘I'm doing things I've never done before’. So my daughter looked at me and I looked at her, I said ‘righto, I'll give it a go’. So that's when I had the tube [peritoneal catheter] put in. (Kevin) For Kevin, problems with the peritoneal catheter meant that he was unable to dialyse and he later reflected that—given the information his doctors had presented him with—choosing dialysis at his age had been a mistake. Kevin, like Mary, found that his inclination to decline dialysis was at odds with the preference of his child and led him to feel pressured to choose dialysis. Kevin was the only participant to regret his dialysis decision, even when taking follow-up interviews into account. All participants considering conservative management shared advancing ages—constituting four of the six ≥80-year olds in the study, yet age appeared intertwined with experiences of health status and social support required (by participants)—and provided by them (to spouses). The dialysis imperative While the decision not to dialyse was dominated by concerns over quality of life, the decision to start dialysis was dominated by concern over length of life. For 13 participants, declining dialysis was not seriously considered an option. Some expressed surprise and disbelief that anyone would choose not to go on dialysis and occasionally this disbelief was accompanied with the disparaging suggestion that choosing not to dialyse was quitting or ‘giving up’ on life. Many felt they had no choice but to dialyse, an understanding that sometimes appeared linked to their clinician's framing of the decision:So that's [deciding against dialysis] just a way for dying … once your kidney goes that's it! (Paul, 69) When they say I've got to go on [dialysis] then I'll work it in, because I've got no choice. It's either that or die [laughs]. (Malcolm, 75) If you don't go on dialysis you die. But I don't, like, I don't really want to go on dialysis …. (Vincent, 72) Despite this acceptance of dialysis as a life-extending mechanism, only two of these 13 participants accepted dialysis without delay. These individuals were strongly influenced by their declining health and accepted the recommendations of their treating physician to start dialysis. The remaining 11 chose to delay beginning dialysis, some for years, but had not excluded dialysis as an option. Delaying dialysis Eleven participants qualified their choice to dialyse with statements such as ‘when the time comes’. Experiencing fewer effects of ESKD, these participants feared the impact of dialysis on their lifestyle and the implications of reversing a decision to start dialysis. The dominant report—the decision to delay—was often grounded in a concern that dialysis would impinge on everyday activities and ideas of recreational travel that were entwined with many participants' expectations of retirement. Participants anticipated that once on dialysis, their social lives would centre more firmly on the home, day trips would be cut short, work and volunteering would be interrupted, and trips overseas would become difficult or impossible. Most participants felt that home-based dialysis would allow them the greatest flexibility to live their lives as they wanted. Yet there was a tacit acknowledgement that life would irrevocably change:Well, I'm hoping it's not going to make much difference except for the, the amount of time you waste, you know …. Get up early and have one, have one before twelve or something, you can, or have one late at night, something like, you have to arrange it yourself. So, it'll be a pain in the butt getting used to it, but I know quite a few guys, a mate of mine, he's dead now, he used to, he said you just got used to it. (Paul, 69) It's more so the inconvenience of it because I'm very active within the community and although I'm retired—nevertheless it's—I suppose in a sense it's an intrusion into the normal daily life. Especially if you're going to be travelling around the place, around New Zealand or overseas, although we don't do that all that much now, we used to do a far bit. (Graham, 69) Well, it's not, not going to change much. We just have to go out every fifth or third day over to [the clinic] for five hours, um, until such time as I'm used to doing it myself. Then the dialysis machine comes here [to own home]. We'll have to get the place plumbed out and sorted out where we're going to put it. (Vincent, 72) In their efforts to maintain the status quo, it was common for participants to take an active role in monitoring their own kidney function and adopt health-enhancing practices such as appropriate diet and exercise to delay the need for dialysis. Many participants believed their kidney function had yet to decline to a level where their clinician felt dialysis was absolutely necessary. Indeed, two participants had been living with CKD stage 5 for over a decade and considered their health to be stable. These individuals had accumulated considerable expertise and self-efficacy in managing their condition and their decision to delay dialysis was perceived as being supported by their healthcare team. In other instances, the purported decision to ‘delay’ dialysis was framed as active resistance to their clinician's recommendation they begin dialysis, for example:I'm not going on dialysis because I think I can manage the situation, I don't know, but it's, that's the plan. In other words, delay it as long as we can so that we just go on and on and on. When that moment arrives you deal with it. It's not going to cause me any great grief I don't think …. I still shower myself, still dress myself all those sorts of things. I have help with drying my lower limbs but, and back, but that's it. I would expect that to continue, so that life just goes on and that we can continue to share in the life of this family and the rest of our family, and share in my grandchildren's lives and all those sorts of things. That's, that's the plan. And, at seventy, there's still a few more years to go on that one. (Marcus, 71) Marcus was uncomfortable with the speed with which he was encouraged to begin dialysis; he resisted his doctor's recommendation and instead turned to managing his condition through lifestyle, medical and alternative treatments. His decision appeared to be about preserving his present quality of life and routine while retaining the option of dialysis if his own management of his condition was not maintaining this. This subset of participants used a variety of strategies to manage their health, including tracking their blood test results and modifying their lifestyle. Yet underpinning their attitude was a belief that they would begin dialysis when their declining health left them with no other option. Among the 11 participants who chose to delay dialysis, an assemblage of factors was critical to assessments of the need to begin dialysis:1 recommendations of clinicians,2 experience of symptoms and 3 clinical indicators of kidney function. These factors took on varying levels of importance between participants and are discussed below: Waiting for a clinical directive Underpinning many participants' narratives was an anticipation that a time would come when their clinician would signal they could no longer delay dialysis. Many awaited the call from their doctor that it was time to start, for example, one participant's narrative placed the clinician as central to the decision to initiate dialysis:At the moment he's sitting on the, on his, hands and saying ‘Well, it doesn't look like it'll be happening until sometime next year’. (Douglas, 83) Such reliance on the clinicians contrasted heavily with the actions of many other participant's self-management of their health and expectations of patient-centred care. In one instance, a decision to begin dialysis was made by a patient's nurse and the participant and his wife remained satisfied that this was the right call:I lost hearing, I'd lost memory, I'd lost sight. I don't know about hearing—I think I could hear alright—but … my memory wasn't good. I was just going downhill quite quickly …. I wouldn't be here if it wasn't for the dialysis …. I had no option at the finish. The girl [nurse], what's her name? Who was it? [Susan] came in and said ‘You'll be on dialysis 8 o'clock on Monday’. And I said ‘Will I?’ and she said ‘Yes’ [laughs]! So that was it. I didn't make that decision …. I'll take everything to the end—but once I'm at the end of the piece of string I'll recognise it. (Barry, 76) Barry had chosen to delay dialysis for as long as possible and during his interview, he told several stories that highlighted the impact his kidney condition was having on his cognitive functioning. While he and his wife had identified dialysis as their preferred treatment path, it was only retrospectively that he was aware of how his decision to delay dialysis for so long was impacting his health. Symptoms and physical decline Participants often either failed to attribute their symptoms to CKD or assigned the cause of their symptoms to another condition or the general ageing process, as evident in the case of a participant who reported his memory had declined prior to beginning dialysis:… and what do they call it, it's a test they do, help, [long pause] give your blood and you, and I forgot to mention, I must have dementia or something. Something—oh, it'll come to me. (Paul, 69) It knocks your, as far as your appetite's concerned—and perhaps I didn't realise with the information that I got—that, that it would do it. I still, I still don't eat as much as I used to, that's for sure. And … I could usually eat everything, just like to, like to enjoy my meals. (Richard, 70) The most common effect of ESKD that participants identified was exhaustion. This was measured in a wide range of ways, from one's ability to get around 18 holes on the golf course to completing domestic chores. Among those who accepted that their tiredness was due to their CKD, most struggled with the loss of motivation and inability to do things they previously took for granted. The absence—or lack of awareness—of symptoms has implications for the timing of dialysis as participants struggled to reconcile their perceptions of health with information that they needed to begin dialysis:Thinking back I felt as though ‘Why, if I feel as I do at the moment and why, as active as I am at the moment, would I think about going on to dialysis?’ It seems ridiculous. (John, 70) For a small number, the decision to begin dialysis was preceded by a marked decline in physical function that led to their acceptance that dialysis was necessary, as Jean explained:So you, you do your best for a while, and I feel okay, so it's alright. See, kidney failure is a thing that just creeps up gradually so you just, you cope with it day by day, not realising that you're getting worse and worse all the time. So I know other people of my age that, I'm not functioning like they are. They're having a good time. I can't mow my lawns and all that sort of stuff, so that's why I'm here [training for dialysis]. (Jean, 66) For these participants, dialysis was viewed as a means of freeing them from their increasingly constrained activities. These individuals hoped dialysis would enable them to become more physically active and carry out the activities that had once been central to their lives, as one participant explained:What do I want to achieve? Just a normal lifestyle, yeah, and I'm getting it. Yeah, once I get my hips right I'm getting it. I'll be right, good as gold again …. Out there, working on the farm, yeah. My son's taken over the farm but I give him a hand. I haven't been able to do anything for the last nearly twelve months now. (Richard, 71) Implicit in this narrative is the expectation that the benefits of dialysis would outweigh the effort of maintaining dialysis treatment. Such views highlight the diversity of patient perceptions by running counter to the dominant narrative of dialysis as a burden. Discussion Our study identified three main themes. The most novel finding of our study is that older patients deliberately delay dialysis initiation as an act of self-efficacy and to manage their own health. Although we did not assess their self-efficacy directly, our data suggest that delaying dialysis arises from the increased confidence and capability of patients to manage their own health,24 rather than the opposite. This was a common finding among participants who had opted for eventual dialysis as well as for those who had not made a clear decision to dialyse or to opt for an interim strategy of conservative treatment. This is in contrast to the perceptions of nephrologists, who often perceive the delaying of dialysis as resistance to dialysis and denial of the seriousness of their condition. This was illustrated in relation to one participant who was described by his clinician as a ‘non-dialyser’ whereas in the qualitative interview, the same man was adamant that he was simply ‘delaying dialysis’. The other two main themes confirmed previous studies that patients prioritise factors that preserve their independence, and that the decision to dialyse is dominated by concern over length of life, as opposed to the decision not to dialyse which is driven by quality of life.25 The implication for health service delivery is that predialysis education in the older age group is a multistaged process where the ultimate decision for dialysis versus conservative care is made by the patient around the time of onset of uraemia. In younger patients, this is regarded as ‘crash starting’, but among elderly, this should be seen as appropriate: a decision made a year in advance is likely to be made in a clinical context that is no longer appropriate at the time of decision-making. This paper suggests the physicians should work with patients to ensure that they are fully equipped to make a robust and shared decision in the future, but not pressure them to make a definitive decision due to potential changes in the older person's status. Older patients who decline dialysis are recognised elsewhere as experiencing greater comorbidities;10 26 therefore, we would expect the decision to dialyse to be affected by one's stage in the life course. Over the course of interviewing patients over time, we observed that decision-making was often a fluid process, rather than occurring at a single point in time, and was heavily influenced by perceptions of oneself as becoming old, social circumstances, life events and health status. We found three indicators that were critical to these perceptions. First, those who did not recognise ESKD symptoms in their day-to-day lives experienced less readiness to begin dialysis than those who experienced declining health. Second, participants anticipated a clear directive would come from their clinician that it was time to begin dialysis and commonly perceived this omission as support for further delay of dialysis. Third, participants who actively managed their condition without dialysis—commonly those for whom the progression of their CKD has been slow, with few symptoms attributable to their CKD—likewise saw stable blood test results as indicators of appropriate delay. These findings suggest that timing of dialysis initiation may be modified by clearer communication between patients (and their carers or supporters) and the clinical team (nephrologists and predialysis educators). These should clearly identify and focus on symptoms and effects of CKD to cultivate patient readiness for the initiation of dialysis and support self-management of their disease. Given their non-specific nature, the clinical team should revisit the symptoms and effects of CKD regularly, and make specific enquires about them during patient follow-up. Predialysis education for this older age group needs to be more explicit with respect to the natural progression of CKD and the long latency between starting predialysis education and the subsequent need to make a shared decision about the type of ESKD management. Patients who described actively self-managing their condition relied heavily on the benchmarks for initiating dialysis communicated by their clinicians. Blood test results and symptoms were effective gauges of the effects of any lifestyle and dietary changes. The cohort of participants in our study felt relatively well at enrolment, in contrast to previous studies that reported denial of the severity of ESKD in the face of frankly uraemic symptoms.27 This situation reflects the clinical reality of CKD care and dialysis decision-making in most predialysis patients, be they younger or older—the majority of patients begin discussions about dialysis at an earlier stage and in a relatively better state of health, compared with when they start dialysis. Our study is the first of its kind to assess older patients' longitudinal experiences of the entire CKD spectrum, and has allowed us to identify the key patient paradigm of delaying dialysis as a strategy to preserve quality of life; anticipating dialysis would impinge on their everyday activities. Clarity around ‘delay’ versus ‘denial’ of dialysis might be improved by better patient understanding of progression of CKD to ESKD, especially in relation to onset of symptoms. In our study, conservative management was seriously considered only by four older participants, for whom in-centre dialysis was their proposed treatment pathway rather than home dialysis, which was precluded by a perceived lack of support to manage home dialysis. The perceived burden of clinic-based treatment was sufficient threat to the quality of life of two of these participants that they declined dialysis. Elsewhere, studies examining patient decision-making processes for dialysis found that where in-centre dialysis was the only option available, this had a significant impact on older patients' perceptions of independence and normal activities of living.9 25 In contrast, previous New Zealand research has shown home-based dialysis to be highly acceptable to older adults, even when patients resided more than 1 hour from specialist services.13 Whereas little difference in quality of life is evident between older patients on haemodialysis and peritoneal dialysis,28 the availability of home dialysis appears to be a positive impact on quality of life, even for older patients.14 This study seeks only to represent the perceptions, experiences and understandings of older patients living with ESKD. By focussing on participants' experiences of decision-making, we have not included the perspectives of clinicians and/or the broad range of factors that influence their interactions with the participants discussed here. This may or may not been seen as an imitation, but addresses an important gap in the literature in contrast the comparatively well-reported attitudes of and experiences of clinicians.25 27 Importantly, the external validity of our study is limited to similarly developed Western nations as New Zealand. Our findings are less generalisable to countries with constrained healthcare resources (eg, developing nations, where access to care rather than patient choice determines healthcare delivery), or with significantly different cultural orientations (eg, familism in context of Confucian cultures such as Taiwan, where filial piety often precludes conservative care for end-stage renal disease (ESRD))29–31 . A strength of our study is its internal validity, insofar as assessments have been undertaken in the context of New Zealand's publicly funded healthcare system—the shared decision to dialyse, or not, is made independently of clinician reimbursement or the ability of patients to pay for dialysis.12 13 The assessment of methodological rigour for qualitative research (as opposed to reporting rigour) is often discussed in the literature but is yet to be standardised.18 32 Notwithstanding, by using a tool of quality appraisal for qualitative research tool such as the Critical Appraisal Skills Programme (CASP) Qualitative Research Checklist,33 our study has a high degree of methodological rigour over the 10 domains of assessment. In summary, this qualitative study has identified key points that underlie the older patient's perspective related to decision-making to manage their ESKD. This adds to an increasing understanding of the importance of clear discussions between the individual, their family and the clinical team involved in their care.3–5 Such an individualised approach should prioritise the modifiable outcomes patients value most, and acknowledge that observed signs and symptoms often reflect the complex interplay between CKD, ageing and associated comorbidities. Prognostic information related to these and other outcomes are generally used to shape rather than dictate treatment decisions.34 An individualised patient-centred approach to care may have more to offer than a traditional disease-based approach to ESKD for many older adults.35 This study along with the prospective longitudinal quality of life data is being generated from our larger study19 will inform the development of decision aids and clinical guidelines that include the older patients' preferences, autonomy and need to remain an active member of their community. Contributors: This was an investigator-initiated and analysed study. RJW, SD and MRM were responsible for the concept of the study. SL undertook the interviews. SL and SD undertook the analyses of the interviews. BMM was responsible for all demographic data collection. Researcher triangulation was conducted whereby SL and SD discussed the preliminary themes with other researchers (MRM, JBWS, RJW), who read the transcripts independently and confirmed that the themes captured the full range of participants' perspectives. All authors (SL, SD, BMM, MRM, JBWS, RJW) contributed to the interpretation of the results and the preparation of the manuscript. RJW is the principal investigator for DOS65, and SD JBWS, MRM and BMM are DOS65 coinvestigators. Funding: The study was supported by Health Research Council of New Zealand (grant number 10/354). Disclaimer: The affiliation with Baxter Healthcare (Asia) of Mark Marshall does not alter the authors’ adherence to PLOS ONE policies on sharing data and materials. Competing interests: MRM is full-time employee of Baxter Healthcare (Asia) Pte, Singapore. Baxter Healthcare has had no input whatsoever into this study. The other authors have no potential or actual conflicts of interest related to this study. Ethics approval: This study was approved by the New Zealand Multiregional Ethics Committee, approval number MEC/10/084. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Tattersall J Dialysis in the over-80s . Age Ageing 2005 ;34 :100 –1 . 10.1093/ageing/afi056 15713853 2 Brown EA , Johansson L Dialysis options for end-stage renal disease in older people . Nephron Clin Pract 2011 ;119 (Suppl 1 ):c10 –13 . 10.1159/000328019 3 Treit K , Lam D , O'Hare AM Timing of dialysis initiation in the geriatric population: toward a patient-centered approach . Semin Dial 2013 ;26 :682 –9 . 10.1111/sdi.12131 24112631 4 Harwood L , Clark AM Dialysis modality decision-making for older adults with chronic kidney disease . J Clin Nurs 2014 ;23 :3378 –90 . 10.1111/jocn.12582 24646195 5 Chandna SM , Da Silva-Gane M , Marshall C Survival of elderly patients with stage 5 CKD: comparison of conservative management and renal replacement therapy . Nephrol Dial Transpl 2011 ;26 :1608 –14 . 10.1093/ndt/gfq630 6 Elliott BA , Gessert CE , Larson PM Shifting responses in quality of life: people living with dialysis . Qual Life Res 2014 ;23 :1497 –504 . 10.1007/s11136-013-0600-9 24343662 7 Madhan K The epidemic of elderly patients with dialysis-requiring end-stage renal disease in New Zealand . N Z Med J 2004 ;117 :U912 .15282624 8 Brown MA , Collett GK , Josland EA CKD in elderly patients managed without dialysis: survival, symptoms, and quality of life . Clin J Am Soc Nephrol 2015 ;10 :260 –8 . 10.2215/CJN.03330414 25614492 9 Morton RL , Snelling P , Webster AC Factors influencing patient choice of dialysis versus conservative care to treat end-stage kidney disease . CMAJ 2012 ;184 :E277 –83 . 10.1503/cmaj.111355 22311947 10 Stringer S , Baharani J Why did I start dialysis? A qualitative study on views and expectations from an elderly cohort of patients with end-stage renal failure starting haemodialysis in the United Kingdom . Int Urol Nephrol 2012 ;44 :295 –300 . 10.1007/s11255-011-0045-4 21850412 11 Llewellyn H , Low J , Smith G Narratives of continuity among older people with late stage chronic kidney disease who decline dialysis . Soc Sci Med 2014 ;114 :49 –56 . 10.1016/j.socscimed.2014.05.037 24911508 12 Thorsteinsdottir B , Swetz KM , Albright RC The ethics of chronic dialysis for the older patient: time to reevaluate the Norms . Clin J Am Soc Nephrol 2015 ;10 :2094 –9 . 10.2215/CJN.09761014 25873266 13 Derrett S , Darmody M , Williams S Older peoples’ satisfaction with home-based dialysis . Nephrology (Carlton) 2010 ;15 :464 –70 . 10.1111/j.1440-1797.2010.01286.x 20609099 14 Rygh E , Arild E , Johnsen E Choosing to live with home dialysis-patients’ experiences and potential for telemedicine support: a qualitative study . BMC Nephrol 2012 ;13 :13 10.1186/1471-2369-13-13 22429705 15 Marshall MR , Walker RC , Polkinghorne KR Survival on home dialysis in New Zealand . PLoS ONE 2014 ;9 :e96847 10.1371/journal.pone.0096847 24806458 16 Ashton T , Marshall MR The organization and financing of dialysis and kidney transplantation services in New Zealand . Int J Health Care Finance Econ 2007 ;7 :233 –52 . 10.1007/s10754-007-9023-x 17638073 17 Tong A , Sainsbury P , Craig J Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups . Int J Qual Health Care 2007 ;19 :349 –57 . 10.1093/intqhc/mzm042 17872937 18 O'Brien BC , Harris IB , Beckman TJ Standards for reporting qualitative research: a synthesis of recommendations . Acad Med 2014 ;89 :1245 –51 . 10.1097/ACM.0000000000000388 24979285 19 Walker R , Derrett S , Campbell J Dialysis outcomes in those aged ≥65 years . BMC Nephrol 2013 ;14 :175 10.1186/1471-2369-14-175 23945009 20 Vest JR , Issel LM Factors related to public health data sharing between local and state health departments . Health Serv Res 2014 ;49 :373 –91 . 10.1111/1475-6773.12138 24359636 21 Murray SA , Kendall M , Carduff E Use of serial qualitative interviews to understand patients’ evolving experiences and needs . BMJ 2009 ;334 :521 –1 . 22 Ritchie J , Lewis J , Nicholls C Qualitative research practice: a guide for social science students and researchers. 2nd Ed. Sage Publications (UK) 2013 . 23 Calvin AO Haemodialysis patients and end-of-life decisions: a theory of personal preservation . J Adv Nurs 2004 ;46 :558 –66 . 10.1111/j.1365-2648.2004.03030.x 15139945 24 Clark NM , Dodge JA Exploring self-efficacy as a predictor of disease management . Health Educ Behav 1999 ;26 :72 –89 . 10.1177/109019819902600107 9952053 25 Tonkin-Crine S , Okamoto I , Leydon GM Understanding by older patients of dialysis and conservative management for chronic kidney failure . Am J Kidney Dis 2015 ;65 :443 –50 . 10.1053/j.ajkd.2014.08.011 25304984 26 Carson RC , Juszczak M , Davenport A Is maximum conservative management an equivalent treatment option to dialysis for elderly patients with significant comorbid disease? Clin J Am Soc Nephrol 2009 ;4 :1611 –19 . 10.2215/CJN.00510109 19808244 27 Greer RC , Ameling JM , Cavanaugh KL Specialist and primary care physicians’ views on barriers to adequate preparation of patients for renal replacement therapy: a qualitative study . BMC Nephrol 2015 ;16 :37 10.1186/s12882-015-0020-x 25885460 28 Brown EA , Johansson L , Farrington K Broadening Options for Long-term Dialysis in the Elderly (BOLDE): differences in quality of life on peritoneal dialysis compared to haemodialysis for older patients . Nephrol Dial Transpl 2010 ;25 :3755 –63 . 10.1093/ndt/gfq212 29 Glass AP , Chen LK , Hwang E A cross-cultural comparison of hospice development in Japan, South Korea, and Taiwan . J Cross Cult Gerontol 2010 ;25 :1 –19 . 10.1007/s10823-009-9108-8 20054707 30 Tzeng HM , Yin CY Family involvement in inpatient care in Taiwan . Clin Nurs Res 2008 ;17 :297 –311 . 10.1177/1054773808324655 18927262 31 Chiu YC What drives patients to sue doctors? The role of cultural factors in the pursuit of malpractice claims in Taiwan . Soc Sci Med 2010 ;71 :702 –7 . 10.1016/j.socscimed.2010.04.040 20646806 32 Côté L , Turgeon J Appraising qualitative research articles in medicine and medical education . Med Teach 2005 ;27 :71 –5 . 10.1080/01421590400016308 16147774 33 Critical Appraisal Skills Programme (CASP) Qualitative Checklist 2013 http://www.casp-uk.net/#!checklists/cb36 (accessed 1 Jan 2015 ). 34 Ladin K , Weiner DE Better informing older patients with kidney failure in an era of patient-centered care . Am J Kidney Dis 2015 ;65 :372 –4 . 10.1053/j.ajkd.2015.01.003 25704042 35 Bowling CB , O'Hare AM Managing older adults with CKD: individualized versus disease-based approaches . Am J Kidney Dis 2012 ;59 :293 –302 . 10.1053/j.ajkd.2011.08.039 22189037

PMC005xxxxxx/PMC5372048.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01252810.1136/bmjopen-2016-012528EpidemiologyCohort Profile15061692168617041730Linking medical and dental health record data: a partnership with the Rochester Epidemiology Project St. Sauver Jennifer L 12Carr Alan B 3Yawn Barbara P 14Grossardt Brandon R 5Bock-Goodner Cynthia M 6Klein Lori L 1Pankratz Joshua J 6Finney Rutten Lila J 12http://orcid.org/0000-0002-1832-7664Rocca Walter A 17 1 Division of Epidemiology, Department of Health Sciences Research, Mayo Clinic, Rochester, Minnesota, USA 2 Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Mayo Clinic, Rochester, Minnesota, USA 3 Department of Dental Specialties, Mayo Clinic, Rochester, Minnesota, USA 4 Department of Research, Olmsted Medical Center, Rochester, Minnesota, USA 5 Division of Biomedical Statistics and Informatics, Department of Health Sciences Research, Mayo Clinic, Rochester, Minnesota, USA 6 Department of Information Technology, Mayo Clinic, Rochester, Minnesota, USA 7 Department of Neurology, Mayo Clinic, Rochester, Minnesota, USACorrespondence to Dr Jennifer L St. Sauver; stsauver.jennifer@mayo.edu2017 29 3 2017 7 3 e0125284 5 2016 19 12 2016 21 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Purpose The purpose of this project was to expand the Rochester Epidemiology Project (REP) medical records linkage infrastructure to include data from oral healthcare providers. The goal of this linkage is to facilitate research studies examining the role of oral health in overall health and quality of life. Participants Eight dental practices joined the REP between 2011 and 2015. The REP study team has linked oral healthcare information with medical record information from local healthcare providers for 31 750 participants who have resided in Olmsted County, Minnesota. Overall, 17 718 (56%) participants are women, 14 318 (45%) are 40 years of age or older and 26 090 (82%) are white. Findings to date A first study using this new information was recently completed. This resource was used to determine whether the 2007 guidelines from the American Heart Association affected prescription rates of antibiotics to patients with moderate-risk cardiac conditions prior to dental procedures. The REP infrastructure was used to identify a series of patients diagnosed with moderate-risk cardiac conditions by the local healthcare providers (n=1351), and to abstract antibiotic prescriptions from dental records both pre-2007 and post-2007. Antibiotic prescriptions prior to dental procedures declined from 62% to 7% following the change in guidelines. Future plans Dental data from participating practitioners will be updated on an annual basis, and new dental data will be linked to patient medical records. In addition, we will continue to invite new dental practices to participate in the REP. Finally, we will continue to use this research infrastructure to investigate associations between oral and medical health, and will present findings at conferences and in the scientific literature. electronic health recordsoral healthdental recordspopulation healthoral systemic healthmedical-dental record linkage ==== Body Strengths and limitations of this study The Rochester Epidemiology Project has linked oral health data from community dental practitioners to electronic health record data from medical providers. This linkage provides a unique opportunity to study associations between oral health and overall health and quality of life. This study includes linked data for 31 750 participants of all ages and both sexes who have resided in Olmsted County, Minnesota. The sample size makes it possible to study a wide range of oral health questions and outcomes. The study does not include all dental providers in Olmsted County, and data are less complete for some segments of the population. Introduction Oral health is increasingly recognised as a key determinant of overall health and quality of life, and several studies have demonstrated that oral health problems may cause, co-occur with or result from other medical conditions.1–3 Oral health problems may lead to a systemic pro-inflammatory state that could contribute to the development of certain medical conditions. For example, persons with chronic periodontitis are at a significantly increased risk of developing cardiovascular diseases compared to persons without this condition,4 and poor periodontal health may worsen diabetes.5 In addition, the systemic dysregulation of inflammatory responses can cause the concomitant development of oral health problems and other medical conditions. For example, patients with rheumatoid arthritis or dermatomyositis are at high risk of also having oral diseases due to the underlying inflammatory aetiologies of these conditions.6 7 Conversely, medications used for several chronic conditions, such as asthma, cardiovascular disease, diabetes and osteoporosis, can change the oral environment and contribute to an increased risk of a number of oral health conditions.5 8–10 Despite mounting evidence pointing to the co-occurrence of oral and systemic conditions, research is limited because of the lack of comprehensive data resources that link oral health information with systemic health information. In the USA, dental providers often function as independent, small-business owners, and have not routinely shared data across sites for research purposes. The increasing recognition of oral health as a key contributor to overall health has prompted recent efforts to share oral health data. The most developed of these partnerships is a large, long-standing collaboration between over 30 schools of dentistry called the ‘Consortium for Oral Health Research and Informatics (COHRI)’.11 This consortium was specifically developed to share electronic oral health data to improve oral health research and education. More recently, six members of COHRI, coordinated by the University of Texas Health Science Center at Houston, have collaborated to share partially de-identified electronic health record data through the ‘BigMouth Dental Data Repository’.12 Partners are able to query data across all sites to rapidly obtain anonymised data summaries, and more detailed, patient-level data with appropriate approvals. Access to these data is currently limited to partnering institutions, and information about non-oral health is limited to data available in the dental electronic health records with no direct links to patients' medical records. In addition, all of the institutions in this collaboration are based in academic centres and large dental schools. Therefore, the patients cared for at these institutions may not represent the full range of dental problems in the community. To address these limitations, the Rochester Epidemiology Project (REP), a community collaboration of healthcare providers in southeastern Minnesota and west central Wisconsin, has established partnerships with community dental practitioners in Olmsted County, Minnesota. After obtaining authorisation from patients, dental and medical health data are systematically linked at the individual patient level to better understand the impact of oral health on overall health and health outcomes in the local population. Cohort description Setting and location The REP was established in 1966 by Dr Leonard T. Kurland, as a collaboration between medical care providers in Olmsted County, Minnesota, to share medical record information for research. Each medical care provider in Olmsted County uses a unit (or dossier) medical record system where all data collected for a person are assembled in one place.13 The REP links these medical records to unique persons residing in the community, and maintains an electronic index of diagnoses and procedures from these records, including hospitalisations, office visits, emergency room visits, surgical procedures and drug prescriptions. The REP allows investigators to follow subjects across all local medical facilities, regardless of where the care was delivered or of insurance status. Thus, using the resources of the REP, investigators can conduct long-term, population-based studies of disease incidence, prevalence, risk and protective factors, outcomes, health services usage and cost-effectiveness. Over 40 medical care providers have participated in this unique collaboration since 1966, and this resource has been used extensively in research studies that have led to over 2400 scientific publications.13 The population of Olmsted County is representative of the population of Minnesota, of the Upper Midwest and of a large segment of the entire US population.14 Details regarding the history of the REP and a list of studies that have used the REP are available at the REP website: http://rochesterproject.org/ Recruitment of dental practitioners Although a wide range of community healthcare providers have partnered to share their data through the REP, local dental practitioners have not historically participated in this collaboration. However, in July 2008, ABC (Chair of the Department of Dental Specialties at Mayo Clinic) conducted a mailed survey among 96 dentists in over 40 practices in southeastern Minnesota. The survey is included in Appendix A, and was designed to assess the interest of the dentists in different types of research. Overall, 45 (47%) dentists responded to the survey, and 42 indicated they were either ‘somewhat’ or ‘very’ interested in research regarding the impact of oral health on general health. With this preliminary information, ABC began an effort in 2009 to develop partnerships with the many interested dental practitioners in the community. He approached the interested dentists through individual meetings, and obtained initial commitments from two dentists in the community to join the REP as new research partners. 10.1136/bmjopen-2016-012528.supp1supplementary appendix Following ABC's initial efforts, the REP team developed specific processes for each new dentist to share data with the REP for research purposes. The first step in each partnership was to establish a data sharing contract and a Health Insurance Portability and Accountability Act (HIPAA) business associate agreement between the dental practitioner and the REP. These agreements explicitly state the rules under which data may be used for research. In particular, the data sharing agreements do not allow the REP, or any investigators that use the REP, to provide details regarding the individual dental practices that participate in the REP (eg, number and types of providers, specific types of services provided, etc). These agreements are in place to ensure that comparisons are not made between practices and that no practice may be singled out as superior or inferior to another practice. Once these agreements were established, dental care providers were then required to obtain Minnesota Research Authorization from their patients.15 Minnesota state law requires that healthcare providers obtain authorisation to use health records for research from all patients who received medical care after 1 January 1997 (Minnesota State privacy law—Statute 144.335, 1997). According to the law, two attempts to obtain this authorisation from each participant must be completed in writing, with at least 60 days between attempts. If the patient gives explicit authorisation, or does not respond after two attempts, then the record can be used for research purposes. Finally, parents must sign the authorisation forms for their children (<18 years), and all children must sign forms for themselves once they become adults (≥18 years). All providers that participate in the REP have implemented systems to comply with this law, and authorisation rates are high. In 2000, only 1.7% of the population refused to allow any of their records to be used for research, and that percentage increased only slightly in 2010 to 2.4%.15 This legal requirement represents a significant burden to providers, and remains the single most important barrier to using oral health data for research in our community. The first community dentist joined the REP in 2011, and implemented the process for obtaining Minnesota Research Authorization beginning in August 2011. Data were first obtained from this practice in March 2013, and were matched to the medical records of persons included in the existing REP research infrastructure. As of May 2015, eight dental practices (30 clinicians) had signed contracts with the REP, collected Minnesota Research Authorization from their patients and began sharing data with the REP. Data from the dental providers are updated annually, and are matched to existing persons in the REP using previously described procedures.16 Complete data from all practices are available beginning in 2012. Obtaining data from the dental practices The REP Administrator (LK) and the REP Programmer (CB) met with dental staff to discuss the type of billing system and the type of dental record used by the practice. The REP team provided a list of desired fields (demographics, billing codes and dates of service) and discussed with office staff the process by which they would record Minnesota Research Authorization status in their systems. All practices chose to record research authorisation as a custom procedure. All participating dental practices used electronic health records, and all captured patient demographic information electronically and billed for their services using American Dental Association (ADA) Current Dental Terminology (CDT) electronic billing codes. However, the type of dental record and billing system varied from site to site. Four sites used Dentrix, three sites used Eaglesoft and one site used a proprietary record. Extraction of the data differed depending on the type of system employed by the practitioner. All data were downloaded at the dental offices by the REP Programmer (CB) on a secure drive, and transported immediately to an REP secure server. Data format differed depending on type of dental record and on the billing system that was used. Data in a comma-separated values (csv) format were simple to work with electronically; however, data were only available in portable document format (pdf) for four sites. These data needed to be extracted via a time-consuming pdf-splitting process before they could be integrated into the REP systems. Linking dental records to the REP research infrastructure All data were standardised prior to incorporation into the REP. Research authorisation status was also extracted from the procedure file and applied to the patient records. ADA categories and CDT code descriptions were applied as part of the standardisation process. Linking medical and dental records was accomplished via the usual REP linkage procedures that have been described previously.16 Briefly, patient records were matched electronically via multiple rounds of matching, where the first three rounds of matching were based on a complete match between the records on at least four of the following data points: patient first and last name, date of birth, sex and social security number. Successive rounds of matching used less stringent criteria, including fuzzy matching of name substrings, use of middle initial and Soundex.17 Matches of less than four data points relied on address or name history for corroboration. If a match had too few data points for a confirmed match, the data were stored but not incorporated into REP until additional demographic data on the patient are acquired. Completeness of the demographic information varied from site to site. For example, some sites captured full names, sex, dates of birth and social security numbers, but other sites captured only patient name and address. Missing data were a significant barrier to linkage. However, ∼89% of all dental patients matched to a person in the REP system. Type and completeness of dental data The type and the completeness of the data in the dental records vary substantially across dental practices, and are highly dependent on the needs of the dentist and the individual patient. Some of the records include a detailed medical history and extensive dental details. Other records are brief and limited only to the reason for the visit. Therefore, it is not possible to electronically obtain all of the pertinent details from these records in a standard way that is simple to access for all research studies. However, these limitations are common to medical records as well, and the goal of the REP research infrastructure is to provide complete electronic information when such information is consistently collected and available (eg, demographic data, billing code information, visit dates and location of the dental record). This information then serves as an index to direct a researcher to the areas of the dental record that are most likely to contain the information needed for a particular study. For example, if a researcher is interested in conducting a study related to dental implants, the REP staff will provide the researcher with a list of patients with a billing code for dental implants, the date of the service and the name of the practice at which the implant was performed. The REP staff will then arrange a time for the researcher to visit the dental practice and review the dental record to abstract the pieces of information pertinent to the study. All of the dental partners have agreed to facilitate these visits and to provide space for a visiting researcher. Therefore, all information in the dental records is accessible to an interested researcher. However, as the types and amount of data collected differ dramatically from practice to practice, it is often necessary for a researcher to review all possible dental and medical records to collect the relevant information for the study. Number of patients with dental data Using the REP Census, we identified all persons who resided in Olmsted County, Minnesota, at any time from 1 January 2013 through 31 December 2013.16 We then identified all persons who received a dental service of any kind between 2012 and 2014. The proportion of persons for whom dental data were available is presented separately by sex, four broad age groups (0–18, 19–39, 40–64 and 65+ years), race (white, black, Asian and other race) and ethnicity (Hispanic). χ2 tests were performed to determine whether the availability of dental information differed by these characteristics. In addition, among persons for whom dental data were available, we describe the frequency of various types of dental services, separately across the four age groups. As of 14 April 2015, 46 973 patients have been asked to sign Minnesota Research Authorization for their dental records, and 42 745 (91%) have granted authorisation for use of their dental record for research. The amount of dental data varies across practices, but all practices have shared at least 3 years of data (2012–2014). Data consist of demographic information (name, sex, date of birth) and dental service data. Dental service data are coded using the ADA and CDT codes, and dates of each service are also available. Between 2012 and 2014, 31 750 patients received nearly 477 000 dental services. Persons with at least some dental data in the REP are similar to persons who do not yet have dental data in the REP (table 1); however, male patients, patients between the ages of 19–39 years, and non-white patients are less likely to have dental data. The median number and types of services received during this time frame are shown by age group in table 2. Diagnostic services were the most common service for all age groups; however, other services differed by age. For example, 74% of children (0–18 years) had at least one preventive service code compared with 50% of older adults (65+ years). Conversely, older adults (65+ years) were more likely to have a removable service code (47%) compared with only 2% of children (0–18 years). Table 1 Comparison of persons with and without dental data included in the Rochester Epidemiology Project* (data from 2012 to 2014) Population Have dental data Have no dental data p Value for comparison† Characteristic Total N N Per cent N Per cent All persons 158 786 31 750 20.0 127 036 80.0 – Sex  Men 75 130 14 032 18.7 61 098 81.3 <0.0001  Women 83 656 17 718 21.2 65 938 78.8 – Age, years  0–18 42 786 9752 22.8 33 034 77.2 <0.0001  19–39 47 270 7680 16.2 39 590 83.8 –  40–64 48 281 10 129 21.0 38 152 79.0 –  65+ 20 449 4189 20.5 16 260 79.5 – Race  White 127 781 27 217 21.3 100 564 78.7 <0.0001  Black 9517 1021 10.7 8496 89.3 –  Asian 8686 1703 19.6 6983 80.4 –  Other‡ 12 802 1809 14.1 10 993 85.9 – Ethnicity  Hispanic 10 579 1794 17.0 8785 83.0 <0.0001  Non-Hispanic 148 207 29 956 20.2 118 251 79.8 – *The population was defined at the 2013 Rochester Epidemiology Project Census. †χ2 test of homogeneity to determine whether the frequency counts are distributed similarly across persons who do or do not have dental data. ‡The other race category includes persons of a race other than white, black or Asian and persons with unknown race. Table 2 Distribution of type of service by age for both sexes and all races and ethnicities combined (data from 2012 to 2014) Age group 0–18 years (N=9752) 19–39 years (N=7680) 40–64 years (N=10 129) 65+ years (N=4189) Type of service N (%)* Med (Q1, Q3)† N (%)* Med (Q1, Q3)† N (%)* Med (Q1, Q3)† N (%)* Med (Q1, Q3)† Diagnostic 8397 (86.1) 5 (3, 8) 6521 (84.9) 5 (3, 7) 9080 (89.6) 6 (3, 9) 3648 (87.1) 5 (3, 8) Preventive 7207 (73.9) 6 (3, 10) 5344 (69.6) 3 (2, 6) 6438 (63.6) 4 (2, 6) 2075 (49.5) 4 (2, 6) Restorative 2418 (24.8) 2 (1, 4) 2626 (34.2) 2 (1, 4) 4668 (46.1) 2 (1, 4) 1817 (43.4) 2 (1, 4) Removable 211 (2.2) 1 (1, 2) 1503 (19.6) 2 (1, 3) 3563 (35.2) 3 (1, 6) 1977 (47.2) 5 (2, 8) Adjunctive general 2080 (21.3) 1 (1, 2) 1581 (20.6) 1 (1, 2) 2887 (28.5) 2 (1, 3) 868 (20.7) 2 (1, 3) O & M surgery‡ 2553 (26.2) 1 (1, 2) 1340 (17.4) 1 (1, 1) 1839 (18.2) 1 (1, 2) 1065 (25.4) 1 (1, 2) Orthodontics 1132 (11.6) 2 (1, 3) 404 (5.3) 1 (1, 2) 377 (3.7) 1 (1, 3) 38 (0.9) 1 (1, 2) Endodontics 262 (2.7) 1 (1, 2) 334 (4.3) 1 (1, 2) 660 (6.5) 1 (1, 2) 204 (4.9) 1 (1, 2) Fixed 2 (<0.1) 1 (1, 1) 20 (0.3) 3 (1, 4) 94 (0.9) 2 (1, 4) 103 (2.5) 2 (1, 4) Prosthetics 6 (0.1) 1 (1, 1) 10 (0.1) 1 (1, 1) 42 (0.4) 1 (1, 1) 18 (0.4) 1 (1, 1) Implant services 15 (0.2) 1 (1, 3) 208 (2.7) 1 (1, 2) 781 (7.7) 2 (1, 3) 391 (9.3) 2 (1, 3) *Among all persons with some dental information, this is the number (and per cent) of persons with a specific dental service. †Among persons who received a service of this type, this is the median number of services received and the 25th centile (Q1) and 75th centile (Q3). ‡Oral and maxillofacial surgery. Findings to date As an initial proof-of-concept project, a team of Mayo Clinic investigators used this resource to determine whether the 2007 guidelines from the American Heart Association affected prescription rates of antibiotics to patients with moderate-risk cardiac conditions prior to dental procedures. They used the existing REP research infrastructure to identify all patients residing in Olmsted County, Minnesota, between 2005 and 2014 with a cardiac condition that carries a moderate or high risk of developing infective endocarditis. The investigators reviewed the dental records for patients that received dental care with a participating dental practitioner (n=1351 patients with 8787 dental visits). They found that antibiotic prescriptions prior to dental procedures declined from 62% to 7% in this group of patients following the change in guidelines.18 This study demonstrates the utility of the expanded REP research infrastructure to address research questions that encompass medical and oral health. Strengths and limitations A major strength of this expansion is the availability of linked medical and dental health data for over 30 000 individuals. This large sample size provides a unique opportunity to address a wide range of oral health research questions. A limitation of this project is the relatively small number of dental partners participating in the REP. In 2014, more than 40 private dental practices were located in Olmsted County; however, only eight currently participate in the REP. We are not able to compare the type of dental data available for these eight practices to dental data for the entire county, because such data are not available. It is possible that the data available from these eight practices are not representative of the type of dental care received by all community members. In addition, the proportion of persons of black race or Hispanic ethnicity with available dental data is lower than the proportion of persons with available medical data, indicating that we are disproportionately missing dental data on our racial/ethnic minority population. Further efforts to engage other community dental practitioners will provide a more complete picture of the oral health for the entire community. In particular, Olmsted County has a single oral health service which provides free or low-cost dental services to the underinsured and uninsured local population, and discussions are ongoing with this service for inclusion in the REP. Collaboration We have successfully expanded a community-based partnership of healthcare providers to include information from eight community dental practitioners for over 30 000 persons. Data are linked at the person level, and it is now possible to identify cohorts of patients who received a particular type of dental service and to determine whether such services are associated with other types of health outcomes. It is also possible to identify persons with a specific disease or condition and assess the use of dental services either before or after the diagnosis. We welcome inquiries regarding the use of this research infrastructure for specific projects, and have a long history of collaboration. Details regarding access to REP data for research are available on our website at: http://www.rochesterproject.org. For further information, please contact us at info@rochesterproject.org. The authors would like to thank our dental partners for their collaboration. The authors would also like to thank Robin Adams for her assistance with manuscript preparation and submission. Contributors: JLSS, ABC, BPY, LJFR and WAR were involved in the conception and design of the study. ABC, CMB-G, LLK and JJP engaged dental practitioners and collected the data. JLSS, BRG, CMB-G and WAR participated in data analysis. All authors contributed to interpretation of the data and critical revision of the manuscript. All authors also approved the final version to be published. Funding: This study was made possible using the resources of the Rochester Epidemiology Project, which is supported by the National Institute on Aging of the National Institutes of Health under award number R01AG034676. This study was also supported by the Robert D. and Patricia E. Kern Mayo Clinic Center for the Science of Healthcare Delivery Population Health Research Program. Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the Center for the Science of Healthcare Delivery. Competing interests: None declared. Ethics approval: This study was approved by the Mayo Clinic and Olmsted Medical Center Institutional Review Boards. Informed consent was waived; however, we did not include data for any patients who had not given permission for their medical or oral health records to be used for research (Minnesota Research Authorization). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Researchers can request access to the data by submitting a request to info@rochesterproject.org. The Rochester Epidemiology Project website provides further information for interested investigators: http://rochesterproject.org/ ==== Refs References 1 Razak PA , Richard KM , Thankachan RP Geriatric oral health: a review article . J Int Oral Health 2014 ;6 :110 –16 .25628498 2 Dutt P , Chaudhary S , Kumar P Oral health and menopause: a comprehensive review on current knowledge and associated dental management . Ann Med Health Sci Res 2013 ;3 :320 –3 . 10.4103/2141-9248.117926 24116306 3 Slade GD Oral health-related quality of life is important for patients, but what about populations? Community Dent Oral Epidemiol 2012 ;40( Suppl 2 ):39 –43 . 10.1111/j.1600-0528.2012.00718.x 22998303 4 Bahekar AA , Singh S , Saha S The prevalence and incidence of coronary heart disease is significantly increased in periodontitis: a meta-analysis . Am Heart J 2007 ;154 :830 –7 . 10.1016/j.ahj.2007.06.037 17967586 5 Tavares M , Lindefjeld Calabi KA , San Martin L Systemic diseases and oral health . Dent Clin North Am 2014 ;58 : 797 –814 . 10.1016/j.cden.2014.07.005 25201543 6 Zalewska A , Waszkiewicz N , Szajda SD Impact of salivary flow and lysozyme content and output on the oral health of rheumatoid arthritis patients . Postepy Hig Med Dosw (Online) 2011 ;65 :40 –5 . 10.5604/17322693.932232 21357993 7 Tanaka TI , Geist SM Dermatomyositis: a contemporary review for oral health care providers . Oral Surg Oral Med Oral Pathol Oral Radiol 2012 ;114 :e1 –8 . 10.1016/j.oooo.2012.07.434 8 Widmer RP Oral health of children with respiratory diseases . Paediatr Respir Rev 2010 ;11 :226 –32 . 10.1016/j.prrv.2010.07.006 21109181 9 Lam OL , Zhang W , Samaranayake LP A systematic review of the effectiveness of oral health promotion activities among patients with cardiovascular disease . Int J Cardiol 2011 ;151 :261 –7 . 10.1016/j.ijcard.2010.11.014 21176980 10 Anil S , Preethanath RS , AlMoharib HS Impact of osteoporosis and its treatment on oral health . Am J Med Sci 2013 ;346 :396 –401 . 10.1097/MAJ.0b013e31828983da 23588259 11 Stark PC , Kalenderian E , White JM Consortium for oral health-related informatics: improving dental research, education, and treatment . J Dent Educ 2010 ;74 :1051 –65 .20930236 12 Walji MF , Kalenderian E , Stark PC BigMouth: a multi-institutional dental data repository . J Am Med Inform Assoc 2014 ;21 :1136 –40 . 10.1136/amiajnl-2013-002230 24993547 13 Rocca WA , Yawn BP , St Sauver JL History of the Rochester Epidemiology Project: half a century of medical records linkage in a US population . Mayo Clin Proc 2012 ;87 :1202 –13 . 10.1016/j.mayocp.2012.08.012 23199802 14 St Sauver JL , Grossardt BR , Leibson CL Generalizability of epidemiological findings and public health decisions: an illustration from the Rochester Epidemiology Project . Mayo Clin Proc 2012 ;87 :151 –60 . 10.1016/j.mayocp.2011.11.009 22305027 15 St Sauver JL , Grossardt BR , Yawn BP Data resource profile: the Rochester Epidemiology Project (REP) medical records-linkage system . Int J Epidemiol 2012 ;41 :1614 –24 . 10.1093/ije/dys195 23159830 16 St Sauver JL , Grossardt BR , Yawn BP Use of a medical records linkage system to enumerate a dynamic population over time: the Rochester Epidemiology Project . Am J Epidemiol 2011 ;173 :1059 –68 . 10.1093/aje/kwq482 21430193 17 Roesch A Matching data using sounds-like operators and SAS® compare functions . In: SAS , ed. SAS Global Forum 2012 , 2012 :1 –11 . 18 DeSimone DC, El Rafei A, Challener DW, et al. Impact of the American Heart Association's 2007 guidelines on the practice of dental prophylaxis for the prevention of infective endocarditis in Olmsted County, MN. Mayo Clin Proc (in press) .

PMC005xxxxxx/PMC5372052.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01256710.1136/bmjopen-2016-012567Evidence Based PracticeResearch15061694172317321706Risk of tuberculosis in patients treated with TNF-α antagonists: a systematic review and meta-analysis of randomised controlled trials Zhang Zheng 1Fan Wei 12Yang Gui 3Xu Zhigao 2Wang June 1Cheng Qingyuan 1Yu Mingxia 13 1 Department of Clinical Laboratory & Center for Gene Diagnosis, Zhongnan Hospital of Wuhan University, Wuhan, Hubei, China 2 Department of Pathology, Zhongnan Hospital of Wuhan University, Wuhan, Hubei, China 3 Department of Clinical Laboratory, Zhongnan Hospital of Wuhan University, Wuhan, Hubei, ChinaCorrespondence to Dr Mingxia Yu; dewrosy520@163.com2017 22 3 2017 7 3 e0125679 5 2016 22 2 2017 24 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives An increased risk of tuberculosis (TB) has been reported in patients treated with TNF-α antagonists, an issue that has been highlighted in a WHO black box warning. This review aimed to assess the risk of TB in patients undergoing TNF-α antagonists treatment. Methods A systematic literature search for randomised controlled trials (RCTs) was performed in MEDLINE, Embase and Cochrane library and studies selected for inclusion according to predefined criteria. ORs with 95% CIs were calculated using the random-effect model. Subgroup analyses considered the effects of drug type, disease and TB endemicity. The quality of evidence was assessed using the Grades of Recommendation, Assessment, Development and Evaluation (GRADE) approach. Results 29 RCTs involving 11 879 patients were included (14 for infliximab, 9 for adalimumab, 2 for golimumab, 1 for etanercept and 3 for certolizumab pegol). Of 7912 patients allocated to TNF-α antagonists, 45 (0.57%) developed TB, while only 3 cases occurred in 3967 patients allocated to control groups, resulting in an OR of 1.94 (95% CI 1.10 to 3.44, p=0.02). Subgroup analyses indicated that patients of rheumatoid arthritis (RA) had a higher increased risk of TB when treated with TNF-α antagonists (OR 2.29 (1.09 to 4.78), p=0.03). The level of the evidence was recommended as ‘low’ by the GRADE system. Conclusions Findings from our meta-analysis indicate that the risk of TB may be significantly increased in patients treated with TNF-α antagonists. However, further studies are needed to reveal the biological mechanism of the increased TB risk caused by TNF-α antagonists treatment. TNF-α antagonistsmeta-analysisRCTs ==== Body Strengths and limitations of this study This meta-analysis evaluated the tuberculosis (TB) risk of all TNF-α antagonists across a variety of conditions in randomised controlled trials (RCTs) with low heterogeneity. In addition to the diseases most commonly treated by TNF-α antagonists (rheumatoid arthritis, ulcerative colitis, ankylosing spondylitis and psoriatic arthritis), the review included studies that involved patients with asthma, sarcoidosis and graft-versus-host disease. The quality of the evidence was assessed using the GRADE approach, which has been recommended for grading evidence by the British Medical Journal since 2006. The relatively short follow-up period in the RCTs might have caused an underestimation of the TB rates. Introduction Tumour necrosis factor-α (TNF-α) is a pleiotropic cytokine that plays a central role in the pathogenesis of rheumatoid arthritis (RA), inflammatory bowel disease (IBD), ankylosing spondylitis (AS) and other immune-mediated or inflammation-related diseases.1 Therefore, it is a critical molecular member in targeted biological interventions,2 and the advent of TNF-α-directed targeted therapies represents a major advance in the treatment and management of conditions such as RA, psoriatic arthritis (PsA) and IBD,3–5 improving the quality of life for these patients.6 Increasingly, evidence indicate that TNF-α antagonists may possess promising therapeutic potential in many TNF-α-mediated diseases. Our previous study showed that TNF-α played a critical role in the occurrence and development of inflammation and tumour, and the TNF-α monoclonal antibody which we prepared as a TNF-α antagonist significantly suppressed the growth of breast cancer in an animal model.7 To date, five TNF-α antagonists have been used in clinical practice: etanercept, adalimumab, infliximab, golimumab and certolizumab pegol. Although their therapeutic efficacy has been confirmed, the side effects of these TNF-α antagonists need to be considered carefully in clinical practice.8 An increased risk of tuberculosis (TB) among patients receiving TNF-α antagonists has been observed,9 and several meta-analyses have evaluated the risk of TB in patients treated with TNF-α antagonists or with specific conditions.10–13 Nevertheless, the association between TNF-α antagonists and an increased risk of TB remains uncertain. With the aim of further clarifying the issue, this meta-analysis compared the risk of TB between TNF-α antagonists treatment and control groups in randomised controlled trials (RCTs) focusing on any disease condition. A secondary objective was to investigate the association of the rate of active TB with the type of medication, the disease condition and the location of study. Materials and methods The review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.14 Inclusion and exclusion criteria We performed a search for all published RCTs that reported TB risk among patients treated with any of the existing five TNF-α antagonists: etanercept (ETN), adalimumab (ADA), infliximab (IFX), golimumab (GOL) and certolizumab pegol (CZP). Studies were selected for inclusion according to predefined inclusion criteria: Participants: Adults (aged 16 years or older) with any disease included in studies of any of the five TNF-α antagonists. Interventions: TNF-α antagonists ETN, ADA, IFX, GOL or CZP with or without standard-care treatment for any medical condition. Comparators: Placebo with or without standard-care treatment or standard-care treatment alone. Outcomes: Diagnosis of TB, TB reactivation, miliary or cavitary TB of the lung or any other body organ. Study design: RCTs. The exclusion criteria included: (1) duplicated studies or studies based on unoriginal data, (2) studies that did not report TB incidence, (3) studies that did not observe TB events and (4) articles not published in English. Data sources and search strategies We systematically searched for reports of trials and systematic reviews up to December 2015 from the following online databases: MEDLINE, Embase and Cochrane Library. No restrictions were imposed with regard to region and time. To identify all RCTs, a highly sensitive search strategy developed on the basis of Cochrane Handbook for Systematic Reviews of Interventions was applied, which combined with the following key terms: ‘etanercept’, ‘adalimumab’, ‘infliximab’, ‘golimumab’, ‘certolizumab’ and ‘TNF-α antagonist’ (The MEDLINE search strategy is provided in online supplementary appendix 1). In addition, the reference lists of all topic-related review articles, reports or meta-analyses were searched for potentially relevant studies. 10.1136/bmjopen-2016-012567.supp1supplementary appendices Selection of studies Two reviewers independently screened the titles and abstracts of all records retrieved by the searches and identified studies that were potentially eligible for inclusion. Full-text versions were obtained, and these were independently assessed for eligibility by two reviewers according to inclusion and exclusion criteria. Disagreements between reviewers at both stages of screening were resolved by discussion and consensus. Data extraction and methodological quality assessment Data extraction was conducted independently by two investigators, and discrepancies were resolved through discussion. For each included study, we extracted essential information, including publication details, sample size, characteristics of trial participants, timing of assessment, interventions/comparisons, incidence cases of TB, performance of TB screening prior to therapy and geographic location of the study classified according to the incidence rate (IR) of TB (WHO, incidence TB estimation, 2014). Countries with an IR ≥40/100 000 are considered as high-incidence TB areas. The methodological quality of all included RCTs was assessed using the Cochrane collaboration's tool. The tool contains seven dimensions: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting and other bias. Studies were considered as low risk of bias when all these key aspects were assessed to be at low risk. Statistical analysis Principal statistical analyses were performed using Review Manager 5.2 software according to the Cochrane handbook. On the basis of events reported by included studies, the number of patients developing TB was compared between the placebo-controlled or standard-care populations and patients receiving at least one dose of TNF-α antagonists. Statistical heterogeneity among results was evaluated by using the I² statistic with the significance level set at 0.1. Meta-analyses were performed using the random-effects model. Results were presented as OR and its 95% CI. An OR >1 suggests a higher risk of TB than the control. Publication bias was tested by funnel plots, Egger's regression method and Begg's rank correlation method, using Stata software (V.11.0, College Station, Texas, USA). To evaluate the influence of all single studies on the pooled outcome, we also performed sensitivity analysis through the leave-one-out approach. Stratified analyses were performed by type of medication, disease being treated and estimated TB rates of studies' geographic locations. Quality of evidence We assessed the quality of evidence using the Grades of Recommendation, Assessment, Development and Evaluation (GRADE) methods.15 GRADEprofiler 3.6 software was applied to create the evidence profile. The GRADE approach categorises the quality of evidence as follows: (1) high quality (further research is extremely unlikely to change the credibility of the pooled results); (2) moderate quality (further research is likely to influence the credibility of pooled results and may change the estimate); (3) low quality (further research is extremely likely to influence the credibility of pooled results and is likely to change the estimate) and (4) very low quality (the pooled results have extreme uncertainty). Results Search results A total of 6843 study records were identified following the search strategy; 2773 references were left after removing duplicates. After title and abstract screening, 187 references progressed to the next stage, in which articles were re-evaluated based on full texts. Ultimately, 27 RCTs met the inclusion criteria and were included in our meta-analysis. In addition, two records were added after checking the references of previous systematic reviews.16 17 The PRISMA flow diagram of study selection is presented in figure 1. Figure 1 PRISMA flow diagram of study selection. Study characteristics and methodological quality The 29 included studies involved a total of 11 879 patients.16–44 The duration of outcome assessment in included studies ranged from 8 weeks to 3 years. Fourteen trials assessed infliximab, two trials assessed golimumab, nine trials assessed adalimumab, one trial assessed etanercept and three trials assessed certolizumab pegol. Thirteen RCTs were in areas with a low IR of TB and eleven in areas with a high incidence; this information was unavailable in the remaining five RCTs (table 1). TB screening was reported in 26 RCTs but was not carried out in 3 trials. A total of 45 TB cases occurred among 7912 patients treated with TNF-α antagonists and only 3 cases developed in 3967 patients in the control groups (see online supplementary appendix 2). The methodological quality assessments of included studies are summarised in online supplementary appendix 3. Table 1 Characteristics of randomised controlled trials included First author Year Disease Timing of assessment Comparison EA Kim16 2007 RA Week 24 PBO vs ADA No Rutgeerts17 2005 UC Week 54 PBO vs IFX Yes Baranauskaite18 2012 PsA Week 16 MTX vs IFX+MTX Yes Barker19 2011 Ps Week 24 MTX vs IFX – Braun20 2002 AS Week 12 PBO vs IFX No Breedveld21 2006 RA Year 2 MTX vs ADA/ADA+MTX – Chen22 2009 RA Week 12 MTX vs ADA+MTX No Colombel23 2010 CD Week 20 AZA vs IFX/IFX+AZA – Couriel24 2009 GvH Month 6 MP vs IFX+MP No Judson25 2014 Sarcoidosis Week 44 PBO vs GOL – Kavanaugh26 2013 RA Week 26 PBO+MTX vs ADA+MTX Yes Kennedy27 2014 RA Week 12 PBO vs ADA No Keystone28 2004 RA Week 52 PBO+MTX vs ADA+MTX No Keystone29 2008 RA Week 52 PBO+MTX vs CZP+MTX Yes Maini30 1999 RA Week 102 DMARDs vs IFX+DMARDs No Nam31 2014 RA Week 78 PBO+MTX vs IFX+MTX No Reich32 2012 Ps Week 12 PBO vs CZP No Schiff33 2014 RA Year 2 ABA+MTX vs ADA+MTX No Schiff34 2008 RA Year 1 PBO+MTX vs IFX+MTX Yes Sieper35 2014 AS Week 28 PBO+NPX vs IFX+NPX Yes Smolen36 2009 RA Week 24 PBO+MTX vs CZP+MTX Yes St Clair37 2004 RA Week 54 PBO+MTX vs IFX+MTX No Suzuki38 2014 UC Week 8 PBO vs ADA No Tam39 2012 RA Month 6 MTX vs IFX+MTX Yes Van Den Bosch40 2002 AS Week 12 PBO vs IFX - van der Heijde41 2007 RA Year 3 MTX vs ETN/ETN+MTX Yes van Vollenhoven42 2011 RA Week 24 PBO+MTX vs ADA+MTX Yes Wenzel43 2009 Asthma Week 76 PBO vs GOL No Westhovens44 2006 RA Week 22 PBO+MTX vs IFX+MTX Yes ABA, abatacept; ADA, adalimumab; AS, ankylosing spondylitis; AZA, azathioprine; CD, Crohn's disease; CZP, certolizumab pegol; DMARDs, disease-modifying anti-rheumatic drugs; EA, endemic area of TB; ETN, etanercept; GOL, golimumab; GvH, graft-versus-host disease; IFX, infliximab; MP, methylprednisolone; MTX, methotrexate; NPX, naproxen; PBO, placebo; Ps, plaque psoriasis; PsA, psoriatic arthritis; RA, rheumatoid arthritis; UC, ulcerative colitis. TB risk and TNF-α antagonists Pooled analysis determined that treatment with TNF-α antagonists was associated with an increased occurrence of TB compared with control groups (OR 1.94 (1.10, 3.44), p=0.02; figure 2). No significant heterogeneity was detected (I²=0%). The funnel plot revealed no obvious asymmetry in distribution, suggesting a low likelihood of publication bias (see online supplementary appendix 4), and this was statistically confirmed by Begg's test (p=0.348) and Egger's regression asymmetry test (p=0.321). Sensitivity analysis using random-effects model suggested that pooled result was not affected substantially by any of the included studies (see online supplementary appendix 5). Figure 2 Meta-analysis of TB risk associated with TNF-α antagonists. TNF-α, tumour necrosis factor-α; TB, tuberculosis. We performed subgroup analyses based on type of medication, disease under treatment and TB rate of the geographic location. In these analyses, the type of drugs was not associated with statistically significant differences in the risk of TB between patients treated with TNF-α antagonists and control groups (IFX: 1.82 (0.82–4.06), ADA: 2.11 (0.73–6.12), CZP: 2.38 (0.42–13.42)) (see online supplementary appendix 6). When grouped for disease, a significantly increased TB risk was associated with anti-TNF-α drugs in RA patients (OR 2.29 (1.09 to 4.78), p=0.03) (figure 3). When analysed according to estimated TB rates of studies' geographic locations, ORs for studies in high or low TB rate areas were 2.39 (95% CI 0.97 to 5.90, p=0.06) and 1.64 (95% CI 0.70 to 3.88, p=0.26), respectively (figure 4). Figure 3 Subgroup analysis of TB risk in RA and AS patients. AS, ankylosing spondylitis; RA, rheumatoid arthritis; TB, tuberculosis. Figure 4 Subgroup analysis of TB risk in high or low TB rate areas. TB, tuberculosis. GRADE profile evidence The results of assessing the quality of evidence are shown in online supplementary appendix 7. The quality for the main result was recommended as ‘low’ by the GRADE system. Discussion TNF-α antagonists have been widely used in many rheumatic diseases due to their considerable therapeutic effects and are promising candidates for future clinical applications in many other relevant diseases.7 24 However, an increased risk of TB has been observed among patients receiving anti-TNF treatments,9 an issue that has been highlighted by WHO in a black box warning for TB and other opportunistic infections. This meta-analysis aimed to consider TB risk in any patient treated with TNF-α antagonists, with the premise that the adverse event profile of TNF-α antagonists would be similar irrespective of the condition being treated. Twenty-nine published RCTs involving 11 879 patients were eventually included. In addition to the diseases most commonly treated with TNF-α antagonists (RA, UC, AS and PsA), this review also included studies that involved patients with asthma, sarcoidosis and Graft-versus-Host disease (GvH). We found that the risk of TB was statistically significantly increased in patients treated with TNF-α antagonists. With patients being treated with any TNF-α antagonist for any disease included, the risk of TB was almost doubled compared with those in normal care or placebo comparator arms. This result is in accordance with previously reported suspicions that TNF-α antagonists could increase TB risk, but differs from the findings of two previous meta-analyses on this topic, which found no significantly increased TB risk among patients with chronic immune-mediated inflammatory diseases or RA treated with different TNF-α antagonists.10 11 One possible reason for this discrepancy may be the relatively small number of patients included in those meta-analyses. In order to take into account the effects of disease condition and the rate of TB in the background population on the pooled results, subgroup analyses were performed. When patients with RA were considered alone, the level of increased risk of TB in RA patients receiving TNF-α antagonists, compared with placebo or normal care groups, was higher than the increased risk among patients in any disease condition. Although it has been reported that RA patients showed an increased risk of TB when compared with the general population,45–47 the potential for anti-TNF drugs to increase this risk further should not be ignored. It was also expected that patients in endemic areas would have a higher risk of TB after treatment with anti-TNF agents. While the difference in TB incidence between anti-TNF treated patients and control groups was not statistically significant (p=0.06), the trend towards higher incidence was enough to suggest the likelihood of a repeatable difference, which indicates that safety studies should include patients from these areas to provide a true profile of the risk of infection. No differences in TB incidence were identified between anti-TNF-treated patients and controls when subgroup analyses were conducted by single drug types. However, it is likely that this is a result of the small number of included patients. TNF-α is an immune mediator that plays a critical role in protective mechanism against infections, especially TB. TNF increases the phagocytic capacity of macrophages and enhances intracellular killing of mycobacterium via the generation of reactive nitrogen and oxygen intermediates, effectively synergising with interferon (IFN)-γ.48 TNF-α is also involved in the pathological changes of latent tuberculous infection (LTBI), especially in maintaining the formation and function of granuloma which prevents mycobacterium from disseminating into the blood.49 These TNF-mediated immune mechanisms may explain the reason for the increased risk of TB in patients receiving anti-TNF agents’ treatment. The results of this review may have direct implications in the management of a large number of patients treated currently with biologics. Therapeutic approaches that include intensive screening and surveillance seem to be advisable when TNF-α antagonists are used. One review of infection risk associated with anti-TNF-α agents suggested that a patient eligible for such treatment should undergo a careful medical history and tests such as the TB skin test (TST) or chest X-ray to assess the risk of TB re-activation.50 Interferon-γ release assay (IGRA) is also established as an alternative to the TST in TB infection diagnosis, especially in the diagnosis of LTBI due to the higher specificity.51 Previous studies have shown that prophylaxis in patients before or during anti-TNF-α therapy with standard anti-TB regimen prevented reactivation effectively.52 53 One study estimated that preventive treatment in patients with LTBI can reduce the risk of reactivation by 65%.10 Some countries have formulated national guidelines to deal with LTBI before anti-TNF agents treatment.54 During the anti-TNF therapy, the patients should also be closely monitored at least once a year to identify reactivation of latent TB or new TB infection. Patients' adherence to isoniazid (INH) treatment is important for preventing the reactivation of latent TB. Screening and surveillance may be of particular importance when TNF-α antagonists are used as part of combined therapies. A previously published systematic review55 reported that, compared with monotherapy, the risk of TB was increased 13-fold when anti-TNF agents were combined with immunosuppressant agents such as methotrexate or azathioprine. Additionally, a recent network meta-analysis and Cochrane overview highlighted the association between different biologics including TNF-α antagonists and higher rates of adverse effects in several diseases. These adverse events included TB reactivation, although the roles of other factors potentially associated with TB reactivation were not fully illuminated.13 Several limitations in this study should be addressed. First, the review identified only a limited number of RCTs, with only two studies about golimumab and one about etanercept. Second, the relatively short follow-up period in the RCTs might have caused an underestimation of TB incidence rates. Third, the meta-analysis was limited to published scientific publications, and the omission of unpublished data from pharmaceutical trials may affect the pooled results. In summary, our results suggest that the risk of TB is doubled when patients with any condition are treated with anti-TNF-α drugs. When anti-TNF-α treatments are considered, the increased risk of TB should be part of the treatment decision-making process. Patients should be screened for LTBI and anti-TB prophylaxis or concomitant treatment should be considered. Further high-quality research regarding the long-term safety of biologics is needed to improve the safety of biological treatment in clinical use. The authors thank the authors of the primary studies and Dr Brian Buckley (Visiting Professor, Wuhan University) for assistance in preparation of the English language manuscript. Contributors: All authors conceived of and designed the study. ZZ and WF performed the literature search, data collection and statistical analysis. GY and ZX assessed the quality of articles. ZZ, JW and QC wrote the paper. MY and WF revised the manuscript. Funding: This study was supported by the National Natural Science Foundation of China (nos. 81472033 and 30901308), the National Science Foundation of Hubei Province (nos. 2013CFB233 and 2013CFB235), the Scientific and Technological Project of Wuhan City (2014060101010045), Hubei Province Health and Family Planning Scientific Research Project (WJ2015Q021) and Seeding Program of the Science and Technology Innovation from Zhongnan Hospital of Wuhan University (ZNPY2016054). Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Bradley JR TNF-mediated inflammatory disease . J Pathol 2008 ;214 :149 –60 . 10.1002/path.2287 18161752 2 Choy EH , Panayi GS Cytokine pathways and joint inflammation in rheumatoid arthritis . N Engl J Med 2001 ;344 :907 –16 . 10.1056/NEJM200103223441207 11259725 3 Baraliakos X , Listing J , Fritz C Persistent clinical efficacy and safety of infliximab in ankylosing spondylitis after 8 years—early clinical response predicts long-term outcome . Rheumatology (Oxford) 2011 ;50 :1690 –9 . 10.1093/rheumatology/ker194 21672969 4 Perrier C , Rutgeerts P Cytokine blockade in inflammatory bowel diseases . Immunotherapy 2011 ;3 :1341 –52 . 10.2217/imt.11.122 22053885 5 Bansback N , Sizto S , Sun H Efficacy of systemic treatments for moderate to severe plaque psoriasis: systematic review and meta-analysis . Dermatology (Basel) 2009 ;219 :209 –18 . 10.1159/000233234 19657180 6 Sánchez-Moya AI , García-Doval I , Carretero G Latent tuberculosis infection and active tuberculosis in patients with psoriasis: a study on the incidence of tuberculosis and the prevalence of latent tuberculosis disease in patients with moderate-severe psoriasis in Spain. BIOBADADERM registry . J Eur Acad Dermatol Venereol 2013 ;27 :1366 –74 . 10.1111/jdv.12011 23134268 7 Yu M , Zhou X , Niu L Targeting transmembrane TNF-alpha suppresses breast cancer growth . Cancer Res 2013 ;73 :4061 –74 . 10.1158/0008-5472.CAN-12-3946 23794706 8 Solovic I , Sester M , Gomez-Reino JJ The risk of tuberculosis related to tumour necrosis factor antagonist therapies: a TBNET consensus statement . Eur Respir J 2010 ;36 :1185 –206 . 10.1183/09031936.00028510 20530046 9 Raval A , Akhavan-Toyserkani G , Brinker A Brief communication: characteristics of spontaneous cases of tuberculosis associated with infliximab . Ann Intern Med 2007 ;147 :699 –702 . 10.7326/0003-4819-147-10-200711200-00006 18025446 10 Ai JW , Zhang S , Ruan QL The risk of tuberculosis in patients with rheumatoid arthritis treated with tumor necrosis factor-alpha antagonist: a meta-analysis of both randomized controlled trials and registry/cohort studies . J Rheumatol 2015 ;42 :2229 –37 . 10.3899/jrheum.150057 26472414 11 Souto A , Maneiro JR , Salgado E Risk of tuberculosis in patients with chronic immune-mediated inflammatory diseases treated with biologics and tofacitinib: a systematic review and meta-analysis of randomized controlled trials and long-term extension studies . Rheumatology (Oxford) 2014 ;53 :1872 –85 . 10.1093/rheumatology/keu172 24821849 12 Xie X , Chen J , Peng Y [Meta-analysis of infection risks of anti-TNF-alpha treatment in rheumatoid arthritis] . Zhong Nan Da Xue Xue Bao Yi Xue Ban 2013 ;38 :722 –36 . 10.3969/j.issn.1672-7347.2013.07.013 23908082 13 Singh JA , Wells GA , Christensen R Adverse effects of biologics: a network meta-analysis and Cochrane overview . Cochrane Database Syst Rev 2011 ;(2) :Cd008794 10.1002/14651858.CD008794.pub2 21328309 14 Moher D , Liberati A , Tetzlaff J Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement . BMJ 2009 ;339 :b2535 10.1136/bmj.b2535 19622551 15 Guyatt GH , Oxman AD , Vist GE GRADE: an emerging consensus on rating quality of evidence and strength of recommendations . BMJ 2008 ;336 :924 –6 . 10.1136/bmj.39489.470347.AD 18436948 16 Kim H-Y , Lee S-K , Song YW A randomized, double-blind, placebo-controlled, phase III study of the human anti-tumor necrosis factor antibody adalimumab administered as subcutaneous injections in Korean rheumatoid arthritis patients treated with methotrexate . APLAR J Rheumatol 2007 ;10 :9 –16 . 10.1111/j.1479-8077.2007.00248.x 17 Rutgeerts P , Sandborn WJ , Feagan BG Infliximab for induction and maintenance therapy for ulcerative colitis . N Engl J Med 2005 ;353 :2462 –76 . 10.1056/NEJMoa050516 16339095 18 Baranauskaite A , Raffayová H , Kungurov NV Infliximab plus methotrexate is superior to methotrexate alone in the treatment of psoriatic arthritis in methotrexate-naive patients: the RESPOND study . Ann Rheum Dis 2012 ;71 :541 –8 . 10.1136/ard.2011.152223 21994233 19 Barker J , Hoffmann M , Wozel G Efficacy and safety of infliximab vs. methotrexate in patients with moderate-to-severe plaque psoriasis: results of an open-label, active-controlled, randomized trial (RESTORE1) . Br J Dermatol 2011 ;165 :1109 –17 . 10.1111/j.1365-2133.2011.10615.x 21910713 20 Braun J , Brandt J , Listing J Treatment of active ankylosing spondylitis with infliximab: a randomised controlled multicentre trial . Lancet 2002 ;359 :1187 –93 . 10.1016/S0140-6736(02)08215-6 11955536 21 Breedveld FC , Weisman MH , Kavanaugh AF The PREMIER study: a multicenter, randomized, double-blind clinical trial of combination therapy with adalimumab plus methotrexate versus methotrexate alone or adalimumab alone in patients with early, aggressive rheumatoid arthritis who had not had previous methotrexate treatment . Arthritis Rheum 2006 ;54 :26 –37 . 10.1002/art.21519 16385520 22 Chen DY , Chou SJ , Hsieh TY Randomized, double-blind, placebo-controlled, comparative study of human anti-TNF antibody adalimumab in combination with methotrexate and methotrexate alone in Taiwanese patients with active rheumatoid arthritis . J Formos Med Assoc 2009 ;108 :310 –19 . 10.1016/S0929-6646(09)60071-1 19369178 23 Colombel JF , Sandborn WJ , Reinisch W Infliximab, azathioprine, or combination therapy for Crohn's disease . N Engl J Med 2010 ;362 :1383 –95 . 10.1056/NEJMoa0904492 20393175 24 Couriel DR , Saliba R , de Lima M A phase III study of infliximab and corticosteroids for the initial treatment of acute graft-versus-host disease . Biol Blood Marrow Transplant 2009 ;15 :1555 –62 . 10.1016/j.bbmt.2009.08.003 19896079 25 Judson MA , Baughman RP , Costabel U Safety and efficacy of ustekinumab or golimumab in patients with chronic sarcoidosis . Eur Respir J 2014 ;44 :1296 –307 . 10.1183/09031936.00000914 25034562 26 Kavanaugh A , Fleischmann RM , Emery P Clinical, functional and radiographic consequences of achieving stable low disease activity and remission with adalimumab plus methotrexate or methotrexate alone in early rheumatoid arthritis: 26-week results from the randomised, controlled OPTIMA study . Ann Rheum Dis 2013 ;72 :64 –71 . 10.1136/annrheumdis-2011-201247 22562973 27 Kennedy WP , Simon JA , Offutt C Efficacy and safety of pateclizumab (anti-lymphotoxin-alpha) compared to adalimumab in rheumatoid arthritis: a head-to-head phase 2 randomized controlled study (The ALTARA Study) . Arthritis Res Ther 2014 ; 16 :467 10.1186/s13075-014-0467-3 25359150 28 Keystone EC , Kavanaugh AF , Sharp JT Radiographic, clinical, and functional outcomes of treatment with adalimumab (a human anti-tumor necrosis factor monoclonal antibody) in patients with active rheumatoid arthritis receiving concomitant methotrexate therapy: a randomized, placebo-controlled, 52-week trial . Arthritis Rheum 2004 ;50 :1400 –11 . 10.1002/art.20217 15146409 29 Keystone E , Heijde Dv , Mason D Jr Certolizumab pegol plus methotrexate is significantly more effective than placebo plus methotrexate in active rheumatoid arthritis: findings of a fifty-two-week, phase III, multicenter, randomized, double-blind, placebo-controlled, parallel-group study . Arthritis Rheum 2008 ;58 :3319 –29 . 10.1002/art.23964 18975346 30 Maini R , St Clair EW , Breedveld F Infliximab (chimeric anti-tumour necrosis factor alpha monoclonal antibody) versus placebo in rheumatoid arthritis patients receiving concomitant methotrexate: a randomised phase III trial. ATTRACT Study Group . Lancet 1999 ;354 :1932 –9 . 10.1016/S0140-6736(99)05246-0 10622295 31 Nam JL , Villeneuve E , Hensor EM Remission induction comparing infliximab and high-dose intravenous steroid, followed by treat-to-target: a double-blind, randomised, controlled trial in new-onset, treatment-naive, rheumatoid arthritis (the IDEA study) . Ann Rheum Dis 2014 ;73 :75 –85 . 10.1136/annrheumdis-2013-203440 23912798 32 Reich K , Ortonne JP , Gottlieb AB Successful treatment of moderate to severe plaque psoriasis with the PEGylated Fab’ certolizumab pegol: results of a phase II randomized, placebo-controlled trial with a re-treatment extension . Br J Dermatol 2012 ;167 :180 –90 . 10.1111/j.1365-2133.2012.10941.x 22413944 33 Schiff M , Weinblatt ME , Valente R Head-to-head comparison of subcutaneous abatacept versus adalimumab for rheumatoid arthritis: two-year efficacy and safety findings from AMPLE trial . Ann Rheum Dis 2014 ;73 :86 –94 . 10.1136/annrheumdis-2013-203843 23962455 34 Schiff M , Keiserman M , Codding C Efficacy and safety of abatacept or infliximab vs placebo in ATTEST: a phase III, multi-centre, randomised, double-blind, placebo-controlled study in patients with rheumatoid arthritis and an inadequate response to methotrexate . Ann Rheum Dis 2008 ;67 :1096 –103 . 10.1136/ard.2007.080002 18055472 35 Sieper J , Lenaerts J , Wollenhaupt J Efficacy and safety of infliximab plus naproxen versus naproxen alone in patients with early, active axial spondyloarthritis: results from the double-blind, placebo-controlled INFAST study, Part 1 . Ann Rheum Dis 2014 ;73 :101 –7 . 10.1136/annrheumdis-2012-203201 23696633 36 Smolen J , Landewé RB , Mease P Efficacy and safety of certolizumab pegol plus methotrexate in active rheumatoid arthritis: the RAPID 2 study. A randomised controlled trial . Ann Rheum Dis 2009 ;68 :797 –804 . 10.1136/ard.2008.101659 19015207 37 St Clair EW , van der Heijde DM , Smolen JS Combination of infliximab and methotrexate therapy for early rheumatoid arthritis: a randomized, controlled trial . Arthritis Rheum 2004 ;50 :3432 –43 . 10.1002/art.20568 15529377 38 Suzuki Y , Motoya S , Hanai H Efficacy and safety of adalimumab in Japanese patients with moderately to severely active ulcerative colitis . J Gastroenterol 2014 ;49 :283 –94 . 10.1007/s00535-013-0922-y 24363029 39 Tam LS , Shang Q , Li EK Infliximab is associated with improvement in arterial stiffness in patients with early rheumatoid arthritis—a randomized trial . J Rheumatol 2012 ;39 :2267 –75 . 10.3899/jrheum.120541 22984272 40 Van Den Bosch F , Kruithof E , Baeten D Randomized double-blind comparison of chimeric monoclonal antibody to tumor necrosis factor alpha (infliximab) versus placebo in active spondylarthropathy . Arthritis Rheum 2002 ;46 :755 –65 . 10.1002/art.511 11920412 41 van der Heijde D , Klareskog L , Landewé R Disease remission and sustained halting of radiographic progression′ with combination etanercept and methotrexate in patients with rheumatoid arthritis . Arthritis Rheum 2007 ;56 :3928 –39 . 10.1002/art.23141 18050208 42 van Vollenhoven RF , Kinnman N , Vincent E Atacicept in patients with rheumatoid arthritis and an inadequate response to methotrexate: results of a phase II, randomized, placebo-controlled trial . Arthritis Rheum 2011 ;63 :1782 –92 . 10.1002/art.30372 21452294 43 Wenzel SE , Barnes PJ , Bleecker ER A randomized, double-blind, placebo-controlled study of tumor necrosis factor-alpha blockade in severe persistent asthma . Am J Respir Crit Care Med 2009 ;179 :549 –58 . 10.1164/rccm.200809-1512OC 19136369 44 Westhovens R , Yocum D , Han J The safety of infliximab, combined with background treatments, among patients with rheumatoid arthritis and various comorbidities: a large, randomized, placebo-controlled trial . Arthritis Rheum 2006 ;54 :1075 –86 . 10.1002/art.21734 16572442 45 Carmona L , Hernández-García C , Vadillo C Increased risk of tuberculosis in patients with rheumatoid arthritis . J Rheumatol 2003 ;30 :1436 –9 .12858438 46 Doran MF , Crowson CS , Pond GR Frequency of infection in patients with rheumatoid arthritis compared with controls: a population-based study . Arthritis Rheum 2002 ;46 :2287 –93 . 10.1002/art.10524 12355475 47 Arkema EV , Jonsson J , Baecklund E Are patients with rheumatoid arthritis still at an increased risk of tuberculosis and what is the role of biological treatments? Ann Rheum Dis 2015 ;74 :1212 –17 . 10.1136/annrheumdis-2013-204960 24608401 48 Bekker LG , Freeman S , Murray PJ TNFα controls intracellular mycobacterial growth by both inducible nitric oxide synthase-dependent and inducible nitric oxide synthase-independent pathways . J Immunol 2001 ;166 :6728 –34 . 10.4049/jimmunol.166.11.6728 11359829 49 Kindler V , Sappino AP , Grau GE The inducing role of tumor necrosis factor in the development of bactericidal granulomas during BCG infection . Cell 1989 ;56 :731 –40 . 10.1016/0092-8674(89)90676-4 2647299 50 Murdaca G , Spanò F , Contatore M Infection risk associated with anti-TNF-alpha agents: a review . Expert Opin Drug Saf 2015 ;14 :571 –82 . 10.1517/14740338.2015.1009036 25630559 51 Danielsen AV , Fløe A , Lillebaek T An interferon-gamma release assay test performs well in routine screening for tuberculosis . Dan Med J 2014 ;61 :A4856 .24947627 52 Carmona L , Gómez-Reino JJ , Rodríguez-Valverde V Effectiveness of recommendations to prevent reactivation of latent tuberculosis infection in patients treated with tumor necrosis factor antagonists . Arthritis Rheum 2005 ;52 :1766 –72 . 10.1002/art.21043 15934089 53 Yun JW , Lim SY , Suh GY Diagnosis and treatment of latent tuberculosis infection in arthritis patients treated with tumor necrosis factor antagonists in Korea . J Korean Med Sci 2007 ;22 :779 –83 . 10.3346/jkms.2007.22.5.779 17982222 54 Hauck FR , Neese BH , Panchal AS Identification and management of latent tuberculosis infection . Am Fam Physician 2009 ;79 :879 –86 .19496388 55 Lorenzetti R , Zullo A , Ridola L Higher risk of tuberculosis reactivation when anti-TNF is combined with immunosuppressive agents: a systematic review of randomized controlled trials . Ann Med 2014 ;46 :547 –54 . 10.3109/07853890.2014.941919 25105206

PMC005xxxxxx/PMC5372053.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01457510.1136/bmjopen-2016-014575Research MethodsProtocol150617301709171317221727Interventions to improve the self-management support health professionals provide for people with progressive neurological conditions: protocol for a realist synthesis http://orcid.org/0000-0002-6956-1100Davies Freya 1Wood Fiona 1Bullock Alison 2Wallace Carolyn 3Edwards Adrian 1 1 Division of Population Medicine, Cardiff University, Cardiff, UK 2 Cardiff Unit for Research and Evaluation in Medical and Dental Education, Cardiff University, Cardiff, UK 3 Faculty of Life Sciences and Education, University of South Wales, Pontypridd, UKCorrespondence to Dr Freya Davies; daviesf9@cardiff.ac.uk2017 20 3 2017 7 3 e0145754 10 2016 29 11 2016 12 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction Supporting self-management among people with long-term conditions is recognised as an important component of healthcare. Progressive neurological conditions (PNCs), for example, Parkinson's disease and multiple sclerosis are associated with problems such as fatigue and cognitive impairment which may make self-management more challenging. Health professionals may need to develop specific skills in order to provide effective self-management support for these patients. The review aims to develop explanatory theories about how health professional-targeted interventions to improve self-management support provision for people with PNCs operate in different circumstances. Methods and analysis A realist synthesis of the evidence is proposed. There are 2 priority questions for the review to address. These relate to the role of a shared concept of self-management support within the healthcare team, and the need to tailor the support provided to the requirements of people with PNCs. Key stakeholders will be involved throughout the process. The initial search strategy uses terms relating to (1) self-management, (2) health professionals and (3) PNCs. Searching, data extraction and synthesis will occur in parallel. Studies will be prioritised for inclusion based on anticipated contribution to generating explanatory theories. Key informant interviews are planned to direct supplementary searches and help further refine the theories developed. Results will be expressed in the form of context–mechanism–outcome configurations. Ethics and dissemination Publication guidelines on realist synthesis will be followed. The results will be published in a peer-reviewed journal and made available to organisations involved in the provision of health professional training. self-managementMEDICAL EDUCATION & TRAININGRoyal College of General Practitionershttp://dx.doi.org/10.13039/100010334SFB 2015-18 ==== Body Strengths and limitations of this study The application of a realist approach to evidence synthesis will lead to theory development about the contexts in which interventions are most likely to succeed. Describing the mechanisms by which existing interventions work will facilitate future theory-driven intervention design and evaluation. The breadth of interventions which might be considered to support self-management may make defining the scope of the review challenging. If evidence available relating to supporting people with progressive neurological conditions is limited the reviewers will need to consider the transferability of knowledge generated in other settings. Introduction People living with long-term conditions make decisions that relate to the management of their condition on a daily basis,1 from choosing how they use their medication to how they plan their activities. Corbin and Strauss2 suggest that self-managing a condition involves three tasks: medical management, role management and emotional management. Health professionals have tended to focus on optimising the medical management of conditions, but there is increasing understanding that the focus of efforts may need to shift towards an approach that encompasses all of these tasks to help people to live well with their condition.3 People often have different definitions of successful self-management compared with their clinicians, with patients emphasising the need for self-management support (SMS) that is relevant to the context of their lives.4 This may be particularly important in progressive neurological conditions (PNCs). PNCs are conditions in which patients experience a progressive deterioration in their functioning (eg, Parkinson's disease and multiple sclerosis). In these conditions successful self-management is not necessarily expected to modify the disease course itself, but may have a significant impact on how well people live with their symptoms. What is SMS? SMS may be delivered directly to patients, for example, via attendance at SMS programmes.5 These programmes may include activities such as information provision, emotional and behavioural management skills, and technical skill development.6 7 Condition-specific self-management programmes for people with a PNC often cover issues such as physical activity, medication adherence, cognitive impairment, depression and fatigue.8 Limitations to these types of approaches have been recognised, including the fact that patients who volunteer to attend such programmes may already be motivated and skilled in self-management,9 while those who may benefit from support most may not access these types of courses.10 If SMS becomes integrated into routine clinical care more patients will have access to support. To encourage this integration, interventions aiming to promote SMS may include indirect components delivered either to individual professionals (such as education and training) or at an organisational level (eg, financial incentives).11 The variety of skills health professionals require to effectively support self-management has been broadly divided into general person-centred skills (such as communication skills), behaviour change skills (eg, motivational interviewing) and organisation/system skills (eg, use of electronic recall systems).12 What is known about training health professionals to support self-management? Supporting self-management is not a straightforward task for clinicians as it requires judgements to be made around patient readiness, professional role boundaries and service expectations. The evidence for training health professionals to support self-management is mixed. While there is some evidence that training health professionals can change clinicians' behaviours,13 others have shown that clinicians failed to apply training in SMS in their routine work.14 Implementation of SMS in routine practice is recognised to be inherently complex, with multiple potential barriers at the levels of the patient, the professional and the wider organisation.11 The need for further research to understand how provider burden can be minimised and self-management programmes can be made more widely acceptable has been recognised.15 Previous suggestions for optimising professional-targeted interventions include involving staff members in the intervention design process; and ensuring that any intervention is seen as professionally desirable, and fits within existing clinical routines.11 16 The context into which an intervention will be delivered should be considered if the intervention design is to be successful. In particular, staff preconceptions about their role in supporting self-management, and its relative importance in relation to other tasks should be addressed.14 Supporting self-management in the context of PNCs Although supporting self-management has been shown to be challenging across a range of settings, supporting people with a PNC may raise particular issues. Depression, cognitive impairment and fatigue are common comorbidities in PNCs and may all make it more challenging for patients to effectively self-manage, and for professionals to know how best to support self-management in these circumstances.17 Professionals working in this area already have a wide remit including providing education and support, symptom management, medication advice, care coordination, and ongoing care planning. High workloads and a lack of time to meet all patient needs have been reported.18 Much of the available research evidence relating to self-management comes from conditions, such as diabetes, where objective measures of disease control which may respond to successful self-management are available. In PNCs the expected outcomes of supporting self-management are likely to be harder for professionals to define and measure. While this may mean that professionals are encouraged to take a more holistic view of supporting self-management than a narrow focus on the medical management of a condition, it may also lead to difficulties in recognising how interventions to support self-management add value to routine clinical care. Professionals are required to make their own judgements about the level of self-management that they might expect their patients to engage in, and the level of support they provide to attempt to facilitate this process. Interventions aiming to improve SMS provision need to influence these decisions. For example, some professionals may worry that expecting people with a PNC to take an active role in self-managing their condition could be excessively burdensome. Training which encourages exploration of the purpose and goals of SMS may work well for this staff group. Others may feel that they lack the time required to provide SMS. In this case, training which provides ideas which can be easily integrated into their current practice may be seen as most valuable. A review approach designed to take into account this type of complexity is therefore required. Methods and analysis Chosen methodology Research into continuing professional development activities has been criticised for focusing only on whether or not interventions work—without attention being paid to the mechanisms by which they have an effect19 or the relevant contextual influences that moderate their effectiveness. The realist synthesis approach has been proposed as an effective method for synthesising evidence from complex interventions which addresses these concerns.20 A realist synthesis uses a theory-driven approach, informed by an acknowledgement that interventions will operate differently when delivered into different contexts. Realist synthesis seeks to unpick what type of intervention works, for which professionals, working in which settings, to what extent and why. This is done through the development of programme theories, developed and refined throughout the review process, which describe how the context into which the intervention is delivered influences how the intervention functions (its ‘mechanism’) in order to produce a range of differing intended and unintended outcomes. We plan to use the realist synthesis approach to review the evidence about interventions which aim to increase or improve the support for self-management provided by health professionals working with people with PNCs. Training health professionals in SMS is by definition a complex intervention, consisting of multiple interacting components21 and therefore well suited to a synthesis approach that acknowledges this complexity. During this review, we will focus specifically on understanding how training in SMS and delivery of support operates at the level of the health professional, rather than at the level of the patient. The approach to searching for evidence in a realist review is more iterative than a traditional review procedure, and allows reviewers to purposively search for and select literature likely to be informative.20 This is likely to be particularly helpful in this review because self-management itself is a complex concept to define, and this also makes a conventional literature search challenging.22 The more inclusive nature of a realist review allows data which may not be indexed under the heading of SMS (but do relate to an important element of SMS) to be included, as researcher judgement on relevance is used in place of strict inclusion/exclusion criteria. Realist reviews operate at the level of transferable programme theories (rather than at the level of a specific intervention). As a result, realist reviewers recognise the transferability of knowledge from other settings and may include evidence from areas that relate to the programme theory (but not necessarily the narrow topic area under review). Again this is likely to be relevant here, to ensure that this review identifies sufficient evidence to be useful and informative. A recent review of systematic reviews of self-management identified only limited evidence related to self-management in PNCs.11 However, evidence on the implementation of SMS for many other patient populations was identified and may be able to provide useful insights where condition-specific literature is sparse. One challenge for the review team will be in trying to decide to what extent knowledge gained from other settings may be transferable to the context of supporting people with PNCs with the challenges discussed above. Significant stakeholder involvement in the review process should help to ensure the relevance to the population of interest. Context of the review The review is part of a larger planned project which forms the basis of a PhD for FD. The programme theories generated during the review process will be used to design a theory-based training intervention. Use of realist reviews for this purpose has been recommended23 and applied in other settings.24 Current stage of review work The iterative nature of realist reviews means it is difficult to prespecify the direction of the review before significant work has already been undertaken to identify and prioritise areas of focus. The authors have attempted to strike a balance in producing this protocol at a point when the review has progressed sufficiently to be able to provide useful detail but not so far into the process to make this an entirely retrospective account. The accompanying online supplementary file indicates the steps in the review process completed at the time of writing and those still anticipated. Initial searching, data extraction and synthesis have all started, with further searching, extraction and synthesis planned. For ease of reading the initial search process is described retrospectively, and the subsequent searches, data extraction and synthesis are described prospectively. Our aim in publishing the protocol at this stage is to add transparency to the synthesis process, especially since the method is open to interpretation. 10.1136/bmjopen-2016-014575.supp1supplementary file The synthesis is being undertaken in parallel with two other pieces of work. The iterative nature of the review will mean that learning collectively from these work streams can usefully inform the direction of further searches and the refinement of the programme theories. An online survey of health professionals working with people with an exemplar PNC mulitple sclerosis (MS) was distributed in April to May 2016 with the aim of getting a snapshot of current practice, future training interests and important barriers in relation to the provision of SMS. Although the primary purpose of the survey data was to help prioritise specific intervention content for the later phase of the work, the data relating to barriers may helpfully inform programme theory development in the synthesis. Interviews with a small group of key informants with experience of training health professionals in skills relating to SMS are planned for October 2016. We will use a convenience sample of contacts made by the stakeholder group from a range of different training backgrounds. These interviews will allow the early developing programme theories from the synthesis to be discussed with the participants and subsequently further refined.25 A clear audit trail will be maintained so that the sources of programme theories remain transparent and these will be clearly reported on. Planned review strategy The planned review strategy was registered on the PROSPERO database (CRD42016035596). The review process will follow the five stages of realist review described by Pawson et al:26 clarification of scope, searching for evidence, appraising evidence and extracting data, synthesis, and dissemination. Clarification of scope A period of reading around the subject was undertaken by the first author (FD) which allowed key recurring themes from the wider literature about SMS to be identified. In the literature relating to training health professionals in SMS, specific SMS skills (and confidence in their use), perceptions of workplace fit and belief in the concept of SMS itself all appeared to be influential factors. Research exploring the implementation of SMS in practice identified issues that included patient-level barriers, the influence of health professional, local multidisciplinary team and wider organisational characteristics. These issues were discussed at an initial stakeholder group meeting in March 2016. Our stakeholder group includes the study authors who are academics from health (two of whom also work clinically as general practitioners), social sciences and education, with interests in SMS and/or postgraduate health professional training. Other members of the group were clinicians working with people with a PNC (MS specialist nurse and occupational therapist), service users with PNCs, a researcher working for a SMS training provider and third sector representation (MS trust). Unlike a traditional systematic review, key stakeholders are consulted throughout the review process from refining the focus of the review to challenging or validating emerging review findings.27 Informed by the group's discussion on priority areas, two key review questions were formulated, with the overarching aim of improving understanding of the circumstances in which health professionals could implement and sustain SMS. Therefore, the scope of the review was planned to include both professionals' experiences of receiving training in supporting self-management and their experiences of applying this training in clinical practice. The initial review questions chosen were: What is the influence of a shared concept of SMS within healthcare teams caring for people with PNCs and how can it be achieved? The first question aimed to examine what professionals thought about SMS, and how this was influenced by training interventions and existing contextual factors (eg, professional role, previous experience and workplace factors). What is known about how SMS can be successfully tailored for people with PNCs? This question aimed to focus on how SMS might need to be provided differently for people with PNCs than for other conditions, and to examine whether training health professionals to adopt a more flexible or tailored approach was important. Although dementia could be classified as a PNC, people with dementia are usually managed by a different healthcare team (old age psychiatry) to people with other PNCs, so for the purpose of this review we have not included dementia within our definition of PNCs.28 Search strategy Iterative searches were planned in line with the realist methodology. The overlap in the searching, extraction and synthesis processes is illustrated by the flow chart in online supplementary file S1. Our initial search strategy, designed with input from an information specialist used three search threads in combination: health professional terms, self-management terms and PNC terms (both relevant MeSH headings and free-text terms; see online supplementary file S2). Search terms relating to self-management were informed by terms used in previous systematic reviews4 11 29 and by terms which existing known papers were indexed under.30 At this stage the aim was to be as inclusive as possible. Therefore, terms relating to goal setting and health coaching were included as these were seen to be important skills related to supporting self-management but which might not be indexed under the term self-management. Although we initially planned to include a fourth search thread of terms relating to education or training, after piloting the searches we noted that relevant papers relating to implementation of SMS interventions were not identified, so we removed this thread from the search. 10.1136/bmjopen-2016-014575.supp2supplementary file The initial search was developed for MEDLINE via Ovid and then adapted for other databases (EMBASE, Cochrane Library, CINAHL, PEDro, ERIC and PsycInfo). The search was limited to English language papers (due to resource constraints) and to papers published in the past 20 years (as the concept of SMS is relatively recent). Following a particularly high recall from a search engine previously found to have a low specificity in relation to this topic (EMBASE),11 additional limitations were placed on the search to ensure only the most relevant papers were retained (non-Organisation for Economic Co-operation and Development (OECD) countries, children, palliative care and diagnosis-related studies were excluded). Details of the search dates and terms used are provided in online supplementary file S2. Initial searches were performed in April to May 2016. Going forward, a grey literature search for relevant websites and policy documents is planned. In addition, forward and backward citation tracking of key papers will be used together with hand searching of relevant journals. Key papers already known to the authors, and identified through initial scoping exercises will also be eligible for inclusion, as will any recommended by members of the stakeholder group. A clear audit trail of the source of included papers will be maintained. The need for and direction of further iterative searching will be informed by the findings of the ongoing synthesis, stakeholder advisory group and key informant interviews as described below. Data extraction Titles will initially be screened for basic relevance by FD. Any titles that are obviously irrelevant will be excluded at this stage. Studies will be excluded if they focus predominantly on: paediatric patients, carers or families, nursing homes/managed care settings, diagnostic or end-of-life period, epidemiology, imaging or testing, measurement instruments, and specific treatments or devices. An abstract screening tool developed by FD and tested in collaboration with FW will then be used to screen the remaining abstracts (see online supplementary file S3 for further detail). The tool will rank papers 1–4 based on their perceived likely relevance to either of the review questions. In brief, the highest ranked abstracts will be those that both related to a PNC and to health professionals' experiences of training in or implementation of SMS. Papers not specific to PNCs will be ranked lower, and those where professional involvement in SMS is unclear will be ranked as least likely to be relevant. Although the tool provides basic guidance on the likely relevance of papers for inclusion, author expertise and judgement will also be applied here to ensure that the tool is flexible enough to ensure potentially highly relevant papers are not deprioritised because they do not meet the predefined criteria. This application of researcher judgement is a key element of the realist approach to literature review which differs significantly from traditional systematic review.20 The full text of all papers ranked of the highest relevance will be sourced and assessed for potential inclusion. Full-text screening of the lower ranked abstracts will be undertaken selectively once data extraction from the initial papers provides further direction. If data saturation for some areas of the review is achieved early in the review process then it is anticipated that including data from these studies is unlikely to provide additional new information. Decisions about saturation will be made collaboratively through discussion among the authors. 10.1136/bmjopen-2016-014575.supp3supplementary file Realist reviewers do not generally rely on traditional quality assessment tools, but instead make judgements on each piece of included evidence based on relevance and rigour.20 At the full-text screening stage, prior to data extraction, the researcher will decide whether the paper can provide information relevant to the research questions. Reasons for exclusion on the basis of relevance will be recorded. The assessment of rigour will be an ongoing process in the data extraction and synthesis phases. The researcher will critically reflect on all evidence during this phase with the aim of safeguarding the inferences made on the basis of individual extracts by ensuring that they are used appropriately.20 A core set of descriptors for each study will be collected including identifiers (author, title, year), type of data (primary evidence, review, opinion piece), patient group details, staff group details, brief description of intervention, relationship with other studies included in the review and setting (country and healthcare setting). Data relevant to the research questions will be extracted in the form of explanatory accounts configured as ‘If–Then’ statements. For example: ‘If self-management is not valued by colleagues Then staff will feel discouraged from applying training in practice’. This approach was successfully used by another realist synthesis project which aimed to inform future training design.24 Extracting data as If–Then configurations rather than as context–mechanism–outcome configurations (the standard expression of realist programme theories) has the advantages of being an accessible way of starting to extract data with a ‘realist lens’, and providing a practical way for partial knowledge to accrue through the review process.24 A single ‘If–Then’ configuration may not contain each element of context, mechanism and outcome but may still be informative for the synthesis. When explanatory accounts derived from one data source are recognised to recur in another, this will be noted. The principles of meta-ethnography31 have been applied during realist synthesis in order to provide clear evidence of the type of data that is used to support the theories developed.32 We will follow this model so that during the extraction process data will be labelled as first order (direct from participants), second order (from study authors' interpretation) or third order (from synthesisers' interpretation of participants and authors' statements). Data synthesis We will take a similar approach to that described by Pearson et al24 to consolidating our initial explanatory accounts into more refined programme theories. The data synthesis process will begin while data extraction is still ongoing and be facilitated by regular discussion between the review team members. Initially the first author will group together apparently linked explanatory accounts. NVivo V.10 (QSR International) will then be used to facilitate movement between the explanatory accounts and original data. A ‘node’ will be created for each group of linked accounts and original data that was used to derive the constituent explanatory accounts will be coded under this node. This will allow the reviewers to look back at the original data when generating a consolidated account, to help ensure that the consolidated account continues to accurately reflect the source material. The consolidation process, which will result in refined explanatory accounts, will be carried out in conjunction with a second author (FW). In addition, throughout this process (once early in the consolidation process and once towards the end to confirm the refined context–mechanism–outcome configurations) two further stakeholder meetings will be held. The stakeholders will have an important role in ensuring that the researchers' interpretations of the literature are seen as both relevant and important by professionals. The stakeholder group will also help to prioritise which of the explanatory accounts are seen as crucially important to continue to pursue and which may be of less immediate relevance. If ‘priority’ programme theories are felt to be described in insufficient detail by the literature identified in the initial searches, supplementary targeted searches of the academic and grey literature will be performed. Key informant interviews are also planned. Individuals with experience of training health professionals using a variety of different approaches, all of which may relate to supporting self-management in some way, will be recruited (4–6 participants anticipated). These interviews will act as another check of the relevance of the theories developed. Trainers may also be able to help fill in the gaps not fully explored within the literature by reflecting on their own experiences, and to indicate whether any important areas have not yet been addressed. During the later stages of the review, once the programme theories are relatively refined, existing middle range theories33 which could help to further our understanding of the programme theories will be sought. There are already a number of candidate middle-range theories known to the authors thought to be potentially relevant to the review (eg, diffusion of innovations and normalisation process theory).34 35 These known theories will be considered along with any substantive theories used within the included papers to explain their findings. If none of these theories prove to be a good explanatory fit, targeted searching for theory will be undertaken.36 Ethics and dissemination Ethical approval is not required for the literature synthesis. However, ethical approval has been obtained for the online survey and supplementary interview data via Cardiff University School of Medicine Research Ethics Committee. The RAMESES publication standards for realist synthesis have been consulted during the planning of the review and will be followed for future publication.37 We will publish the synthesis in a peer-reviewed journal and make the findings available to relevant interested bodies including third sector organisations. We also aim for the theories to be useful to those designing training in SMS for health professionals, to help to identify what may be likely to work and where. Owing to the relatively limited data expected to be available that are specific to the clinical area (PNCs) and the intervention (improving SMS provision by health professionals), we recognise that some of the theories developed during the synthesis may be partially or weakly supported. We will be fully transparent about the level of evidence available to support each theory developed to allow readers to draw their own conclusions about the relevance of the developed theories to their own contexts of interest. The authors would like to thank the members of the stakeholder group for their time and contributions to the study. The authors also thank Mala Mann for assistance with developing the initial search strategies. Contributors: FD planned the synthesis approach with input from AB, AE, CW and FW. All authors participated in the initial stakeholder event to define the direction of the review. FD prepared the first draft of the protocol which was reviewed and critically revised by the other authors. Funding: The realist synthesis has been supported by a grant from the Royal College of General Practitioners Scientific Foundation Board (SFB 2015-18). Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. ==== Refs References 1 Lorig KR , Holman HR : Self-management education: history, definition, outcomes, and mechanisms . Ann Behav Med 2003 ;26 :1 –7 .12867348 2 Corbin J , Strauss A : Unending Work and Care: Managing Chronic Illness at Home . San Francisco: Jossey-Bass, 1988. 3 Morgan HM , Entwistle VA , Cribb A We need to talk about purpose: a critical interpretive synthesis of health and social care professionals’ approaches to self-management support for people with long-term conditions . Health Expect 2016 ; doi:10.1111/hex.12453 10.1111/hex.12453 4 Boger E , Ellis J , Latter S Self-management and self-management support outcomes: a systematic review and mixed research synthesis of stakeholder views . PLoS ONE 2015 ;10 :e0130990 10.1371/journal.pone.0130990 26162086 5 Rogers A , Kennedy A , Bower P The United Kingdom Expert Patients Programme: results and implications from a national evaluation . Med J Aust 2008 ;189 :S21 .19143586 6 de Silva D Evidence: helping people help themselves . London : The Health Foundation , 2011 . 7 Fraser R , Ehde D , Amtmann D Self-management for people with multiple sclerosis: report from the first international consensus conference, November 15, 2010 . Int J MS Care 2013 ;15 :99 –106 . 10.7224/1537-2073.2012-044 24453769 8 Plow MA , Finlayson M , Rezac M A scoping review of self-management interventions for adults with multiple sclerosis . PM R 2011 ;3 :251 –62 . 10.1016/j.pmrj.2010.11.011 21402370 9 Barlow J , Edwards R , Turner A The experience of attending a lay-led, chronic disease self-management programme from the perspective of participants with multiple sclerosis . Psychol Health 2009 ;24 :1167 –80 . 10.1080/08870440802040277 20204986 10 Hibbard JH , Gilburt H Supporting people to manage their health: an introduction to patient activation . London : King's Fund , 2014 . 11 Taylor SJ , Pinnock H , Epiphaniou E A rapid synthesis of the evidence on interventions supporting self-management for people with long-term conditions: PRISMS–Practical systematic Review of Self-Management Support for long-term conditions . Health Serv Deliv Res 2014 ;2 :53 10.3310/hsdr02530 12 Battersby M , Lawn S , Wells L An analysis of training and information options to support chronic disease prevention and self-management in primary health care . Adelaide : Flinders University , 2008 . 13 Jones F , Gage H , Drummond A Feasibility study of an integrated stroke self-management programme: a cluster-randomised controlled trial . BMJ Open 2016 ;6 :e008900 10.1136/bmjopen-2015-008900 14 Kennedy A , Rogers A , Bowen R Implementing, embedding and integrating self-management support tools for people with long-term conditions in primary care nursing: a qualitative study . Int J Nurs Stud 2014 ;51 :1103 –13 . 10.1016/j.ijnurstu.2013.11.008 24373719 15 McKenna S , Jones F , Glenfield P Bridges self-management program for people with stroke in the community: a feasibility randomized controlled trial . Int J Stroke 2015 ;10 :697 –704 . 10.1111/ijs.12195 24256085 16 Ong BN , Rogers A , Kennedy A Behaviour change and social blinkers? The role of sociology in trials of self-management behaviour in chronic conditions . Sociol Health Illn 2014 ;36 :226 –38 . 10.1111/1467-9566.12113 24528304 17 Davies F , Wood F , Brain K The transition to secondary progressive multiple sclerosis: an exploratory qualitative study of health professionals’ experiences . Int J MS Care 2016 ;18 :257 –64 . 10.7224/1537-2073.2015-062 27803641 18 Axelrod L , Gage H , Kaye J Workloads of Parkinson's specialist nurses: implications for implementing national service guidelines in England . J Clin Nurs 2010 ;19 :3575 –80 . 10.1111/j.1365-2702.2010.03279.x 21083781 19 Cervero RM , Gaines JK The impact of CME on physician performance and patient health outcomes: an updated synthesis of systematic reviews . J Contin Educ Health Prof 2015 ;35 :131 –8 . 10.1002/chp.21290 26115113 20 Pawson R Evidence-based policy: a realist perspective . London : SAGE Publications , 2006 . 21 Craig P , Dieppe P , Macintyre S Developing and evaluating complex interventions: the new Medical Research Council guidance . BMJ 2008 ;337 :a1655 10.1136/bmj.a1655 18824488 22 Panagioti M , Richardson G , Murray E Reducing Care Utilisation through Self-management Interventions (RECURSIVE): a systematic review and meta-analysis . Health Serv Deliv Res 2014 ;2 :54 10.3310/hsdr02540 23 Fletcher A , Jamal F , Moore G Realist complex intervention science: applying realist principles across all phases of the Medical Research Council framework for developing and evaluating complex interventions . Evaluation 2016 ;22 :286 –303 .27478401 24 Pearson M , Brand S , Quinn C Using realist review to inform intervention development: methodological illustration and conceptual platform for collaborative care in offender mental health . Implement Sci 2015 ;10 :134 10.1186/s13012-015-0321-2 26415961 25 Manzano A The craft of interviewing in realist evaluation . Evaluation 2016 ;22 :342 –60 . 10.1177/1356389016638615 26 Pawson R , Greenhalgh T , Harvey G Realist review—a new method of systematic review designed for complex policy interventions . J Health Serv Res Policy 2005 ;10 (Suppl 1 ):21 –34 . 10.1258/1355819054308530 16053581 27 Rycroft-Malone J , McCormack B , Hutchinson AM Realist synthesis: illustrating the method for implementation research . Implement Sci 2012 ;7 :33 10.1186/1748-5908-7-33 22515663 28 National Audit Office . Services for people with neurological conditions: progress review . London : National Audit Office , 2015 . 29 Mudge S , Kayes N , McPherson K Who is in control? Clinicians’ view on their role in self-management approaches: a qualitative metasynthesis . BMJ Open 2015 ;5 :e007413 10.1136/bmjopen-2014-007413 30 Yank V , Laurent D , Plant K Web-based self-management support training for health professionals: a pilot study . Patient Educ Couns 2013 ;90 :29 –37 . 10.1016/j.pec.2012.09.003 23031610 31 Atkins S , Lewin S , Smith H Conducting a meta-ethnography of qualitative literature: lessons learnt . BMC Med Res Methodol 2008 ;8 :1 10.1186/1471-2288-8-21 18215293 32 Punton M , Vogel I , Lloyd R Reflections from a realist evaluation in progress: scaling Ladders and Stitching Theory . Centre for Development Impact Practice Paper , 2016 :18 https://opendocs.ids.ac.uk/opendocs/handle/123456789/11254 33 Merton R On theoretical sociology: five essays, old and new . New York : The Free Press , 1967 . 34 Rogers EM Diffusion of innovations . 5th edn New York : Free Press , 2003 . 35 May C , Finch T Implementing, embedding, and integrating practices: an outline of normalization process theory . Sociology 2009 ;43 :535 –54 . 10.1177/0038038509103208 36 Booth A , Carroll C Systematic searching for theory to inform systematic reviews: is it feasible? Is it desirable? Health Info Libr J 2015 ;32 :220 –35 . 10.1111/hir.12108 26095232 37 Wong G , Greenhalgh T , Westhorp G RAMESES publication standards: realist syntheses . BMC Med 2013 ;11 :21 10.1186/1741-7015-11-21 23360677

PMC005xxxxxx/PMC5372054.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-013786corr110.1136/bmjopen-2016-013786corr11506CorrectionCorrection 2017 16 3 2017 7 3 e013786corr1Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ ==== Body Bell BG, Campbell S, Carson-Stevens A, et al. Understanding the epidemiology of avoidable significant harm in primary care: protocol for a retrospective cross-sectional study. BMJ Open 2017;7: e013786. doi:10.1136/bmjopen-2016-013786. This article has been corrected since it was first published. Several instances of ‘GP’ have been expanded for clarity. The title of reference number 26 has been corrected to ‘The preliminary development and testing of a global trigger tool to detect error and patient harm in primary-care records’.

PMC005xxxxxx/PMC5372055.txt

==== Front BMJ Open Diabetes Res CareBMJ Open Diabetes Res CarebmjdrcbmjdrcBMJ Open Diabetes Research & Care2052-4897BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjdrc-2017-00039410.1136/bmjdrc-2017-000394Clinical Care/Education/Nutrition/Psychosocial Research15061866Safety and efficacy of saxagliptin for glycemic control in non-critically ill hospitalized patients Garg Rajesh Schuman Brooke Hurwitz Shelley Metzger Cheyenne Bhandari Shreya Division of Endocrinology, Diabetes, and Hypertension, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, USACorrespondence to Dr Rajesh Garg; rgarg@partners.org2017 29 3 2017 5 1 e00039423 1 2017 2 3 2017 9 3 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective To evaluate whether saxagliptin is non-inferior to basal-bolus insulin therapy for glycemic control in patients with controlled type 2 diabetes mellitus (T2DM) admitted to hospital with non-critical illnesses. Research design and methods This was an open-label, randomized controlled clinical trial. Patients received either saxagliptin or basal-bolus insulin, both with correctional insulin doses. The main study outcome was the mean daily blood glucose (BG) after the first day of randomization. Results Of 66 patients completing the study, 33 (age 69±10 years, 40% men) were randomized to saxagliptin and 33 (age 67±10 years, 52% men) to basal-bolus insulin therapy. The mean daily BG was 149.8±22.0 mg/dL in the saxagliptin group and 146.9±30.5 mg/dL in the insulin group (p=0.59). With an observed group difference of 2.9 mg/dL and an a priori margin of 20 mg/dL, inferiority of saxagliptin was rejected in favor of non-inferiority (p=0.007). There was no significant difference in the percentage of high or low BG values. The insulin group received a higher number of insulin injections (2.3±1.7/day vs 1.2±1.9/day; p<0.001) as well as a higher daily insulin dose (13.3±12.9 units/day vs 2.4±3.3 units/day; p<0.001) than did the saxagliptin group. Continuous BG monitoring showed that glycemic variability was lower in the saxagliptin group as compared to the insulin group. Patient satisfaction scores were similar in the two groups. Conclusions We conclude that saxagliptin use is non-inferior to basal-bolus insulin in non-critically ill hospitalized patients with T2DM controlled on 0–2 oral agents without insulin. Saxagliptin use may decrease glycemic variability in these patients. Trial registration number NCT02182895. Inpatient Diabetes ManagementAstraZenecahttp://dx.doi.org/10.13039/100004325 ==== Body Significance of this study What is already known about this subject? Oral anti-diabetic agents are not currently recommended for treatment of diabetes in the hospital setting. Previous studies have shown sitagliptin use with basal insulin to be non-inferior to basal bolus insulin therapy. What are the new findings? DPP-4 inhibitors can be used safely and effectively without insulin in a subgroup of hospitalized patients with well-controlled diabetes before admission. Use of DPP-4 inhibitors may decrease the variability in glucose levels among hospitalized patients. Use of DPP-4 inhibitors may reduce insulin use and increase comfort for patients. How might these results change the focus of research or clinical practice? DPP-4 inhibitors, either alone or with basal insulin, should be allowed for inpatient glycemic control. Further research needs to investigate the glycemic control in hospitalized patients with mild diabetes without any treatment. Impact of non-insulin agent use on clinical outcomes in hospitalized patients needs further research. Introduction The relationship between inpatient hyperglycemia and poor clinical outcomes has been demonstrated in several observational studies.1–5 Treatment of hyperglycemia is associated with decreased mortality and morbidity among hospitalized patients.6–8 On the basis of these data, good glycemic control in hospitalized patients has been emphasized by many professional organizations.9 10 The current American Diabetes Association (ADA) guidelines recommend insulin as the preferred treatment for hospitalized patients.11 It is recommended that most critically ill patients should receive insulin infusion therapy and non-critically ill patients should receive basal-bolus insulin therapy in the hospital. Non-insulin hypoglycemic agents are not recommended due to multiple contraindications against many of these agents in acutely ill patients. For example, insulin secretagogues can cause hypoglycemia due to poor and unreliable nutritional intake in a hospitalized patient. Metformin and sodium glucose transporter (SGLT)-2 inhibitors are contraindicated in the presence of renal insufficiency or in patients at risk of developing renal insufficiency. Thiazolidinediones take a long time to act and are contraindicated in the presence of congestive heart failure or hepatic dysfunction. Moreover, many non-insulin agents including metformin and GLP-1 agonists can cause gastrointestinal side effects that are undesirable in an already sick hospitalized patient. None of the above contraindications apply to dipeptidyl peptidase-4 (DPP-4) inhibitors. If effective, DPP-4 inhibitors may be preferable over insulin because of the low risk of hypoglycemia. Hypoglycemia is associated with increased mortality and morbidity among hospitalized patients,12–14 and ADA guidelines strongly recommend avoiding hypoglycemia.11 DPP-4 inhibitors may also reduce glycemic variability that is associated with poor clinical outcomes.15 16 Therefore, we conducted a study with the aim of testing the safety and efficacy of DPP-4 inhibitor, saxagliptin, in non-critically ill hospitalized patients with type 2 diabetes mellitus (T2DM). Study design and methods This was an open label randomized controlled clinical trial (ClinicalTrials.gov identifier NCT02182895) conducted at a single center. The Partners HealthCare Institutional Review Board approved the study protocol, and all participants provided a written informed consent. Patients older than age 18 years with T2DM and HbA1C ≤7.5% on a ≤1 non-insulin hypoglycemic agent or HbA1C ≤7.0% on ≤2 non-insulin hypoglycemic agents were enrolled into the study after admission to the hospital for a non-critical illness. HbA1c was measured at the time of admission (unless available within the past 3 months) in all patients with diabetes as a standard of care. Exclusion criteria included admission to the intensive care unit (ICU), a history of diabetic ketoacidosis or hyperosmolar state, insulin treatment before admission to hospital, unable to take oral food or medications, systemic steroid use, pregnancy or breastfeeding, a history of pancreatitis or active gallbladder disease, end-stage renal disease on dialysis, hypersensitivity to saxagliptin or another contraindication against saxagliptin, and inability to provide an informed consent. Eligible participants were randomized by computer-generated numbers to one of the two groups: (1) DPP-4 inhibitor therapy: saxagliptin and (2) standard therapy: basal-bolus insulin regimen. The study statistician generated the randomization scheme using the web site randomization.com and kept it hidden. The investigators were unaware of the treatment assignment until the participant had signed the consent form and was determined to be eligible for the study. The DPP-4 inhibitor therapy group received saxagliptin 5 mg daily except for patients with eGFR <50 mL/min or using strong CYP3A4/5 inhibitors (eg, atazanavir, clarithromycin, indinavir, itraconazole, nefazodone, nelfinavir, ritonavir, saquinavir, telithromycin) who received saxagliptin 2.5 mg daily. Patients in the standard therapy group received basal-bolus insulin treatment at a starting dose of 0.5 units/kg/day, given half as insulin glargine once daily and half as insulin aspart divided into three equal doses before meals. However, a lower insulin dose was allowed at the discretion of the treating team if oral intake was poor or unpredictable. The goal of therapy was to maintain a fasting blood glucose (BG) concentration between 70 and 140 mg/dL and all other BG values <180 mg/dL. The doses of insulin were adjusted daily by 10–20% to achieve these goals as per the standard practice. In addition, both groups received the correctional sliding scale insulin therapy with insulin aspart before each meal and bedtime starting with 1 unit at BG >150 mg/dL and increasing by 1 unit for each 50 mg increment. Point of care BG levels were monitored before meals and at bedtime as per current standard practice. As a safety measure, two consecutive BG values >200 mg/dL in the saxagliptin arm led to withdrawal from the study and a switch to the basal-bolus insulin regimen. Patients were also withdrawn from the study if transferred to the ICU, started on systemic glucocorticoids, or became unable to take oral meals. The primary outcome of this study was the mean daily BG level obtained by point of care testing during study days 2–5. For those discharged before day 5, data were collected until the time of discharge. For those staying longer than 5 days, the study was stopped on day 6 and patients were switched to standard care with basal-bolus insulin therapy. Secondary outcomes included proportion of BG readings in 70–140 mg/dL range, average dose and number of insulin injections, incidence of hypoglycemia (BG <70 mg/dL), incidence of severe hypoglycemia (BG <50 mg/dL), incidence of hyperglycemia (BG >200 mg/dL), treatment failure with DPP-4 inhibitor, and length of hospital stay. All patients enrolled in the study were asked if they would agree to a continuous glucose monitoring (CGM) insertion. A subset of patients agreed and underwent CGM (iPRO2, Medtronic) to obtain data for glycemic variability. Glucose SD, mean amplitude of glycemic excursions (MAGE) and continuous overlapping net glycemic action (CONGA) were derived from the CGM data using software developed by Hill et al.17 18 A well-validated inpatient diabetes treatment satisfaction questionnaire (DTSQ-IP), items 1–16, was administered before the time of discharge.19 This questionnaire had 16 items that were scored on a scale of 0–6. For all the items except for items 2 and 3, ‘0’ indicates lowest satisfaction and ‘6’ indicates highest satisfaction. For items 2 and 3, ‘0’ indicates highest satisfaction and ‘6’ indicates lowest satisfaction. Statistical analysis With the mean daily BG as the main outcome, a sample size of 33 randomized to each treatment arm was determined to achieve at least 80% power to test the null hypothesis of saxagliptin inferiority relative to basal-bolus insulin, with 20 mg/dL as the a priori non-inferiority margin, and a within-group SD of 31 mg/dL.20 The number enrolled was 74 because of dropouts and missing data for some. Along with the non-inferiority hypothesis test for the mean daily BG, the upper limit of a one-sided 95% CI around the observed difference between arms was presented. The mean daily BG level during hospital included days 2–5. Day of enrollment into the study was defined as hospital day 1. All continuous data were summarized as mean±SD and categorical data as number with percent. The Wilcoxon-Mann-Whitney test was used to compare continuous variables, and the χ2 test was used to compare categorical variables. Data were analyzed using software SAS V.9.4. Results Of 74 patients who signed an informed consent form, 2 withdrew consent before the first dose of the study drug and 6 were discharged within 24 hours of enrollment into the study (figure 1). Complete data were available for 33 patients in the saxagliptin group and 33 patients in the basal-bolus insulin group. Figure 1 Patient flow through the study. Baseline patient characteristics are shown in table 1. All patients had well-controlled diabetes on 0–2 oral agents before admission to hospital and had relatively mild hyperglycemia at the time of randomization. The majority of patients were admitted under a surgical service. There were no differences in any of the baseline variables between the two groups. The main study outcomes are shown in table 2. The primary outcome was not different between the two groups, and the non-inferiority criteria for saxagliptin were satisfied. With an observed group difference of 2.9 mg/dL and an a priori non-inferiority margin of 20 mg/dL, the null hypothesis of inferiority was rejected in favor of the alternative hypothesis of non-inferiority (p=0.007). Further, the upper limit of a one-sided 95% CI around the difference between group means was 14.2 mg/dL, well within the non-inferiority margin. Table 1 Baseline characteristics Saxagliptin group N=33 Basal-bolus insulin group N=33 p Value Age, years 69±10 67±10 0.16 Gender, N (%) 0.32  Male 13 (40) 17 (52)  Female 20 (60) 16 (48) Race, N (%) 0.23  White 26 (79) 30 (91)  Other 7 (21) 3 (9) Admitting service, N (%) 0.16  Medicine 6 (18) 11 (33)  Surgery 27 (82) 22 (67) Duration of diabetes, years 7.5±8.5 7.5±5.8 0.64 Mean A1c, % 6.6±0.5 6.5±0.5 0.69 Preadmission diabetes medications, N (%) 0.06  None 8 (24) 5 (15)  Metformin 22 (67) 18 (55)  SU 2 (6) 10 (30)  Other 1 (3) 0 Weight, kg 97.7±32.9 92.4±21.2 0.54 BMI 34.5±12.2 32.6±6.5 0.91 Admission blood glucose, mg/dL 156.1±78.9 152.8±52.5 0.52 Prerandomization blood glucose, mg/dL 154.6±37.3 154.8±54.6 0.97 Serum creatinine 1.02±0.34 1.10±0.35 0.26 Presence of infection, N (%) 4 (12) 8 (24) 0.20 BMI, body mass index. Table 2 Outcome variables Saxagliptin group N=33 Basal-bolus insulin group N=33 p Value* Mean BG on day 1 mg/dL 154.8±28.2 156.0±32.1 1.0 Mean daily BG day 2-5 mg/dL† 149.8±22.0 146.9±30.5 0.59 BGs in 70–140, N (%) 102 (42) 105 (37) 0.16 Number of patient days with at least one BG >200, N (%) 14 (15) 22 (19) 0.27 BGs >200, N (%) 17 (6) 31 (10) 0.16 Number of patient days with at least one BG <70 1 1 BGs <70 (%) 0.4 0.3 NS Number of patient days with at least one BG <50 0 0 BGs <50 (%) 0 0 LOS, days 8.0±14.6 6.4±5.4 0.87 Mean daily insulin dose, units Total 2.4±3.3 13.3±12.9 <0.001 Basal 0 6.1±7.2 Bolus 2.4±3.3 7.4±7.1 <0.001 Mean number of injections per day 1.2±1.9 2.3±1.7 <0.001 DTSQ IP composite score (range 0–84) 70±12 75±11 0.19 DTSQ IP Q2+Q3 score (range 0–12) 3.9±3.7 2.5±3.0 0.12 *Tests of superiority include the Wilcoxon rank sum test and χ2 test. †Main study outcome for the non-inferiority test. With group difference of 2.9 mg/dL and a priori margin of 20 mg/dL, we reject inferiority of saxagliptin in favor of non-inferiority (p=0.007). BG, blood glucose. Other glycemic control indices were also similar between the two groups. However, insulin use, dose as well as the number of insulin injections, was significantly higher in the basal-bolus insulin group as compared to the saxagliptin group. Seven patients in the saxagliptin group and eight patients in the insulin group had BG >180 mg/dL at the time of randomization. Among these patients, the mean daily BG values (days 2–5) were 163.4±19.0 in the saxagliptin group and 161.7±38.0 in the insulin group (p=0.86). The groups had similar BG levels on day 1. However, over the course of the study, the saxagliptin group showed a downward trend in BG levels while the control group showed a decrease on day 2 and then stable BG values despite an increasing dose of insulin (figure 2). One patient in the saxagliptin group was withdrawn from the study and switched to basal bolus insulin due to two consecutive BG values >200 mg/dL. For this patient, data before switching to insulin were included in analysis. One patient in each group had one episode of BG <70 mg/dL. No patient in either group developed severe hypoglycemia. Figure 2 Mean blood glucose levels and insulin use during the course of the study. Continuous BG monitoring data were available in 36 patients, 20 in the saxagliptin group and 16 in the control group. Glucose SD and MAGE were lower in the saxagliptin group as compared to the insulin group (table 3). CONGA values were similar in the two groups. Table 3 Indices of glycemic variability Saxagliptin group N=20 Basal-Bolus insulin group N=16 p Value Glucose SD 1.13±0.47 1.61±0.73 0.03 MAGE 2.72±1.60 3.93±2.03 0.05 CONGA 7.38±1.46 7.56±1.58 0.86 CONGA, continuous overlapping net glycemic action; MAGE, mean amplitude of glycemic excursions. There was no difference in DTSQ-IP scores between the groups. The saxagliptin and control groups rated their diabetes inpatient treatment between 80% and 90% of the satisfaction score with no significant difference between the average scores. Similarly, there was no significant difference in inpatient dissatisfaction (blood sugars being unacceptably high or low) between the two groups. No patient needed additional surgery or was transferred to the ICU while enrolled into the study. Five patients in the saxagliptin group had known heart disease, and they were watched closely for heart failure. No new-onset heart failure or worsening of heart failure was observed in these patients. Discussion This study shows that saxagliptin is non-inferior to basal-bolus insulin for glycemic control in non-critically ill hospitalized patients with well-controlled diabetes before admission. There is no risk of hypoglycemia associated with saxagliptin use. Moreover, glycemic variability is lower with saxagliptin therapy than with basal-bolus insulin therapy. The study is important because current clinical practice guidelines recommend insulin for management of all hyperglycemia in hospitalized patients.11 We estimate that a substantial number of hospitalized patients with diabetes may be eligible for treatment with non-insulin agents. Avoiding insulin in these patients can make the inpatient diabetes management much simpler for the nursing staff and may also alleviate anxiety for the patients. The frequency of BG testing may be also be decreased in these patients. Overall, this may save nursing time and translate into lower hospitalization costs while improving patient comfort. A previous study by Umpierrez et al20 showed sitagliptin use to be non-inferior to basal-bolus insulin in hospitalized patients. Their study enrolled patients irrespective of their baseline HbA1c or previous insulin treatment. In that study, patients with baseline BG >180 mg/dL had higher mean daily BG levels during hospital stay when treated with sitagliptin alone as compared to treatment with insulin. Therefore, a subset of patients remained poorly controlled when treated with DPP-4 inhibitor alone irrespective of their baseline diabetes control. In a recent, multicenter, randomized controlled trial, the same group of investigators have shown that sitagliptin along with basal insulin is able to achieve glycemic control that is non-inferior to the basal-bolus insulin therapy.21 Thus, even in patients with relatively high BG levels at baseline, DPP-4 inhibitor use may obviate the need for nutritional insulin. We restricted our patient population to participants with milder hyperglycemia, who were most likely to respond to DPP-4 inhibitor therapy alone, using clinical criteria often used in the outpatient setting. Moreover, it has been shown that glycemic control prior to treating hyperglycemia in hospitalized patients can predict response to therapy.22 Only one patient was labeled as a non-responder in the saxagliptin group, thus demonstrating that this strategy was highly successful. Our study complements data generated by Umpierrez and colleagues on the safety and efficacy of DPP-4 inhibitors in the hospital setting and adds data on the glycemic variability. Our results are also consistent with the recently suggested algorithm for the use of DPP-4 inhibitors in patients with moderate hyperglycemia at admission.23 Patients in the basal-bolus insulin group showed no change in BG levels over the study period despite increasing insulin doses. This may be due to lower than recommended starting doses of insulin and a lag in escalating the insulin doses as the patients' oral intake improved. In this study, insulin doses were advised by the research team according to protocol but often cut back by the primary teams due to fear of hypoglycemia. Practically, fear of hypoglycemia (on the part of physicians) is one of the common reasons for insulin underdosing and relevant to our study because DPP-4 inhibitors do not cause hypoglycemia. Moreover, it is not uncommon for physicians to be cautious while adjusting insulin doses in the inpatient setting because oral intake is often unpredictable. Owing to the high risk of hypoglycemia with insulin, an increase in insulin doses is often based on the previous day's high BG numbers. This is a reasonable approach, even though it leads to overall high BG levels during the hospital stay. However, no dose adjustment is needed if DPP-4 inhibitor is used instead on insulin. Moreover, lower glycemic variability may be an additional advantage of using these agents in the inpatient setting as previous studies have demonstrated an association between higher glycemic variability and poor clinical outcomes irrespective of mean BG levels in hospitalized patients.15 16 The trend of lower BG levels during the course of study in our saxagliptin group remains unexplained; it may be due to the accumulating effect as the subsequent doses of saxgliptin were administered after day 1. One limitation of our study is the inability to evaluate clinical outcomes and to compare hospital complications between the two groups. Length of hospital stay was not different between the saxagliptin group and the basal-bolus insulin group. There has been some discussion whether the benefits of good glycemic control in acutely ill patients are the result of lower BG levels or a direct effect of insulin.24 If insulin has a direct effect independent of glucose levels, DPP-4 inhibitor therapy may not be able to decrease the risk of complications. A large, prospective, randomized, multicenter clinical trial would be necessary to investigate this hypothesis. Another limitation of our study is the lack of a placebo group. After completion of the study, we noticed that many of the enrolled participants had acceptable BG levels (<180 mg/dL) at the time of randomization. It is possible that their BG levels would have remained in acceptable range without any treatment. However, all these patients had known T2DM and as per current standards of clinical care, all of them should have received basal-bolus insulin therapy as inpatients. Nevertheless, physicians did not give insulin doses proposed by the study team. Insulin underdosing is one of the practical problems of inpatient diabetes management and a limitation of our study. We think future studies should include a placebo group to evaluate whether patients with baseline characteristics similar to those enrolled in this study may be left alone without any anti-diabetic treatment. It would also be worth comparing saxagliptin alone with a basal plus approach, used in the sitaglipitin studies. Another limitation of our study is the small sample size and unequal distribution of surgical and medical patients, making it hard to perform subgroup analyses, because the majority of our patients were admitted under surgical specialties and the results are not necessarily applicable to medical patients. In conclusion, this study demonstrates that saxagliptin use is safe and effective when compared with basal-bolus insulin in a subgroup of non-critically ill hospitalized patients with well-controlled T2DM on 0–2 oral agents without insulin use before admission. Saxagliptin use may decrease glycemic variability in these patients. Thus, DPP-4 inhibitors may be an alternative to insulin use in a subgroup of hospitalized patients with diabetes. We propose that non-critically ill hospitalized patients with HbA1c <7% on a ≤1 oral anti-diabetic agent or HbA1c <7.5% on a ≤2 oral anti-diabetic agent should be treated with a DPP-4 inhibitor as a first step. Fasting BG should be monitored, and basal insulin should be added if BG >140 mg/dL. Multiple daily insulin injections are unnecessary in the majority of these patients. Contributors: RG designed the study, obtained funding, led the program, supervised the data collection, analyzed the data, and wrote the manuscript; BS, CM, and SB helped in conducting the study, collected the data, and reviewed the manuscript; SH conducted all statistical analysis and collaborated on writing the manuscript. Funding: This work was supported by AstraZeneca Company through an investigator-initiated research grant. Competing interests: RG received research support from AstraZeneca for this study. Other authors have no conflict of interest. Patient consent: Obtained. Ethics approval: Partners Institutional Review Board. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Ishihara M , Kojima S , Sakamoto T Acute hyperglycemia is associated with adverse outcome after acute myocardial infarction in the coronary intervention era . Am Heart J 2005 ;150 :814 –20 . 10.1016/j.ahj.2004.12.020 16209987 2 Garg R , Bhutani H , Alyea E Hyperglycemia and length of stay in patients hospitalized for bone marrow transplantation . Diabetes Care 2007 ;30 :993 –4 . 10.2337/dc06-2563 17259474 3 Umpierrez GE , Isaacs SD , Bazargan N Hyperglycemia: an independent marker of in-hospital mortality in patients with undiagnosed diabetes . J Clin Endocrinol Metab 2002 ;87 :978 –82 . 10.1210/jcem.87.3.8341 11889147 4 Williams LS , Rotich J , Qi R Effects of admission hyperglycemia on mortality and costs in acute ischemic stroke . Neurology 2002 ;59 :67 –71 . 10.1212/WNL.59.1.67 12105309 5 Ascione R , Rogers CA , Rajakaruna C Inadequate blood glucose control is associated with in-hospital mortality and morbidity in diabetic and nondiabetic patients undergoing cardiac surgery . Circulation 2008 ;118 :113 –23 . 10.1161/CIRCULATIONAHA.107.706416 18591441 6 Krinsley JS Translating evidence into practice in managing inpatient hyperglycemia . J Hosp Med 2007 ;2 (Suppl 1 ):13 –19 . 10.1002/jhm.181 7 Umpierrez GE , Smiley D , Jacobs S Randomized study of basal-bolus insulin therapy in the inpatient management of patients with type 2 diabetes undergoing general surgery (RABBIT 2 surgery) . Diabetes Care 2011 ;34 :256 –61 . 10.2337/dc10-1407 21228246 8 Murad MH , Coburn JA , Coto-Yglesias F Glycemic control in non-critically ill hospitalized patients: a systematic review and meta-analysis . J Clin Endocrinol Metab 2012 ;97 :49 –58 . 10.1210/jc.2011-2100 22090269 9 Moghissi ES , Korytkowski MT , DiNardo M American Association of Clinical Endocrinologists and American Diabetes Association consensus statement on inpatient glycemic control . Diabetes care 2009 ;32 :1119 –31 . 10.2337/dc09-9029 19429873 10 Umpierrez GE , Hellman R , Korytkowski MT Management of hyperglycemia in hospitalized patients in non-critical care setting: an endocrine society clinical practice guideline . J Clin Endocrinol Metab 2012 ;97 :16 –38 . 10.1210/jc.2011-2098 22223765 11 American Diabetes Association . 13. Diabetes Care in the Hospital . Diabetes Care 2016 ;39 :S99 –S104 . 10.2337/dc16-S016 26696689 12 D'Ancona G , Bertuzzi F , Sacchi L Iatrogenic hypoglycemia secondary to tight glucose control is an independent determinant for mortality and cardiac morbidity . Eur J Cardiothorac Surg 2011 ;40 :360 –6 . 10.1016/j.ejcts.2010.11.065 21256761 13 Garg R , Hurwitz S , Turchin A Hypoglycemia, with or without insulin therapy, is associated with increased mortality among hospitalized patients . Diabetes Care 2013 ;36 :1107 –10 . 10.2337/dc12-1296 23248192 14 Turchin A , Matheny ME , Shubina M Hypoglycemia and clinical outcomes in patients with diabetes hospitalized in the general ward . Diabetes Care 2009 ;32 :1153 –7 . 10.2337/dc08-2127 19564471 15 Farrokhi F , Chandra P , Smiley D Glucose variability is an independent predictor of mortality in hospitalized patients treated with total parenteral nutrition . Endocr Pract 2014 ;20 :41 –5 . 10.4158/EP13131.OR 24013986 16 Mendez CE , Mok KT , Ata A Increased glycemic variability is independently associated with length of stay and mortality in noncritically ill hospitalized patients . Diabetes Care 2013 ;36 :4091 –7 . 10.2337/dc12-2430 24170754 17 Hill NR , Oliver NS , Choudhary P Normal reference range for mean tissue glucose and glycemic variability derived from continuous glucose monitoring for subjects without diabetes in different ethnic groups . Diabetes Technol Ther 2011 ;13 :921 –8 . 10.1089/dia.2010.0247 21714681 18 https://www.phc.ox.ac.uk/research/technology-outputs/easygv [article online] . 19 Sampson MJ , Singh H , Dhatariya KK Psychometric validation and use of a novel diabetes in-patient treatment satisfaction questionnaire . Diabet Med 2009 ;26 :729 –35 . 10.1111/j.1464-5491.2009.02754.x 19573123 20 Umpierrez GE , Gianchandani R , Smiley D Safety and efficacy of sitagliptin therapy for the inpatient management of general medicine and surgery patients with type 2 diabetes: a pilot, randomized, controlled study . Diabetes Care 2013 ;36 :3430 –5 . 10.2337/dc13-0277 23877988 21 Pasquel FJ , Gianchandani R , Rubin DJ Efficacy of sitagliptin for the hospital management of general medicine and surgery patients with type 2 diabetes (Sita-Hospital): a multicentre, prospective, open-label, non-inferiority randomised trial . Lancet Diabetes Endocrinol 2017 ;5 :125 –33 . 10.1016/S2213-8587(16)30402-8 27964837 22 Pasquel FJ , Gomez-Huelgas R , Anzola I Predictive value of admission hemoglobin A1c on inpatient glycemic control and response to insulin therapy in medicine and surgery patients with type 2 diabetes . Diabetes Care 2015 ;38 :e202 –203 . 10.2337/dc15-1835 26519335 23 Nauck MA , Meier JJ Sitagliptin plus basal insulin: simplifying in-hospital diabetes treatment? Lancet Diabetes Endocrinol 2017 ;5 :83 –5 . 10.1016/S2213-8587(16)30401-6 27964838 24 Garg R , Chaudhuri A , Munschauer F Hyperglycemia, insulin, and acute ischemic stroke: a mechanistic justification for a trial of insulin infusion therapy . Stroke 2006 ;37 :267 –73 . 10.1161/01.STR.0000195175.29487.30 16306459

PMC005xxxxxx/PMC5372056.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01435810.1136/bmjopen-2016-014358Public HealthProtocol1506172416991715Evidence on health-promoting lifestyle practices and information and communication technologies: scoping review protocol Joseph-Shehu Elizabeth M 12Ncama Busisiwe P 1 1 School of Nursing and Public Health, University of KwaZulu-Natal, Howard College, Durban, South Africa 2 Faculty of Health Sciences, National Open University of Nigeria, Jabi, Abuja, NigeriaCorrespondence to Elizabeth M Joseph-Shehu; ejoseph-shehu@noun.edu.ng; lizjoe26@gmail.com2017 29 3 2017 7 3 e01435820 9 2016 10 12 2016 25 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction Information and communication technologies (ICTs) play a key role in improving health and maintaining health promoting behaviours. ICTs are therefore one potential solution for promoting healthy lifestyles. In addition, they can assist in the reduction and control of the menace of both communicable and non-communicable diseases. This study will map evidence of interventions that demonstrate the effect of ICTs on health-promoting lifestyle practices that can prevent and control diseases. It is anticipated that this study will help identify areas where there is need for primary research. Methods and analysis The following electronic databases will be searched: PsycArticle (EBSCO), PsycINFO (EBSCO), Science direct, PubMed, Medline (EBSCO) and Google Scholar. The study will be conducted in two stages: the first stage will map out the studies descriptively while the second stage will map the additional inclusion criteria of quality assessment. Two independent reviewers will undertake the data extraction. Relevant outcomes of the studies will be analysed thematically using NVIvo computer software. The authors will code all evidence independently. Thereafter the authors will critically cross-examine the relationship of the research questions to the emerging themes from the selected articles. The authors hope to find a large number of studies on health-promoting lifestyles that encompass six-subscales of health-promoting activities (nutrition, stress management, interpersonal relation, self-actualisation, health responsibility, physical activity) and ICT. Dissemination This study will be presented in conferences related to health promotion and health-promoting lifestyles. It will also be disseminated in print and electronically. Trial registration number CRD42016042568. Information and communication technologiesHealth promoting lifestyle profileDisease preventionSystematic review protocol ==== Body Strengths and limitations of this study There are limited studies that link health-promoting lifestyle practices (HPLPs) to information and communication technologies. There are gaps in studies that address all the subscales of the health-promoting lifestyle profile. The study will be restricted to literature published only in the English language. Only literature published from January 2007 to July 2016 will be included in the study. The outcomes of the study will be effective in that they will sustain HPLPs. Introduction Globally, health promotion is used as an umbrella concept for both prevention of disease and promotion of health.1 It is a means of enabling people to have control over the determinants of health and hence achieve maximum health.2 Health promotion views health as a product of daily life and explicitly lists some prerequisites for health.3 Zhang et al4 emphasised the need to improve lifestyle in maintaining a healthy life. Health-promoting lifestyle behaviours focus on enhancing, sustaining and increasing an individual's level of well-being, self-actualisation and personal fulfilment5 6 which brings about optimal health. Information and communication technologies (ICTs) have helped to improved health promotion interventions. According to While and Dewsbury,7 the US National Broadband Plan identified a key role for ICTs in improving health and healthcare through enhancing care delivery, coordination and engagement with patients. The use of ICTs in disease treatment, preventive vaccination, medical appointments, and medication for long-term self-management has had a positive influence on healthcare delivery.8 Despite the benefits of ICTs in all spheres of life, there are adverse effects associated with the use of ICTs if there are no checks and balances on the implementation of the programmes. For example, the increase in the development of technology has made many academic staff adopt sedentary lifestyles which make them physically inactive.9 10 Inactivity has been linked to most of the non-communicable diseases (NCDs) that threaten public health globally.11 NCDs account for 80% of the adult disease burden in developed countries and 70% in middle-income countries, while almost 50% of the adult disease burden is attributed to NCDs in high mortality regions in the world.12 Boutayeb and Boutayeb12 have further documented that by 2020 NCDs will be causing seven out of every 10 deaths in lower and middle income countries. These only relate to physical inactivity; in addition, there is evidence that other components of the health-promoting lifestyle profile (nutrition, stress management, interpersonal relation, self-actualisation, health responsibility, physical activity) also contribute greatly to health promotion and disease prevention. Bully et al13 documented that recent studies have attributed health-promoting lifestyle practices (HPLPs) and moderate alcohol intake to reductions in mortality rates related to NCDs by 50% and increases in life expectancy by more than 11 years. Despite the availability of literature and campaigns on the benefits of health-promoting behaviour, there are still gaps relating to HPLPs that can increase the life expectancy and quality of life of individuals. The healthcare systems pay little attention to lifestyle factors. The focus of healthcare systems is mostly on curative rather than on preventive health. All federal institutions in Nigeria are now beneficiaries of the National Health Insurance Scheme, but most of the care rendered by this scheme is curative rather than preventive. Little is known about the use of ICTs for comprehensive HPLPs that can contribute greatly to the prevention and control of non-communicable and also some communicable diseases among adults in low and middle income countries such as Nigeria. Hence, there is a need to map available evidence through a scoping review. Research objectives The main aim of the review is to establish and map out evidence of interventions that relate to ICTs and effective and efficient HPLPs that can prevent and control diseases as well as to identify areas required to conduct primary research. The specific objectives of the study include the following: To review evidence that links the use of ICT in HPLPs that improve the quality of life and increase life expectancy To review evidence where HPLPs in combination with ICT improve health status (physical and mental health) To review evidence relating to the incorporation of ICT into health-promoting lifestyle interventions that can sustain HPLPs. The study will provide answers to the following questions: What evidence exists that links the use of ICT to HPLPs to improve the quality of life and to increase life expectancy? What evidence exists that HPLPs in combination with ICT improve physical and mental health? Is there evidence that incorporating ICT into health-promoting lifestyle intervention will sustain behavioural modification that will last? Methods/designs This study protocol was registered and published with the International Prospective Register of Systematic Reviews (PROSPERO) (registration number CRD42016042568). Search strategy and study selection The following electronic databases will be searched for the study: PsycArticle (EBSCO), PsycINFO (EBSCO), Science direct, PubMed, Medline (EBSCO) and Google Scholar. The search strategy will include all studies that addressed the intervention irrespective of the type of studies and study designs. All peer-reviewed studies and grey literature that addresses the research questions will be selected in the study. Only articles published in English between 2007 to date will be used. The literature search results will be uploaded to EndNote X7. The EndNote X7 software will be used to find and remove duplicates. Criteria for inclusion Articles published in the English language Literature published from January 2007 to July 2016 Studies with focus on well adults, workers, professionals Evidence from published relevant interventions Studies reporting on intervention(s) such as any of the subscales of the health-promoting profile (nutrition, physical activity, stress management, interpersonal relationships, self-actualisation and health responsibility) and ICT Review articles including: systematic reviews, meta-analysis, scoping reviews, peer-reviewed journal articles and rapid reviews Grey literature sources such as: documents from government and non-governmental organisations and academic dissertations All types of study designs such as cohort studies, cross-sectional studies, qualitative studies, quantitative studies, randomised control trial studies, quasi-experimental study designs and pilot studies. Exclusion criteria Intervention that does not include any form of ICT Studies focusing on diseases, youth, students or children Studies not including all or any of the subscales of the health-promoting lifestyle profile Studies reporting on either alcoholism or cigarette smoking Non-English publications Articles published before January 2007 and after July 2016 Articles not focusing on adults. Keywords search (table 1) will be combined into a phrase including Boolean (AND, OR) terms, as follows: (health promoting lifestyle profile OR health promoting lifestyle behaviour OR wellness OR nutrition OR physical activity OR interpersonal relationships OR health responsibility OR stress management OR self-actualisation, and information and communication technology OR ICT OR mobile phone OR text messages OR SMS OR the internet). Table 1 Electronic search records Date Keyword searched Search engine used Number of publications retrieved Updated records of the number of publications identified during each session of the literature search and date will be kept by all the members of the research team using the information in table 1. Data extraction Numerical summary and thematic analysis will be employed to extract background information from the selected studies. In order to answer the research questions as guided by Population, Interventions, Comparison and Outcome (PICO) (table 2), the reviewers will collectively design a data-chronicling form to determine the text words, and themes to include and extract. They will also collectively develop the data-charting form. Information and data relevant to answer the research questions will be determined by the reviewers collectively. Update recording of data relevant to the study. Relevant data will be extracted from all the eligible studies by two independent reviewers in duplicate. Data to be extracted include: (i) author (s) names; (ii) year of publication; (iii) study design and/or methodology; (iv) study population; (v) intervention(s); (vi) study setting; (vii) aim of the study; (viii) geographic location of the study; (ix) primary and secondary outcome of the study; (x) conclusions. Information specific to HPLPs and description of the intervention (ICT) will also be extracted. If necessary, a third reviewer will be consulted in order to reach consensus. The researchers will collectively carry out a thematic analysis to extract relevant outcomes using NVivo software. Table 2 Population, Interventions, Comparison, Outcome and Study setting (PICOS) framework for determination of the eligibility of the review questions Criteria Determinants Population The population for this study will be healthy adults, workers, well individuals. Patients, children, students, adolescents (youth will be excluded from this study) Interventions Health-promoting lifestyle profile (nutrition, physical activity, interpersonal relationship, health responsibility, stress management, self-actualisation) and information and communication technologies (ICT, mobile phone, text messages, SMS, internet, etc) Comparison Health-promoting lifestyle profile intervention without ICT Outcomes Effective and sustaining health-promoting lifestyle practices and health status (normal weight, normal blood pressure, normal blood sugar and good mental and physical health) Study setting Nigeria and developing countries are the focus. Owing to the scarcity of literature on health-promoting lifestyle profile intervention and ICT, we proposed not to limit the scope of this review by study setting. The rationale for this is to have a good number of publications globally for comparison Analysis and results Narrative synthesis of the outcomes of the selected articles will be presented in the study. The researchers will employ the following steps in analysing and reporting findings from the study. Author (s) names, year of publication, study design and/or methodology, study population, intervention(s), study setting, aim of the study, geographic location of the study and outcome of the study will be examined. Findings from the studies on HPLPs of adults, evidence on effective and efficient HPLP intervention using ICT; evidence on effective and efficient HPLP intervention using ICT will be compared with evidence on effective and efficient HPLC intervention without ICT. The researchers will code all the findings independently. Thereafter the researchers will collectively assess the themes and conduct a critical appraisal of their relationship to the research questions. Furthermore, the authors will also examine the meaning of the findings in relation to the aim of the study as well as its implications for research, practice and policy. Discussion This study will generate evidence that will help to describe the link between health-promoting lifestyle profiles and ICT in achieving a behavioural modification that will endure and enable the individual to have a quality of life that will, in turn, increase his or her life expectancy. Health-promoting lifestyle profiles and ICTs have received little research attention. Findings from studies on just two subscales, nutrition and physical activity, have shown that if individuals maintain a healthy nutritional diet and participate in low to moderate physical activity their quality of life is improved. Therefore, one can argue that having a good healthy lifestyle that encompasses all of the six subclasses of the health-promoting lifestyle profile will have much more impact. However, measures to increase our understanding of the roles of these other subclasses of health-promoting lifestyle profiles, and their relationship to improving the quality of life and increasing life expectancy, have not been given much attention. This study will help researchers to address these issues and also describe the relationship between the phenomena comprehensively. Effective and efficient health-promoting lifestyles that prevent and control diseases (both non-communicable and communicable) is an area that needs more attention. We are not aware if there are any empirical studies or reviews that have been done in low and middle income countries that answer the specific questions outlined in the current proposal. Therefore, findings from this study will contribute to the body of knowledge on health-promoting lifestyle profiles and ICTs which will have a positive impact on research, practice and policy in the area of health promotion that should improve the quality of life of individuals and increase their life expectancy. Due to the paucity of data in the literature regarding this topic, we intend to include all studies from anywhere in the world that address our types of interventions and population. Our argument is that if an individual has been practising a health-promoting lifestyle effectively and efficiently, such an individual will not have issues with medication or dietary compliance. Findings from this scoping review will play a vital role in the primary, secondary and tertiary prevention of diseases. In addition, evidence generated from this study will provide the building blocks for effective and efficient HPLPs that will help in the primary prevention of disease, and also add to the existing literature in health promotion and ICT. Lastly it will also outline aspects of the study that require meta-analysis and primary empirical study. Contributors: EMJ-S: conception (generator of the review) and design; writing the manuscript. BPN: critical revision of the manuscript and administrative and technical support. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: The manuscript is a protocol for a scoping review and there are no data currently. We would be happy to share the dataset after the completion of the study. ==== Refs References 1 Nilsson P , Andersson IH , Ejlertsson G Workplace health resources based on sense of coherence theory . Int J Workplace Health Manag 2012 ;5 :156 –67 . 10.1108/17538351211268809 2 Torp S , Vinje HF Is workplace health promotion research in the Nordic countries really on the right track? Scand J Public Health 2014 ;42 (15 Suppl):74 –81 . 10.1177/1403494814545106 3 Potvin L , Jones CM Twenty-five years after the Ottawa Charter: the critical role of health promotion for public health . Can J Public Health 2011 :102 ;244 –8 .21913576 4 Zhang SC , Wei CN , Fukumoto K A comparative study of health-promoting lifestyles in agricultural and non-agricultural workers in Japan . Environ Health Prev Med 2011 ;16 :80 –9 . 10.1007/s12199-010-0167-9 21432222 5 Tsai YC , Liu CH Factors and symptoms associated with work stress and health-promoting lifestyles among hospital staff: a pilot study in Taiwan . BMC Health Serv Res 2012 ;12 :199 10.1186/1472-6963-12-199 22799278 6 Nassar OS , Shaheen AM Health-promoting behaviours of university nursing students in Jordan . Health 2014 ;6 :2756 10.4236/health.2014.619315 7 While A , Dewsbury G Nursing and information and communication technology (ICT): a discussion of trends and future directions . Int J Nurs Stud 2011 ;48 :1302 –10 . 10.1016/j.ijnurstu.2011.02.020 21474135 8 Cocosila M , Archer N Adoption of mobile ICT for health promotion: an empirical investigation . Electronic Markets 2010 ;20 :241 –50 . 10.1007/s12525-010-0042-y 9 Pirincci E , Rahman S , Durmuş A Factors affecting health-promoting behaviours in academic staff . Public health 2008 ;122 :1261 –3 . 10.1016/j.puhe.2008.04.006 18849056 10 Turkmen M , Ozkan A , Murat K Investigation of the relationship between physical activity level and healthy life-style behaviors of academic staff . Educ Res Rev 2015 ;10 :577 –81 . 10.5897/ERR2015.2083 11 de Rezende LF , Rabacow FM , Viscondi JY Effect of physical inactivity on major noncommunicable diseases and life expectancy in Brazil . J Phys Act Health 2015 ;12 :299 –306 . 10.1123/jpah.2013-0241 24769913 12 Boutayeb A , Boutayeb S The burden of non communicable diseases in developing countries . Int J Equity Health 2005 ;4 :2 10.1186/1475-9276-4-1 15651987 13 Bully P , Sánchez Á , Zabaleta-del-Olmo E Evidence from interventions based on theoretical models for lifestyle modification (physical activity, diet, alcohol and tobacco use) in primary care settings: a systematic review . Prev Med 2015 ;76 (Suppl):S76 –93 . 10.1016/j.ypmed.2014.12.020 25572619

PMC005xxxxxx/PMC5372058.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01481910.1136/bmjopen-2016-014819EpidemiologyProtocol150616921683Secondary prevention and cognitive function after stroke: a study protocol for a 5-year follow-up of the ASPIRE-S cohort http://orcid.org/0000-0001-8834-2539Rohde Daniela 1Williams David 2Gaynor Eva 3Bennett Kathleen 1Dolan Eamon 4Callaly Elizabeth 5Large Margaret 6Hickey Anne 1 1 Division of Population Health Sciences, Royal College of Surgeons in Ireland, Dublin, Ireland 2 Department of Geriatric and Stroke Medicine, Royal College of Surgeons in Ireland and Beaumont Hospital, Dublin, Ireland 3 Department of Medicine, Royal College of Surgeons in Ireland, Dublin, Ireland 4 Department of Geriatric Medicine, Connolly Hospital Blanchardstown, Dublin, Ireland 5 Department of Geriatric Medicine, Mater Misericordiae University Hospital, Dublin, Ireland 6 Clinical Research Centre, Royal College of Surgeons in Ireland, Beaumont Hospital, Dublin, IrelandCorrespondence to Daniela Rohde; danielamrohde@rcsi.ie2017 27 3 2017 7 3 e01481919 10 2016 22 12 2016 6 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction Cognitive impairment is common following stroke and can increase disability and levels of dependency of patients, potentially leading to greater burden on carers and the healthcare system. Effective cardiovascular risk factor control through secondary preventive medications may reduce the risk of cognitive decline. However, adherence to medications is often poor and can be adversely affected by cognitive deficits. Suboptimal medication adherence negatively impacts secondary prevention targets, increasing the risk of recurrent stroke and further cognitive decline. The aim of this study is to profile cognitive function and secondary prevention, including adherence to secondary preventive medications and healthcare usage, 5 years post-stroke. The prospective associations between cognition, cardiovascular risk factors, adherence to secondary preventive medications, and rates of recurrent stroke or other cardiovascular events will also be explored. Methods and analysis This is a 5-year follow-up of a prospective study of the Action on Secondary Prevention Interventions and Rehabilitation in Stroke (ASPIRE-S) cohort of patients with stroke. This cohort will have a detailed assessment of cognitive function, adherence to secondary preventive medications and cardiovascular risk factor control. Ethics and dissemination Ethical approval for this study was granted by the Research Ethics Committees at Beaumont Hospital, Dublin and Connolly Hospital, Dublin, Mater Misericordiae University Hospital, Dublin, and the Royal College of Surgeons in Ireland. Findings will be disseminated through presentations and peer-reviewed publications. secondary preventionmedication adherencecardiovascular risk factorsHealth Research Boardhttp://dx.doi.org/10.13039/100010414SPHeRE 2013/1 ==== Body Introduction Stroke is one of the leading causes of death and disability worldwide,1 with over half of patients exhibiting signs of cognitive impairment 6 months post-stroke.2 Deficits in cognitive function are associated with poor functional outcomes, increased vulnerability and poorer quality of life, and can increase disability and levels of dependency, potentially leading to a greater burden on carers and the healthcare system.3 4 While there is evidence to support early rehabilitation interventions post-stroke, less is known about the reduction of longer term stroke-related disability.5 The chronic phase of stroke accounts for a considerable proportion of the total costs of stroke care, with measures to improve long-term outcomes potentially having a substantial impact on the economic burden of stroke.6 Cardiovascular risk factor management and secondary prevention Several risk factors, including hypertension, dyslipidaemia, diabetes, obesity and smoking, are associated with both cardiovascular disease and cognitive impairment. A study of 355 patients with stroke aged over 75 years reported that the presence of three or more cardiovascular risk factors increased the risk of dementia during a mean follow-up of 3.8 years.7 Secondary preventive treatment, including antihypertensive and anticoagulant medications, has been associated with a reduced risk of cognitive impairment 6 months post-stroke.2 Similarly, the use of anticoagulant, antiplatelet and antihypertensive medications was associated with a reduced risk of cognitive impairment up to 7 years post-stroke in a sample of patients with stroke from the South London Stroke Register.8 Targeting cardiovascular risk factors may reduce both the risk of recurrent vascular events and the risk of cognitive decline.2 7 9 Secondary prevention and rehabilitation are thus essential to maximising health and well-being post-stroke, particularly as recurrent strokes account for up to 50% of all strokes, indicating unsuccessful secondary prevention.10 However, there are limited longer term follow-up data on patients with stroke, and a lack of information regarding the adequacy of longer term post-stroke secondary prevention.11 12 The Action on Secondary Prevention Interventions and Rehabilitation in Stroke (ASPIRE-S) study found a high prevalence of cognitive impairment and cardiovascular risk factors at 6 months post-stroke.2 4 11 While three-quarters of the sample were on antihypertensive therapy, almost two-thirds had blood pressure above the recommended target of 140/90 mm Hg. Similarly, while 95% of patients were on lipid-lowering medications, only approximately three-quarters had total cholesterol or low-density lipoprotein (LDL) cholesterol at European guideline targets.11 Medication adherence Effective secondary stroke prevention is contingent on consistent adherence to prescribed secondary preventive medications.13 Non-adherence is associated with adverse outcomes, including rehospitalisation, recurring vascular events and death, as well as increased costs of care.14 15 However, medication adherence is often poor, with up to 50% of patients discontinuing secondary preventive medications 2 years post-stroke.16 A recent systematic review reported an estimated rate of non-adherence to secondary preventive medications among stroke survivors of 30.9% (95% CI 26.8% to 35.3%).17 A number of factors were found to be related to non-adherence in individual studies, including disability, reduced cognitive function, polypharmacy and concerns about treatment.17 However, this review also noted substantial heterogeneity in the inclusion criteria and definitions and measurements of adherence, with several studies excluding patients with stroke with evidence of cognitive impairment. Medication adherence and cognitive function Adherence to medication regimens requires a number of cognitive skills that are affected by cognitive impairment, including instructions for medication taking and accessing and scheduling medications.18 Suboptimal adherence, in turn, adversely impacts secondary prevention targets, increasing the risk of recurrent stroke and further cognitive decline.2 However, there is a scarcity of data on longer term treatment continuation and adherence rates in patients with stroke, with several studies of adherence either not including cognitive assessments, or focusing exclusively on adherence to antithrombotic medications by patients with stroke and atrial fibrillation.11 13 19–22 Further, previous studies of stroke outcomes are limited by relatively short follow-ups and small sample sizes, and may not reflect contemporary outcomes or treatments.23 Thus, while secondary prevention may be related to a reduced risk of cognitive impairment in patients with stroke, few studies have examined the association between adherence to secondary preventive medications and cognitive impairment, and it is unclear whether suboptimal adherence affects the risk of later cognitive impairment. Further, there are little data available at present regarding the use of healthcare resources, outcomes and costs post-stroke, or how these may be impacted by cognitive impairment or inadequate secondary prevention.24 The identification of medication adherence patterns, factors associated with adherence, and the impact on clinical and cognitive outcomes can inform the development of policies and interventions focused on improving medication management.25 Given the growing health, social and economic burden of cognitive impairment and dementia, longer term follow-up studies are needed to test these associations over extended periods of time.3 20 Aims and objectives The aim of this study is to profile cognitive function and secondary prevention, including healthcare usage and adherence to secondary preventive medications, at 5 years post-stroke. Specifically, the objectives are: To provide a detailed description of cognitive function and secondary prevention, including medication adherence at 5 years post-stroke; To investigate risk factor management, including blood pressure and cholesterol control 5 years post-stroke; To ascertain rates of recurrent stroke, other cardiovascular events and death in the 5 years post-stroke; To explore the prospective associations between cognition, cardiovascular risk factors and adherence to secondary preventive medications from 6 months to 5 years post-stroke; To explore levels of agreement between self-reported medication adherence assessed using a validated scale, pill counts and adherence estimates using pharmacy prescription refill data, in this cohort of patients with stroke; To explore longer term stroke rehabilitation and healthcare usage, and their associations with cognition and secondary prevention 5 years post-stroke; To examine the longer term costs of stroke, in terms of healthcare usage and quality of life, and their association with cognition and secondary prevention 5 years post-stroke. A secondary aim is to explore carer well-being and the reciprocal associations between changes in cognitive function of patients with stroke and emotional distress and vulnerability in carers or family members.4 Methods and analysis Study design This is a 5-year follow-up of a prospective observational study of the ASPIRE-S cohort of patients with stroke.2 4 11 The ASPIRE-S study recruited patients with acute stroke in 2011–2012. A total of 256 patients and their family members or carers were last followed up at 6 months post-stroke. This cohort will be followed up again, 5 years post-stroke (2016–2017), with a comprehensive assessment of clinical and cognitive measures. A detailed health assessment will be conducted, including cognitive function, adherence to secondary prevention and cardiovascular risk factor control. Data will be collected using a combination of clinical and laboratory measurements using standard data collection forms, interviewer and self-completion questionnaires. Sample size All participants from the original study will be eligible to participate. On the basis of an estimated rate of attrition of 30% (including death and other loss to follow-up), it is anticipated that c. 180 patients and their family members or carers will be reassessed. Outcomes Cardiovascular risk factor control Secondary prevention will be assessed using a number of clinical measures as part of a health assessment, including height and weight, blood pressure, pulse, anticoagulation, blood analysis and breath analysis (table 1). Analysis of serum total cholesterol, calculated LDL cholesterol and blood glucose will be carried out on all participants. Diabetic patients will also have glycated haemoglobin assessed. Table 1 Clinical measures to be collected as part of a health assessment Clinical measure Details Blood pressure Three measurements will be taken with the average of the last two readings used for analysis. Twenty-four-hour BP monitoring will also be performed where possible. Pulse Where the pulse is found to be irregular, an ECG will be arranged to confirm atrial fibrillation. Anticoagulation Record of the last three INR measurements to assess the safety of anticoagulation medication, where relevant Blood analysis To assess serum total cholesterol, HDL cholesterol and calculated LDL cholesterol, glucose and HbA1c Weight and height To assess BMI Breath analysis Carbon monoxide monitoring to confirm non-smoking status in former smokers BMI, body mass index; BP, blood pressure; HbA1c, glycated haemoglobin; HDL, high-density lipoprotein; INR, international normalised ratio; LDL, low-density lipoprotein. Secondary preventive medications Reported medications at the time of follow-up will be recorded as part of the interviewer-administered patient questionnaire, and confirmed with prescription refill data where possible. Medication adherence There is no consensus on the best measure to define medication adherence.14 Combining prescription refill data with self-reported measures of adherence is considered essential to capturing the full extent to which patients are (non-)adherent.15 Thus, adherence to secondary preventive medications will be assessed using a number of methods: Prescription refills Participants who are eligible for the General Medical Services (GMS) scheme will be asked for permission to access records of monthly dispensed medications from the Irish Health Service Executive Primary Care Reimbursement Service (HSE-PCRS) database. Access to the GMS scheme, which provides free general practitioner (GP) and hospital visits, and prescription medications at minimal cost, is means-tested for persons aged <70 years, with a substantially higher income threshold for the over 70 s,26 such that over 90% of those over 70 in Ireland are included in the scheme. As a result, members of the scheme are representative of the Irish population of over-70 s, but this is not the case for those aged under 70, where women and those in lower socioeconomic groups are over-represented. The HSE-PCRS pharmacy claims database provides details on monthly dispensed medications for each individual in the scheme.26 Prescription refills provide an objective estimate of medication adherence, circumventing the problem of incorrect self-reporting of medication adherence.14 16 25 27 Prescription refills can be used to calculate metrics such as the proportion of days covered—the sum of the days supplied for medications within each medication class, divided by the number of days in the study period.14 15 This method has been adopted by another recent study on adherence to secondary preventive medications post-stroke.21 Self-reported medication adherence Self-reported medication adherence will be assessed using the Medication Adherence Report Scale (MARS-5),28 a five-point scale that assesses both intentional and unintentional non-adherence. The eight-item Morisky Medication Adherence Scale (MMAS),29 which assesses a number of medication-taking behaviours, will also be included. Pill counts The research team will conduct pill counts of the remaining tablets in each medication or blister pack, to assess to what extent patients have taken medications as prescribed. In addition to providing a more comprehensive account of medication-taking, the use of several methods to estimate adherence will allow exploration of the levels of agreement between self-reported medication adherence, pill counts and adherence estimates using pharmacy prescription refill data, which has not previously been reported for patients with stroke. Stroke recurrence and other cardiovascular events In order to ascertain stroke recurrence or other cardiovascular events, participants will be asked if they have experienced a stroke or other cardiovascular event since the last assessment. Stroke and other cardiovascular event occurrence will also be ascertained from participants' hospital medical notes, based on recording of diagnosis or other mention in the record. Cognitive function Cognitive function will be assessed using a number of standardised, validated instruments, including the Montreal Cognitive Assessment (MoCA),30 the National Institute of Neurological Disorders and Stroke (NINDS) 30 min test battery,31 and the Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE)32 (table 2). Table 2 Cognitive assessments Assessment Details Patient MoCA30 A rapid screening instrument for global cognitive function, which has exhibited good sensitivity and specificity for mild cognitive impairment NINDS 30 mintest protocol31 A 30 min neuropsychological assessment battery that includes: Semantic fluency (Animal Naming) Phonemic fluency (Controlled Oral Word Association Test) Digit Symbol-Coding Verbal Learning Test Centre for Epidemiologic Studies—Depression Scale NPI-Q Trail Making Test Carer/family member IQCODE32 An informant-reported assessment of cognitive function IQCODE, Informant Questionnaire on Cognitive Decline in the Elderly; MoCA, Montreal Cognitive Assessment; NINDS, National Institute of Neurological Disorders and Stroke; NPI-Q, Neuropsychiatric Inventory, Questionnaire Version. Rehabilitation and healthcare usage A number of questions on use of physiotherapy, occupational therapy, psychology, and speech and language therapy in the past 12 months, and the perceived need for these services, will be included in the patient assessment. Questions on healthcare usage, adapted from The Irish Longitudinal Study on Ageing,44 will also be included. These questions will assess use of GP and nursing services, inpatient and outpatient hospital care, use of mental health services, home help and home care, and meals-on-wheels in the past 12 months. In addition, the Timed Up and Go test, a performance-orientated mobility assessment tool used to identify individuals at increased risk of falls,33 and a number of questions on falls adapted from the Irish Longitudinal Study on Ageing, will be included. Covariates A range of demographic information, including age, education, occupational status, marital status and living arrangements, will be collected as part of the patient and family member/carer assessments. The patient assessment will also include questions relating to lifelong learning, social participation and current smoking. Potential barriers to medication adherence will be explored using the Beliefs about Medicines Questionnaire (BMQ),28 which assesses concerns and perceived necessity of medications. Beliefs of patients with stroke about medicines have been reported to be associated with non-adherence, with non-adherent patients scoring lower on the positive beliefs dimensions of the BMQ (necessity and benefit) and higher on negative beliefs (concern, overuse and harm).34 The Frenchay Aphasia Screening Test (FAST)35 will be used to screen for communication problems that may affect performance on cognitive assessments. Quality of life and health and well-being status will be assessed using the ICEpop CAPability measure for Older people (ICECAP-O),36 the EQ-5D37 and the Stroke Specific Quality of Life Scale (SSQOL).38 Finally, the Nottingham Extended Activities of Daily Living Scale,39 which assesses activities that may be important to patients with stroke who have been discharged home, and the Vulnerable Elders Scale (VES),40 a simple function-based tool for identifying older persons who may be at risk of health deterioration, will be included. Family member and carer assessments In addition to the patient assessment, family members or carers will be asked to complete a family member or carer assessment. This assessment will involve a self-completion questionnaire that includes the IQCODE32 and the Neuropsychiatric Inventory, Questionnaire Version (NPI-Q) from the NINDS test battery, as well as a number of questions regarding the extent to which patients receive assistance with taking medications. The use of informant report in addition to patient assessments will facilitate triangulation of measures of cognition and increase the validity of the findings. Furthermore, informant report is important in cases where patients receive help from carers with medication taking, as there is a lack of data on factors affecting medication adherence in patients relying on others for medication management.41 The family member and carer assessment will also include a measure of anxiety (the Hospital Anxiety and Depression Scale—Anxiety (HADS-A))42 and depression (the Centre for Epidemiologic Studies Depression Scale (CES-D)),43 and vulnerability (VES).40 These measures were included in the original ASPIRE-S study, which reported substantial levels of anxious and depressive symptoms among carers, linked to anxiety, depressive symptoms and cognitive impairment in the patients with stroke.4 By repeating the measures of emotional distress and vulnerability included in the original ASPIRE-S study, we will be able to consider carer well-being over time, and explore the associations between changes in cognitive impairment in patients with stroke and emotional distress and vulnerability in carers or family members. Follow-up procedure Each potential participant's consultant physician will be contacted to request permission to contact patients and their GPs. Efforts will be made to identify deceased participants prior to contact being made, using, for example, hospital and GP records, as well as a widely used website of death notifications (RIP.ie). Participants will be sent a study information pack by post, containing a cover letter, patient and family member/carer information leaflets, and a stamped addressed postcard that can be returned by participants wishing to opt out of the study. The cover letter will advise patients and family members that they will be contacted by phone by the research team to discuss their participation in the study, unless they return the postcard or inform their consultant that they do not wish to be contacted. Participants will then be contacted by phone to discuss the study and to ascertain their interest in participating in the follow-up. If participants agree, a meeting will be arranged with two members of the research team, in the participants' own home, at a hospital or another location convenient to participants, and written consent will be obtained. In cases where significant cognitive deficits are suspected, either by the patient's GP or members of the research team conducting the follow-up phone calls, contact will be made with the patient's family member or carer, who will be requested to be present at the time of interview. The research team will attempt to carry out individual interviews with all patients. Assisted self-interviews, in which the participant can answer most of the questions but requires some help from a family member or carer, may also be used. This procedure has been used by follow-up waves of the Irish Longitudinal Study on Ageing.44 If a participant is unable to complete an interview (cognitively or physically), the family member or carer will still be eligible to complete the family member/carer assessment. Data analysis Data will be analysed and reported using descriptive statistics, including means and SDs for normally distributed data, medians and IQRs for non-normally distributed data, and frequencies and percentages, as appropriate. Adjusted associations between medication (non-)adherence, (uncontrolled) cardiovascular risk factors and cognitive function will be assessed using multivariate regression models. The predicted sample size of 180 and an estimated prevalence of cognitive impairment of 50% will permit the inclusion of ∼9 variables in multivariate analyses.45 Covariates that are associated with each outcome at the p<0.10 level will be included in multivariate models. Agreement between measures of medication adherence will be assessed using the κ statistic.26 Ethics and dissemination Treatment of study participants Participants will continue to receive standard care throughout the study period from their hospital consultant or GP. If participants become distressed or otherwise unwell during the interview or health assessment, the interview or assessment will be terminated immediately, and the interviewer will call back within 24 hours to see how the participant is. The family member/carer and GP or consultant will be informed that the patient became upset during the course of the interview and, if appropriate, that the patient would like to talk to a member of the medical team. Although potential harm is expected to be minimal, some participants may experience temporary and localised discomfort, mild pain or bruising as a result of venepuncture or blood pressure assessment. Dissemination Findings from this study will be disseminated through presentations in academic fora and to relevant policymaker, practitioner and stakeholder audiences, and through peer-reviewed publications. Limitations This study has a number of limitations, including the reliance on self-report or carer report to ascertain usage of healthcare services. Unfortunately, owing to the complex system of public/private healthcare provision in Ireland, and the lack of universal patient identifiers, there is currently no alternative method available to collect this information. While every effort will be made to follow-up all patients still alive from the original study, patients with more severe cognitive decline may be more likely to be lost to follow-up, which could lead to an underestimation of the prevalence of cognitive impairment in this cohort. Since this is a follow-up study, the sample size will be based on the availability of participants, rather than a statistical power calculation. It is possible that the analyses for some associations of interest will be underpowered. While we will aim to assess and adjust for a number of known confounders, owing to the observational nature of this study, there may be other unknown or unmeasured confounding factors. However, given the fact that longer term follow-up studies of patients with stroke are relatively rare, and no such studies exist in Ireland, recalling this cohort provides a unique opportunity to explore longer term secondary prevention and cognitive function in patients with stroke. Conclusion Given the potential public health burden of cognitive impairment after stroke, identification of modifiable risk and protective factors to avert or delay cognitive decline is paramount. Delaying cognitive impairment could allow many individuals to reach the end of their natural lifespan before crossing the threshold for dementia.46 While effective vascular risk factor management through secondary prevention may reduce the burden of post-stroke cognitive impairment,2 7 8 there is a lack of data on longer term treatment continuation and adherence rates in patients with stroke. Similarly, there are currently no Irish data on the costs of cognitive impairment or inadequate secondary prevention post-stoke. This study will provide valuable information on the trajectories of cognitive impairment, medication adherence and cardiovascular risk factors post-stroke, and will indicate how cognitive function and secondary prevention are related to each other, as well as to stroke or other cardiovascular event recurrence and healthcare usage, over time. These findings can help inform future health policy and service planning regarding the longer term management of patients with stroke. Twitter: Follow Daniela Rohde @dannifromdublin Contributors: DR, DW, KB, EG, EC, ED, ML and AH conceived and designed the study. DR drafted and edited the manuscript. DW, KB, EG, EC, ED, ML and AH critically revised the manuscript. All authors approved the final draft. Funding: This work was supported by the Health Research Board (grant number SPHeRE 2013/1). The original ASPIRE-S research was supported by a Health Research Award from the Irish Health Research Board (grant number 1404/7400). The Health Research Board (HRB) supports excellent research that improves people's health, patient care and health service delivery. Competing interests: All authors have completed the Unified Competing Interest form at http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: DW is an Advisory Board Member for Boehringer Ingelheim, Daiichi Sankyo, Bristol Myers Squibb and Bayer, and has received personal fees for this outside the submitted work. DW is Speaker Honorarium for Boehringer Ingelheim, and has received personal fees for this outside the submitted work. Ethics approval: This study was approved by the Research Ethics Committees at Beaumont Hospital (ref. 16/26), Connolly Hospital Blanchardstown (28/11/2016) and the Royal College of Surgeons in Ireland (REC 1355). Provenance and peer review: Not commissioned; externally peer reviewed. ==== Refs References 1 McElwaine P , McCormack J , Harbison J , on behalf of the National Stroke Programme Audit Steering Group IHF, and HSE . National stroke audit . Dublin Irish Heart Foundation and HSE , 2015 . 2 Mellon L , Brewer L , Hall P Cognitive impairment six months after ischaemic stroke: a profile from the ASPIRE-S study . BMC Neurol 2015 ;15 :31 10.1186/s12883-015-0288-2 25879880 3 Brainin M , Tuomilehto J , Heiss WD Post-stroke cognitive decline: an update and perspectives for clinical research . Eur J Neurol 2015 ;22 :229 –38 , e13–6 10.1111/ene.12626 25492161 4 Atteih S , Mellon L , Hall P Implications of stroke for caregiver outcomes: findings from the ASPIRE-S study . Int J Stroke 2015 ;10 :918 –23 . 10.1111/ijs.12535 26061711 5 Wolfe CD , Crichton SL , Heuschmann PU Estimates of outcomes up to ten years after stroke: analysis from the prospective South London Stroke Register . PLoS Med 2011 ;8 :e1001033 10.1371/journal.pmed.1001033 21610863 6 Smith S , Horgan F , Sexton E The future cost of stroke in Ireland: an analysis of the potential impact of demographic change and implementation of evidence-based therapies . Age Ageing 2013 ;42 :299 –306 . 10.1093/ageing/afs192 23302602 7 Allan LM , Rowan EN , Firbank MJ Long term incidence of dementia, predictors of mortality and pathological diagnosis in older stroke survivors . Brain 2011 ;134 (Pt 12) :3716 –27 . 10.1093/brain/awr273 22171356 8 Douiri A , McKevitt C , Emmett ES Long-term effects of secondary prevention on cognitive function in stroke patients . Circulation 2013 ;128 :1341 –8 . 10.1161/CIRCULATIONAHA.113.002236 23935013 9 Richard E , Jongstra S , Soininen H Healthy ageing through internet counselling in the elderly: the HATICE randomised controlled trial for the prevention of cardiovascular disease and cognitive impairment . BMJ Open 2016 ;6 :e010806 10.1136/bmjopen-2015-010806 10 Hankey GJ Secondary stroke prevention . Lancet Neurol 2014 ;13 :178 –94 . 10.1016/S1474-4422(13)70255-2 24361114 11 Brewer L , Mellon L , Hall P Secondary prevention after ischaemic stroke: the ASPIRE-S study . BMC Neurol 2015 ;15 :216 10.1186/s12883-015-0466-2 26492943 12 Van Schaik SM , Van den Berg-Vos RM , Weinstein HC Limited efficacy of a long-term secondary prevention program in ischemic stroke and transient ischemic attack patients . J Stroke Cerebrovasc Dis 2015 ;24 :1378 –82 . 10.1016/j.jstrokecerebrovasdis.2015.02.023 25804570 13 Murphy SJ , Coughlan CA , Tobin O Continuation and adherence rates on initially-prescribed intensive secondary prevention therapy after Rapid Access Stroke Prevention (RASP) service assessment . J Neurol Sci 2016 ;361 :13 –18 . 10.1016/j.jns.2015.12.009 26810509 14 Ho PM , Bryson CL , Rumsfeld JS Medication adherence: its importance in cardiovascular outcomes . Circulation 2009 ;119 :3028 –35 . 10.1161/CIRCULATIONAHA.108.768986 19528344 15 Kronish IM , Ye S Adherence to cardiovascular medications: lessons learned and future directions . Prog Cardiovasc Dis 2013 ;55 :590 –600 . 10.1016/j.pcad.2013.02.001 23621969 16 Glader EL , Sjölander M , Eriksson M Persistent use of secondary preventive drugs declines rapidly during the first 2 years after stroke . Stroke 2010 ;41 :397 –401 . 10.1161/STROKEAHA.109.566950 20075360 17 Al AlShaikh S , Quinn T , Dunn W Predictive factors of non-adherence to secondary preventative medication after stroke or transient ischaemic attack: a systematic review and meta-analyses . Eur Stroke J 2016 ;1 :65 –75 . 10.1177/2396987316647187 29900404 18 Campbell NL , Boustani MA , Skopelja EN Medication adherence in older adults with cognitive impairment: a systematic evidence-based review . Am J Geriatr Pharmacother 2012 ;10 :165 –77 . 10.1016/j.amjopharm.2012.04.004 22657941 19 Bushnell CD , Olson DM , Zhao X Secondary preventive medication persistence and adherence 1 year after stroke . Neurology 2011 ;77 :1182 –90 . 10.1212/WNL.0b013e31822f0423 21900638 20 O'Carroll R , Whittaker J , Hamilton B Predictors of adherence to secondary preventive medication in stroke patients . Ann Behav Med 2011 ;41 :383 –90 . 10.1007/s12160-010-9257-6 21193977 21 Sjölander M , Eriksson M , Glader E-L Inequalities in medication adherence to statin treatment after stroke: a nationwide observational study . Eur Stroke J 2016 ;1 :101 –107 . 10.1177/2396987316646026 22 Björck F , Renlund H , Svensson PJ Warfarin persistence among stroke patients with atrial fibrillation . Thromb Res 2015 ;136 :744 –8 . 10.1016/j.thromres.2015.07.028 26254195 23 Crichton SL , Bray BD , McKevitt C Patient outcomes up to 15 years after stroke: survival, disability, quality of life, cognition and mental health . J Neurol Neurosurg Psychiatr 2016 ;87 :1091 –8 . 10.1136/jnnp-2016-313361 24 Wren MA , Gillespie P , Smith S Towards earlier discharge, better outcomes, lower costs: stroke rehabilitation in Ireland . Dublin Economic and Social Research Institute , 2014 . 25 Sanfélix-Gimeno G , Rodríguez-Bernal CL , Hurtado I Adherence to oral anticoagulants in patients with atrial fibrillation-a population-based retrospective cohort study linking health information systems in the Valencia region, Spain: a study protocol . BMJ Open 2015 ;5 :e007613 10.1136/bmjopen-2015-007613 26 Richardson K , Kenny RA , Peklar J Agreement between patient interview data on prescription medication use and pharmacy records in those aged older than 50 years varied by therapeutic group and reporting of indicated health conditions . J Clin Epidemiol 2013 ;66 :1308 –16 . 10.1016/j.jclinepi.2013.02.016 23968693 27 Hansen RA , Kim MM , Song L Comparison of methods to assess medication adherence and classify nonadherence . Ann Pharmacother 2009 ;43 :413 –22 . 10.1345/aph.1L496 19261962 28 Horne R , Weinman J , Hankins M The beliefs about medicines questionnaire: the development and evaluation of a new method for assessing the cognitive representation of medication . Psychol Health 1999 ;14 :1 –24 . 10.1080/08870449908407311 29 Morisky DE , Ang A , Krousel-Wood M Predictive validity of a medication adherence measure in an outpatient setting . J Clin Hypertens (Greenwich) 2008 ;10 :348 –54 . 10.1111/j.1751-7176.2008.07572.x 18453793 30 Nasreddine ZS , Phillips NA , Bedirian V The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment . J Am Geriatr Soc 2005 ;53 :695 –9 . 10.1111/j.1532-5415.2005.53221.x 15817019 31 Hachinski V , Iadecola C , Petersen RC National Institute of Neurological Disorders and Stroke-Canadian Stroke Network vascular cognitive impairment harmonization standards . Stroke 2006 ;37 :2220 –41 . 10.1161/01.STR.0000237236.88823.47 16917086 32 Jorm AF , Jacomb PA The Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE): socio-demographic correlates, reliability, validity and some norms . Psychol Med 1989 ;19 :1015 –22 . 10.1017/S0033291700005742 2594878 33 Podsiadlo D , Richardson S The timed “Up & Go”: a test of basic functional mobility for frail elderly persons . J Am Geriatr Soc 1991 ;39 :142 –8 . 10.1111/j.1532-5415.1991.tb01616.x 1991946 34 Sjölander M , Eriksson M , Glader EL The association between patients’ beliefs about medicines and adherence to drug treatment after stroke: a cross-sectional questionnaire survey . BMJ Open 2013 ;3 :e003551 10.1136/bmjopen-2013-003551 35 Enderby PM , Wood VA , Wade DT The Frenchay Aphasia Screening Test: a short, simple test for aphasia appropriate for non-specialists . Int Rehabil Med 1987 ;8 :166 –70 . 10.3109/03790798709166209 2440825 36 Grewal I , Lewis J , Flynn T Developing attributes for a generic quality of life measure for older people: preferences or capabilities? Soc Sci Med 2006 ;62 :1891 –901 10.1016/j.socscimed.2005.08.023 .16168542 37 EuroQol Group . EuroQol—a new facility for the measurement of health-related quality of life . Health Policy 1990 ;16 : 199 –208 . 10.1016/0168-8510(90)90421-9 10109801 38 Williams LS , Weinberger M , Harris LE Development of a stroke-specific quality of life scale . Stroke 1999 ;30 :1362 –9 . 10.1161/01.STR.30.7.1362 10390308 39 Nouri F , Lincoln N An extended activities of daily living scale for stroke patients . Clin Rehabil 1987 ;1 :301 –5 . 10.1177/026921558700100409 40 Saliba D , Elliott M , Rubenstein LZ The Vulnerable Elders Survey: a tool for identifying vulnerable older people in the community . J Am Geriatr Soc 2001 ;49 :1691 –9 . 10.1046/j.1532-5415.2001.49281.x 11844005 41 De Simoni A , Mant J , Sutton S Adherence to medication in stroke survivors dependent on caregivers . Br J Gen Pract 2015 ;65 :e789 –e91 . 10.3399/bjgp15X687589 26500329 42 Zigmond AS , Snaith RP The hospital anxiety and depression scale . Acta Psychiatr Scand 1983 ;67 :361 –70 . 10.1111/j.1600-0447.1983.tb09716.x 6880820 43 Radloff LS The CES-D scale: a self-report depression scale for research in the general population . Appl Psychol Meas 1977 ;1 :385 –401 . 10.1177/014662167700100306 44 Nolan A , O'Regan C , Dooley C The over 50s in a changing Ireland: economic circumstances, health and well-being . Dublin : The Irish Longitudinal Study on Ageing , 2014 . 45 Vittinghoff E , McCulloch CE Relaxing the rule of ten events per variable in logistic and Cox regression . Am J Epidemiol 2007 ;165 :710 –18 . 10.1093/aje/kwk052 17182981 46 Lautenschlager NT , Anstey KJ , Kurz AF Non-pharmacological strategies to delay cognitive decline . Maturitas 2014 ;79 :170 –3 . 10.1016/j.maturitas.2014.07.012 25231337

PMC005xxxxxx/PMC5372059.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01393210.1136/bmjopen-2016-013932Health PolicyResearch150617031681169917241725Global systematic review of Indigenous community-led legal interventions to control alcohol Muhunthan Janani 12Angell Blake 13Hackett Maree L 1Wilson Andrew 24Latimer Jane 1Eades Anne-Marie 1Jan Stephen 16 1 The George Institute for Global Health, Sydney, Australia 2 The Australian Prevention Partnership Centre, Australia 3 The Poche Centre for Indigenous Health, Sydney Medical School, University of Sydney, Sydney, Australia 4 Menzies Centre for Health Policy, University of Sydney, Sydney, Australia 5 The University of Central Lancashire, Preston, UK 6 Sydney Medical School, University of Sydney, AustraliaCorrespondence to Janani Muhunthan; jmuhunthan@georgeinstitute.org.au2017 27 3 2017 7 3 e01393219 8 2016 25 11 2016 17 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives The national and subnational governments of most developed nations have adopted cost-effective regulatory and legislative controls over alcohol supply and consumption with great success. However, there has been a lack of scrutiny of the effectiveness and appropriateness of these laws in shaping the health-related behaviours of Indigenous communities, who disproportionately experience alcohol-related harm. Further, such controls imposed unilaterally without Indigenous consultation have often been discriminatory and harmful in practice. Setting, participants and outcome measures In this systematic review of quantitative evaluations of Indigenous-led alcohol controls, we aim to investigate how regulatory responses have been developed and implemented by Indigenous communities worldwide, and evaluate their effectiveness in improving health and social outcomes. We included articles from electronic databases MEDLINE, EMBASE, CINAHL, PsycINFO and Web of Science from inception to December 2015. Results Our search yielded 1489 articles from which 18 met the inclusion criteria. Controls were implemented in rural and remote populations of high-income nations. Communities employed a range of regulatory options including alcohol rationing, prohibition of sale, importation or possession, restrictions on liquor sold, times of sale or mode of sale, Indigenous-controlled liquor licensing, sin tax and traditional forms of control. 11 studies reported interventions that were effective in reducing crime, injury deaths, injury, hospitalisations or lowering per capita consumption. In six studies interventions were found to be ineffective or harmful. The results were inconclusive in one. Conclusions Indigenous-led policies that are developed or implemented by communities can be effective in improving health and social outcomes. Indigenous healthAlcohol controlPublic health law ==== Body Strengths and limitations of this study This systematic review is the first to explore the effectiveness of legal responses to alcohol misuse and alcohol-related harm that have been designed and implemented by Indigenous communities. Summarising the evidence on effective and ineffective community-led policy models could provide valuable insights into local policy innovation to promote the health of marginalised populations. Effective controls could encourage nation-states and subnational governments to facilitate, through a menu of regulatory instruments, powerful and acceptable community-led approaches to Indigenous alcohol policy worldwide. This review was limited to published literature and did not include grey literature. Introduction Governments in most developed nations have adopted some form of regulatory and legislative control over alcohol supply and consumption through initiatives such as licensing systems, taxation, a minimum age of purchase and penalties for drink drivers. These measures have proven beneficial to the health of populations worldwide, however, Indigenous communities across the world remain disproportionately affected by harmful alcohol use, casting doubt over the effectiveness of these measures in reducing the burden of alcohol-related deaths, chronic disease and disability in Indigenous populations. There have been limited evaluations of the effectiveness of such legal measures and their appropriateness in shaping alcohol-related behaviours in Indigenous communities. In part, the disproportionate burden of alcohol harm borne by Indigenous communities around the world has resulted from the stressful experiences of discrimination, colonialism, dispossession and economic and social marginalisation of Indigenous peoples in many countries.1 2 Public health academics hypothesise that legal measures may be ineffective in reducing this burden in Indigenous populations because policy models lack cultural acceptability, due to inadequate investment in the social determinants of health (eg, education, transport and employment) and that these policies fail to achieve adequate levels of enforcement and accountability in these communities.3 4 The United Nations (UN) Expert Mechanism on the Rights of Indigenous Peoples has mandated that Indigenous participation in decision-making on the full spectrum of matters that affect their lives forms the fundamental basis for the enjoyment of human rights.5 Further, the UN has cautioned member states that a failure to ensure these rights can lead to further marginalisation and inequities among Indigenous people.5 Yet peak Indigenous health bodies such as Australia's National Aboriginal Community Controlled Health Organisation attribute the ongoing crisis in Indigenous health to a lack of cohesive public policy and the lack of power that Indigenous people have had in influencing public policy decisions that affect their social and emotional and well-being.6 In Australia, public health academics have argued that compulsory alcohol restrictions which have been imposed on Indigenous communities by unilateral state action (in some areas for decades), have constituted “an affront to Indigenous self-determination” and that without community support, these interventions are “an overly simplistic solution to a complex problem” and risk the revival of a protectionist past.7 In the USA, following a move to nullify local alcohol laws established by Alaska Native tribal councils by the Federal Government, alcohol abuse and alcohol-related mortality escalated for the next two decades until this centralised approach was reversed.8 Local Indigenous knowledge and systems of governance, which could provide insights into policy innovation and a sustainable shift in social norms, remain neglected areas of public health research to date. There is potential for community-led legal initiatives to be used as preventative health tools for communities affected by the harmful use of alcohol and disproportionately affected by chronic conditions and their risk factors. While community leadership is the ideal, it is not always clear how this concept is applied, in particular how community involvement in the design and enforcement of these programmes interacts with formal government regulatory input. Such community input could involve the engagement of local leaders in developing local solutions through innovative initiatives that use tools such as community partnerships, local taxes, restrictions or incentives. Lessons regarding the use and effectiveness of such measures have not been synthesised previously in the published literature. In light of emerging interest in building countries' legal infrastructure to promote health, this study aimed to investigate how Indigenous communities use public health law mechanisms to control alcohol and prevent its misuse and to what extent controls are effective in achieving improvements in health and social outcomes.9 Methods A systematic review of published literature was undertaken. Databases MEDLINE, EMBASE, CINAHL, PsycINFO and Web of Science, were searched from inception to 18 December 2015 to identify published quantitative evaluations of Indigenous-led alcohol controls. The WHO definition was followed in identifying Indigenous populations.10 Globally, there is a lack of consensus regarding the terminology used to describe Indigenous populations. In this review, we have mirrored the study authors' use of Indigenous terminology. Inclusion criteria Studies had to meet four criteria for inclusion in this review. First, studies had to examine an Indigenous population. The search strategy (see online supplementary file 1) for this component was adapted from previously published systematic reviews.11 12 Second, studies had to examine legal interventions (including traditional forms of law) that were community-led. The authors defined community-led in terms of development and/or implementation (either in terms of development and/or implementation). In order to be included on the basis of development, communities needed to have developed (in isolation or in collaboration with other stakeholders such as governments) the nature and scope of the alcohol control. To be included on the basis of implementation, it was necessary that alcohol controls were implemented or governed by the community or community representatives. Studies were included if controls were government-facilitated with supporting regulatory controls. The authors acknowledge that regulatory responses to control alcohol often require the resources and regulatory power of governments and their agencies. Therefore, it is open to conjecture whether these controls can be genuinely community led. Third, studies had to examine interventions designed to reduce the harms of alcohol consumption. Finally, studies had to quantitatively evaluate the effectiveness of legal interventions to control alcohol in improving one or more health or social outcomes. Qualitative articles, as they lacked a quantitative measure of effectiveness, were excluded from this review. Studies were not excluded on the basis of language, the methods of the quantitative studies or their outcome measures. 10.1136/bmjopen-2016-013932.supp1supplementary file Data extraction Study review and selection were independently undertaken by two authors (JM and BA). Abstracts, titles and keywords of the studies returned from the search were screened for compatibility with the inclusion criteria. Once studies were identified for potential inclusion, full texts were reviewed. Reference lists of included papers were independently reviewed by two authors (JM and BA) for studies that may warrant inclusion. Data were extracted from the studies using a form developed for the review in consultation with the authorship team. Data extracted included country of origin, Indigenous population and size, category of alcohol control, legal instruments employed, communities' use of research evidence or population data to inform the control, data collected and time horizon of the evaluation (table 1). Any disagreements throughout this process were resolved by arbitration with reviewers MLH and SJ. Table 1 Data extracted from evaluations meeting inclusion criteria Population Characteristics of evaluation Mechanism of alcohol control Effectiveness of control Indigenous population Country of origin Setting (urban or remote) Size of study population Study design Time horizon of evaluation Use of control population Study limitations Generalisability of findings Policy implications Type of alcohol control mechanism Legal instruments employed Use of research evidence and population data to inform scope and nature of control Outcomes reported Effectiveness of control in achieving prevention outcomes Results The review found 18 studies spanning 1975 to 2014. Populations studied All studies came from high-income nations, the USA (n=10), Australia (n=6), Canada (n=1) and Greenland (n=1; see table 2). Except for the Greenlandic study which was nationwide, populations studied were rural or remote communities. Table 2 Quantitative evaluations of alcohol controls meeting inclusion criteria Author (year) Country of origin (Indigenous population) Population size Category of alcohol control Legal instruments Health and social outcomes reported Data collected Time Horizon of evaluation May (1975)19 United States (Native American) 12 000 (Native American) Prohibition of sale, importation or possession Tribally imposed prohibition laws Arrests Tribal Police data 3 years (June–July 1969, June–July 1970, June–July 1971) Schechter (1986)20 Greenland (Greenlandic Inuit) 50 000 (80% Indigenous) Alcohol rationing Rationing ordinance passed by the Greenland Council following public plebiscite Alcohol sales; crime Alcohol sales or consumption data 6 years (1978–1984) Gallaher et al (1992)27 USA (Native American) 123 000 (all residents of New Mexico) Prohibition of sale, importation or possession Tribally imposed prohibition laws Unintentional injury Cause-specific mortality data; Hospital records or medical examiner reports; Traffic accident report files (containing location of Indigenous pedestrian deaths) 10 years (January 1 1980 to December 31 1989) Lee (1993)25 USA (Alaska Native) 8 Indigenous villages (of 57 in the region) Restrictions on liquor sold, times of sale or mode of sale; traditional forms of control Local option law Crime; Intentional injury (self-harm, attempted suicide or suicide) Crime data for serious or minor offences 5 years (1983–1987) Chiu et al (1997)16 USA (Alaska Native) 4000 (61% Indigenous) Prohibition of sale, importation or possession Local option law Social or health service usage Alcohol-related outpatient visit records 33 months (November 1993 through July 1996) Landen et al (1997)24 USA (Alaska Native) Dry villages: 63 419 person-years (93% Indigenous) Wet villages: 38 867 person-years (55% Indigenous)*† Prohibition of sale, importation or possession Local option law Unintentional injury Cause-specific mortality data 3 years (1990-1993) Landen (1997)23 USA (American Indian) Wet reservations: (1) <5000; (2) 5000–10 000; (3) <5000 (4) 5000–10 000 (5) 5000–10 000 Dry reservations: (6/7) 10 000 –15 000 10 000–15 000 (8) 5000–10 000 Prohibition of sale, importation or possession Tribally imposed prohibition laws Alcohol-related mortality American Indian mortality data by county 11 years (1979–1990) Douglas (1998)17 Australia (Aboriginal and or Torres Strait Islander—primarily Kija and Djaru language groups) ∼1200 with 3000 from surrounding towns (63% Indigenous) Restrictions on liquor sold, times of sale or mode of sale Regulation instituted by state or territory liquor licensing authorities Crime; Social or health service usage Alcohol sales or consumption data 3 years (1991–1994) d'Abbs (1998)29 Australia (Aboriginal and Torres Strait Islander) 7 Indigenous communities (of total of 8 with licensed clubs) Indigenous-controlled liquor licensing Not stated Alcohol consumption Alcohol sales or consumption data 1 year (1994–1995) Berman et al (2000)22 USA (Alaska Native) 29 000 (26 000 in control group) Indigenous-controlled liquor licensing; Prohibition of sale, importation or possession Local option law Injury Legal determinations classifying injury deaths (data not available to determine whether alcohol-related) 13 years (1980–1993) Gray et al (2000)15 Australia (Aboriginal and Torres Strait Islander) ∼2700 (all residents of Tennant Creek) Restrictions on liquor sold, times of sale or mode of sale Regulation instituted by state or territory liquor licensing authorities Alcohol consumption; hospital admissions; admissions to local women's refuge and sober up shelter; crime; Alcohol sales data, health and social service admissions data; local police data 4 years (1994–1998) Ellis (2003)28 USA (American Indian) McKinley County: 43 000 Indigenous Fremont: 7000 Indigenous Restrictions on liquor sold, times of sale or mode of sale; local excise tax; enhanced law enforcement Local option law Crime; Mortality (motor vehicle accident mortality, homicide, suicide and alcohol-induced causes); Motor vehicle accidents Alcohol sales or consumption data; adolescent substance use data; Hospital records or medical examiner reports; Traffic accident report files; crime data for serious or minor offences 21 years (1974–1995) for annual mortality rates for selected substance abuse-related causes; 1 year (1989–96) for traffic crash rates. Wood and Gruenewald (2006)8 USA (Alaska Native) Dry villages: 165 191 person-years (108 906 with and 56 285 without local police presence) Wet villages: 67 906 person-years (45 655 person-years with and 22 251 person-years without local police presence)*† Prohibition of sale, importation or possession Local option law Motor vehicle accidents; Intentional injury (self-harm, attempted suicide or suicide); Intentional injury (self-harm, attempted suicide or suicide) Serious injury data obtained from state trauma registries; data pertaining to police presence (number of months that a village had a police service used as an indicator of police presence) 10 years (1991–2000) Hogan et al (2006)18 Australia (Aboriginal and Torres Strait Islander) Not stated Restrictions on liquor sold, times of sale or mode of sale Regulations instituted by state or territory liquor licensing authorities Alcohol sales; crime; Social or health service usage Alcohol sales or consumption data; Hospital records or medical examiner reports; Admissions to local Sobering Up shelter; Crime data for serious or minor offences 1 year (April 2002—June 2003) Margolis et al (2008)13 Australia (Aboriginal and Torres Strait Islander) 4 Indigenous communities Restrictions on liquor sold, times of sale or mode of sale Alcohol Management Plan (AMP) Injury; Social or health service usage Royal Flying Doctor Service trauma retrieval data 8 years pre and 2 years post-AMP (1 January 1995–24 November 2005) Wood (2011)21 Canada (First Nation) 23 Indigenous communities Prohibition of sale, importation or possession Local option law Crime data for serious or minor offences 21 years (1986—2006) Margolis et al (2011)14 Australia (Aboriginal and Torres Strait Islander) Community A 1129 (1059, 94% Indigenous), community B 1101 (1028, 93% Indigenous), community C 599 (541, 90% Indigenous), community D 644 (580, 90% Indigenous). Prohibition of sale, importation or possession; Restrictions on liquor sold, times of sale or mode of sale Alcohol Management Plan (AMP) Injury; Social or health service usage Royal Flying Doctor Service trauma retrieval data 14.5 years (1 January 1996—31 July 2010) Berman (2014)26 USA (Alaska Native) 178 Indigenous communities Prohibition of sale, importation or possession Local option law Intentional injury (self-harm, attempted suicide or suicide) Cause-specific mortality data 27 years (1980-2007) *Person-years: Some studies used person-years to quantify the populations of wet and dry villages where communities changed alcohol status one or more times during the period of the study. For example, in Landen's study, multiple wet and dry villages changed alcohol status. Thus, each month a village was dry, its population contributed one-twelfth of a person-year to the dry total. A similar method was used for wet villages. †. Study designs All study designs were of before and after, cross-sectional or time-series analyses using primarily secondary data (table 2). Two of the included Australian studies are linked before and after studies evaluating the impact of alcohol controls in four communities following the introduction of two state government-facilitated supply-reduction strategies, one in 2002–2003 and the other in 2008.13 14 Eight studies evaluated health or social outcomes before and after alcohol controls were introduced.13–20 Two of those studies also evaluated health or social outcomes after the controls were repealed in their study populations.19 20 Six studies used one or more other populations as a control group to compare ‘wet’ villages (those without a restrictive alcohol law) with ‘damp’ villages (those with some restrictions in place but where alcohol could still be bought in specific containers, quantities and venues) or ‘dry’ villages (those with laws prohibiting the availability of alcohol).8 17 21–24 Two studies from the US compared health outcomes between one or more communities with different forms of governance of alcohol control.25 26 One compared traditional forms of control with communities that had implemented local option laws (government-facilitated opt-in alcohol controls that could be voted in by residents via local referendum).25 The other compared outcomes between communities that had alcohol prohibited under federal law, state law and local option laws.26 One study compared Native and non-Native American populations.27 One study evaluated a package of Indigenous-led interventions implemented at different points in time and used a time-series analysis to evaluate health and social outcomes.28 Three studies employed a comparison of Indigenous communities and state or national averages across health and social outcomes.17 28 29 In their use of comparators, some studies fell into more than one of the categories discussed above. Time horizon of evaluation The length of follow-up investigations ranged between 1 and 27 years (table 2). Categories of alcohol control A range of regulatory options were designed and implemented by local communities (table 2). The majority of controls involved prohibition of the sale, importation or possession, restrictions on liquor sold, times of sale or mode of sale, Indigenous-controlled liquor licensing and enhanced law enforcement. Other forms of regulatory control included a local excise tax, alcohol rationing and traditional forms of control. Under Greenland's alcohol rationing system, each individual was entitled to a sheet of 72 points of rationing coupons per month. Alcohol was priced according to strength or size of alcohol purchase (eg, one point for one beer and 24 points for three-quarters of a litre of hard liquor).20 Traditional forms of control were observed in Alaskan communities where the ‘Yupi'it Nation’ sovereignty movement (in opposition to government-facilitated local option laws) had gained popularity. Interventions involved a range of group-oriented responses including ignoring antisocial behaviour, presenting the transgression in a dramatised dance form in the men's house (kashgi), community leaders resolving the dispute through consultation, banishment from the village and finally, killing the offender.25 Indigenous representation in decision-making Indigenous-led controls reviewed fell into four categories. First, controls that were conceived and implemented by the community included the traditional forms of control discussed previously.25 Second, government-facilitated community-led controls included prohibition and restriction controls that were facilitated by local option laws that allowed communities to vote on their desired form of alcohol control (eg, that alcohol could be bought and sold, imported, sold and imported or that possession was to be prohibited) through legally held referendums.22 A third model involved ‘community coalitions’ backed by government intervention. In an Australian study in the Kimberley region, this involved the Alcohol Action Advisory Committee consisting of police, health, Aboriginal Legal Services and faith-based organisations as well as individual members of the community who together lobbied the state Director of Liquor Licensing to impose alcohol restrictions.17 A fourth model of governance involved government initiated community partnerships such as Queensland Australia's Meeting Challenges, Making Choices Committee. The Committee involved community justice groups and Indigenous Elders tasked with developing alcohol management plans for their communities in partnership with government agencies.13 14 Use of evidence in informing controls The extent to which research evidence or population data was used by communities to inform the development and implementation of controls is unclear as most studies (n=17) did not report this. One study reported that research evidence pertaining to the severity of local burden and conditions was available to tribal leaders and the public.28 In that study, research evidence and population health data, in combination with the local knowledge of tribal leaders, helped to inform decision makers about the various dimensions of the substance abuse problem, provide indicators for the ongoing monitoring of progress and safeguard against unintended impacts on vulnerable sectors of the community through the development of programmes (eg, an alcohol crisis centre, alcohol server training and KICK-IT, an adolescent programme targeting at-risk youth). Enforcement of alcohol controls One study investigated the impact of the brief legalisation of alcohol in a large American Indian tribe on the Great Plains where the possession and use of alcohol had never been legal. That study estimated that arrests for intoxication declined immediately after legalisation.19 The findings suggested that the enforcement of prohibition controls did not necessarily reduce incidences of harmful alcohol use. For example, authors hypothesised that drinkers were less likely to rapidly consume alcohol in attempts to circumvent prohibition controls when alcohol was legal. Another study from the US, reported that despite prohibition on the sale and importation of alcohol, 117 convictions for bootlegging were enforced by the Alaska Alcoholic Beverage Control Board over the 3-year evaluation.24 Sixteen studies did not report quantitative data regarding the legal enforcement of the alcohol controls. However, the authors noted several challenges, many specific to particular categories of control. Alcohol rationing With regards to alcohol rationing, alcohol coupons were legally non-transferable but the proper legal use of the coupons was seldom enforced. Identification was not requested and the point sheets were not printed with a marker of personal identification.20 Prohibition and restriction controls With regards to the prohibition category, ‘dry’ communities that bordered ‘wet’ communities often found a persisting problem in the movement of heavy drinkers in ‘dry’ communities to other towns where alcohol was legally available.19 27 In New Mexico, this resulted in limited options for intoxicated persons to travel home safely and deaths due to hypothermia or pedestrian accidents resulting from travelling the large distances between towns.27 In studies that evaluated interventions in the prohibition or restriction categories, providing adequate police services in rural areas was problematic.8 Challenges included difficulties imposed by extreme weather and an absence of roads connecting communities to major towns. In one study, this precluded the deployment of trained state-certified police officers to the most isolated Alaska Native villages. Only a few communities had their own local departments that employed fully certified police officers. Other villages were served by non-certified para-professional Village Public Safety Officers. However, many villages went months without the presence of an officer due to extremely high rates of attrition of officers. Some villages had no local police presence and were instead served by state troopers on an as-needed basis by air or river.8 The presence of illegal smuggling of alcohol into communities, and then unlicensed sales to drinkers presented a challenge to enforcement for other communities. In one study, authors cited anecdotal reports and unpublished consultation data pointing to rising rates of sly grogging in some communities in Cape York, Queensland.14 However, one study reported that stronger enforcement measures had brought about increased efficiency. Code enforcement had become increasingly diligent and in combination with the enforcement of other liquor laws, had the effect of reducing the frequency of inappropriate alcohol sales. The measures also led to increased community cohesion and engagement with alcohol sellers. When hours for drive-up alcohol sales were restricted, some alcohol sellers voluntarily closed their drive-up windows.28 Traditional forms of control With regard to traditional forms of control, a significant issue was that the village social structure did not support a village member assuming a formal, external, authoritarian role. Having to interfere with personal relationships or arrest one's friends and relatives was not compatible with village life.25 Effectiveness of interventions In 11 of the 18 studies reviewed, the Indigenous-led alcohol controls employed were effective in achieving improvements in one or more health outcomes (eg, reductions in disease, injury or crime) (table 3).8 13–17 20–22 24 28 Interventions were classified as effective if the controls were associated with improvements within the same population or in comparison with other populations, except where authors deemed the results inconclusive due to methodological or reporting biases. Table 3 Effect of legal interventions to control alcohol on health and social outcomes Author (year) Main results May (1975)19 Arrests declined for the 2-month period of alcohol legalisation by 30% (182 in June-July of 1969 to 126 in June-July of 1970). Following the repeal, arrests rose by over 30% to 189 in June-July of 197.1 Data pertaining to individual arrests was unavailable. However, at an aggregate level in 1969 and 1970, 88.7% of Native American arrests within and bordering the reservations were for alcohol intoxication and driving while intoxicated. Schechter (1986)20 Overall importation and consumption declined significantly. Number of drinks imported fell from 47 million in 1978 to 35.2 million in 1979, 30.5 million in 1980 and 36 million in 1981. Litres of pure alcohol consumed dropped from 513 627 in 1978 to 406 856 in 1979, 346 384 in 1980 and 436 066 in 1981. National crime rate dropped markedly particularly violent crimes (murder, attempted murder and assault). After the repeal of alcohol rationing, consumption and importation rose by 60% and incidences of crime increased significantly. Gallaher et al (1992)27 Over 50% excess mortality from all unintentional injuries among Native Americans resulted from hypothermia and pedestrian-motor vehicle crashes. New Mexico Native Americans nearly eight times more likely to die of hypothermia compared with other New Mexico residents. At death, 90% of those Native Americans tested were highly intoxicated (median blood alcohol concentrations of 0.24 and 0.18 mg/dL for pedestrian and hypothermia deaths, respectively). Most deaths occurred at off-reservation sites in border towns and on roads leading back to the reservation. Lee (1993)25 Of 9882 reported incidents of crime overall, Nation villages (employing traditional forms of control) reported 34.5% of the total, non-Nation villages (employing local option controls) reported 65.5% of the total. Rates for felonies and misdemeanors were lower in Nation villages, with the exception of liquor violations and drunk in public and protective custody incidents. In non-Nation villages, incidence of strongarm rape were 6.9 times higher, non-aggravated assault were 3.8 times higher, burglary 2.9 times higher and sexual assault 5 times higher. Nation villages reported more protective custody and drunk-in-public incidents than non-Nation villages, yet serious reported crime were lower. Chiu et al (1997)16 Substantial decrease in the number of alcohol-related outpatient visits when the ban on possession and importation was imposed compared with baseline. When the ban was lifted, outpatient visits increased; when the ban was reimposed, the number of outpatient visits again decreased. Interrupted time-series analysis confirmed that the alcohol ban, its lifting and its reimposition had a statistically significant and negative effect on the number of alcohol-related outpatient visits (p<0.05). A significantly higher number of visits were made during the two non-ban periods (November 1993-October 1994) and November 1995-February 1996) compared with the two ban periods (p<0.05). Landen et al (1997)24 Of 302 injury deaths, blood alcohol concentrations (BACs) were available for 200 deaths (66.2%). Of these, 130 (65.0%) had a BAC greater than or equal to 17 mmol/L (greater than or equal to 80 mg/dL) and were therefore considered alcohol-related. Total injury mortality rate was greater among Alaska Natives from wet villages (rate ratio (RR), 1.6; 95% CI 1.3 to 1.2). This difference was not present for non-natives (RR, 1.1; 95% CI 0.3 to 3.8). For Alaska Natives, the alcohol-related injury mortality was greater among residents of wet villages (RR, 2.7; 95% CI 1.9 to 3.8) than among residents of dry villages. The strength of this association was greatest for deaths due to motor vehicle injury, homicide and hypothermia. Landen (1997)23 No significant differences across reservations with prohibition controls and those where alcohol was legal. The average age-adjusted mortality rate found that the mortality rate was higher (n=158) than prohibition (n=138) reservations. Douglas (1998)17 Decrease in alcohol consumption observed for each of the 2 years following the intervention. Overall, incidence of crime declined. Alcohol-related presentations to the hospital and presentations resulting from domestic violence decreased relative to the equivalent quarterly period prior to the intervention. Short-term fluctuations were observed, particularly with domestic violence, where presentations (of lesser severity) became more request during several quarters. Emergency evacuations as a result of injury showed a marked decrease. D'Abbs (1998)29 Mean consumption levels in standard drink terms, the corresponding equivalents were 5.8 standard drinks per day for female drinkers and 9.3 for male drinkers. Among both male and female drinkers, the overall mean consumption levels were ∼50% above the level designated as harmful. In one community, mean consumption lay in the responsible range and the other in the hazardous range. In all others, male and females were above the harmful level. Total beer sales in the seven clubs (1994-1995) amounted to 882 259 L. Assuming a retail price of $3.50 per 375 mL can of full-strength beer and $3.00 per can of light beer, the total retail turnover would amounted to ∼$8.1 million. Berman et al (2000)22 Injury deaths generally lower during periods when alcohol sales, importation or possession were restricted than when no restrictions were in place (wet). More restrictive controls (dry) significantly reduced homicides but had no effect on suicide rates; less restrictive control options (damp) reduced suicides but had no effect on homicide rates. Accident and homicide death rates fell, on average, by 74 and 66 per 100 000, respectively, for the 89 communities that banned sale and importation or possession. Sixty-one small communities that did not change control status under the law showed no significant changes over time in accident or homicide death rates. The decline in overall injury death rates was much greater in communities with less restrictive options (127 compared to 48 per 100 000). However, death rates were higher in these communities while they were wet, with the discrepancy statistically significant for suicides. Gray et al (2000)15 Annual per capita consumption declined by 19.4% in the 2 years following the introduction of alcohol restrictions. Hospital admissions for acute alcohol-related conditions declined. Restrictions were circumvented by a shift to fortified wine purchases. Fortified wine purchases increased by 570% (573 L) offsetting 14% of the mean quarterly decline of 4173 L of cask wine immediately following the restrictions. Purchases at Aboriginal-controlled licensed clubs (also not covered by the restrictions) increased by 55.7% from 2801 L to 1799 L of pure alcohol and offsetting 20% of the mean quarterly decline of 3002 L. Ellis (2003)28 Following the introduction of restrictions on liquor sold, times of sale or mode of sale, a local excise tax and enhanced law enforcement measures, from 1974 to 1995, McKinley County's (MC) motor vehicle accident mortality rate declined by 60% and was matched by similar declines in mortality from homicide (58%), suicide (59%), alcohol-induced causes (30%) and drug-induced causes (50%). From 1989 to 1995, alcohol-related arrests declined 42% in Gallup, and protective custody detentions were cut in half. Between 1982 and 1995, traffic crashes had declined 32% in MC. All declines experienced in MC exceeded similar trends for New Mexico and the nation. Wood (2006)8 Villages prohibiting alcohol had lower age-adjusted rates of serious injury resulting from assault, motor vehicle collisions. Dry villages with a local police presence had a lower age-adjusted rate of serious injury caused by assault. Local prohibition was associated with lower rates of assault injuries. Local police presence was associated with lower rates of assault injuries. Contrary to expectations, there was no difference in the age-adjusted rate of injury attributed to self-harm for wet vs dry isolated Alaska Native villages. Rates of serious injury caused by assault were 36% higher in villages during periods of police absence than when police were present. Hogan et al (2006)18 Over 12 months the reduction in trading hours was accompanied by decreased levels of alcohol-related harm. However, the regulation of container size was undermined by a shift to cheap cask port with sales of this product increasing by 1000%. A one-third reduction was observed in instances of drunkenness and breaches of the 2 km law, as were Protective Custodies. Ambulance services received 25% less alcohol-related call-outs and selected presentation to the emergency department of Alice Springs were reduced by 19%. Alcohol-related assaults were 13% lower. There was a nearly 20% increase in alcohol-related offences, especially criminal damage and disturbances and indications of more acute conditions being admitted to Alice Springs Hospital. However, the author's reanalysis of these findings found that the evaluators did not provide sufficient data to ascertain whether these were significant reductions of chance phenomena. Margolis et al (2008)13 Overall reduction (2 years vs 2 years before the AMP was implemented) was on average 51.9% (Community (A) 44.8% (B) 54.6% (C) 66% (D) 42.2%). Retrieval rates for all other causes did not reveal any statistically significant change. Serious injury resulted in 798 retrievals during the observation period. There was a significant (p=0.021) decrease in injury after the introduction of AMP. Wood (2011)21 Wet communities in Nunavut recorded rates of violent crime that were higher than dry communities. Relative to dry communities, wet communities’ overall sexual assault rate was 1.48 (95% CI 1.38 to 1.60) times higher, the serious assault rate was 2.10 (95% CI 1.88 to 2.35) times higher and the homicide rate was 2.88 (95% CI 1.18 to 8.84) times higher. Dry communities were safer than wet communities but still reported rates of violence that were higher than national rates, including a serious assault rate that was double the national rate (3.25 per 1000 vs 1.44 per 1000) and a sexual assault rate that was at least seven times as high as the national rate (7.58 per 1000 vs 0.88 per 1000). Homicide, the rarest violent offence, was relatively more frequent in wet communities than in dry communities (RR=2.88; 95% CI 1.18 to 8.84). Margolis et al (2011)14 After alcohol restrictions were introduced in 2002-2003, retrievals for serious injury dropped initially, then increased in the 2 years before further restrictions in 2008 (average increase, 2.34 per 1000 per year). This trend reversed in the 2 years after the 2008 restrictions (average decrease 7.97 per 1000 per year). There was a statistically significant decreasing time trend in serious-injury retrieval rates in each of the four communities for the period 2 years before the 2002-2003 restrictions, 2 years before the 2008 restrictions and the final 2 years of observations (2009-2010) (p <0.001 for all four communities combined). Overall, serious injury retrieval rates dropped from 30 per 1000 in 2008 to 14 per 1000 in 2010, and the proportions of serious-injury retrievals decreased significantly for all four communities. Berman (2014)26 Suicide rates were higher in communities prohibiting alcohol importation under state law, but the effect was not significant after controlling for other community characteristics. More remote communities, those with fewer non-Natives and those with evidence of cultural divides had higher suicide risks. Communities with higher incomes, more married couples and traditional elders had lower risks. There was a strong association of community characteristics with the choice of alcohol status, consistent with the hypothesis that it is endogenous. Communities choosing alcohol control by referendum were generally larger, with a higher percentage of Alaska Native residents and more remote. Communities with lower median incomes were more likely to choose prohibition. Young men's suicide risks were significantly higher (p<0.01) when alcohol was prohibited under the state local option law. The association between adoption of any local alcohol control option and suicide was even stronger (p=0.01). However, communities using federal Indian law to ban alcohol had significantly lower suicide risks (p<0.5). Seven of these studies involved the evaluation of one category of alcohol control.8 13 16 17 20 21 24 The remaining four studies involved evaluations of multiple categories of alcohol control.14 15 22 28 Prohibition of alcohol possession, sale or importation was the most frequently implemented form of alcohol control among those that were effective (n=6).8 14 16 21 22 24 Five studies reported less crime in the communities studied (either in comparison to before the controls were implemented, or in comparison to other populations without similar controls).8 17 20 21 28 Three studies reported fewer injury deaths.22 24 28 Four studies reported fewer injuries in the communities studied.8 13 14 17 One reported that annual per capita consumption of pure alcohol declined, as well as hospital admissions for acute alcohol-related conditions.15 Some alcohol controls were effective in achieving only some of the health outcome measures included in the evaluations.17 22 In a US study of communities that went ‘dry’, by prohibiting the sale, importation and possession of alcohol, they achieved reductions in homicides but there was no effect on rates of suicide. Communities that became ‘damp’ achieved reductions in suicides, but there was no effect on homicides.22 In an Australian study, a package of interventions implemented in a single community including reduced trading hours for take-away outlets and restrictions on the type of alcohol sold at specific times of the day, achieved reductions in criminal charges imposed by the Halls Creek Police Station. Reductions were also achieved in aeromedical retrieval evacuations resulting from serious injury requiring hospital treatment and as such, could not be managed within the study communities. However, despite fluctuations in the data for domestic violence presentations to Halls Creek District Hospital, there was no discernible or statistically significant evidence of an increasing or decreasing trend in hospitalisations over a 2-year evaluation period.17 Controls also had unintended consequences. In Tennant Creek in Australia, despite overall reductions in per capita consumption, restrictions were circumvented with a shift to fortified wine purchases which remained accessible when sold in containers less than or equal to 1125 mL. In addition, unlawful entries to dwellings significantly increased following the imposition of alcohol restrictions.15 In six studies, alcohol controls were ineffective (table 3).18 19 23 26 27 29 In one study in New Mexico, US, there had been an increase in hypothermia and pedestrian deaths in response to restrictions that deemed alcohol possession and sale to be illegal.27 In another US study involving 178 Alaska Native communities, the prohibition on importation was shown to be ineffective in preventing suicide, the primary outcome of the study. This finding raised questions about whether alcohol was the causal factor in risk of suicide or a comorbid response to depression or other mental health problems.26 One US study found no significant differences in alcohol-related mortality between reservations where alcohol possession and use was prohibited and those where it was legal.23 In another US study, the short-lived (2-month) legalisation of alcohol possession and consumption saw a decline in arrests for driving while intoxicated.19 Arrests significantly increased following a return to prohibition.19 In the Northern Territory in Australia, Aboriginal-controlled licensed clubs failed to lower persistently high consumption levels. Aboriginal-controlled licensed clubs, a term often used interchangeably with Aboriginal Social Clubs and wet canteens are specific community venues that encourage responsible drinking patterns (eg, through day caps, restrictions on hours of sale or bans on problem-drinkers) and reinforce community standards. In communities with these clubs, per capita consumption remained 183% higher than the territory average in males and 76% higher in females.29 Another Australian study within the same territory found that implementation of a suite of measures including restrictions on time of sale in take-away outlets, the removal of liquor in containers >2 L and the provision of only light beer in bars before noon was not associated with a significant reduction in quarterly wholesale sales of pure alcohol over the 12-month trial period. Rather, drinkers shifted from one product to another product of equal price and purchased the same amount of pure alcohol. There was a shift in consumption from cask wine to 2 L cask port which was priced at 28 cents (AUD) per 10 mL, the same price as alcohol in the 4 and 5 L casks that became unavailable under the trial restrictions.18 In a study of Alaska Natives, the results were inconclusive.25 The data showed that villages that had maintained traditional forms of social control had less crime (including felonies and misdemeanours, with the exception of liquor violations as well as drunk in public and protective custody incidents). However, the authors reasoned that this may be linked to reporting mechanisms—the abolition of VPSO positions, avoiding contact with state troopers, relying on traditional mechanisms with no formal records of crime or choosing to ignore certain behaviours. Another possibility that was highlighted is that such communities are more culturally cohesive as they have retained traditional (Yup'ik) values to a greater extent than communities with ‘Western mechanisms’ of social control developed by state and federal governments and operationalised by authorities (eg, ‘judges, magistrates, district courts and marshals’) external to Indigenous social and cultural systems.28 Discussion This review found 18 quantitative evaluations of Indigenous-led alcohol controls. The controls were implemented in rural and remote populations in high-income countries. Prohibition on the sale, importation or possession of alcohol was the predominant category of control. Eleven of the 18 studies found one or more alcohol controls (predominantly prohibition) to be effective in improving a broad range of health and social outcomes including reduced rates of hospitalisations, injury and crime. The findings of this review add to qualitative research that has explored the widespread and under-recognised agency of Indigenous communities to engage in effective health governance.30 31 The studies reveal a widespread preference for strong, legal responses to public health problems such as harmful alcohol use which confirm findings of qualitative work exploring community views regarding Indigenous-led alcohol controls.15 32 In addition, they suggest that many such policy models characterised by community representation and leadership in the development and implementation of alcohol controls are effective in improving health outcomes and deserve serious consideration by governments who can facilitate opportunities for Indigenous populations globally to actively participate in the advocacy, design and implementation of public health law. Nevertheless, the results reveal the complexity involved in implementing Indigenous-led alcohol controls such that populations benefit (in terms of health and social outcomes and other social goods including community engagement and capacity building) and vulnerable members of the community are not further disadvantaged. Many of the studies reviewed reported unintended impacts on vulnerable members of communities including those experiencing substance addiction, unemployment, poverty, women and children and other at-risk populations. This was particularly significant for communities that were not geographically isolated from communities where alcohol was legally available. The availability of alcohol in other regions carried the potential to undermine controls in neighbouring ‘dry’ or ‘damp’ communities by facilitating heavy drinking in neighbouring areas without alcohol restrictions or the illegal importation of alcohol into ‘dry’ areas. For example, the introduction of prohibition, the strongest form of alcohol control carried health risks to individuals experiencing substance abuse and travelling long distances to neighbouring areas to obtain alcohol. Health risks included hypothermia due to excessively cold weather conditions, hunger and thirst and pedestrian deaths due to being struck by a vehicle driven by an intoxicated driver or exposure to risk as an intoxicated pedestrian.27 This information problem is pervasive in under-resourced and marginalised communities.33 However, the experience of McKinley County in the US, where sizable declines (60%) in the rate of motor vehicle accident mortality and (40%) in alcohol-related arrests were achieved, may suggest that communities could benefit from being able to access and use research evidence and public health surveillance systems to inform local alcohol policy.28 Another significant challenge faced by communities was maintaining the sustainability of the controls, which in many cases were repealed and reinstated on multiple occasions. This occurred due to demographic change (eg, a higher proportion of non-Indigenous voters, who were less likely to support alcohol restrictions), local business and economic interests, unpopularity of the control and inadequate resourcing to combat unlicensed alcohol supply (bootlegging), unsafe home brewing and crime (eg, theft) committed with the intent of obtaining possession of alcohol.17 20 22 24 27 Prohibition controls that were implemented through local elections in the US and Canada could be repealed easily by local referendum and as such, many communities changed their alcohol status from year to year, highlighting the challenges of sustainability.16 21 An additional challenge to the sustainability of controls is their enforcement. Few studies reported data (eg, arrests, imposition of penalties) on the extent to which controls were enforced. Such measures often do not reveal the true extent of compliance. However, to the extent that it is accessible, the reporting of these data could work to better inform and prepare communities as alcohol control models evolve and adapt to community need. The responsibility of governments to protect the health of populations, particularly that of marginalised populations is almost universally recognised.34 35 Further, national and subnational governments are considerably better resourced than Indigenous communities to meet this aim and as such, hold a unique position to contribute in the key areas highlighted in this review.34 35 The findings suggest that a first step could be taken in remedying the information problem by providing access to public health evidence and population data at the outset of the consultation and design process. This could enable a more targeted, informed and effective approach, help to identify potential unintended impacts on vulnerable individuals who may be disproportionately affected and identify measurable indicators for the monitoring and evaluation of the control. In addition, governments could provide support to ensure local decision makers have access to a context-specific menu of available legal avenues that communities can consider for potential effectiveness, acceptability and equity. Specifically, this could involve legal information pertaining to the availability of alcohol within the communities and in surrounding populations, legal status of the control, enforcement measures, the scope of the control, models for immediate and stepwise implementation and consideration of time horizons (eg, trial periods or sunset provisions). In addition, this could include the provision of case studies and evaluations of controls in other communities. A key strength of this study is its comprehensiveness. No limitations were placed on the basis of language, study type, type of health or social outcome, measure of effectiveness or Indigenous population. Further, we included traditional forms of control that did not involve government intervention. One limitation is that we focused exclusively on the peer reviewed literature as it there are no comprehensive global holdings of the grey literature in this area. Conclusions Public health law can be a powerful prevention tool capable of alleviating the devastating effects of alcohol and alcohol-related harms, particularly where strengthened by community based and culturally sensitive enforcement mechanisms. Advocates and Indigenous health scholars have underscored the importance of Indigenous peoples retaining a place at the forefront of decision-making in a manner that is reflective of Indigenous rights including, but not exclusive to those enshrined in the UN Declaration on the Rights of Indigenous Peoples and distinct from the almost universal history of paternalistic alcohol control.5 36 The findings of this review indicate that community-led alcohol controls characterised by their development and/or implementation by Indigenous communities globally have been shown to be effective in improving health and social outcomes. Twitter: Follow JANANI MUHUNTHAN @JanChildRights Contributors: JM wrote the first draft and led the writing process. JM and BA independently conducted the systematic review with any uncertainties about inclusion or exclusion of empirical studies resolved in discussion with SJ and MLH. All authors contributed to the conceptualisation and design of the paper. All authors contributed to the refinement of the manuscript. Funding: This research was supported by a Doctoral Scholarship (held by JM) from The Australian Prevention Partnership Centre, funded by NHMRC, the Australian Government Department of Health, NSW Ministry of Health, ACT Health and the HCF Research Foundation. SJ was in receipt of a NHMRC Senior Research Fellowship. MH was in receipt of a National Heart Foundation Future Leader Fellowship (100034). Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Brady M Ethnography and understandings of Aboriginal drinking . J Drug Issues 1992 ;22 :699 –712 . 10.1177/002204269202200316 2 Gray D , Saggers S 6.7 The Evidence Base for Responding to Substance Misuse in Indigenous Minority Populations. Preventing harmful substance use: the evidence base for policy and practice 2005 :381 . 3 Lander F , Gray D , Wilkes E The Alcohol Mandatory Treatment Act: evidence, ethics and the law . Med J Aust 2015 ;203 :47 –9 . 10.5694/mja15.00173 26126570 4 Wood DS Explanations of employment turnover among Alaska village public safety officers . J Crim Justice 2002 ;30 :197 –215 . 10.1016/S0047-2352(01)00136-2 5 United Nations . Implementing the UN Declaration on the Rights of Indigenous Peoples Handbook for Parliamentarians: United Nations 2014 http://www.un.org/esa/socdev/publications/Indigenous/Handbook/EN.pdf. Accessed 30 June, 2016 . 6 National Aboriginal Community Controlled Health Organisation (NACCHO) . Aboriginal Health: NACCHO 2016 http://www.naccho.org.au/about/aboriginal-health/. 30 June, 2016. Secondary National Aboriginal Community Controlled Health Organisation (NACCHO). Aboriginal Health: NACCHO 2016 http://www.naccho.org.au/about/aboriginal-health/. 30 June, 2016 . 7 Calladine K Liquor restrictions in Western Australia . Indigenous Law Bulletin 2009 ;7 11:23 10.1136/jech.2006.057455 8 Wood DS , Gruenewald PJ Local alcohol prohibition, police presence and serious injury in isolated Alaska Native villages . Addiction 2006 ;101 :393 –403 . 10.1111/j.1360-0443.2006.01347.x 16499512 9 Marks-Sultan G , Tsai FJ , Anderson E National public health law: a role for WHO in capacity-building and promoting transparency . Bull World Health Organ 2016 ;94 :534 –9 . 10.2471/BLT.15.164749 27429492 10 Health Organization . Indigenous populations: World Health Organization 2016 http://www.who.int/topics/health_services_indigenous/en/. (accessed 30 June) . 11 Arnold M , Moore SP , Hassler S The burden of stomach cancer in indigenous populations: a systematic review and global assessment . Gut 2014 ;63 :64 –71 . 10.1136/gutjnl-2013-305033 24153248 12 Angell B , Muhunthan J , Eades A-M The health-related quality of life of Indigenous populations: a global systematic review . Qual Life Res 2016 ;25 :2161 –78 . 10.1007/s11136-016-1311-9 27165149 13 Margolis SA , Ypinazar VA , Muller R The impact of supply reduction through alcohol management plans on serious injury in remote indigenous communities in remote Australia: a ten-year analysis using data from the royal flying doctor service . Alcohol Alcohol 2008 ;43 :104 –10 . 10.1093/alcalc/agm152 17923455 14 Margolis SA , Ypinazar VA , Muller R Increasing alcohol restrictions and rates of serious injury in four remote Australian Indigenous communities . Med J Aust 2011 ;194 :503 –6 .21644892 15 Gray D , Saggers S , Atkinson D Beating the grog: an evaluation of the Tennant Creek liquor licensing restrictions . Aust N Z J Public Health 2000 ;24 :39 –44 . 10.1111/j.1467-842X.2000.tb00721.x 10777977 16 Chiu AY , Perez PE , Parker RN Impact of banning alcohol on outpatient visits in Barrow, Alaska . JAMA 1997 ;278 :1775 –7 . 10.1001/jama.1997.03550210073042 9388156 17 Douglas M Restriction of the hours of sale of alcohol in a small community: a beneficial impact . Aust N Z J Public Health 1998 ;22 :714 –19 . 10.1111/j.1467-842X.1998.tb01476.x 9848970 18 Hogan E , Boffa J , Rosewarne C What price do we pay to prevent alcohol-related harms in Aboriginal communities? The Alice Springs trial of liquor licensing restrictions . Drug Alcohol Rev 2006 ;25 :207 –12 .16753643 19 May PA Arrests, alcohol, and alcohol legalization among an American Indian tribe . Plains Anthropologist 1975 :129 –34 . 20 Schechter EJ Alcohol rationing and control systems in Greenland. Contemp . Drug Probs 1986 ;13 :587 . 21 Wood DS Alcohol controls and violence in Nunavut: a comparison of wet and dry communities . Int J Circumpolar Health 2011 ;70 :19 –28 . 10.3402/ijch.v70i1.17801 21329578 22 Berman M , Hull T , May P Alcohol control and injury death in Alaska Native communities: wet, damp and dry under Alaska's local option law . J Stud Alcohol 2000 ;61 :311 –19 . 10.15288/jsa.2000.61.311 10757142 23 Landen MG Alcohol-related Mortality and Tribal Alcohol Legislation . J Rural Health 1997 ;13 :38 –44 . 10.1111/j.1748-0361.1997.tb00832.x 10167764 24 Landen MG , Beller M , Funk E Alcohol-related injury death and alcohol availability in remote Alaska . JAMA 1997 ;278 :1755 –8 . 10.1001/jama.1997.03550210053038 9388152 25 Lee N Differential deviance and social control mechanisms among two groups of Yup'ik Eskimo . Am Indian Alsk Native Ment Health Res 1993 ;5 :57 –72 . 10.5820/aian.0502.1993.57 8130314 26 Berman M Suicide among young Alaska Native men: community risk factors and alcohol control . Am J Public Health 2014 ;104 (Suppl 3 ):S329 –35 . 10.2105/AJPH.2013.301503 24754505 27 Gallagher MM , Fleming DW , Berger LR Pedestrians and hypothermia deaths among Native Americans in New Mexico . JAMA 1992 ;267 :1345 –8 . 10.1001/jama.1992.03480100051029 1740855 28 Ellis BH Mobilizing communities to reduce substance abuse in Indian country . J Psychoactive Drugs 2003 ;35 :89 –96 . 10.1080/02791072.2003.10399999 12733764 29 d'Abbs PH Out of sight, out of mind? Licensed clubs in remote aboriginal communities . Aust NZ J Public Health 1998 ;22 :679 –84 . 10.1111/j.1467-842X.1998.tb01469.x 30 Clough AR , Margolis SA , Miller A Alcohol control policies in Indigenous communities: a qualitative study of the perceptions of their effectiveness among service providers, stakeholders and community leaders in Queensland (Australia) . Int J Drug Policy 2016 ;36 :67 –75 . 10.1016/j.drugpo.2016.06.015 27518836 31 Boffa J , Rosewarne C Alcohol and Alice Springs: meeting the needs and wishes of the community . Indigenous Law Bulletin 2003 ;5 :11 . 32 Brady M , Byrne J , Henderson G ‘Which Bloke Would Stand Up for Yalata?’: the Struggle of an Aboriginal Community to Control the Availability of Alcohol . Australian Aboriginal Studies 2003 ;62 –71 . 33 Abimbola S The information problem in global health . 34 Gracey M , King M Indigenous health part 1: determinants and disease patterns . Lancet 2009 ;374 :65 –75 . 10.1016/S0140-6736(09)60914-4 19577695 35 King M , Smith A , Gracey M Indigenous health part 2: the underlying causes of the health gap . Lancet 2009 ;374 :76 –85 . 10.1016/S0140-6736(09)60827-8 19577696 36 d'Abbs P Widening the gap: the gulf between policy rhetoric and implementation reality in addressing alcohol problems among Indigenous Australians . Drug and Alcohol Review 2015 ;34 :461 –6 . 10.1111/dar.12299 26179138

PMC005xxxxxx/PMC5372060.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01447210.1136/bmjopen-2016-014472Intensive CareResearch150617071707Transfusion of red blood cells in patients with traumatic brain injuries admitted to Canadian trauma health centres: a multicentre cohort study Boutin Amélie 12Moore Lynne 12Lauzier François 13Chassé Michaël 13English Shane 4Zarychanski Ryan 5McIntyre Lauralyn 4Griesdale Donald 6Fergusson Dean A 4http://orcid.org/0000-0001-5675-8791Turgeon Alexis F 13 1 Population Health and Optimal Health Practices Research Unit (Trauma—Emergency—Critical Care Medicine), CHU de Québec—Université Laval Research Centre, Université Laval, Québec, Québec, Canada 2 Department of Social and Preventive Medicine, Université Laval, Québec, Québec, Canada 3 Department of Anesthesiology and Critical Care Medicine, Division of Critical Care Medicine, Université Laval, Québec, Québec, Canada 4 Clinical Epidemiology Unit, Ottawa Hospital Research Institute, Ottawa, Ontario, Canada 5 Department of Internal Medicine, Sections of Critical Care Medicine, of Haematology and of Medical Oncology, University of Manitoba, Winnipeg, Manitoba, Canada 6 Department of Anesthesia, University of British Columbia, Vancouver, British Columbia, CanadaCorrespondence to Dr Alexis F Turgeon; alexis.turgeon@fmed.ulaval.ca2017 29 3 2017 7 3 e01447226 9 2016 17 2 2017 27 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Background Optimisation of healthcare practices in patients sustaining a traumatic brain injury is of major concern given the high incidence of death and long-term disabilities. Considering the brain's susceptibility to ischaemia, strategies to optimise oxygenation to brain are needed. While red blood cell (RBC) transfusion is one such strategy, specific RBC strategies are debated. We aimed to evaluate RBC transfusion frequency, determinants of transfusions and associated clinical outcomes. Methods We conducted a retrospective multicentre cohort study using data from the National Trauma Registry of Canada. Patients admitted with moderate or severe traumatic brain injury to participating hospitals between April 2005 and March 2013 were eligible. Patient information on blood products, comorbidities, interventions and complications from the Discharge Abstract Database were linked to the National Trauma Registry data. Relative weights analyses evaluated the contribution of each determinant. We conducted multivariate robust Poisson regression to evaluate the association between potential determinants, mortality, complications, hospital-to-home discharge and RBC transfusion. We also used proportional hazard models to evaluate length of stay for time to discharge from ICU and hospital. Results Among the 7062 patients with traumatic brain injury, 1991 patients received at least one RBC transfusion during their hospital stay. Female sex, anaemia, coagulopathy, sepsis, bleeding, hypovolemic shock, other comorbid illnesses, serious extracerebral trauma injuries were all significantly associated with RBC transfusion. Serious extracerebral injuries altogether explained 61% of the observed variation in RBC transfusion. Mortality (risk ratio (RR) 1.23 (95% CI 1.13 to 1.33)), trauma complications (RR 1.38 (95% CI 1.32 to 1.44)) and discharge elsewhere than home (RR 1.88 (95% CI 1.75 to 2.04)) were increased in patients who received RBC transfusion. Discharge from ICU and hospital were also delayed in transfused patients. Conclusions RBC transfusion is common in patients with traumatic brain injury and associated with unfavourable outcomes. Trauma severity is an important determinant of RBC transfusion. Prospective studies are needed to further evaluate optimal transfusion strategies in traumatic brain injury. TransfusionTraumatic Brain InjuryRed Blood CellInstitute of Health Services and Policy Researchhttp://dx.doi.org/10.13039/501100000037 ==== Body Strengths and limitations of this study This study was conducted using a large data set from the inclusive Canadian trauma systems. We explored a large spectrum of a priori conceptual models of the potential relationship between transfusions and clinical variables. Although residual confounding remains a concern, we optimised control for potential bias. We, however, could not perform any evaluation of the impact of haemoglobin thresholds on red blood cell transfusions. Introduction Traumatic brain injury is a major cause of death and long-term disabilities.1–6 Caring for the most severe cases, usually admitted to the intensive care unit (ICU), requires significant and costly healthcare resources.1 7 8 Care provided to this specific ICU population is the subject of significant research with the aim to establish best practice. Since brain oxygenation is associated with secondary brain injuries,9 10 increasing cerebral oxygen delivery by maintaining higher levels of haemoglobin, through transfusion of red blood cells, has been suggested.9 11 The presumed benefits of red blood cell transfusion, however, might be counterbalanced by potential harms due to inflammatory, thrombotic and immunological effects of transfusion, many of which are not well understood.10 12 Even though optimal transfusion strategies in this population are uncertain and clinical equipoise exists,4 10 13–18 blood products are frequently administered to patients who have sustained a traumatic brain injury, with approximately a third receiving red blood cells according to our previous systematic review.2 19 Significant practice variation is also observed.19 20 Furthermore, little is known regarding potential determinants of transfusions, most cohort studies on transfusion in traumatic brain injury having reported no data or only descriptive statistics of potential determinants.19 Also, red blood cell transfusion has been shown to be associated with ICU and hospital lengths of stay in various population,21–24 even after adjustment for age, trauma or pathology severity and comorbidities. One trial and a subgroup of a large trial conducted specifically in patients with traumatic brain injury25 26 have failed to show significant differences in patient-oriented outcomes between transfusion strategies (transfusion at a high haemoglobin level compared with a low haemoglobin level). Best practices to guide optimal use of red blood cell transfusion in this population remain unclear19 20 and further analyses of the current transfusions practices, their determinants and associated outcomes are needed to best inform the design of a future clinical trial. The objective of our study was to estimate the frequency of red blood cell transfusion in patients with traumatic brain injury across Canadian health centres. Our secondary aim was to explore potential determinants of and clinical outcomes associated with transfusion. Finally, we wanted to examine effect modification by age, comorbidities and severity of traumatic brain injury on these associations. Methods Study design, setting and participants We conducted a retrospective cohort study of patients with a moderate to severe traumatic brain injury admitted to Canadian trauma centres. The study was reported in accordance to the Strengthening of the Reporting of Observational Studies in Epidemiology (STROBE) statement.27 Patients were identified in the Canadian National Trauma Registry, which is maintained by the Canadian Institute for Health Information (see online supplementary appendix A for eligibility criteria of the registry). Records from the registry were linked to the Discharge Abstract Database, also maintained by the Canadian Institute for Health Information. The Discharge Abstract Database does not include data from the province of Québec. 10.1136/bmjopen-2016-014472.supp1supplementary appendices Adult patients (aged ≥18 years old) with a moderate or severe traumatic brain injury defined using International Classification of Diseases codes (ICD-10 S06), and Glasgow Coma Scale (GCS) scores (<13) or intubated on admission, admitted to one of the centres participating to the registry (see online supplementary appendix A) between April 2005 and March 2013 were eligible. Since data pertaining to transfusions were collected from the Discharge Abstract Database, patients whose records could not be linked to the database were excluded. Variables Our primary objective and exposure of interest was the frequency of red blood cell transfusion, which was available in the Discharge Abstract Database as a dichotomous variable indicating if a patient had received or not received a transfusion at any time during hospitalisation. Our secondary objectives were to identify the potential determinants of transfusion. Variables related to age, sex, brain injury severity (GCS scores), extracerebral injuries (measured by the Abbreviated Injury Scale (AIS)28), on-admission comorbidities such as anaemia, bleeding, coagulopathy, sepsis, hypovolemic shock, ischaemic heart disease and cerebrovascular disease (identified with ICD-10 codes according to the literature,29 30 list provided in online supplementary appendix B), and surgical procedures (identified using Canadian Classification of Health Interventions (CCI) codes, list provided in online supplementary appendix C) were extracted. Those variables were also potential confounding factors in the analysis of clinical outcomes. Our tertiary objectives were to evaluate the potential effect of red blood cell transfusion on clinical outcomes, that is, mortality, complications (defined previously in the literature31 32 using ICD-10 codes, list provided in online supplementary appendix D), discharge destinations (home or elsewhere), length of hospital stay, length of ICU stay and duration of mechanical ventilation. When a patient had been admitted to more than one centre for the same injury (eg, transferred from one hospital to another), we aggregated data from both hospital stays and considered it as one trauma hospital stay. For descriptive purposes and when considering cluster effects related to hospitals, we used data from the centre with the highest level of care. Statistical analysis Primary objective: frequency of red blood cell transfusion We computed the overall incident proportion of patients with traumatic brain injury that received at least one red blood cell transfusion and its 95% CI. Secondary objective: potential determinants We conducted multivariable analyses to identify baseline factors associated with red blood cell transfusion, including age, sex, comorbidities, traumatic brain injury severity (GCS and head AIS), and serious extracerebral injuries. We constructed robust Poisson models,33 with log link and random intercept at the centre level to take into account the variation in transfusion practices across centres. The final model included all potential determinants identified. We computed relative weight analyses based on logistic models to evaluate of relative strength of each identified variable associated with transfusion, individually and when grouped according to the nature of the variable (pertaining to patient characteristics or trauma characteristics). The analysis uses variable transformations and creates sets of predictors that are orthogonal to one another in order to compute standardised weights,34 and allowed evaluating the importance of each predictor and their contribution relative to each other in the variation explained by the model. A SAS macro created by Tonidandel in collaboration with Breton was used to compute the relative weights. Considering that surgical procedures or interventions associated with bleeding may have occurred after transfusion or may be intermediate factors of multiple causal paths, we did not include them in our main models, but rather constructed alternative models adding an intervention indicator. Tertiary objective: patients’ clinical outcomes We constructed a robust generalised linear mixed model with a Poisson distribution, log link and random intercept at the centre level to evaluate the association between red blood cell transfusion and mortality, presence of at least one trauma complication or hospital-to-home discharge, while adjusting for the previously identified baseline factors (ie, potential determinants of transfusion and confounders). We evaluated the association between red blood cell transfusion and lengths of ICU and hospital stay with crude mean differences and as the distribution of length of stay was right-skewed, we used a log-transformation and computed adjusted geometric mean ratios in survivors. Since we considered a short length of stay better than a longer stay and death can shorten hospitalisation, lengths of stay analyses were primarily limited to survivors. In order to avoid the exclusion of patients who died, with length of stay contingent on survival,35 a secondary analysis considered death at discharge in the model. Deaths were equivalent to indefinite stay (the longest ICU stay or hospital stay observed,36 37 ie, 400 and 620 days respectively, with censoring at the end of observation). We constructed proportional hazard models of time to discharge from ICU or hospital, censoring death.37 We computed the inverse of HRs of ICU and hospital discharge to obtain a measure that would be >1 for unfavourable outcome (lower hazard of being discharged alive; longer stay), and <1 for favourable outcome (higher hazard of being discharged; shorter stay). We evaluated association between transfusions and mechanical ventilation duration using multivariable mixed log-linear model, taking into account the cluster effect at centre level. We obtained adjusted geometric mean ratios and their 95% CIs. Effect modification/stratified analyses We conducted exploratory stratified analysis according to age, comorbidities and traumatic brain injury severity (moderate (GCS 9–12) or severe (GCS 3–8)), these factors being considered potential effect modifiers of the association between red blood cell transfusion and clinical outcomes by the study expert committee. We considered specific groups of baseline comorbidities (ischaemic heart disease, cerebrovascular disease, sepsis and septic shock, hypovolemic shock, bleeding, coagulopathy, anaemia identified using ICD-10 codes) in order to account, at least partially, for clinical heterogeneity of the pathological processes associated with different groups of comorbidities. Sensitivity analyses In a sensitivity analysis, lengths of stays were evaluated among survivors only and we reconducted the analyses for binary outcomes after excluding patients who died within 24, 48 and 72 hours of trauma since most withdrawal of life-sustaining therapy happened in a short window after trauma6 and death would therefore not be related to transfusion practices, but other factors such as goals of care for which we had no information. Treatment of missing data Multiple imputation was used for the treatment of missing GCS data. Ten data sets per imputation process were created. We used Markov chain Monte Carlo method, based on a multivariate normal model with 1000 chain iterations. Since GCS is an ordinal variable, we created dummy variables, imputed their values, and transformed them back into an ordinal scale by comparing them to the observed distribution of GCS scores.38 39 All analyses were conducted using SAS statistical software packages (V.9.3, SAS Institute, Cary, North Carolina, USA). A type I error of 5% was considered. Results Among the 58 251 patients that could be linked to Discharge Abstract Database data, 7062 were adult patients with moderate or severe traumatic brain injury or intubated on arrival (see figure 1—flow diagram and table 1). Most included patients were admitted in trauma centres situated in Ontario (n=3327; 47.1%) or Alberta (n=1809; 25.6%); others were admitted in British Columbia (n=1313; 18.6%), Nova Scotia (n=415; 5.9%), New Brunswick (n=152; 2.2%) and Saskatchewan (n=46; 0.7%). No data from Prince Edward Island were available. Trauma admissions from Manitoba and Newfoundland could not be linked to the Discharge Abstract Database data due to invalid or missing health card numbers. Hospital admissions from the province of Québec do not contribute to the Discharge Abstract Database and therefore could not be linked. Most patients were admitted to a level I or II trauma centre (95.8%). Mean age was 48.7 years old (SD 21.7) and most patients were men (73.2%). Median ISS was 26 (IQR 25–36). More than half of included patients were intubated on arrival (55.2%). Most patients (78.0%) did not have any identified comorbidities recorded in the registry. The majority of patients sustained three or more injuries (n=5047, 71.5%). Table 1 Characteristics of participants Variables N (%) Transfused with RBC n (%) Total 7062 (100.0) 1991 (28.2) Age  18–55 4336 (61.4) 1346 (67.6)  56–65 823 (11.7) 252 (12.7)  66–75 700 (9.9) 174 (8.7)  ≥75 1203 (17.0) 219 (11.0)  Male gender 5168 (73.2) 1368 (68.7) GCS on admission  Moderate (9–12) 1509 (21.4) 278 (14.0)  Severe (3–8) 1788 (25.3) 505 (25.4)  Missing 3765 (53.3) 1208 (60.7) Intubated on arrival 3899 (55.2) 1249 (62.7) Extracerebral trauma of serious or greater severity 3339 (47.3) 1417 (71.2) Serious injury (AIS≥3) by body region  Head 6682 (94.6) 1839 (92.4)  Face and neck 598 (8.5) 237 (11.9)  Thorax and abdomen 2512 (35.6) 1140 (57.3)  Spinal 655 (9.3) 286 (14.4)  Upper extremities 479 (6.8) 257 (12.9)  Lower extremities 1001 (14.2) 635 (31.9)  Other 43 (0.6) 20 (1.0) Invasive interventions* 4201 (59.5) 1695 (85.1) Number of comorbidities†  0 5505 (78.0) 1543 (77.5)  1 1214 (17.2) 334 (16.8)  ≥2 343 (4.9) 114 (5.7) Other specific comorbidities‡  Ischaemic heart disease 172 (2.4) 45 (2.3)  Cerebrovascular disease 114 (1.6) 39 (2.0)  Anaemia 314 (4.5) 218 (10.9)  Coagulopathy 89 (1.3) 53 (2.7)  Sepsis 79 (1.1) 39 (2.0)  Hypovolemic shock 554 (7.8) 276 (13.9)  Bleeding/haemorrhage 1701 (24.1) 611 (30.7) Direct admission 3586 (50.8) 1030 (51.7) Mechanism of injury  MVC 3162 (44.8) 1237 (62.1)  Fall 2828 (40.1) 543 (27.3)  Other 1072 (15.2) 211 (10.6) *See online supplementary appendix C for specific codes. †Based on previous publication,33 see online supplementary appendix B. ‡For specific comorbidities, the total of percentages may exceed 100% since they are not mutually exclusive. AIS, Abbreviated Injury Severity Score; GCS, Glasgow Coma Scale; MVC, motor vehicle collision; RBC, red blood cell. Figure 1 Flow diagram. *Admissions to more than one hospital for the same injury (transfers) were combined to consider only one trauma admission. Red blood cell transfusion frequency Overall, 1991 patients (28.2%; 95% CI 27.2% to 29.3%) received at least one red blood cell transfusion during their hospital stay. The proportion of patients receiving a red blood cell transfusion varied considerably across centres, 0–43%. Level I and II trauma centre admissions were associated with the highest frequencies (28.9% and 31.0%, respectively), while level III–IV had 3.4% of patients transfused. Red blood cell transfusion also varied between provinces from 16.6% to 34.9%. Potential determinants Female sex, age, anaemia, coagulopathy, sepsis, bleeding, hypovolemic shock, presence of other comorbidities, GCS <9, serious trauma to the face, thorax–abdomen, spine, upper and lower limbs were significantly associated with greater frequency of red blood cell transfusion according to our robust Poisson regression model (table 2). Table 2 Adjusted* risk ratios for the association between determinants of red blood cell transfusion Effect Risk ratio (95% CI) Sex 1.16 (1.10 to 1.23) Age  18–55 1.00  56–65 1.14 (1.04 to 1.24)  66–75 0.96 (0.84 to 1.09)  ≥75 0.81 (0.72 to 0.90) Comorbidities  0 1.00  1 1.11 (1.02 to 1.21)  ≥2 1.66 (1.40 to 1.97) Other specific comorbidities considered  Ischaemic heart disease 1.19 (0.93 to 1.53)  Cerebrovascular disease 1.17 (0.95 to 1.43)  Anaemia 2.10 (1.81 to 2.43)  Coagulopathy 1.37 (1.08 to 1.74)  Sepsis 1.57 (1.23 to 2.01)  Hypovolemic shock 1.33 (1.19 to 1.47)  Bleeding/haemorrhage 1.12 (1.03 to 1.22) Brain trauma severity  Moderate TBI (GCS 9–12) 0.78 (0.70 to 0.87)  Serious head injury (AIS≥3) 0.93 (0.81 to 1.07) Serious extracerebral injury by anatomic region  Face 1.36 (1.22 to 1.51)  Thorax and abdomen 1.76 (1.55 to 2.00)  Spine 1.24 (1.08 to 1.43)  Upper extremities 1.25 (1.12 to 1.40)  Lower extremities 1.88 (1.75 to 2.02)  Others 1.48 (0.96 to 2.29) *Adjusted for all covariates in the table. AIS, Abbreviated Injury Severity Score; GCS, Glasgow Coma Scale; TBI, traumatic brain injury. Measures of association between determinants and transfusions remained similar when an indicator of interventions with known bleeding risk was added in the robust Poisson model, with the exception of a reduction of the effect size of serious trauma of the lower extremities. In such model, the indicator of interventions was associated with a threefold risk of receiving a red blood cell transfusion (RR 3.00; 95% CI 2.64 to 3.41). The magnitude of the variation in the mean frequency of red blood cell transfusion among centres was estimated to 10% (95% CI 1% to 67%). The intraclass correlation coefficient, calculated on the covariance parameter for centre clusters (0.96±0.30) indicated that 61% of the total variation in red blood cell transfusion was accounted for by variation in practice across centres after adjustment for patient-level risk factors. In the relative weights analyses, serious extracerebral injuries were the most important determinant of transfusion accounting for 61% of red blood cell transfusion logit variation explained by the model (figure 2). Anaemia on arrival was the most important baseline factor related to the risk of being transfused, with a relative weight of 14% (figure 2). Interventions that confer potential bleeding risk accounted for 37% of the variation when considered in the model. Figure 2 Potential determinants of red blood cell transfusion (the percentage represents the relative weight of each determinant in terms of the proportion of variation explained by a variable in the model). Clinical outcomes Mortality Patients who were transfused with at least one unit of red blood cells had a significantly greater risk of death (adjusted RR 1.23; 95% CI 1.13 to 1.33) compared with patients not transfused. The presence of comorbidities and traumatic brain injury severity were not significant modifiers of the association. Associations between red blood cell transfusion and mortality were stronger in patients ≤55 years than their older counterparts (table 3). Table 3 Adjusted* risk ratios and HRs of outcomes according to the transfusion status N (%) Mortality, adjusted risk ratios (95%CI) Complications, adjusted risk ratios (95%CI) ICU stay, adjusted HRs† (95%CI) Hospital stay, adjusted HRs† (95%CI) Discharged home, adjusted risk ratios (95% CI) Overall 7062 1.23 (1.13 to 1.33) 1.38 (1.32 to 1.44) 1.63 (1.52 to 1.75) 1.56 (1.45 to 1.67) 0.53 (0.49 to 0.57) Strata Age  18–55 4336 (61.40) 1.54 (1.37 to 1.73) 1.55 (1.44 to 1.66) 1.06 (0.85 to 1.33) 1.67 (1.53 to 1.81) 0.53 (0.48 to 0.58)  56–65 823 (11.65) 1.10 (0.94 to 1.28) 1.26 (1.15 to 1.38) 1.79 (1.65 to 1.94) 1.24 (1.02 to 1.50) 0.46 (0.30 to 0.71)  66–75 700 (9.91) 1.00 (0.84 to 1.19) 1.18 (1.05 to 1.32) 1.38 (1.14 to 1.68) 1.45 (1.15 to 1.83) 0.55 (0.30 to 1.02)  ≥75 1203 (17.03) 0.94 (0.78 to 1.13) 1.13 (1.01 to 1.26) 1.31 (1.04 to 1.66) 1.26 (1.01 to 1.58) 0.60 (0.35 to 1.03) GCS  Moderate 1788 (25.32) 1.21 (0.98 to 1.49) 1.55 (1.37 to 1.75) 1.80 (1.60 to 2.02) 1.61 (1.43 to 1.80) 0.53 (0.44 to 0.63)  Severe 1509 (21.37) 1.23 (1.12 to 1.36) 1.33 (1.27 to 1.39) 1.53 (1.40 to 1.67) 1.53 (1.40 to 1.67) 0.52 (0.47 to 0.59) Comorbidities  0 5505 (77.95) 1.27 (1.16 to 1.39) 1.42 (1.36 to 1.49) 1.67 (1.55 to 1.81) 1.60 (1.48 to 1.74) 0.51 (0.47 to 0.56)  1 1214 (17.19) 1.18 (0.98 to 1.42) 1.33 (1.19 to 1.48) 1.51 (1.28 to 1.79) 1.41 (1.20 to 1.66) 0.51 (0.37 to 0.69)  ≥2 343 (4.86) 0.97 (0.74 to 1.26) 1.07 (0.92 to 1.24) 1.29 (0.95 to 1.74) 1.35 (0.99 to 1.82) 0.87 (0.50 to 1.51) Specific comorbidities  Anemic 314 (4.45) 1.00 (0.72 to 1.40) 1.00 (0.88 to 1.13) 1.17 (0.87 to 1.57) 0.95 (0.71 to 1.28) 0.65 (0.48 to 0.88)  Non-anemic 6748 (95.55) 1.23 (1.14 to 1.33) 1.38 (1.33 to 1.44) 1.62 (1.51 to 1.75) 1.55 (1.43 to 1.67) 0.52 (0.48 to 0.57)  Septic 79 (1.12) 0.36 (0.09 to 1.46) 1.11 (0.87 to 1.42) 0.90 (0.54 to 1.50) 2.11 (1.27 to 3.51) 1.52 (0.72 to 3.19)  Non-septic 6983 (98.88) 1.23 (1.14 to 1.34) 1.37 (1.31 to 1.42) 1.61 (1.49 to 1.73) 1.50 (1.39 to 1.61) 0.52 (0.48 to 0.57)  Bleeding 1701 (24.09) 1.15 (0.93 to 1.42) 1.18 (1.08 to 1.30) 1.36 (1.20 to 1.55) 1.47 (1.29 to 1.67) 0.67 (0.55 to 0.82)  Non-bleeding 5361 (75.91) 1.25 (1.15 to 1.36) 1.44 (1.39 to 1.50) 1.68 (1.55 to 1.83) 1.52 (1.40 to 1.65) 0.50 (0.46 to 0.54) *Adjusted for sex, age, comorbidities, GCS, head AIS, square of max extracerebral AIS. †With death considered as censored at the longest length of stay; value >1 indicates that patients were at lower risk of being discharge alive (longer stays) and <1 at higher risk (shorter stays). AIS, Abbreviated Injury Severity Score; GCS, Glasgow Coma Scale. Complications A high proportion of patients developed complications (1361 (68%) transfused; 2386 (47%) non-transfused; adjusted RR 1.38 95% CI 1.32 to 1.44). The strength of the association was greater in younger patients as well as those with moderate traumatic brain injury severity or having no comorbidity (table 3). Transfusions in non-anaemic or non-bleeding patients on admission were associated with higher risks of complications. The association was not significant in patients with anaemia, and was weaker in patients who were bleeding on admission. Discharge destination Patients who were transfused were less often discharged home (RR 0.53; 95% CI 0.49 to 0.57). This association was significant in patients who did not have sepsis on arrival, but was not in patients with sepsis. Patients who were bleeding on arrival showed a weaker effect of transfusion on the risk of being discharged home (table 3). Length of stay In survivors, ICU length of stay was longer (crude mean difference: 18.63 days (95% CI 16.64 to 20.62)) in patients who were transfused red blood cells than those who were not. After applying a correction to the length of stay for patients who died, we obtained ICU stay HRs of 1.63 (95% CI 1.52 to 1.75) between patients who were transfused and those who were not, meaning that patients who were transfused had longer lengths of stay in ICU (table 3). The HR for hospital length of stay was 1.56 (95% CI 1.45 to 1.67), indicating a significantly longer length of stay in patients who were transfused compared with those who were not. ICU lengths of stay were longer following transfusion in strata of non-anaemic, non-septic or non-bleeding patients. No significant differences in length of stay between transfused and non-transfused were observed in patients with anaemia or sepsis. A weaker association between transfusion and ICU length of stay was seen in patients who were bleeding on admission. Hospital lengths of stay were longer following transfusion in non-anaemic patients, but not in anaemic patients (table 3). Patients who were transfused were ventilated for a crude mean of 4.87 more days (95% CI 4.28 to 5.46) than those who were not. The difference remained significant in adjusted models, with adjusted geometric mean ratios of 1.87 (95% CI 1.75 to 1.99). Sensitivity analyses All unfavourable outcomes were significantly more frequent in patients who were transfused than those who were not. The same observations were made even when adding an indicator of interventions with bleeding risk for adjustment. Excluding patients who died within 24, 48 or 72 hours, in other words limiting the analysis to those who survived at 24, 48 or 72 hours following admission, yielded stronger associations between transfusions and outcomes. Results were similar in analyses of complete observations and multiple imputations. Discussion Across trauma hospitals in Canada, close to one-third of patients with a moderate or severe traumatic brain injury received a red blood cell transfusion during their hospital stay. Large variation in the frequency of transfusion was present across centres mainly explained by variation in practice across centres. Red blood cell transfusion was more common among patients with extracerebral trauma. After adjusting for confounding factors, patients who received at least one red blood cell transfusion experienced worse clinical outcomes, such as higher mortality, longer length of stay, more complications and discharge elsewhere than home. The frequency of red blood cell transfusion in our cohort (28%; 95% CI 27% to 29%) was lower than the estimation from our recent systematic review (36%; 95% CI 28% to 44%).20 In both studies, frequencies varied greatly between healthcare settings. Since centre variation remained high when evaluating determinants of red blood cell transfusion, regional practice variations may account for a large part of variations observed in absence of clear consensus. Variability in practices across centres has been previously observed in a large cohort of patients with traumatic brain injury looking at withdrawal of life-sustaining therapy,6 another domain in which data is scarce. As we observed previously in our systematic review,19 patients with anaemia on admission were at a higher risk of receiving red blood cell transfusion. This is likely due to the fact that haemoglobin usually drops over an ICU stay and red blood cell transfusions are administered with the intention of increasing haemoglobin level and thus oxygen delivery.24 40–43 Active bleeding on arrival and hypovolemic shock, which can also result from major rapid blood loss, were associated with red blood cell transfusion. Although such conditions are often treated with volume replacement using crystalloids,44 45 early red blood cell transfusion is frequently used as it is included in developed protocols of care for massive transfusion following evidence of benefit from early transfusion of blood products in these situations.44 Considering the high frequency of major bleeding from serious thoraco-abdominal wounds and lower limb fractures,46 47 it is unsurprising that those type of injuries were also associated with higher incidence proportions of red blood cell transfusion. Sepsis was also associated with transfusions in our cohort. It is another factor associated with loss of erythrocytes and also with the limited ability to extract oxygen from blood. In this context, higher haemoglobin levels have been advocated in this population.48 However, no strong evidence supports this approach.49 50 Transfusions were also more frequent in patients with coagulopathy. Even if red blood cell preparations do not include coagulation factors and cannot treat coagulopathy, patients presenting with the latter condition tend to experience prolonged bleeding. Early red blood cell transfusion following trauma might beneficially reduce the risk of coagulopathy.51 Furthermore, trauma-induced coagulopathy increases the risk of major bleeding and use of 1:1:1 platelet:fresh frozen plasma:red blood cell has been advocated by some experts in the prevention or treatment of trauma-induced coagulopathy. However, few evidences support the use of such strategy.52 53 When added to the model, surgical interventions were associated with transfusions. Trauma surgeries can lead to important blood loss, or be initiated in order to control haemorrhage. We observed worse outcomes in patients who received at least one red blood cell transfusion, which is similar to the results of previous studies.54–57 When we stratified analyses by anaemic status, the effect of transfusion on complications, and ICU and hospital length of stay were different. Specifically, transfused patients in the non-anaemic strata had significantly higher risk of worse outcome and no such relationship was observed in patients who were anaemic on admission. This is similar to the observations of previous studies,54 57 and illustrates again the importance of considering haemoglobin levels when evaluating the association between red blood cell and patient-oriented outcomes in the context of traumatic brain injury. A recent trial26 failed to observe significant differences over mortality and neurological outcome but was underpowered and not designed to evaluate equivalency of such outcomes. Further analyses of the progressive decline in haemoglobin during ICU stay, pretransfusion haemoglobin level and its impact on optimal transfusion strategies are needed. Similarly, sepsis was also a modifying factor of the effect of red blood cell transfusion on clinical outcomes. According to our observations, red blood cell transfusion showed neither significant benefit nor disadvantage over clinical outcomes when patients were in sepsis (except for hospital length of stay), but tended to be associated with worse outcomes in patients who did not have sepsis diagnosed on arrival. According to previous studies comparing liberal to restrictive transfusion strategies in populations of septic patients, similar outcomes were observed for both strategies49 leading clinical guidelines to support restrictive use of red blood cell in the treatment of severe sepsis and septic shock.58 As mentioned earlier, patients with sepsis often show decreasing levels of haemoglobin and a reduced ability to extract oxygen from blood. The apparent equivalency between groups of patients who did or did not receive red blood cell in this stratum could be related to important haemoglobin depletion, requiring further investigation of haemoglobin levels. Strengths and limitations Level I and II trauma centres across the country are well represented in the registry, and patients with moderate or severe traumatic brain injury are rarely treated outside highly specialised centres.59–61 Therefore, undercoverage from patients being treated in trauma centres that are not included in the registry or in non-trauma centres should not have had a major effect on our frequency estimate. However, we could not link 21.9% of the National Trauma Registry to the Discharge Abstract Database. Considering that the province of Québec does not contribute to the Discharge Abstract Database of the National Trauma Registry, and that its population represents 23% of Canada's population, we can estimate that our study population represents trauma patients in Canada, with the exception of those from the province of Québec. Considering that training of physicians and certifications of trauma centres are managed at a national Canadian level, healthcare delivery is thus likely to be consistent across provinces. More so, referral and triage being independent in each Canadian province, the distribution of trauma severity would thus not be affected by the absence of data from one province. Our study was inclusive and we had a large sample size, and therefore high power. Our statistical plan was strong, with advanced biostatistics methods and conceptual rigour. We used multilevel models, with random effects. This method allowed to account for potential differences in local practices (between centres) in our evaluation of red blood cell transfusions and its potential determinants, as well as associated outcomes. We explored a large spectrum of a priori conceptual models of the potential relationship between transfusions and clinical variables, which we explored with multiple sensitivity and subgroup analyses. Although residual confounding remains a concern, we optimised control for potential bias through multivariate analyses. Secondary data use has the risk of measurement bias/misclassification and incompleteness of data. However, data quality is monitored by The Canadian Institute for Health Information62 assuring a high reliability of data collection and encoding63 and limiting the potential for such errors. It is possible that incomplete assessment of comorbidities, extracerebral injuries and interventions contributed to the underestimation of their frequencies and effect sizes. For example, comorbidity capture depends on what was noted in the hospital chart and may not represent an exhaustive account of all comorbidities present in a given patient at the time of admission. Even with the inclusion of comorbidities as potential confounders in our models of patient-oriented outcomes, residual confounding is possibly present. Similarly, complications are reported differently across trauma centres, screening and coding practices may vary greatly, causing incomplete outcome measurement. Any under-reporting should not be systematically linked to transfusion practices, but may bias our estimates. Furthermore, we had limited information on traumatic brain injury severity, as GCS scores were often missing and the registry do not contain data on clinical findings such as pupillary reflex, which is another indicator of severity and prognostic variable in traumatic brain injury. Although we endeavoured to overcome this limitation with multiple imputation and included information on severity of head trauma (head AIS) in our models, adjustment might still be imperfect. We did not have data on the quantity of red blood cell units transfused or number of transfusion events. As such, a dose–response relationship could not be evaluated. We also did not have information on haemoglobin levels during hospitalisation, which would have been a more precise measure of the status of the patient than simply the presence of anaemia on admission. We could not evaluate the effect of transfusion strategies (transfusion at different haemoglobin levels). Owing to the use of secondary data, we could not properly identify and control timing and reasons for identified surgical interventions, so both causal mechanisms are plausible (ie, interventions could increase the risk of transfusions and transfusions might increase the risk of interventions as well). For this reason, we kept adjustment to baseline variables. However, when we added the indicator of interventions, our results remained similar. Furthermore, the registry does not contain information on timing of complications either and thus, we cannot exclude that complications might have contributed to decision to transfuse rather than resulted from it. We were also limited on variables to control for by data available in the registry (eg, data on decisions over the goals of care are not reported), resulting in potential residual confounding. Overall, due to our limited ability to control for factors related to the need or decision to transfuse and clinical outcomes (such as severity of trauma, comorbidities, complications directly related with the trauma, decisions on intensity of care, other interventions), we cannot exclude confounding by indication. Conclusion In our multicentre cohort study, we observed that red blood cell transfusions were frequent in patients with traumatic brain injury and varied considerably between centres. Trauma severity and invasive interventions are important determinants of transfusion decisions. More so, red blood cell transfusion was associated with increased mortality, prolonged ICU and hospital length of stay and adverse events. However, these associations are potentially confounded considering that transfusions are correlated to worse clinical baseline prognosis, which we could not fully account for in the analyses. Prospective trials are required to better understand the impact of different transfusion strategies on clinical outcomes in which to inform future trials of optimal transfusion strategy for traumatic brain injury. The authors want to thank Caroline Léger PhD, Michele Shemilt MSc and Marie-Joëlle Cossi PhD for their administrative help with this study and manuscript. Contributors: AB, LMo, FL, MC, RZ, DAF and AFT were involved in the conception and design. AB, LMo and AFT were involved in the acquisition. AB, LMo, FL, MC, SE, RZ, LMc, DG DAF and AFT were involved in the interpretation of data. AB, LMo and AFT drafted the manuscript. AB, LMo, FL, MC, SE, RZ, LMc, DG, DAF and AFT were involved in revising the manuscript and approved the version published. Funding: AB is recipient of a Frederick Banting and Charles Best Canada Graduate Scholarships Doctoral Award from the Canadian Institutes of Health Research (CIHR). LM, AFT and RZ are/were recipients of New Investigator Awards from the CIHR. AFT and FL are supported by the Traumatology Research Consortium of the Fonds de Recherche du Québec—Santé (FRQS). FL is a recipient of a salary support Award from the FRQS. AFT is the Canada Research Chair in Critical Care Neurology and Trauma. Competing interests: None declared. Ethics approval: The study received ethics approval by CHU de Québec—Université Laval. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Maas AI , Stocchetti N , Bullock R Moderate and severe traumatic brain injury in adults . Lancet Neurol 2008 ;7 :728 –41 . 10.1016/S1474-4422(08)70164-9 18635021 2 Ruttan L , Martin K , Liu A Long-term cognitive outcome in moderate to severe traumatic brain injury: a meta-analysis examining timed and untimed tests at 1 and 4.5 or more years after injury . Arch Phys Med Rehabil 2008 ;89 (12 Suppl):S69 –76 . 10.1016/j.apmr.2008.07.007 19081444 3 Andelic N , Sigurdardottir S , Schanke AK Disability, physical health and mental health 1 year after traumatic brain injury . Disabil Rehabil 2010 ;32 :1122 –31 . 10.3109/09638280903410722 20113311 4 McIntyre A , Mehta S , Janzen S A meta-analysis of functional outcome among older adults with traumatic brain injury . NeuroRehabilitation 2013 ;32 :409 –14 . 10.3233/NRE-130862 23535806 5 Centers for Disease Control and Prevention . Report to congress on traumatic brain injury in the United States: epidemiology and rehabilitation . Atlanta, GA : National Center for Injury Prevention and Control; Division of Unintentional Injury Prevention , 2014 . 6 Turgeon AF , Lauzier F , Simard JF Mortality associated with withdrawal of life-sustaining therapy for patients with severe traumatic brain injury: a Canadian multicentre cohort study . CMAJ 2011 ;183 :1581 –8 . 10.1503/cmaj.101786 21876014 7 Fu TS , Jing R , McFaull SR Health & economic burden of traumatic brain injury in the Emergency Department . Can J Neurol Sci 2016 ;43 :238 –47 . 10.1017/cjn.2015.320 26842131 8 Hyder AA , Wunderlich CA , Puvanachandra P The impact of traumatic brain injuries: a global perspective . NeuroRehabilitation 2007 ;22 :341 –53 .18162698 9 Kramer AH , Zygun DA Anemia and red blood cell transfusion in neurocritical care . Crit Care 2009 ;13 :R89 10.1186/cc7916 19519893 10 Bellapart J , Boots R , Fraser J Physiopathology of anemia and transfusion thresholds in isolated head injury . J Trauma Acute Care Surg 2012 ;73 :997 –1005 . 10.1097/TA.0b013e318265cede 22922968 11 Sena MJ , Rivers RM , Muizelaar JP Transfusion practices for acute traumatic brain injury: a survey of physicians at US trauma centers . Intensive Care Med 2009 ;35 :480 –8 . 10.1007/s00134-008-1289-z 18854976 12 Raghavan M , Marik PE Anemia, allogenic blood transfusion, and immunomodulation in the critically ill . Chest 2005 ;127 :295 –307 . 10.1378/chest.127.1.295 15653997 13 Carson JL , Carless PA , Hebert PC Transfusion thresholds and other strategies for guiding allogeneic red blood cell transfusion . Cochrane Database Syst Rev 2012 ;(4) :CD002042 10.1002/14651858.CD002042.pub3 22513904 14 Carson JL , Grossman BJ , Kleinman S Red blood cell transfusion: a clinical practice guideline from the AABB* . Ann Intern Med 2012 ;157 :49 –58 . 10.7326/0003-4819-157-1-201206190-00429 22751760 15 Utter GH , Shahlaie K , Zwienenberg-Lee M Anemia in the setting of traumatic brain injury: the arguments for and against liberal transfusion . J Neurotrauma 2011 ;28 :155 –65 . 10.1089/neu.2010.1451 20954887 16 Kramer AH , Le Roux P Red blood cell transfusion and transfusion alternatives in traumatic brain injury . Curr Treat Options Neurol 2012 ;14 :150 –63 . 10.1007/s11940-012-0167-8 22314930 17 Roberts DJ , Zygun DA Anemia, red blood cell transfusion, and outcomes after severe traumatic brain injury . Crit Care 2012 ;16 :154 10.1186/cc11489 22979948 18 Mirski MA , Frank SM , Kor DJ Restrictive and liberal red cell transfusion strategies in adult patients: reconciling clinical data with best practice . Crit Care 2015 ;19 :202 10.1186/s13054-015-0912-y 25939346 19 Boutin A , Chasse M , Shemilt M Red blood cell transfusion in patients with traumatic brain injury: a systematic review and meta-analysis . Transfus Med Rev 2016 ;30 :15 –24 . 10.1016/j.tmrv.2015.08.004 26409622 20 Desjardins P , Turgeon AF , Tremblay MH Hemoglobin levels and transfusions in neurocritically ill patients: a systematic review of comparative studies . Crit Care 2012 ;16 :R54 10.1186/cc11293 22471943 21 Veenith T , Sharples L , Gerrard C Survival and length of stay following blood transfusion in octogenarians following cardiac surgery . Anaesthesia 2010 ;65 :331 –6 . 10.1111/j.1365-2044.2009.06225.x 20148816 22 Chant C , Wilson G , Friedrich JO Anemia, transfusion, and phlebotomy practices in critically ill patients with prolonged ICU length of stay: a cohort study . Crit Care 2006 ;10 :R140 10.1186/cc5054 17002795 23 Andersen J , Sharkey W , Schwartz ML Injury severity score, head injury, and patient wait days: contributions to extended trauma patient length of stay . J Trauma 1992 ;33 :219 –20 .1507284 24 Marik PE , Corwin HL Efficacy of red blood cell transfusion in the critically ill: a systematic review of the literature . Crit Care Med 2008 ;36 :2667 –74 . 10.1097/CCM.0b013e3181844677 18679112 25 McIntyre LA , Fergusson DA , Hutchison JS Effect of a liberal versus restrictive transfusion strategy on mortality in patients with moderate to severe head injury . Neurocrit Care 2006 ;5 :4 –9 .16960287 26 Robertson CS , Hannay HJ , Yamal JM Effect of erythropoietin and transfusion threshold on neurological recovery after traumatic brain injury: a randomized clinical trial . JAMA 2014 ;312 :36 –47 . 10.1001/jama.2014.6490 25058216 27 von Elm E , Altman DG , Egger M The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies . PLoS Med. 2007 ;4 :e296 10.1371/journal.pmed.0040296 17941714 28 Gennarelli TA , Wodzin E AIS 2005: a contemporary injury scale . Injury 2006 ;37 :1083 –91 . 10.1016/j.injury.2006.07.009 17092503 29 Charlson ME , Pompei P , Ales KL A new method of classifying prognostic comorbidity in longitudinal studies: development and validation . J Chronic Dis 1987 ;40 :373 –83 .3558716 30 Moore L , Lavoie A , Le Sage N Using information on preexisting conditions to predict mortality from traumatic injury . Ann Emerg Med 2008 ;52 :356 –64.e2 . 10.1016/j.annemergmed.2007.09.007 18282639 31 Moore L , Lauzier F , Stelfox HT Complications to evaluate adult trauma care: an expert consensus study . J Trauma Acute Care Surg 2014 ;77 :322 –9 ; discussion 9–30 10.1097/TA.0000000000000366 25058261 32 Moore L , Lauzier F , Stelfox HT Validation of complications selected by consensus to evaluate the acute phase of adult trauma care: a multicenter cohort study . Ann Surg 2015 ;262 :1123 –9 . 10.1097/SLA.0000000000000963 25243558 33 Zou G A modified poisson regression approach to prospective studies with binary data . Am J Epidemiol 2004 ;159 :702 –6 .15033648 34 Tonidandel S , LeBreton JM Determining the relative importance of predictors in logistic regression: an extension of relative weight analysis . Organ Res Methods 2010 ;13 :767 –81 . 35 Rosenbaum PR The place of death in the quality of life . Stat Sci 2006 ;21 :313 –6 . 36 Halpern SD , ed. ICU length of stay: a new approach to a challenging outcome . Toronto : Canadian Critical Care Forum , 2012 . 37 Brock GN , Barnes C , Ramirez JA How to handle mortality when investigating length of hospital stay and time to clinical stability . BMC Med Res Methodol 2011 ;11 :144 10.1186/1471-2288-11-144 22029846 38 Moore L , Hanley JA , Turgeon AF A multiple imputation model for imputing missing physiologic data in the national trauma data bank . J Am Coll Surg 2009 ;209 :572 –9 . 10.1016/j.jamcollsurg.2009.07.004 19854396 39 Moore L , Hanley JA , Lavoie A Evaluating the validity of multiple imputation for missing physiological data in the national trauma data bank . J Emerg Trauma Shock 2009 ;2 :73 –9 . 10.4103/0974-2700.44774 19561964 40 Corwin HL , Gettinger A , Pearl RG The CRIT Study: anemia and blood transfusion in the critically ill--current clinical practice in the United States . Crit Care Med 2004 ;32 :39 –52 . 10.1097/01.CCM.0000104112.34142.79 14707558 41 Rawn J The silent risks of blood transfusion . Curr Opin Anaesthesiol 2008 ;21 :664 –8 . 10.1097/ACO.0b013e32830f1fd1 18784496 42 Chargé S, ed. Clinical guide to transfusion medicine. 5th edn. Canada: Canadian Blood Services, 2013 . 43 Spinelli E , Bartlett RH Anemia and transfusion in critical care: physiology and management . J Intensive Care Med 2016 ;31 :295 –306 . 10.1177/0885066615571901 25693602 44 Juffermans NP , Walsh TS Transfusion in the intensive care unit. Switzerland: Springer International Publishing, 2015 . 45 McCullough JJ Transfusion medicine . Chichester, UK : Wiley-Blackwell , 2012 . 46 Evans JA , van Wessem KJ , McDougall D Epidemiology of traumatic deaths: comprehensive population-based assessment . World J Surg 2010 ;34 :158 –63 . 10.1007/s00268-009-0266-1 19882185 47 Yucel N , Lefering R , Maegele M Trauma Associated Severe Hemorrhage (TASH)-Score: probability of mass transfusion as surrogate for life threatening hemorrhage after multiple trauma . J Trauma 2006 ;60 :1228 –36 ; discussion 36–7 10.1097/01.ta.0000220386.84012.bf 16766965 48 Juffermans NP , Walsh TS Transfusion in the intensive care unit . Switzerland : Springer International Publishing , 2015 10.1007/978-3-319-08735-1 49 Holst LB , Haase N , Wetterslev J Lower versus higher hemoglobin threshold for transfusion in septic shock . N Engl J Med 2014 ;371 :1381 –91 . 10.1056/NEJMoa1406617 25270275 50 Marik PE Early management of severe sepsis: concepts and controversies . Chest 2014 ;145 :1407 –18 . 10.1378/chest.13-2104 24889440 51 Brown JB , Cohen MJ , Minei JP Pretrauma center red blood cell transfusion is associated with reduced mortality and coagulopathy in severely injured patients with blunt trauma . Ann Surg 2015 ;261 :997 –1005 . 10.1097/SLA.0000000000000674 24670858 52 Hallet J , Lauzier F , Mailloux O The use of higher platelet: RBC transfusion ratio in the acute phase of trauma resuscitation: a systematic review . Crit Care Med 2013 ;41 :2800 –11 . 10.1097/CCM.0b013e31829a6ecb 23982024 53 Holcomb JB , Tilley BC , Baraniuk S Transfusion of plasma, platelets, and red blood cells in a 1:1:1 vs a 1:1:2 ratio and mortality in patients with severe trauma: the PROPPR randomized clinical trial . JAMA 2015 ;313 :471 –82 . 10.1001/jama.2015.12 25647203 54 Elterman J , Brasel K , Brown S Transfusion of red blood cells in patients with a prehospital Glasgow Coma Scale score of 8 or less and no evidence of shock is associated with worse outcomes . J Trauma Acute Care Surg 2013 ;75 :8 –14 ; discussion .10.1097/TA.0b013e318298492e 23778432 55 Yang CJ , Hsiao KY , Su IC The association between anemia and the mortality of severe traumatic brain injury in emergency department . J Trauma 2011 ;71 :E132 –E5 . 10.1097/TA.0b013e31820ea36b 21427611 56 Warner MA , O'Keeffe T , Bhavsar P Transfusions and long-term functional outcomes in traumatic brain injury . J Neurosurg 2010 ;113 :539 –46 . 10.3171/2009.12.JNS091337 20113158 57 Salim A , Hadjizacharia P , DuBose J Role of anemia in traumatic brain injury . J Am Coll Surg 2008 ;207 :398 –406 . 10.1016/j.jamcollsurg.2008.03.013 18722946 58 Dellinger RP , Levy MM , Rhodes A Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2012 . Crit Care Med 2013 ;41 :580 –637 . 10.1097/CCM.0b013e31827e83af 23353941 59 Institut National d'Excellence en Santé et en Services sociaux (INESSS) . Avis sur les critères de triage préhospitalier en traumatologie . Rapport rédigé par André Lavoie en collaboration avec Gilles Bourgeois et Jean Lapointe ETMIS , 2013 :1 –46 . 60 Sasser SM , Hunt RC , Faul M , Centers for Disease Control and Prevention . Guidelines for field triage of injured patients—recommendations of The National Expert Panel on Field Triage, 2011 . MMWR Recomm Rep 2012 ;61 (RR-1) :1 –20 . 61 Kuimi BL , Moore L , Cisse B Access to a Canadian provincial integrated trauma system: a population-based cohort study . Injury 2015 ;46 :595 –601 . 10.1016/j.injury.2015.01.006 25640590 62 Canadian Insitute for Health Information . National trauma registry comprehensive data set—data quality summary for external users . Ottawa, Ontario : Canadian Institute for Health Information—Institut canadian de l'information sur la santé (CIHI-ICIS) , 2014 . 63 Canadian Insitute for Health Information . CIHI data quality study of the 2009–2010: discharge abstract database . CIHI-ICIS , 2012 (cited 2012). https://secure.cihi.ca/free_products/Reabstraction_june19revised_09_10_en.pdf

PMC005xxxxxx/PMC5372063.txt

==== Front RMD OpenRMD OpenrmdopenrmdopenRMD Open2056-5933BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR 28405470rmdopen-2016-00035310.1136/rmdopen-2016-000353Vasculitis1506Original articleIs vascular endothelial growth factor a useful biomarker in giant cell arteritis? Goodfellow Nicola 1Morlet Julien 1Singh Surjeet 2Sabokbar Afsie 1Hutchings Andrew 3Sharma Vanshika 1Vaskova Jana 1Masters Shauna 1Zarei Allahdad 4Luqmani Raashid 1the TABUL InvestigatorsLee Ellen Loban Amanda Ellis Christopher Gillett Mike Bradburn Mike Schmidt Wolfgang Dasgupta Bhaskar Diamantopoulos Andreas McNally Eugene Piper Jennifer Forrester-Barker Wulf Hamilton Willie McDonald Brendan Pease Colin Salmon John Wailoo Allan Wolfe Konrad Fathers Keri Marcus-Wan Leo Farrar Nicola Manhas Varun Scott Connor Sullivan Nicky Brown Denise Bicknell Gareth Kliskey Karolina Gray David Mahmood Samiya Marie Morgan Ann Ahmed Ifzal Ehrenstein Michael Davies Bleddyn Raza Karim Mant David Williamson Lyn Gilbert Kate Travis Simon Sterne Jonathan Pease Colin Carruthers David Klocke Rainer Saravanan Vadivelu Makkuni Damodar Geraldes Ruth Diamantopoulos Andreas Neumann Thomas Pendleton Adrian Ahmed Khalid Hull Richard Chakravarty Kuntal Lanyon Peter Chan Antoni Raj Nicholas Borg Frances Molloy Eamonn Magliano Malgorzata Wright David 1 Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, Oxford, UK 2 Nuffield Department of Surgical Sciences, University of Oxford, Oxford, UK 3 London School of Hygiene and Tropical Medicine, London, UK 4 Botnar Research Centre, University of Oxford, Oxford, UKCorrespondence to Professor Raashid Luqmani; raashid.luqmani@ndorms.ox.ac.uk2017 29 3 2017 3 1 e00035324 8 2016 21 12 2016 3 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives To assess the performance of circulating vascular endothelial growth factor (VEGF) levels as a tool for diagnosing giant cell arteritis (GCA) in a cohort of patients referred for assessment of suspected GCA. Methods We selected 298 patients recruited to the multicentre study Temporal Artery Biopsy versus Ultrasound in diagnosis of suspected GCA (TABUL). In a random subset of 26 biopsy-proven GCA cases and 26 controls, serum from weeks 0, 2 and 26 was analysed for VEGF concentration using ELISA. VEGF concentration at week 0 was used to generate a receiver-operating characteristic curve and thereby identify a cut-off for an abnormal result which was used to analyse the full patient cohort. Sections of paraffin-embedded temporal artery were stained by immunohistochemistry for VEGF. Results The mean (95% CI) VEGF concentration at week 0 was 873 pg/mL (631 to 1110) in 26 patients versus 476 pg/mL (328 to 625) in 26 controls (p=0.017). This difference was not observed at any other time point. The optimal cut-off of 713 pg/mL was applied to the whole patient cohort (n=298), yielding sensitivity of 32% and specificity of 85%. This was not improved by combination with any clinical parameters. When patients with biopsy-proven GCA were compared with controls, sensitivity was 58% and specificity remained 85%. Sections of biopsy from biopsy-positive GCA showed intense staining in the adventitia which was not seen in controls. Conclusions Serum VEGF concentration predicts biopsy positivity but is not useful for differentiating clinical cases of GCA from controls. Further studies into VEGF as a prognostic marker and therapeutic target are warranted. Trial registration number NCT00974883; Post-results. Giant Cell ArteritisSystemic vasculitisInflammation ==== Body Key messages What is already known about this subject? Vascular endothelial growth factor (VEGF), mainly derived from macrophages and giant cells, promotes angiogenesis and increased vascular permeability in health and in disease. Circulating levels of VEGF are increased in patients with giant cell arteritis (GCA). What does this study add? We have shown that levels of circulating VEGF are significantly higher in patients with newly diagnosed GCA compared with controls, especially if their temporal artery histology is positive, but could not be used to successfully differentiate between cases and controls. How might this impact on clinical practice? There is not enough evidence to use VEGF as a diagnostic test, either alone or in combination with clinical features suggesting the diagnosis. The role of VEGF as a prognostic marker or potential therapeutic target could be explored. Introduction Giant cell arteritis (GCA) is the most common primary systemic vasculitis1 2 typically presenting with headache, constitutional symptoms and visual disturbance.3 Diagnosis is challenging,4 relying on clinical judgement and imperfect yet potentially invasive tests including temporal artery biopsy.5 Elevation of inflammatory markers (erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP)) are non-specific and can be misleading.6 Furthermore, permanent visual loss affecting up to 20%7 requires timely diagnosis. Most patients present to their general practitioner; therefore, a serological biomarker that could be tested in primary care, could aid the diagnosis of GCA thereby avoiding more invasive, costly and time-consuming tests. Pentraxin-3 is a peripheral marker upregulated in GCA, and associated with optic nerve ischaemia.8 Pentraxins are members of the acute phase reactant superfamily. Pentraxin-3 is synthesised in response to vascular injury,9 and levels correlate with disease activity in small-vessel vasculitides.10 11 Therefore, it holds promise as a potential biomarker in GCA. Vascular endothelial growth factor (VEGF) promotes angiogenesis and increased vascular permeability in health and in disease.12 There are five variants of which VEGF-A is the predominant in adults,13 with higher levels in serum than in plasma.14 In GCA, neoangiogenesis occurs throughout the vessel wall, correlating with increased tissue expression of VEGF.15 The main source of VEGF is CD68 macrophages and giant cells within the vessel wall,15 16 and this local production of VEGF is mirrored by raised serum levels of VEGF in patients with GCA compared with controls.8 Furthermore, allelic variants in VEGF confer different susceptibility to development of GCA17 18 and certain polymorphisms are associated with increased risk of ischaemic complications.18 19 VEGF may function in GCA either to compensate for ischaemia in the thickened artery,20 or to drive inflammation by increasing new blood vessels which are the primary expressers of adhesion molecules involved in recruitment of inflammatory cells.16 Regardless of its role, VEGF represents an attractive candidate biomarker for the diagnosis of GCA. We evaluated the performance of VEGF-A as a diagnostic tool for GCA within a large cohort study of patients referred with suspected GCA (Temporal Artery Biopsy vs ULtrasound study (TABUL), HTA 08/64/01). We also undertook a preliminary investigation of pentraxin-3 as a candidate diagnostic marker for GCA. Materials and methods Study population We used samples collected during the TABUL study, n=298 total serum samples, and n=40 total plasma samples. Written consent was obtained according to the Declaration of Helsinki (BMJ 1996; 313:1448) and ethical approval was gained from Berkshire Research Ethics Committee (09/H0505/132). TABUL recruited patients referred to secondary care for investigation and management of suspected GCA. Patients with a previous diagnosis of GCA or polymyalgia rheumatica, and those who had received glucocorticoids for longer than 7 days, were excluded. Among the 298 participants, 202 patients had a final diagnosis of GCA (based on expert consensus using clinical judgement, histology, biochemistry and imaging), and 96 patients were given a diagnosis other than GCA (controls). Clinical data were collected during the study including age, sex, presenting symptoms, glucocorticoid duration at baseline and laboratory test results (ESR and CRP). Research bloods were taken at baseline (week 0) and at week 2 and month 6 of follow-up. Serum samples were prepared as per standard protocols and frozen at −80°C within 4 hours. No samples had been thawed prior to use in this study. Enzyme-linked immunosorbent assay Serum samples were analysed in duplicate using a high-sensitivity ELISA kit (R&D systems, Quantikine Human VEGF-165 kit, assay range 31.2–2000 pg/mL) according to the manufacturers' instructions. Each plate included a standard curve of known VEGF concentrations, and calculations were performed using a four-parameter logistical curve fitting. In a random subset of 26 biopsy-proven GCA cases and 26 controls, sera from weeks 0, 2 and 26 were analysed for VEGF concentration using ELISA. VEGF concentration at week 0 was used to generate a receiver-operating characteristic curve and identify a cut-off for abnormal results, which was used to analyse the full patient cohort. For pentraxin-3 measurements, plasma samples were used. Samples from 20 randomly selected biopsy-proven GCA patients and 20 controls were analysed in duplicate using a high-sensitivity kit, according to the manufacturer's instructions (R&D systems, Quantikine Human Pentraxin-3 kit, assay range 0.3–20 ng/mL). Immunohistochemistry Paraffin-embedded temporal artery samples were obtained from the TABUL study for five patients; three had a positive biopsy and two were control patients. Sections were cut at 4 μm and mounted on adhesive glass slides for staining. Slides were deparaffinised in xylene and rehydrated through a graded series of 100–50% ethanol. Endogenous peroxidase activity was blocked by 3% hydrogen peroxide. Heat-induced antigen retrieval for VEGF-A was performed by citrate buffer (10 mM anhydrous citric acid, 0.05% Tween 20, pH 6.0). Non-specific reactivity was blocked in buffer solution containing 3% filtered bovine serum albumin (BSA). Representative sections from each patient were incubated with 1:100 dilution of rabbit anti-human VEGF polyclonal IgG antibody (ABCAM). Negative control sections were incubated with non-immune rabbit IgG (R&D Systems). Secondary biotinylated goat anti-rabbit antibody (Vector Laboratories) was added at 1:250 dilution before incubation with avidin-biotin-peroxidase (ABC, Vectastain Elite kit, Vector Laboratories). Staining was developed with DAB substrate kit (Vector Laboratories) and counterstained with haematoxylin (Vector Laboratories). Images were captured using a microscope (Zeiss Imager M1) connected to a camera (Zeiss Axiocam). For general morphological analysis, serial sections were stained with Mayer's H&E. Sample size calculation For the VEGF subset, sample size was based on previous analysis21 with mean values in the control group of 362 pg/mL (SD=178 pg/mL) and mean values in the GCA group of 1145 pg/mL. Therefore, to give an α of 0.01 and a power of 0.9, the required size was 26 per group. For the pentraxin-3 study, the sample size was based on Baldini et al8 where pentraxin-3 in controls was 3.97±0.28 ng/mL and in patients with GCA, it was 23.3±4.06 ng/mL. Therefore, to give an α of 0.01 and a power of 0.9, the required size was 20 per group. Subgroup analysis In 26 cases and 26 controls, we undertook a subanalysis of glucocorticoid duration, grouping patients according to duration of steroid treatment, since the data were insufficient to allow for overall exposure to be calculated. Statistical analyses Statistical analysis was undertaken in GraphPad Prism V.6 and SPSS. Unless stated, significance was tested with the Mann-Witney U test for non-normally distributed data, with Bonferroni correction where appropriate for multiple comparisons. Receiver-operating characteristic curves, contingency table analysis, correlation and linear regression were performed in Prism V.6. Classification and regression tree analysis was performed in SPSS Statistics V.23 according to its standard protocols, using an input of VEGF concentration (high/low) and final diagnosis alongside other parameters including ESR, CRP and individual clinical features. Results From 298 available patients in the TABUL cohort (see figure 1), a randomly selected subset of 26 biopsy-proven cases of GCA were compared with 26 control patients who did not have GCA (Not GCA). These groups had an identical sex split (62% women each) but the mean age was slightly higher in the GCA group compared with controls (76, SD=7.3 vs 66, SD=11.3; p=0.003). The serum concentration of VEGF was measured at baseline (week 0), week 2 and week 26 (figure 2A). The mean (95% CI) baseline VEGF concentration was 873 pg/mL (631 to 1110) in GCA compared with 476 pg/mL (328 to 625) in controls (p=0.017). However, this difference was not observed at any later time point. Figure 1 Distribution of patients amongst diagnostic groups in this study. TAB=temporal artery biopsy, US=temporal artery ultrasound, +=positive result compatible with GCA, −=negative result. Figure 2 A. Serum VEGF is raised in patients with biopsy-proven GCA compared to controls at week 0, but this is not seen at any other time-point; B. Serum VEGF concentration is not dependent on duration of glucocorticoid therapy at the point of serum sampling. Graphs show mean ±SEM, significance tested with Mann-Whitney U test with Bonferroni correction; C. The receiver-operating characteristic curve for VEGF concentration at week 0 generates an optimal cut off of 713pg/ml Mean serum VEGF concentration in GCA patients fell from 873 pg/mL at week 0 to 393 pg/mL at week 2. It was hypothesised that this rapid decline in VEGF concentration may represent glucocorticoid responsiveness. Since patients in TABUL had been taking glucocorticoids for between 0 and 7 days prior to their enrolment, it was possible that those with a longer duration of glucocorticoid therapy would have lower VEGF concentrations. This was tested with a subanalysis of VEGF concentration according to duration of glucocorticoid therapy and final diagnosis, but there was no association between glucocorticoid exposure and VEGF concentration (figure 2B). It was not possible to analyse overall glucocorticoid exposure because the data were insufficient for this purpose. The data from week 0 were used to fit a receiver-operating characteristic curve (figure 2C) which had an area under the curve of 0.73. The optimal cut-off for an abnormal result was 713 pg/mL, providing a sensitivity of 65% and specificity of 88%. Sera from week 0 were analysed for all available TABUL patients, whose diagnosis was made by positive biopsy, positive ultrasound or purely clinical parameters (figure 1). VEGF concentration was higher in GCA (regardless of means of diagnosis) when compared with control patients but this difference was not significant (597 vs 464 pg/mL, p=0.09, see figure 3). When patients were separated into subgroups according to the means of diagnosis, the VEGF concentration at week 0 was significantly higher only for patients with a positive biopsy (804 vs 464 pg/mL, p=0.001). Diagnoses for control patients were as follows: 67% non-specific headache, 1% temporomandibular dysfunction, 11% cervical spondylosis, 1% granulomatosis with polyangiitis (GPA), 1% other vasculitis, 19% other diagnosis (mostly acute infections). Figure 3 VEGF concentration at week 0 is significantly raised in biopsy-proven GCA (TAB+) but not in Ultrasound-positive (US+) GCA or in clinical diagnoses of GCA. Graph shows mean ± 95% CI. Significance tested with Mann-Whitney U test with Bonferroni correction. We correlated VEGF levels with traditional acute phase reactants CRP and ESR (figure 4). There was a moderate positive correlation for VEGF with CRP (r=0.50; 95% CI 0.40 to 0.58; p<0.001) and a weak positive correlation of VEGF with ESR (r=0.25; 95% CI 0.14 to 0.26; p<0.001). Figure 4 Mild or moderate positive correlation between measurements of A. VEGF and ESR, B. VEGF and CRP at week 0. When the cut-off of 713 pg/mL was used to dichotomise patient data into normal and abnormal results, the sensitivity of VEGF concentration at week 0 was 34% and specificity was 85%. When patients were subgrouped according to diagnostic method, specificity remained stable but sensitivity fell to 25% for clinical GCA diagnoses rising to 58% for biopsy-positive cases (table 1). Correlation and regression tree analysis undertaken to determine whether VEGF concentration could be useful in combination with other parameters including CRP, ESR, headache, jaw or tongue claudication and sight loss (temporary or permanent) failed to reveal any combination that improved the diagnostic performance of the test in this cohort. Table 1 VEGF concentration predicts biopsy positivity but is not useful to identify other diagnostic subgroups. Performance of VEGF expressed as sensitivity and specificity based on data dichotomised by the cut-off of 713 pg/mL, subgrouped according to method of diagnosis Parameter All patients Clinically diagnosed GCA Ultrasound-positive GCA Biopsy-proven GCA Sensitivity 34% 25% 34% 58% Specificity 85% 85% 85% 85% GCA, giant cell arteritis; VEGF, vascular endothelial growth factor. It is not known whether VEGF concentration in the serum reflects local VEGF expression in the temporal artery. Therefore, we used immunohistochemistry to look for local VEGF expression in biopsy-positive cases of GCA, compared with control sections from patients who were not diagnosed with GCA. Representative images in figure 5 show that in this small group of samples, the VEGF staining in biopsy-positive GCA patients was much stronger than in controls, and was predominantly in the tunica adventitia, although some staining was seen in the media. Figure 5 VEGF is upregulated in the arterial wall of patients with biopsy-proven GCA but not in controls. Representative images of VEGF immunohistochemistry, all taken at 10× magnification. A–C from control patient; A. H&E, B. control IgG, C. αVEGF; arrow=tunica adventitia. D–F from GCA patient; D. H&E, E. control IgG, F. αVEGF; arrow=internal elastic lamina For analysis of pentraxin-3 as a diagnostic tool, we used randomly selected patients from TABUL and compared controls with biopsy-positive GCA patients, on the basis that this was the group which had a significant difference in VEGF concentration. Groups had similar proportions of women (60% in GCA group, 55% in control) but age again was slightly higher in the GCA group (75, SD 6.3) compared with the controls (69, SD 10.2, p=0.03). Plasma samples were analysed from week 0, 2 and 26. In contrast to the findings of Baldini et al, no significant difference was observed between patients with GCA and controls at any of the three time points tested (figure 6), therefore this marker was not tested in any further groups. Figure 6 Pentraxin 3 concentrations in the plasma of biopsy-positive GCA patients and controls are indistinguishable at weeks 0, 2, and 26. Graph shows mean ± SEM. Discussion We hypothesised that VEGF might be a diagnostic marker in GCA, but the evidence from the current study was not conclusive evidence, although we have been able to demonstrate a strong association between biopsy-positive GCA and elevated serum levels of VEGF. Serum concentration of VEGF at first presentation of patients with suspected GCA has some predictive power for a positive temporal artery biopsy. However, it does not have good diagnostic performance in separating all cases of GCA, regardless of diagnostic method, from controls. There are several potential explanations for this discrepancy. It may be that a positive temporal artery biopsy reflects a more vigorous inflammatory process with increased local and systemic expression of factors including VEGF; therefore, a higher likelihood of measuring an abnormal VEGF concentration in the peripheral blood. Indeed, patients with a positive biopsy are more likely to have raised inflammatory markers, and more likely to develop visual complications.22 23 Alternatively, it may be that patients with a positive temporal artery biopsy have a more robust diagnosis, therefore are more likely to have a high inflammatory response including high VEGF concentrations. Patients contributing to this study had been receiving glucocorticoids for up to 7 days at the point of blood collection. Twenty-three per cent controls and 22% GCA patients had not received any glucocorticoids at baseline and the remainder had received at least one dose. The duration of glucocorticoid exposure did not appear to affect VEGF concentrations in this study. However, it is possible that VEGF measurements may be more diagnostically accurate if taken prior to any glucocorticoid treatment. For most patients, this would involve testing blood samples taken at their general practitioners at the point of initial presentation. This may be a worthwhile future study. One limitation of this study is the potential instability of VEGF after freeze–thawing. All samples used had one freeze–thaw cycle: they were frozen at acquisition then defrosted only when used in this study. One paper has compared samples from four patients which were either analysed fresh or after various cycles of freeze–thawing, and found that even after one cycle, the degradation in serum signal was on average 67% (range 30–90%).24 However, three other studies have failed to observe any difference in serum VEGF after freeze–thawing up to 10 times in 30 patients,25 20 patients26 and 9 patients,27 respectively. Therefore, it is not clear whether the VEGF measured in serum stored at −80°C is an accurate reflection of VEGF concentration at the time of sampling. We selected controls for the study on the basis that they were presenting with a suspected diagnosis of GCA, but on further investigation, they did not have the condition. A significant limitation to the study is the absence of samples from other diseases including infection, malignancy and other inflammatory conditions. Further work could examine the comparative levels across these diseases; we speculate that the potential lack of difference between VEGF levels among different inflammatory conditions might make it difficult to use as a true diagnostic marker. There is debate over the choice of plasma or serum for the measurement of VEGF concentration. VEGF is released from platelets on clotting28 and it has been argued that this would make plasma the sample of choice for measuring extracellular VEGF.29 This is a potential limitation given our choice of serum for its measurement. However, it is not clear whether intracellular or extracellular VEGF is more important in the inflammatory process in GCA. It remains unclear what contribution clotting would have had on the levels of VEGF measured or whether this effect was different between patients and controls. Furthermore, it was not feasible to measure the VEGF concentration in plasma as this was collected in far fewer patients in the parent TABUL study and therefore would have significantly reduced the power of this study. However, we would suggest that further studies are warranted comparing VEGF levels in serum and plasma in patients with GCA and controls. We tested pentraxin-3 in this cohort, but there was no difference between biopsy-positive GCA cases and controls. This was surprising given that Baldini et al reported a sixfold upregulation of pentraxin-3 in GCA compared with healthy controls or controls with rheumatoid arthritis.8 Concentration of pentraxin-3 in controls for this study (3.35 ng/mL) was similar to Baldini et al (3.97 ng/mL) but concentration in GCA patients (5.48 ng/mL) was significantly lower than in Baldini et al (23.3 ng/mL). This suggests that the groups of GCA patients in the two studies are different. Baldini et al studied patients with established GCA for between 0.25 and 125 months, whereas TABUL studied an inception cohort of new diagnoses. However, if pentraxin-3 levels increase over time then there ought to have been an increase over the 6 months of sample collection in TABUL, which was not seen. Samples for this study underwent one freeze–thaw cycle, but Baldini et al do not comment on whether their samples were analysed fresh or after freeze–thaw. However, there is evidence that freeze–thawing does not affect measurement of pentraxin-3,30 so this is probably not the cause of the discrepancy. It would be informative to study a third population to confirm whether there is or is not an upregulation in pentraxin-3 measurements in patients with GCA. A challenge in managing GCA is accurate recognition or prediction of relapse. Despite early treatment with high-dose glucocorticoids, 34–79% relapse.31–34 We cannot predict which patients are likely to relapse; identification of relapse is complicated by presentation with non-specific symptoms coupled with normal laboratory tests in up to 20%.32 It is possible that VEGF may be a useful marker for the identification of relapsing patients, and this should be tested in future studies. We report raised tissue expression and systemic concentration of VEGF, related to positive temporal artery biopsy, which in turn predicts more severe disease. VEGF may be part of a useful healing response to the vascular inflammation and injury, alternatively it may be driving the inflammatory process. If the latter is the case, then it provides a more specific therapeutic target in comparison to the current standard of care. There is an established role for anti-VEGF therapy in other diseases including neovascularising ophthalmological conditions (such as macular degeneration, proliferative diabetic retinopathy, retinal vein occlusion35–37) and cancers including colorectal, breast, ovarian and cervical.38 Therefore, a number of agents are available which have already been approved for use in humans, providing possible novel therapeutic options in GCA. Further studies should focus on delineating the role of VEGF in GCA to explore its usefulness as a prognostic and therapeutic target. In conclusion, we report that serum levels of VEGF, a molecule with a potential role in the pathogenesis of GCA, are elevated in patients with newly diagnosed GCA, but its performance as diagnostic marker is currently unclear and should be further investigated. The following individuals contributed to domains of TABUL. Statistics: Ellen Lee, Amanda Loban, Christopher Ellis, Mike Gillett, Mike Bradburn; Ultrasound training and expert review: Wolfgang Schmidt, Bhaskar Dasgupta, Andreas Diamantopoulos, Eugene McNally, Jennifer Piper; Data management: Wulf Forrester-Barker; Primary Care expertise: Willie Hamilton; Design and conduct: Brendan McDonald, Colin Pease, John Salmon, Allan Wailoo, Konrad Wolfe; Study coordination: Keri Fathers, Leo Marcus-Wan, Nicola Farrar, Varun Manhas, Connor Scott, Nicky Sullivan, Denise Brown, Gareth Bicknell, Karolina Kliskey, Wulf Forrester-Barker; Data entry: David Gray, Samiya Mahmood, Ann-Marie Morgan, Ifzal Ahmed; Trial Steering Committee: Michael Ehrenstein, Bleddyn Davies, Karim Raza, David Mant; Data Monitoring Committee: Lyn Williamson, Kate Gilbert, Simon Travis, Jonathan Sterne; Principal Investigators: Colin Pease, David Carruthers, Rainer Klocke, Vadivelu Saravanan, Damodar Makkuni, Ruth Geraldes, Andreas Diamantopoulos, Thomas Neumann, Adrian Pendleton, Khalid Ahmed, Richard Hull, Kuntal Chakravarty, Peter Lanyon, Antoni Chan, Nicholas Raj, Frances Borg, Eamonn Molloy, Malgorzata Magliano, David Wright. Collaborators: TABUL Investigators; Ellen Lee; Amanda Loban; Christopher Ellis; Mike Gillett; Mike Bradburn; Wolfgang Schmidt; Bhaskar Dasgupta; Andreas Diamantopoulos; Eugene McNally; Jennifer Piper; Wulf Forrester-Barker; Willie Hamilton; Brendan McDonald; Colin Pease; John Salmon; Allan Wailoo; Konrad Wolfe; Keri Fathers; Leo Marcus-Wan; Nicola Farrar; Varun Manhas; Connor Scott; Nicky Sullivan; Denise Brown; Gareth Bicknell; Karolina Kliskey; David Gray; Samiya Mahmood; Ann-Marie Morgan; Ifzal Ahmed; Michael Ehrenstein; Bleddyn Davies; Karim Raza; David Mant; Lyn Williamson; Kate Gilbert; Simon Travis; Jonathan Sterne; Colin Pease; David Carruthers; Rainer Klocke; Vadivelu Saravanan; Damodar Makkuni; Ruth Geraldes; Andreas Diamantopoulos; Thomas Neumann; Adrian Pendleton; Khalid Ahmed; Richard Hull; Kuntal Chakravarty; Peter Lanyon; Antoni Chan; Nicholas Raj; Frances Borg; Eamonn Molloy; Malgorzata Magliano; David Wright. Contributors: NG designed, executed and analysed the ELISA work and prepared the manuscript. JM conducted pilot experiment upon which the work is based and executed the immunohistochemistry with AZ. SS was the trial coordinator for the parent TABUL study. AS supervised JM and provided scientific design and expertise, and edited the manuscript. AH was the statistician for parent TABUL study, also helped with statistics for this study. VS helped to coordinate sample collection and storage. JV took over from SS as TABUL trial coordinator. SM was the research nurse for parent TABUL study, and helped coordinate sample collection and storage. AZ helped JM with immunohistochemistry. RL was the principal investigator for TABUL, was supervisor for NG, and was involved in design of work and editing of manuscript. Funding: This work was supported by the HTA (grant number 08/64/01 to RL), the Oxfordshire Health Research Services Committee (grant number 1175 to NG) and the NIHR (Academic Clinical Fellowship to NG). Competing interests: None declared. Ethics approval: Ethics approval was obtained from Berkshire Research Ethics Committee (09/H0505/132). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 De Smit E , Palmer AJ , Hewitt AW Projected worldwide disease burden from giant cell arteritis by 2050 . J Rheumatol 2015 ;42 :119 –25 . 10.3899/jrheum.140318 25362658 2 Smeeth L , Cook C , Hall AJ Incidence of diagnosed polymyalgia rheumatica and temporal arteritis in the United Kingdom, 1990–2001 . Ann Rheum Dis 2006 ;65 :1093 –8 . 10.1136/ard.2005.046912 16414971 3 Dasgupta B , Borg FA , Hassan N BSR and BHPR guidelines for the management of giant cell arteritis . Rheumatology (Oxford) 2010 ;49 :1594 –7 . 10.1093/rheumatology/keq039a 20371504 4 Smetana GW , Shmerling RH Does this patient have temporal arteritis? JAMA 2002 ;287 :92 –101 . 10.1001/jama.287.1.92 11754714 5 Davies C , Frost B , Eshan O Temporal artery biopsy…who needs one? Postgrad Med J 2006 ;82 :476 –8 . 10.1136/pgmj.2005.043646 16822927 6 Kermani TA , Schmidt J , Crowson CS Utility of erythrocyte sedimentation rate and C-reactive protein for the diagnosis of giant cell arteritis . Semin Arthritis Rheum 2012 ;41 :866 –71 . 10.1016/j.semarthrit.2011.10.005 22119103 7 Borg FA , Salter VL , Dasgupta B Neuro-ophthalmic complications in giant cell arteritis . Curr Allergy Asthma Rep 2008 ;8 :323 –30 . 10.1007/s11882-008-0052-4 18606086 8 Baldini M , Maugeri N , Ramirez GA Selective up-regulation of the soluble pattern-recognition receptor pentraxin 3 and of vascular endothelial growth factor in giant cell arteritis: relevance for recent optic nerve ischemia . Arthritis Rheum 2012 ;64 :854 –65 . 10.1002/art.33411 21989653 9 Maugeri N , Rovere-Querini P , Baldini M Translational mini-review series on immunology of vascular disease: mechanisms of vascular inflammation and remodelling in systemic vasculitis . Clin Exp Immunol 2009 ;156 :395 –404 . 10.1111/j.1365-2249.2009.03921.x 19309348 10 van Rossum AP , Pas HH , Fazzini F Abundance of the long pentraxin PTX3 at sites of leukocytoclastic lesions in patients with small-vessel vasculitis . Arthritis Rheum 2006 ;54 :986 –91 . 10.1002/art.21669 16508993 11 Fazzini F , Peri G , Doni A PTX3 in small-vessel vasculitides: an independent indicator of disease activity produced at sites of inflammation . Arthritis Rheum 2001 ;44 :2841 –50 . 10.1002/1529-0131(200112)44:12<2841::AID-ART472>3.0.CO;2-6 11762945 12 Dvorak HF , Brown LF , Detmar M Vascular permeability factor/vascular endothelial growth factor, microvascular hyperpermeability, and angiogenesis . Am J Pathol 1995 ;146 :1029 –39 .7538264 13 Byrne AM , Bouchier-Hayes DJ , Harmey JH Angiogenic and cell survival functions of vascular endothelial growth factor (VEGF) . J Cell Mol Med 2005 ;9 :777 –94 . 10.1111/j.1582-4934.2005.tb00379.x 16364190 14 Lee JK , Hong YJ , Han CJ Clinical usefulness of serum and plasma vascular endothelial growth factor in cancer patients: which is the optimal specimen? Int J Oncol 2000 ;17 :149 –52 .10853032 15 Kaiser M , Younge B , Björnsson J Formation of new vasa vasorum in vasculitis. Production of angiogenic cytokines by multinucleated giant cells . Am J Pathol 1999 ;155 :765 –74 . 10.1016/S0002-9440(10)65175-9 10487834 16 Cid MC , Cebrián M , Font C Cell adhesion molecules in the development of inflammatory infiltrates in giant cell arteritis: inflammation-induced angiogenesis as the preferential site of leukocyte-endothelial cell interactions . Arthritis Rheum 2000 ;43 :184 –94 . 10.1002/1529-0131(200001)43:1<184::AID-ANR23>3.0.CO;2-N 10643715 17 Boiardi L , Casali B , Nicoli D Vascular endothelial growth factor gene polymorphisms in giant cell arteritis . J Rheumatol 2003 ;30 :2160 –4 .14528511 18 Enjuanes A , Benavente Y , Hernandez-Rodríguez J Association of NOS2 and potential effect of VEGF, IL6, CCL2 and IL1RN polymorphisms and haplotypes on susceptibility to GCA--a simultaneous study of 130 potentially functional SNPs in 14 candidate genes . Rheumatology (Oxford) 2012 ;51 :841 –51 . 10.1093/rheumatology/ker429 22258388 19 Rueda B , Lopez-Nevot MA , Lopez-Diaz MJ A functional variant of vascular endothelial growth factor is associated with severe ischemic complications in giant cell arteritis . J Rheumatol 2005 ;32 :1737 –41 .16142870 20 Cid MC , Hernandez-Rodríguez J , Esteban MJ Tissue and serum angiogenic activity is associated with low prevalence of ischemic complications in patients with giant-cell arteritis . Circulation 2002 ;106 :1664 –71 . 10.1161/01.CIR.0000030185.67510.C0 12270860 21 Morlet J , Mahoney D , Zarei A Diagnosing giant cell arteritis: the challenges and potential solutions . Rheumatology (Oxford) 2014 ;53 (suppl 1 ):188 10.1093/rheumatology/keu129.011 22 Gonzalez-Gay MA , Garcia-Porrua C , Llorca J Biopsy-negative giant cell arteritis: clinical spectrum and predictive factors for positive temporal artery biopsy . Semin Arthritis Rheum 2001 ;30 :249 –56 . 10.1053/sarh.2001.16650 11182025 23 Duhaut P , Pinede L , Bornet H Biopsy proven and biopsy negative temporal arteritis: differences in clinical spectrum at the onset of the disease. Groupe de Recherche sur l'Artérite à Cellules Géantes . Ann Rheum Dis 1999 ;58 :335 –41 . 10.1136/ard.58.6.335 10340957 24 Kisand K , Kerna I , Kumm J Impact of cryopreservation on serum concentration of matrix metalloproteinases (MMP)-7, TIMP-1, vascular growth factors (VEGF) and VEGF-R2 in Biobank samples . Clin Chem Lab Med 2011 ;49 :229 –35 . 10.1515/CCLM.2011.049 21118050 25 Lee JE , Kim SY , Shin SY Effect of repeated freezing and thawing on biomarker stability in plasma and serum samples . Osong Public Health Res Perspect 2015 ;6 :357 –62 . 10.1016/j.phrp.2015.11.005 26835245 26 Hetland ML , Christensen IJ , Lottenburger T Circulating VEGF as a biological marker in patients with rheumatoid arthritis? Preanalytical and biological variability in healthy persons and in patients . Dis Markers 2008 ;24 :1 –10 . 10.1155/2008/707864 18057530 27 Guo GH , Dong J , Yuan XH Clinical evaluation of the levels of 12 cytokines in serum/plasma under various storage conditions using evidence biochip arrays . Mol Med Rep 2013 ;7 :775 –80 . 10.3892/mmr.2013.1263 23291902 28 Webb NJ , Bottomley MJ , Watson CJ Vascular endothelial growth factor (VEGF) is released from platelets during blood clotting: implications for measurement of circulating VEGF levels in clinical disease . Clin Sci 1998 ;94 :395 –404 . 10.1042/cs0940395 9640345 29 Ferrero S Serum and plasma vascular endothelial growth factor levels in testicular cancer patients . Ann Oncol 2004 ;15 :989 –90 . 10.1093/annonc/mdh228 15151961 30 Bastrup-Birk S , Skjoedt MO , Munthe-Fog L Pentraxin-3 serum levels are associated with disease severity and mortality in patients with systemic inflammatory response syndrome . PLoS ONE 2013 ;8 :e73119 10.1371/journal.pone.0073119 24039869 31 Alba MA , Garcia-Martínez A , Prieto-González S Relapses in patients with giant cell arteritis: prevalence, characteristics, and associated clinical findings in a longitudinally followed cohort of 106 patients . Medicine (Baltimore) 2014 ;93 :194 –201 . 10.1097/MD.0000000000000033 25181312 32 Kermani TA , Warrington KJ , Cuthbertson D Disease relapses among patients with giant cell arteritis: a prospective, longitudinal cohort study . J Rheumatol 2015 ;42 :1213 –17 . 10.3899/jrheum.141347 25877501 33 Labarca C , Koster MJ , Crowson CS Predictors of relapse and treatment outcomes in biopsy-proven giant cell arteritis: a retrospective cohort study . Rheumatology (Oxford) 2016 ;55 : 347 –56 . 10.1093/rheumatology/kev348 26385368 34 Martinez-Lado L , Calviño-Diaz C , Piñeiro A Relapses and recurrences in giant cell arteritis: a population-based study of patients with biopsy-proven disease from northwestern Spain . Medicine (Baltimore) 2011 ;90 :186 –93 . 10.1097/MD.0b013e31821c4fad 21512412 35 Sarwar S , Bakbak B , Sadiq MA Fusion Proteins: Aflibercept (VEGF Trap-Eye) . Dev Ophthalmol 2016 ;55 :282 –94 . 10.1159/000439008 26501481 36 Klein A , Loewenstein A Therapeutic monoclonal antibodies and fragments: bevacizumab . Dev Ophthalmol 2016 ;55 :232 –45 . 10.1159/000431199 26502311 37 Smith AG , Kaiser PK Therapeutic monoclonal antibodies and fragments: ranibizumab . Dev Ophthalmol 2016 ;55 :246 –51 . 10.1159/000431200 26501149 38 Arjaans M , Schröder CP , Oosting SF VEGF pathway targeting agents, vessel normalization and tumor drug uptake: from bench to bedside . Oncotarget 2016 ;7 21247 –58 . 10.18632/oncotarget.6918 26789111

PMC005xxxxxx/PMC5372065.txt

==== Front BMJ Open Diabetes Res CareBMJ Open Diabetes Res CarebmjdrcbmjdrcBMJ Open Diabetes Research & Care2052-4897BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjdrc-2016-00033910.1136/bmjdrc-2016-000339Clinical Care/Education/Nutrition/Psychosocial Research15061866Successful long-term weight loss among participants with diabetes receiving an intervention promoting an adapted Mediterranean-style dietary pattern: the Heart Healthy Lenoir Project Embree Genevieve G R 1Samuel-Hodge Carmen D 23Johnston Larry F 3Garcia Beverly A 3Gizlice Ziya 3Evenson Kelly R 4DeWalt Darren A 5Ammerman Alice S 23Keyserling Thomas C 35 1 Ambulatory Care Physician, Durham VA Medical Center, Durham, North Carolina, USA 2 Department of Nutrition, Center for Health Promotion and Disease Prevention, Gillings School of Global Public Health, University of North Carolina, Chapel Hill, North Carolina, USA 3 Center for Health Promotion and Disease Prevention, University of North Carolina, Chapel Hill, North Carolina, USA 4 Department of Epidemiology, Gillings School of Global Public Health, University of North Carolina, Chapel Hill, North Carolina, USA 5 Division of General Medicine and Clinical Epidemiology, School of Medicine, University of North Carolina, Chapel Hill, North Carolina, USACorrespondence to Dr Thomas C Keyserling; jato@med.unc.edu2017 29 3 2017 5 1 e0003397 10 2016 2 2 2017 20 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective To examine weight change by diabetes status among participants receiving a Mediterranean-style diet, physical activity, and weight loss intervention adapted for delivery in the southeastern USA, where rates of cardiovascular disease (CVD) are disproportionately high. Research design and methods The intervention included: Phase I (months 1–6), an individually tailored intervention promoting a Mediterranean-style dietary pattern and increased walking; Phase II (months 7–12), option of a 16-week weight loss intervention for those with BMI≥25 kg/m2 offered as 16 weekly group sessions or 5 group sessions and 10 phone calls, or a lifestyle maintenance intervention; and Phase III (months 13–24), weight loss maintenance intervention for those losing ≥8 pounds with all others receiving a lifestyle maintenance intervention. Weight change was assessed at 6, 12, and 24-month follow-up. Results Baseline characteristics (n=339): mean age 56, 77% female, 65% African-American, 124 (37%) with diabetes; mean weight 103 kg for those with diabetes and 95 kg for those without. Among participants with diabetes, average weight change was −1.2 kg (95% CI −2.1 to −0.4) at 6 months (n=92), −1.5 kg (95% CI −2.9 to −0.2) at 12 months (n=96), and −3.7 kg (95% CI −5.2 to −2.1) at 24 months (n=93). Among those without diabetes, weight change was −0.4 kg (95% CI −1.4 to 0.6) at 24 months (n=154). Conclusions Participants with diabetes experienced sustained weight loss at 24-month follow-up. High-risk US populations with diabetes may experience clinically important weight loss from this type of lifestyle intervention. Trial registration number NCT01433484. Dietary InterventionEthnic DisparitiesPhysical Activity Intervention(s)Weight Loss ProgramNational Heart, Lung, and Blood Institutehttp://dx.doi.org/10.13039/1000000505P50 HL105184Centers for Disease Control and Preventionhttp://dx.doi.org/10.13039/100000030U48/DP001944 ==== Body Significance of this study What is already known about this subject? Though not well studied in the USA, particularly among low-income and minority populations, a Mediterranean-style dietary pattern can reduce the risk for cardiovascular disease (CVD) among those with and without type 2 diabetes and has been shown in some studies to be associated with maintenance of weight loss. What are the new findings? We developed and evaluated a lifestyle and weight loss intervention for low-income and minority residents of eastern North Carolina, with a major focus on implementing a culturally adapted Mediterranean-style dietary pattern while also promoting physical activity. Among study participants with diabetes, the intervention was well received and associated with sustained weight loss at 24-month follow-up. How might these results change the focus of research or clinical practice? This study demonstrates high acceptability of a Mediterranean-style dietary pattern among low-income and minority populations residing in eastern North Carolina and appears to be promising as a weight loss intervention for participants with diabetes. This dietary pattern and type of intervention should be evaluated in randomized controlled trials to further assess the intervention’s effectiveness at improving CVD risk factors, yielding sustained weight loss, and reducing CVD events. Introduction Weight loss, defined by the American Diabetes Association as a sustained reduction of 5% of initial body weight,1 is recommended for overweight and obese patients with type 2 diabetes, while others have noted that weight loss as small as 2 kg appears to provide clinical benefit.2 However, weight loss is difficult to achieve and harder to sustain; and in general, individuals with diabetes have a harder time losing weight and maintaining weight loss, when compared with those without diabetes.3 This is especially true for high risk groups such as African-American women with type 2 diabetes.4 For example, in the Action for Health in Diabetes (Look AHEAD) trial,5 among female participants with diabetes, the year-1 weight loss was 6.8% for African-Americans compared with 9.1% for non-Hispanic whites. Thus, improved approaches for promoting weight loss are needed for patients with diabetes and especially for groups at higher risk for obesity and diabetes such as African-American women. Some have argued that a variety of weight loss dietary patterns are acceptable if they lead to weight loss.6 However, as patients with type 2 diabetes are at very high risk for cardiovascular disease (CVD), consideration should be given to advocating a weight loss dietary pattern that also reduces CVD risk. When a Mediterranean diet pattern, supplemented with olive oil or nuts, was evaluated in the Prevención con Dieta Mediterránea (PREDIMED) randomized trial,7 there was a 30% reduction in CVD risk among participants with and without diabetes. This contrasts with the lower fat weight loss diet evaluated in Look AHEAD,8 which led to significant weight loss but did not reduce CVD risk. Attention to dietary pattern as a critical component of weight loss diets may be even more important for populations at very high risk for CVD, such as minority populations and those residing in the ‘stroke belt’ of the southeastern USA, where CVD rates are substantially higher compared with national levels.9 Thus, we developed and evaluated a lifestyle and weight loss intervention for residents of eastern North Carolina, with a major focus on implementing a culturally adapted Mediterranean-style dietary pattern while also promoting physical activity. In this paper, we report the intervention's effect on diet and physical activity behaviors, CVD risk factors, and weight loss through 24 months of follow-up by participants' diabetes status. Research design and methods Study overview The Heart Healthy Lenoir (HHL) Project was a collaborative research effort designed to reduce CVD risk and disparities in CVD risk among Lenoir County, North Carolina residents, as previously described.10 11 It was funded by the National Heart, Lung, and Blood Institute (NHLBI) as part of an initiative with the National Cancer Institute ‘to develop and test multilevel interventions to reduce health disparities.’12 This paper focuses on participants who took part in the HHL lifestyle study,11 one of three coordinated HHL studies which also included a study to improve high blood pressure (BP) management at local practices and a study examining associations between genetic markers and change in CVD risk factors. These studies were conducted in Lenoir County because of its location in the ‘stroke belt’9 of eastern North Carolina, where rates of CVD are higher than state and national averages13 and because it has a large minority population (40% African-American) that experiences disproportionally higher rates of CVD.14 The HHL lifestyle study was designed and conducted with input from a local community advisory committee10 and was approved and monitored by the University of North Carolina's Institutional Review Board. Data were collected between 20 September, 2011 and 7 November, 2014. The lifestyle study consisted of three phases as shown in figure 1, which depicts the three sequential phases of the study and the number of participants with and without diabetes who took part in each component of the intervention. Phase I, which lasted 6 months and was the same for all study participants, focused on improving diet quality and increasing physical activity. In Phase II, which also lasted 6 months, participants with a body mass index (BMI) ≥25 kg/m2 were offered an intensive weight loss intervention while those with a BMI<25 kg/m2 and those who declined the weight loss intervention received a maintenance of lifestyle intervention. In Phase III, participants who lost ≥8 lbs (3. 6 kg) at the conclusion of Phase II were invited to take part in a year-long, randomized controlled trial (RCT), comparing a more intensive and less intensive maintenance of weight loss intervention; a year-long maintenance of lifestyle intervention was given to those who did not take part in the RCT. As prior RCTs conducted by our research team15–18 have shown that similar formats of lifestyle and weight loss interventions are effective in improving lifestyle and achieving weight loss among low socioeconomic status participants, we did not include a control group for Phases I and II of this study. Furthermore, our community advisory committee strongly encouraged a study design in which all participants received ‘active treatment.’ Figure 1 Study flow diagram. Participants The enrollment goal for this study, based on having an adequate sample for the Phase III maintenance of weight loss RCT, was 350 with about 150 participants recruited from the community and 200 from the high BP study, as previously described.11 19 Criteria for screening from the community were age ≥18 years and interest in improving lifestyle behaviors to reduce CVD risk. Screening criteria for the high BP study were age ≥18, being an established patient at a participating practice, and systolic BP≥150 mmHg when assessed during routine care within the prior 12 months. After obtaining verbal informed consent, research staff conducted phone interviews to screen potential participants.11 If eligibility criteria were met, participants were invited to an enrollment visit at a central research office or at participating clinics. Written informed consent was obtained prior to collecting baseline measures. Participants attending the enrollment visit for the high BP study were invited to also take part in the lifestyle study until 200 agreed to do so. Intervention Phase I (months 1–6)—lifestyle intervention This intervention focused on dietary and physical activity behaviors and did not address other aspects of lifestyle relevant to CVD risk reduction, such as smoking cessation. Further, the same intervention was given to those with and without diagnosed diabetes. (All participants with diabetes had type 2 diabetes). The lifestyle intervention was originally developed by Ammerman et al20 21 and subsequently revised to emphasize carbohydrate quality as an important component of a heart-healthy diet.15 Consistent with the evolving literature suggesting frequent consumption of food with high-quality fats (polyunsaturated and monounsaturated fats primarily from plant sources and fish) is also important in reducing CVD risk in those with and without diabetes,7 22–26 this study's dietary intervention was further modified to include a major focus on improving dietary fat quality. These changes rendered the HHL lifestyle intervention dietary pattern very similar to the PREDIMED study's nut intervention arm diet;7 therefore, the dietary pattern was termed ‘Med-South’ because of its intended use in the southeastern USA. The intervention format and content have been previously published and described in detail.11 Phase I included four monthly sessions delivered by a trained counselor administered as hour-long individual counseling sessions or 90 min group sessions given at a centrally located research office or participating clinic. Participants chose their preferred counseling format. About 75% of counseling time was devoted to dietary behaviors with the rest focusing on physical activity. Dietary counseling included culturally relevant content to improve fat quality (such as increasing consumption of nuts, full-fat salad dressings and mayonnaise, and vegetable oils), increase fruit and vegetable consumption, promote fish and poultry intake while reducing red and processed meat intake, and minimize consumption of sugar-sweetened beverages, high-sugar-content desserts, and snacks. Physical activity counseling focused mostly on walking with a recommended goal of at least 7500 steps/day or ≥30 min/day of physical activity on ≥5 days/week. Spouses and friends were invited to attend the counseling intervention sessions. When participants could not physically attend counseling sessions, telephone counseling was offered. Participants also received a pedometer and activity logs to self-monitor physical activity as well as a listing of local community resources that promoted healthy eating (eg, farmers' markets) and physical activity (eg, community parks). Those participants who were co-enrolled in the high BP study received a home BP monitor and were instructed to measure their BP at least three times per week. They also received monthly phone calls for a year, primarily promoting BP medication adherence. During counseling sessions, participants worked with their counselor to create individually tailored action plans to improve dietary and physical activity behaviors. Dietary and physical activity tips were tailored to problematic lifestyle behaviors assessed on the baseline lifestyle questionnaire. Dietary tips included recipe suggestions from a southern-style cookbook that was given to all participants. At the beginning of sessions 2–4, the counselor and participants reviewed progress made towards previously stated goals. Phase II (months 7 through 12)—weight loss and maintenance of lifestyle interventions Participants with a BMI≥25 kg/m2 could choose to take part in the weight loss intervention. Those who were not eligible for the weight loss intervention (BMI<25 kg/m2) and those who declined the intervention received a maintenance of lifestyle intervention consisting of three phone calls, as previously described.11 The weight loss intervention was offered in two formats over ∼16 weeks: 16 weekly group sessions as previously tested,16 18 27 or five group sessions plus 10 phone contacts (combination intervention), as recently described.11 The major modification from the previously tested weight loss intervention was the focus on the Med-South dietary pattern and addition of newer evidence-based behavioral components (eg, daily self-weighing).28 Phase III (13–24 months)—weight loss and lifestyle maintenance interventions Participants who took part in the Phase II weight loss intervention and lost ≥8 lbs (3.6 kg) were invited to take part in the maintenance of weight loss RCT. All other study participants received brief, quarterly maintenance of lifestyle intervention phone calls (similar to Phase II), as previously described.11 For the maintenance of weight loss RCT, participants were randomized 1:1 to receive either 36 phone contacts (24 weekly calls over 6 months followed by 12 biweekly calls over 6 months; more intensive intervention) or 18 phone contacts (12 biweekly calls over 6 months followed by 6 monthly calls over 6 months; less intensive intervention). Measures Outcome measures were assessed at baseline, 6, 12, and 24 months. Previously validated questionnaires were administered to assess lifestyle change including the Dietary Risk Assessment (DRA) which assessed overall diet quality,20 29 the brief Block fruit and vegetable questionnaire,30 a dietary fat quality screener,31 and a questionnaire to assess walking and overall physical activity.32 33 In addition, the SF-12 (SF-12 instrument, Quality Metric, Lincoln, RI) was administered to assess quality of life. Weight was assessed by electronic scale (Seca 874, Seca, Hanover, MD) as the average of two measures to the closest tenth pound. Height was measured with a portable stadiometer at baseline only. BP was calculated as the average of three measurements recorded at 60 s intervals (Omron HEM-907XL, Omron Healthcare, Lake Forest, IL) after being seated for 5 min. Blood lipids and glycated hemoglobin (A1c) were assessed by a commercial laboratory (LabCorp, Burlington, North Carolina, USA). At follow-up measurement visits, questionnaires were administered to assess acceptability of the intervention and adverse outcomes. Participants received compensation for measurement visits: $40 for enrollment, $25 for 6 and 12-month visits, and $30 for the 24-month visit. Sample size and statistical analysis The enrollment goal of 350 participants was based on having a sufficient sample (N=100) for the embedded RCT of weight loss maintenance.11 In addition, the overall sample of 350 was considered sufficient to describe the primary objective of the lifestyle intervention to improve diet quality at 6-month follow-up, and the major secondary objectives of improved diet quality, physical activity, and weight loss at 12 and 24-month follow-up. Sample characteristics were summarized using descriptive statistics, with subgroups by diabetes status and race. Outcomes were assessed using pre–post changes by diabetes status and race with paired t-tests for continuous outcomes, McNemar's tests for binary outcomes, and χ2 tests for subgroup analysis when appropriate. Data are reported for returnees at follow-up without imputation for missing data. As described,11 among all study participants, weight loss was substantially less than anticipated and observed in our prior weight loss studies.16 18 Only 27 participants overall (eight with diabetes) took part in the maintenance of weight loss RCT. Given this small number, weight loss results are reported in the aggregate and not separately for those who took part in the RCT. Because age, race, sex, education, and baseline weight are potential confounders for weight loss, a linear regression analysis was conducted adjusting for these variables. Participants who became pregnant, had bariatric surgery, or were diagnosed with cancer (excluding non-melanoma skin cancer or localized breast or prostate cancer diagnosed by screening tests) were excluded from analysis (eight overall, four with diabetes). SAS V.9.3 was used for analysis. Results Baseline characteristics As outlined in detail elsewhere,11 of 642 individuals assessed as eligible for this study, 366 (57%) attended the enrollment visit and 339 (53%) completed all baseline measurements and comprised the study sample, including 134 recruited from the community and 205 from the clinic-based high BP study. Participants' baseline characteristics, by diabetes status then further categorized by race, are outlined in table 1. Table 1 Baseline characteristics: overall, by diabetes status then by race Characteristics Diabetes No Diabetes Overall All African-American White All African-American White n=339 n=124 n=89 n=34 n=215 n=130 n=83 Demographics Age, mean (SE) 56 (0.6) 59 (0.9) 59 (1.1) 61 (1.7) 54 (0.8) 51 (1.1) 57 (1.2) Female 260 (77) 93 (75) 70 (79) 22 (65) 167 (78) 111 (85) 54 (65) Race  African-American 219 (65) 89 (72) 130 (61)  White 117 (35) 34 (28) 83 (39) Education, years  ≤8 (middle school or less) 16 (5) 7 (6) 4 (5) 3 (9) 9 (4) 7 (5) 2 (2)  9–11 (some high school) 45 (13) 25 (20) 22 (25) 2 (6) 20 (9) 13 (10) 7 (8)  12 (high school graduate) 128 (38) 46 (37) 36 (40) 10 (29) 82 (38) 58 (45) 24 (29)  13–15 (some college) 79 (23) 24 (19) 16 (18) 8 (24) 55 (26) 29 (22) 25 (30)  16 (college graduate) 49 (14) 14 (11) 9 (10) 5 (15) 35 (16) 17 (13) 17 (21)  >16 (graduate school) 22 (7) 8 (7) 2 (2) 6 (18) 14 (7) 6 (5) 8 (10) Education: high school or less 189 (56) 78 (63) 62 (70) 15 (44) 111 (52) 78 (60) 33 (40) Marital status  Married or living with a partner 159 (47) 51 (41) 32 (36) 19 (56) 108 (50) 50 (39) 57 (69)  Other 180 (53) 73 (59) 57 (64) 15 (44) 107 (50) 80 (62) 26 (31) Currently have health insurance 251 (74) 94 (76) 68 (76) 25 (74) 157 (73) 88 (68) 67 (81) Current employment  Working full time 124 (37) 25 (20) 18 (20) 7 (21) 99 (46) 64 (49) 33 (40)  Working part time 42 (12) 17 (14) 14 (16) 2 (6) 25 (12) 17 (13) 8 (10)  Do not work due to health reasons 69 (20) 38 (31) 26 (29) 12 (35) 31 (14) 17 (13) 14 (17)  Retired 53 (16) 26 (21) 18 (20) 8 (24) 27 (13) 8 (6) 19 (23)  Other 51 (15) 18 (14) 13 (15) 5 (15) 33 (15) 24 (19) 9 (11) Annual household income  <$10 000 62 (20) 31 (29) 24 (32) 6 (18) 31 (16) 26 (22) 5 (7)  $10 000 to<$20 000 64 (21) 28 (26) 23 (31) 5 (15) 36 (18) 22 (19) 14 (18)  $20 000 to<$40 000 84 (28) 29 (27) 20 (27) 9 (27) 55 (28) 40 (34) 14 (18)  $40 000 to<$60 000 33 (11) 9 (8) 3 (4) 6 (18) 24 (12) 12 (10) 12 (16)  $60 000 to <$80 000 27 (9) 9 (8) 4 (5) 5 (15) 18 (9) 9 (8) 9 (12)  ≥$80 000 34 (11) 2 (2) 0 (0) 2 (6) 32 (15) 10 (8) 22 (29) CVD and risk factors for CVD Known coronary heart disease 49 (14) 28 (23) 20 (23) 8 (24) 21 (10) 10 (8) 11 (13) Known cardiovascular disease 62 (18) 33 (27) 23 (26) 10 (29) 29 (13) 14 (11) 15 (18) Hypertension 291 (86) 121 (98) 88 (99) 32 (94) 170 (79) 107 (82) 63 (76) Cholesterol  High (≥240 mg/dL) 187 (56) 82 (67) 57 (65) 24 (71) 105 (49) 53 (42) 52 (63)  Borderline (200–239 mg/dL) 46 (14) 8 (7) 7 (8) 1 (3) 38 (18) 26 (21) 11 (13)  Desirable (<200 mg/dL) 102 (30) 33 (27) 24 (27) 9 (27) 69 (33) 48 (38) 20 (24) Diabetes 124 (37) Current cigarette smoker 54 (16) 21 (17) 17 (19) 4 (12) 33 (15) 20 (15) 13 (16) Packs of cigarettes smoked per day, mean (SE) for current smokers 0.7 (0.1) 0.6 (0.1) 0.6 (0.1) 0.8 (0.2) 0.7 (0.1) 0.6 (0.1) 0.9 (0.2) Taking BP lowering medications 260 (77) 114 (92) 83 (93) 30 (88) 146 (68) 93 (72) 53 (64) Lifestyle DRA total score 27.8 (0.3) 28.5 (0.5) 28.3 (0.6) 28.9 (0.7) 27.4 (0.4) 27.2 (0.5) 27.7 (0.7) Fat quality screener score 15.5 (0.2) 15.5 (0.2) 15.4 (0.3) 15.7 (0.3) 15.4 (0.2) 15.3 (0.2) 15.7 (0.4) Fruit and vegetable servings per day 3.4 (0.1) 3.7 (0.2) 3.8 (0.2) 3.3 (0.3) 3.3 (0.1) 3.1 (0.2) 3.5 (0.2) Walking time (min/wk) 91 (11.3) 67 (12.8) 68 (15.6) 54 (20.5) 105 (16.1) 122 (24.9) 80 (14.3) Total physical activity time (min/wk) 149 (14.0) 112 (18.1) 112 (20.5) 99 (36.0) 171 (19.3) 176 (27.1) 161 (26.3) Physiological Weight, kg 98 (1.4) 103 (2.2) 103 (2.4) 104 (4.7) 95 (1.7) 99 (2.3) 90 (2.5) BMI, kg/m2 36 (0.5) 38 (0.8) 38 (0.9) 38 (1.6) 35 (0.7) 37 (0.9) 32 (0.9) Systolic blood pressure, mm Hg 135 (1.2) 136 (2.0) 138 (2.4) 131 (3.6) 134 (1.5) 136 (2.1) 133 (2.2) Diastolic blood pressure, mm Hg 82 (0.7) 81 (1.1) 82 (1.4) 78 (1.9) 83 (0.8) 84 (1.1) 81 (1.2) HbA1c, % 6.6 (0.1) 7.9 (0.2) 8.0 (0.2) 7.6 (0.3) 5.8 (0.0) 5.8 (0.0) 5.7 (0.0) HbA1c, mmol/mol 49 (1.1) 63 (2.2) 64 (2.2) 60 (3.3) 40 (0.0) 40 (0.0) 39 (0.0) Total cholesterol, mg/dL 193 (2.3) 188 (4.1) 188 (5.0) 187 (7.2) 196 (2.7) 193 (3.2) 200 (4.8) HDL cholesterol, mg/dL 54 (0.8) 51 (1.4) 55 (1.6) 43 (2.1) 56 (1.0) 58 (1.3) 53 (1.6) Data are means (SE) or n (%). BMI, body mass index; BP, blood pressure; CVD, cardiovascular disease; DRA, dietary risk assessment; HDL, high-density lipoprotein. A total of 124 (37%) participants had diabetes. The overall average age was 56 years; participants with diabetes were older on average than those without diabetes (59 vs 54 years). Men, particularly African-American men, were underrepresented in the sample. More than half of participants did not have any college education. Those without diabetes and whites were more likely to be married or living with a partner compared with others. Most participants had health insurance (74%) and this did not vary by diabetes status or race. Those with diabetes were more likely to be unemployed due to health reasons (31% vs 14%) and less likely to be currently employed full-time or part-time (34% vs 58%). Median annual household income was <$40 000 overall and <$20 000 among participants with diabetes. In terms of CVD risk factors, reported rates of hypertension were very high in the study (86% overall), which may be largely attributed to how participants were selected. Participants with diabetes were more likely to report a history of hypertension (98% compared with 79%) and African-Americans reported higher rates of hypertension than whites within each subgroup. Participants with diabetes also reported higher rates of prior coronary heart disease and CVD. At baseline, overall diet quality and fat quality were similar across diabetes status and race, while participants with diabetes had slightly higher fruit and vegetable consumption (average 3.7 compared with 3.3 servings/day). Participants without diabetes reported much higher total walking and total activity time at baseline (mean 105 and 171 min per week, respectively, compared with 67 and 112 min per week, respectively, among those with diabetes). Participants with diabetes weighed more on average than those without diabetes (103 kg compared with 95 kg). Systolic BP was similar across all categories, with a somewhat higher average for African-Americans than whites (138 mmHg compared with 131 mmHg) among participants with diabetes. Among those with diabetes, African-American participants had higher hemoglobin A1c percentage at baseline than whites (8.0% compared with 7.6%). Outcomes Figure 1 depicts the three sequential phases of the study and the number of participants with and without diabetes that took part in each component of the intervention and returned for follow-up. Follow-up rates at 6, 12, and 24 months were ∼75% and were similar between participants with and without diabetes. Lifestyle outcomes are shown in table 2. Table 2 Change in lifestyle outcomes by diabetes status and race from baseline to 6, 12, and 24 months Outcome N Phase 1 N Phase 2 N Phase 3 Baseline to 6 months Baseline to 12 months Baseline to 24 months Mean, 95% CI Mean, 95% CI Mean, 95% CI Dietary DRA total score 235 4.4 (3.7 to 5.0)*** 227 3.3 (2.5 to 4.0)*** 226 3.0 (2.3 to 3.6)*** Diabetes (all) 86 4.2 (3.1 to 5.3)*** 86 3.0 (1.8 to 4.2)*** 84 2.0 (1.0 to 3.1)***  African-American 60 4.4 (2.9 to 5.9)*** 63 3.8 (2.3 to 5.3)*** 62 2.5 (1.2 to 3.7)***  White 25 4.0 (2.6 to 5.4)*** 22 1.1 (−0.6 to 2.8) 21 0.7 (−1.3 to 2.7) No diabetes (all) 149 4.4 (3.6 to 5.3)*** 141 3.4 (2.5 to 4.4)*** 142 3.5 (2.7 to 4.3)***  African-American 95 4.7 (3.7 to 5.7)*** 93 4.1 (3.0 to 5.3)*** 96 3.3 (2.3 to 4.4)***  White 52 3.9 (2.5 to 5.4)*** 47 2.0 (0.3 to 3.7)* 45 3.7 (2.5 to 5.0)*** Fat quality screener score 229 1.4 (1.0 to 1.8)*** 225 1.0 (0.6 to 1.3)*** 224 0.7 (0.3 to 1.1)*** Diabetes (all) 84 1.3 (0.7 to 2.0)*** 85 1.2 (0.6 to 1.8)*** 83 0.3 (−0.3 to 1.0)  African-American 58 1.2 (0.4 to 2.0)** 62 1.5 (0.8 to 2.2)*** 61 0.4 (−0.3 to 1.1)  White 25 1.6 (0.4 to 2.7)** 22 0.4 (−0.3 to 1.2) 21 0.1 (−1.2 to 1.5) No Diabetes (all) 145 1.4 (1.0 to 1.9)*** 140 0.8 (0.4 to 1.3)*** 141 0.8 (0.4 to 1.3)***  African-American 92 1.5 (1.0 to 2.1)*** 92 1.1 (0.6 to 1.6)*** 95 0.8 (0.3 to 1.4)**  White 51 1.2 (0.5 to 2.0)** 47 0.4 (−0.4 to 1.2) 45 0.9 (−0.1 to 1.9) Fruit and vegetable servings per day 249 0.3 (0.1 to 0.5)* 253 0.5 (0.3 to 0.8)*** 250 0.4 (0.2 to 0.6)*** Diabetes (all) 93 −0.1 (−0.5 to 0.4) 98 0.2 (−0.3 to 0.7) 96 0.1 (−0.4 to 0.5)  African-American 67 −0.1 (−0.7 to 0.5) 75 0.2 (−0.4 to 0.8) 74 0.1 (−0.4 to 0.6)  White 25 0.1 (−0.4 to 0.7) 22 0.0 (−0.6 to 0.6) 21 0.1 (−0.6 to 0.7)  No Diabetes (all) 156 0.5 (0.3 to 0.7)*** 155 0.8 (0.5 to 1.1)*** 154 0.6 (0.3 to 0.9)***  African-American 101 0.5 (0.2 to 0.8)** 103 0.8 (0.4 to 1.1)*** 103 0.6 (0.2 to 0.9)**  White 53 0.5 (0.0 to 0.9)* 51 0.8 (0.3 to 1.2)*** 50 0.6 (0.2 to 1.1)** Summary score for drinks, desserts, snacks 236 1.1 (0.9 to 1.3)*** 229 1.3 (1.1 to 1.6)*** 228 1.1 (0.8 to 1.3)*** Diabetes (all) 87 1.1 (0.7 to 1.4)*** 87 1.2 (0.8 to 1.6)*** 85 0.9 (0.5 to 1.3)***  African-American 61 1.1 (0.7 to 1.6)*** 64 1.3 (0.8 to 1.7)*** 63 1.1 (0.7 to 1.6)***  White 25 1.0 (0.5 to 1.5)*** 22 1.1 (0.3 to 1.8)** 21 0.1 (−0.7 to 1.0) No diabetes (all) 149 1.1 (0.8 to 1.5)*** 142 1.4 (1.1 to 1.8)*** 143 1.2 (0.8 to 1.5)***   African-American 95 1.3 (0.9 to 1.8)*** 93 1.7 (1.2 to 2.1)*** 96 1.3 (0.9 to 1.8)***  White 52 0.7 (0.2 to 1.2)** 48 0.9 (0.4 to 1.4)*** 46 0.8 (0.3 to 1.3)** Physical activity Walking time, min/wk 249 64 (19 to 109)** 253 71 (28 to 113)** 250 22 (−13 to 56) Diabetes (all) 93 101 (17 to 184)* 98 126 (58 to 194)*** 96 62 (11 to 113)*  African-American 67 119 (18 to 221)* 75 136 (52 to 220)** 74 75 (13 to 137)*  White 25 71 (−81 to 223) 22 117 (14 to 219)* 21 38 (−36 to 112) No diabetes (all) 156 42 (−9 to 94) 155 36 (−17 to 89) 154 −3 (−49 to 42)   African-American 101 29 (−45 to 103) 103 19 (−53 to 91) 103 −17 (−82 to 47)  White 53 63 (8 to 118)* 51 68 (−3 to 138) 50 22 (−22 to 66) Total physical activity time, min/wk 249 97 (36 to 158)** 253 83 (30 to 136)** 250 48 (−7 to 103) Diabetes (all) 93 120 (19 to 221)* 98 109 (29 to 188)** 96 62 (−8 to 131)  African-American 67 136 (17 to 255)* 75 127 (32 to 223)** 74 89 (3 to 175)*  White 25 106 (−94 to 305) 22 74 (−63 to 212) 21 −5 (−91 to 81) No Diabetes (all) 156 83 (6 to 159)* 155 67 (−3 to 137) 154 40 (−39 to 118)  African-American 101 86 (−17 to 189) 103 37 (−42 to 116) 103 28 (−76 to 132)  White 53 79 (−33 to 192) 51 125 (−15 to 265) 50 58 (−55 to 171) Significance level: *p<0.05; **p<0.01; ***p<0.001. DRA, dietary risk assessment; min, minute; wk, week. Overall diet quality, as assessed by the DRA total score, improved by about 4 points at 6-month follow-up for all participants and each of the subgroups. Improvement in DRA total score was maintained at 12 and 24 months compared with baseline among all subgroups except white participants with diabetes. Fat quality score improved by 1.4 points on average at 6 months with no significant difference between subgroups by race or diabetes status, but this improvement was attenuated over time for participants with diabetes. Only participants without diabetes reported a statistically significant increase in fruit and vegetable servings per day at 6, 12, and 24-month follow-up. The improvement in the summary score for drinks, desserts, and snacks was higher among African-American participants with and without diabetes compared with whites. Increased walking time was sustained at 24 months among participants with diabetes and among African-Americans with diabetes. Physiological outcomes are shown in table 3. Table 3 Change in physiological outcomes by diabetes status and race from baseline to 6, 12, and 24 months Outcome n Phase 1 n Phase 2 n Phase 3 Baseline to 6 months Baseline to 12 months Baseline to 24 months Mean, 95% CI Mean, 95% CI Mean, 95% CI Systolic BP, mm Hg 249 −6.4 (−8.7 to −4.1)*** 251 −6.2 (−9.0 to −3.3)*** 250 −7.3 (−9.9 to −4.6)*** Diabetes (all) 93 −7.1 (−11.3 to −3.0)*** 97 −5.8 (−10.9 to −0.6)* 96 −7.4 (−11.9 to −3.0)**  African-American 67 −8.6 (−13.9 to −3.2)** 74 −6.4 (−12.7 to −0.1)* 74 −9.3 (−14.6 to −4.0)***  White 25 −3.8 (−9.6 to 2.0) 22 −4.1 (−13.2 to 5.0) 21 −0.4 (−8.7 to 8.0) No diabetes (all) 156 −5.9 (−8.7 to −3.2)*** 154 −6.4 (−9.7 to −3.1)*** 154 −7.1 (−10.5 to −3.8)***   African-American 101 −5.6 (−8.9 to −2.2)** 102 −7.3 (−11.6 to −3.1)*** 103 −7.9 (−12.2 to −3.5)***  White 53 −6.5 (−11.3 to −1.7)** 51 −4.3 (−9.5 to 0.8) 50 −5.6 (−10.7 to −0.5)* Diastolic BP, mm Hg 249 −3.7 (−4.9 to −2.5)*** 251 −5.0 (−6.4 to −3.6)*** 250 −6.7 (−8.3 to −5.2)*** Diabetes (all) 93 −4.4 (−6.4 to −2.3)*** 97 −5.6 (−8.2 to −3.0)*** 96 −7.2 (−10.1 to −4.3)***  African-American 67 −4.6 (−7.2 to −2.1)*** 74 −5.8 (−8.8 to −2.7)*** 74 −8.2 (−11.7 to −4.7)***  White 25 −4.1 (−7.6 to −0.7)* 22 −4.7 (−10.0 to 0.6) 21 −3.3 (−8.3 to 1.7) No diabetes (all) 156 −3.4 (−4.8 to −1.9)*** 154 −4.6 (−6.2 to −3.0)*** 154 −6.4 (−8.2 to −4.7)***  African-American 101 −2.8 (−4.8 to −0.9)** 102 −4.8 (−7.0 to −2.7)*** 103 −6.5 (−8.9 to −4.0)***  White 53 −4.3 (−6.5 to −2.1)*** 51 −3.9 (−6.2 to −1.6)*** 50 −6.3 (−8.5 to −4.1)*** Weight, kg 248 −0.7 (−1.2 to −0.3)** 250 −1.7 (−2.5 to −1.0)*** 247 −1.6 (−2.5 to −0.8)*** Diabetes (all) 92 −1.2 (−2.1 to −0.4)** 96 −1.5 (−2.9 to −0.2)* 93 −3.7 (−5.2 to −2.1)***  African-American 66 −1.1 (−2.0 to −0.3)** 73 −1.0 (−2.5 to 0.4) 71 −3.1 (−4.6 to −1.5)***  White 25 −1.6 (−4.0 to 0.7) 22 −3.3 (−6.9 to 0.4) 21 −5.5 (−9.9 to −1.2)* No diabetes (all) 156 −0.4 (−0.9 to 0.1) 154 −1.8 (−2.7 to −1.0)*** 154 −0.4 (−1.4 to 0.6)   African-American 101 −0.7 (−1.4 to 0.0) 102 −2.2 (−3.4 to −1.1)*** 103 −0.9 (−2.2 to 0.4)  White 53 0.2 (−0.5 to 0.8) 51 −1.0 (−2.3 to 0.4) 50 0.9 (−0.8 to 2.5) ≥5% weight loss, % 248 9.3 (5.6 to 12.9) 250 23.2 (17.9 to 28.5) 247 23.1 (17.8 to 28.4) Diabetes (all) 92 10.9 (4.5 to 17.3) 96 20.8 (12.7 to 29) 93 34.4 (24.7 to 44.1)   African-American 66 12.1 (4.1 to 20.2) 73 19.2 (10.0 to 28.4) 71 32.4 (21.3 to 43.5)  White 25 8.0 (0.0 to 18.8) 22 27.3 (8.3 to 46.2) 21 38.1 (16.9 to 59.3) No diabetes (all) 156 8.3 (4.0 to 12.7) 154 24.7 (17.8 to 31.5) 154 16.2 (10.4 to 22.1)   African-American 101 11.9 (5.5 to 18.3) 102 26.5 (17.8 to 35.1) 103 18.5 (10.9 to 26.0)  White 53 1.9 (0.0 to 5.6) 51 19.6 (8.6 to 30.6) 50 10.0 (1.6 to 18.4) HbA1c, %† 217 0.01 (−0.09 to 0.12) 220 −0.07 (−0.20 to 0.06) Diabetes (all) 80 −0.11 (−0.38 to 0.17) 84 −0.30 (−0.63 to 0.02)   African-American 60 −0.10 (−0.45 to 0.25) 66 −0.33 (−0.72 to 0.06)  White 19 −0.20 (−0.60 to 0.20) 17 −0.28 (−0.85 to 0.28) No diabetes (all) 137 0.08 (0.04 to 0.11)*** 136 0.07 (0.01 to 0.13)*   African-American 87 0.10 (0.05 to 0.14)*** 88 0.08 (0.01 to 0.16)*  White 48 0.05 (0.00 to 0.10) 47 0.05 (−0.03 to 0.13) HbA1c, mmol/mol† 217 0.1 (−1.0 to 1.3) 220 −0.8 (−2.2 to 0.7) Diabetes (all) 80 −1.2 (−4.2 to 1.9) 84 −3.3 (−6.9 to 0.2)   African-American 60 −1.1 (−4.9 to 2.7) 66 −3.6 (−7.9 to 0.7)  White 19 −2.2 (−6.6 to 2.2) 17 −3.1 (−9.3 to 3.1) No diabetes (all) 137 0.9 (0.4 to 1.2)*** 136 0.8 (0.1 to 1.4)*   African-American 87 1.1 (0.5 to 1.5)*** 88 0.9 (0.1 to 1.7)*  White 48 0.5 (0.0 to 1.1) 47 0.5 (−0.3 to 1.4) Total cholesterol, mg/dL‡ 221 −3.2 (−7.0 to 0.7) 211 −3.1 (−7.8 to 1.7) Diabetes (all) 84 0.4 (−6.7 to 7.6) 83 −4.2 (−13.1 to 4.6)   African-American 66 −0.4 (−8.3 to 7.6) 65 −7.1 (−17.5 to 3.4)  White 17 6.9 (−9.5 to 23.3) 17 9.1 (−6.6 to 24.8) No diabetes (all) 137 −5.4 (−9.8 to −1.1)* 128 −2.3 (−7.7 to 3.0)   African-American 89 −6.8 (−11.8 to −1.8)** 83 −2.4 (−8.9 to 4.2)  White 47 −2.5 (−11.0 to 6.0) 44 −2.2 (−11.6 to 7.2) HDL cholesterol, mg/dL‡ 220 −1.6 (−2.8 to −0.4)* 211 −1.0 (−2.3 to 0.3) Diabetes (all) 84 −2.0 (−4.0 to 0.0)* 83 −2.1 (−4.4 to 0.3)   African-American 66 −1.4 (−3.7 to 0.8) 65 −2.2 (−4.8 to 0.5)  White 17 −3.9 (−8.7 to 0.8) 17 −1.8 (−7.6 to 4.0) No diabetes (all) 136 −1.3 (−2.8 to 0.2) 128 −0.3 (−1.7 to 1.2)   African-American 88 −1.6 (−3.6 to 0.4) 83 −0.1 (−2.1 to 2.0)  White 47 −0.9 (−3.1 to 1.2) 44 −0.6 (−2.6 to 1.3) Significance level: *p<0.05; **p<0.01; ***p<0.001. †HbA1c was not measured at 24 months. ‡Lipids were not measured at 6 months. BP, blood pressure; HbA1c, glycated hemoglobin; HDL, high-density lipoprotein. Among all participants, there was a statistically significant reduction in systolic BP of about 6 to 7 mmHg across all follow-up time points. The reduction was similar for those with diabetes and larger for African-Americans with diabetes compared with whites. Outcomes for diastolic BP were similar. Of note, there was very little change in participants' use of blood measure medication at follow-up (data not shown). For those with diabetes, the number reporting use of BP medication was unchanged at 6-month, one less at 12-month, and two less at 24-month follow-up. For those without diabetes, use increased by one at 6, two at 12, and two at 24-month follow-up. There was a trend toward significant improvement in hemoglobin A1c at 12 months among participants with diabetes (−0.30% (95% CI −0.63 to 0.02)/−3.3 mmol/mol (95% CI −6.9 to 0.2), p=0.07). Overall, there was a small decrease in total and HDL cholesterol. Weight loss was a major outcome of interest. Participants with diabetes had significant, sustained, and progressive weight loss compared with baseline at 6, 12, and 24 months (mean −1.2 kg, −1.5 kg, and −3.7 kg, respectively) across all intervention groups. White participants with diabetes had greater weight loss than African-American participants with diabetes. Weight change for participants with diabetes at 12-month follow-up (data not shown) by intervention group selected at the start of the HHL weight program (start of Phase II) was: −3.9 kg (95% CI −7.4 to −0.4, p=0.03) for the group weight loss program (n=17); −2.6 kg (95% CI −5.0 to −0.2, p=0.03) for the combination weight loss program (n=27); and –0.2 kg (95% CI −2.0 to 1.6, p=0.81) for the maintenance of lifestyle intervention program (n=52). At 24-month follow-up (data not shown), it was −5.2 kg (95% CI −9.6 to −0.8, p=0.2) for the group weight loss program (n=18); −2.2 kg (95% CI −4.6 to 0.1, p=0.06) for the combination weight loss program (n=25); and –3.8 kg (95% CI −5.9 to −1.8, p<0.001) for the maintenance of lifestyle intervention program (n=50). After adjusting for age, race, sex, education, and baseline weight, only participants with diabetes had significant sustained weight loss at 24 months (p<0.0001) and participants with diabetes had significantly more weight loss on average than participants without diabetes (p=0.01). Figure 2 depicts weight change (loss and gain) at follow-up by diabetes status with cut points for weight change from baseline of ≥2.5, ≥5, and ≥7.5%. Consistent with mean weight loss, a greater proportion of participants with diabetes demonstrated substantially greater weight loss than gain at these time points, especially at 24 months. At 24 months, 18 participants (20%) lost more than 7.5% body weight while none gained this amount. For those without diabetes, the percentages of those who lost and gained weight at 24 months were similar. Figure 2 N and percentage of participants, by diabetes status, for weight change from baseline to follow-up at 6, 12, and 24 months. Weight change is shown for three cut points (≥2.5, ≥5, and ≥7.5% of baseline weight). N, number. From baseline to 24-month follow-up, 32 participants (34%) with diabetes lost at least 5% body weight compared with 25 participants (16%) without diabetes. Percentages for ≥5% weight loss were similar between African-Americans and whites with diabetes; however, more African-American participants without diabetes achieved ≥5% weight loss at 24 months than whites without diabetes (18% compared with 10%, respectively). Percentages of participants with diabetes who achieved ≥5% weight loss at 24 months, by Phase II intervention group, were (data not shown): 44% (95% CI 21 to 68) for the group weight loss program (n=18); 24% (95% CI 7 to 41) for the combination weight loss program (n=25); and 36% (95% CI 22 to 50) for the maintenance of lifestyle intervention program. At the conclusion of Phase I, 85 of 88 (97%) participants with diabetes and 151 of 154 (98%) without diabetes either strongly agreed or agreed that they would recommend the lifestyle program to others. After Phase II, among participants completing the acceptability survey with diabetes, all in the group weight loss program (n=12) and 16 of 19 (84%) in the combination weight loss program were satisfied or very satisfied with the intervention. Participants without diabetes were similarly satisfied. In addition, no adverse outcomes were attributed to the intervention. Conclusions This study evaluated a lifestyle and weight loss intervention promoting a Mediterranean-style diet in a way that was designed to be appealing to residents of the southeastern USA residing in the ‘stroke belt,’ where the population's CVD risk is very high. Though this type of dietary intervention has been carefully evaluated in Europe,7 it has not been adapted for and assessed in low-income and minority US populations. In this study, the intervention was well received by participants and, compared with baseline, there was improvement in self-reported lifestyle behaviors, BP, and weight, with substantially greater sustained weight loss observed for participants with diabetes. Among participants with diabetes, our intervention did not achieve the same degree of weight loss documented in the Diabetes Prevention Program (DPP) study,34 which was > 5 kg at 24 months (20% of participants were African-American, average age 51). However, in the small subgroup of participants (n=25) who selected the group-based weight loss option, weight loss at 24 months was >5 kg, as it was for all of the white participants with diabetes. In essence, among participants selecting the group-based weight loss format, weight loss at 24 months in our low-income and high-minority population was similar to that observed in the DPP. Moreover, weight loss among all participants with diabetes was greater than in most weight loss studies enrolling disadvantaged population groups.4 Different from the PREDIMED study, which reported weight loss <0.5 kg at 24 months follow-up in all study arms (all participants, not stratified by diabetes status),35 participants in our study with diabetes who selected the lifestyle only option (which focused on diet quality and physical activity, but not weight loss) lost 3.8 kg at 24-month follow-up. A possible explanation for weight loss in this group of our study, as compared with PREDIMED, is difference in baseline diet, with more poor quality carbohydrates and processed food in the American diet compared with the European diet36 and particularly so in the southeastern USA.37–40 In this setting of excess intake of poor quality carbohydrates, a change to a Mediterranean-style, unrestricted fat diet, may have positive metabolic and appetite-suppressing effects, as recently outlined by Ludwig.41 Although the dietary pattern advocated in this study was similar to the PREDIMED intervention arms (especially the nut intervention arm),7 we recommended regular consumption of vegetable oils (high in polyunsaturated and monounsaturated fats) as opposed to four tablespoons of extra virgin olive oil per day as tested in the PREDIMED olive oil intervention arm. In this regard, it is worth noting a recent publication assessing outcomes in two large cohort studies which indicated diets higher in polyunsaturated and monounsaturated fats are associated with a reduction in CVD mortality and total mortality.26 As illustrated in figure 2, among all participants with diabetes, including the majority who did not elect to take part in the weight loss intervention offered in Phase II of the intervention, weight loss increased steadily over the 24-month study period. This weight loss began during Phase I, which focused on diet quality and physical activity, but not specifically on weight loss, and continued at about the same rate during Phases II and III. The observed pattern of weight loss in this study was different from that observed in most other weight loss studies, which is usually greatest weight loss in the short term (6–12 months) with attenuation of weight loss beyond 1 year.42 In a recent meta-analysis of weight loss among named weight loss diet programs, only one of 10 programs achieved greater weight loss at 12 months than at 6 months.6 Even in programs with intensive interventions that include maintenance of weight loss components such as the DPP43 and Look AHEAD,8 there was attenuation in weight loss at 18 and 24 months, respectively. A possible explanation for the observed sustained weight loss in our study was that all participants received some level of intervention over 24 months and unlike most weight loss studies, which try to achieve weight loss over 6 months, our approach was to focus on weight loss over a longer time frame. Additionally, the Mediterranean dietary pattern may have contributed to the sustained weight loss over 24 months, as this dietary pattern has previously been shown to be associated with maintenance of weight loss.44 This study has several limitations including a pre–post design without a control group. While the observed changes may be due to the intervention, they could also be due to other factors, including secular trends. However, sustained weight loss among adults is uncommon. For example, in the control groups of weight loss RCTs such as DPP and Look AHEAD, average weight loss at 24-month follow-up was <1kg.8 34 With regard to secular trends in North Carolina during the time frame of this study (2011 to 2014), there were no changes in the rates of overweight and obesity as assessed by the Centers for Disease Control and Prevention's behavioral risk factor surveillance system.45 Another major limitation is the sample size of this study, especially with regard to outcomes by intervention groups (group weight loss, combination weight loss, or maintenance of lifestyle intervention) and race. Further, because weight loss was less than expected at the end of Phase II, as discussed in detail elsewhere,11 the sample size for the ‘embedded’ RCT of weight loss maintenance was so small that we did not undertake a formal analysis of outcomes for this RCT. Another limitation is that lifestyle outcomes were self-reported and may have been exaggerated due to social desirability reporting bias. Finally, our findings may not be generalizable to populations different from the sample enrolled from one community in eastern North Carolina. The cost-effectiveness of an intervention is a very important consideration for community-based and clinic-based weight loss programs. Without a control group, we did not undertake a cost-effectiveness analysis, which is another limitation of this study. However, we have previously reported a cost-effectiveness analysis for the group-based intervention format that we tested in Phase II, comparing it with a delayed intervention control group.46 In that study, in which the weight loss intervention was considered cost-effective (assuming weight loss could be sustained over time), weight loss at 6-month follow-up was 3.7 kg in the intervention group, very similar to the weight loss of 3.9 kg observed at the completion of Phase II for our participants with diabetes for our participants with diabetes who received the group-based intervention. The study has several strengths. First, is the relatively unselected sample (few exclusion criteria were applied and no run-in period) which enhances its generalizability; second, it used a design that mimicked real-world situations, allowing participants to choose between two weight loss intervention formats or to focus on lifestyle change without weight loss as a goal; third, the follow-up was 74% at 24 months, which is a larger follow-up percentage and longer follow-up interval than reported in many weight loss studies; and fourth, physiological outcomes were obtained using standardized objective measures. In this study, with a relatively unselected sample, largely minority and of lower socioeconomic status, the tested Mediterranean-style dietary pattern, a pattern associated with substantial reduction in CVD risk7 was very well received. Among participants with diabetes, there was sustained improvement in self-reported lifestyle behaviors, BP, and weight change at 24-month follow-up. Though the study did not have a control group, our findings confirm the acceptability of a Mediterranean-style dietary pattern among this very high-risk population and suggest that this dietary pattern may be associated with sustained weight loss. Further study of interventions promoting this dietary pattern is warranted in high-risk US populations with diabetes, including RCTs that assess intermediate outcomes and CVD events. We give special thanks to our Community Advisory Committee who provided helpful guidance with this project and to our study participants, whose willing participation made this study possible. Contributors: GGRE and TCK drafted the manuscript. CDS-H designed the study and developed the weight loss intervention. LFJ and ZG had full access to all the data and carried out the analysis. BAG supervised data collection. KRE developed the physical activity intervention. DAD acquired funding. ASA and TCK acquired funding and designed the study. All authors revised the manuscript and approved the final draft. TCK is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Funding: This research was supported by National Institutes of Health (NIH) grant 5P50 HL105184 to the University of North Carolina Center for Health Promotion and Disease Prevention (HPDP) with subcontract to the Brody School of Medicine, East Carolina University. Other support was provided by Centers for Disease Control and Prevention (CDC) cooperative agreement No. U48/DP001944 to HPDP (a CDC Prevention Research Center). Competing interests: None declared. Ethics approval: University of North Carolina at Chapel Hill biomedical IRB. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: The data sets generated and/or analyzed during the current study are being prepared for web access on the BioLINCC repository, available at: https://biolincc.nhlbi.nih.gov/home/. Until available on this repository, the data are available from the corresponding author on reasonable request. ==== Refs References 1 American Diabetes Association . Foundations of care and comprehensive medical evaluation. Sec. 3 in standards of medical care in diabetes–2016 . Diabetes Care 2016 ;39 (Suppl 1 ):S23 –35 . 10.2337/dc16-S006 26696676 2 American Diabetes Association . Foundations of care: education, nutrition, physical activity, smoking cessation, psychosocial care, and immunization. Sec. 4 in standards of medical care in diabetes–2015 . Diabetes Care 2015 ;38 ():S20 –30 . 10.2337/dc15-S007 25537702 3 Franz MJ , Boucher JL , Rutten-Ramos S Lifestyle weight-loss intervention outcomes in overweight and obese adults with type 2 diabetes: a systematic review and meta-analysis of randomized clinical trials . J Acad Nutr Diet 2015 ;115 :1447 –63 . 10.1016/j.jand.2015.02.031 25935570 4 Harvey JR , Ogden DE Obesity treatment in disadvantaged population groups: where do we stand and what can we do? Prev Med 2014 ;68 :71 –5 . 10.1016/j.ypmed.2014.05.015 24878585 5 Wadden TA , West DS , Neiberg RH One-year weight losses in the Look AHEAD study: factors associated with success . Obesity (Silver Spring) 2009 ;17 :713 –22 . 10.1038/oby.2008.637 19180071 6 Johnston BC , Kanters S , Bandayrel K Comparison of weight loss among named diet programs in overweight and obese adults: a meta-analysis . JAMA 2014 ;312 :923 –33 . 10.1001/jama.2014.10397 25182101 7 Estruch R , Ros E , Salas-Salvado J Primary prevention of cardiovascular disease with a Mediterranean diet . N Engl J Med 2013 ;368 :1279 –90 . 10.1056/NEJMoa1200303 23432189 8 Wing RR , Bolin P , Brancati FL , Look Ahead Research Group . Cardiovascular effects of intensive lifestyle intervention in type 2 diabetes . N Engl J Med 2013 ;369 :145 –54 . 10.1056/NEJMoa1212914 23796131 9 Howard G , Labarthe DR , Hu J Regional differences in African Americans’ high risk for stroke: the remarkable burden of stroke for Southern African Americans . Ann Epidemiol 2007 ;17 :689 –96 . 10.1016/j.annepidem.2007.03.019 17719482 10 Pitts SB , Vu MB , Garcia BA A community assessment to inform a multilevel intervention to reduce cardiovascular disease risk and risk disparities in a rural community . Fam Community Health 2013 ;36 :135 –46 . 10.1097/FCH.0b013e31828212be 23455684 11 Keyserling TC , Samuel-Hodge CD , Pitts SJ A community-based lifestyle and weight loss intervention promoting a Mediterranean-style diet pattern evaluated in the stroke belt of North Carolina: the Heart Healthy Lenoir Project . BMC Public Health 2016 ;16 :732 10.1186/s12889-016-3370-9 27495295 12 Cooper LA , Ortega AN , Ammerman AS Calling for a bold new vision of health disparities intervention research . Am J Public Health 2015 ;105 (Suppl 3 ):S374 –6 . 10.2105/AJPH.2014.302386 25905830 13 Centers for Disease Control and Prevention . Interactive Atlas of Heart Disease and Stroke. http://nccd.cdc.gov/DHDSPAtlas/ (accessed 29 Aug 2016 ). 14 Mozaffarian D , Benjamin EJ , Go AS , Writing group members . Heart disease and stroke statistics-2016 update: a report from the American Heart Association . Circulation 2016 ;133 :e38 –360 . 10.1161/CIR.0000000000000350 26673558 15 Keyserling TC , Samuel Hodge CD , Jilcott SB Randomized trial of a clinic-based, community-supported, lifestyle intervention to improve physical activity and diet: the North Carolina enhanced WISEWOMAN project . Prev Med 2008 ;46 :499 –510 . 10.1016/j.ypmed.2008.02.011 18394692 16 Samuel-Hodge CD , Johnston LF , Gizlice Z Randomized trial of a behavioral weight loss intervention for low-income women: the Weight Wise Program . Obesity (Silver Spring) 2009 ;17 :1891 –9 . 10.1038/oby.2009.128 19407810 17 Kraschnewski JL , Keyserling TC , Bangdiwala SI Optimized probability sampling of study sites to improve generalizability in a multisite intervention trial . Prev Chronic Dis 2010 ;7 :A10 .20040225 18 Samuel-Hodge CD , Garcia BA , Johnston LF Translation of a behavioral weight loss intervention for mid-life, low-income women in local health departments . Obesity (Silver Spring) 2013 ;21 :1764 –73 . 10.1002/oby.20317 23408464 19 Halladay JR , Donahue KE , Hinderliter AL The Heart Healthy Lenoir project–an intervention to reduce disparities in hypertension control: study protocol . BMC Health Serv Res 2013 ;13 :441 10.1186/1472-6963-13-441 24156629 20 Ammerman AS , Haines PS , DeVellis RF A brief dietary assessment to guide cholesterol reduction in low-income individuals: design and validation . J Am Diet Assoc 1991 ;91 :1385 –90 .1939975 21 Ammerman AS , DeVellis BM , Haines PS Nutrition education for cardiovascular disease prevention among low income populations: description and pilot evaluation of a physician-based model . Patient Educ Couns 1992 ;19 :5 –18 . 10.1016/0738-3991(92)90098-4 1298949 22 Jakobsen MU , O'Reilly EJ , Heitmann BL Major types of dietary fat and risk of coronary heart disease: a pooled analysis of 11 cohort studies . Am J Clin Nutr 2009 ;89 :1425 –32 . 10.3945/ajcn.2008.27124 19211817 23 Mozaffarian D , Micha R , Wallace S Effects on coronary heart disease of increasing polyunsaturated fat in place of saturated fat: a systematic review and meta-analysis of randomized controlled trials . PLoS Med 2010 ;7 :e1000252 10.1371/journal.pmed.1000252 20351774 24 Siri-Tarino PW , Sun Q , Hu FB Meta-analysis of prospective cohort studies evaluating the association of saturated fat with cardiovascular disease . Am J Clin Nutr 2010 ;91 :535 –46 . 10.3945/ajcn.2009.27725 20071648 25 Sofi F , Abbate R , Gensini GF Accruing evidence on benefits of adherence to the Mediterranean diet on health: an updated systematic review and meta-analysis . Am J Clin Nutr 2010 ;92 :1189 –96 . 10.3945/ajcn.2010.29673 20810976 26 Wang DD , Li Y , Chiuve SE Association of specific dietary fats with total and cause-specific mortality . JAMA Intern Med 2016 ;176 :1134 –45 . 10.1001/jamainternmed.2016.2417 27379574 27 Samuel-Hodge CD , Garcia BA , Johnston LF Rationale, design, and sample characteristics of a practical randomized trial to assess a weight loss intervention for low-income women: The Weight-Wise II Program . Contemp Clin Trials 2012 ;33 :93 –103 . 10.1016/j.cct.2011.08.009 21930244 28 Steinberg DM , Tate DF , Bennett GG Daily self-weighing and adverse psychological outcomes: a randomized controlled trial . Am J Prev Med 2014 ;46 :24 –9 . 10.1016/j.amepre.2013.08.006 24355668 29 Jilcott SB , Keyserling TC , Samuel-Hodge CD Validation of a brief dietary assessment to guide counseling for cardiovascular disease risk reduction in an underserved population . J Am Diet Assoc 2007 ;107 :246 –55 . 10.1016/j.jada.2006.11.006 17258961 30 Block G , Gillespie C , Rosenbaum EH A rapid food screener to assess fat and fruit and vegetable intake . Am J Prev Med 2000 ;18 :284 –8 . 10.1016/S0749-3797(00)00119-7 10788730 31 Kraschnewski JL , Gold AD , Gizlice Z Development and evaluation of a brief questionnaire to assess dietary fat quality in low-income overweight women in the southern United States . J Nutr Educ Behav 2013 ;45 :355 –61 . 10.1016/j.jneb.2012.10.008 23340242 32 Giles-Corti B , Timperio A , Cutt H Development of a reliable measure of walking within and outside the local neighborhood: RESIDE's Neighborhood Physical Activity Questionnaire . Prev Med 2006 ;42 :455 –9 . 10.1016/j.ypmed.2006.01.019 16574208 33 Jones SA , Evenson KR , Johnston LF Psychometric properties of the modified RESIDE physical activity questionnaire among low-income overweight women . J Sci Med Sport 2014 ;18 :37 –42 . 10.1016/j.jsams.2013.12.007 24462117 34 Knowler WC , Barrett-Connor E , Fowler SE Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin . N Engl J Med 2002 ;346 :393 –403 . 10.1056/NEJMoa012512 11832527 35 Estruch R , Martinez-Gonzalez MA , Corella D Effect of a high-fat Mediterranean diet on bodyweight and waist circumference: a prespecified secondary outcomes analysis of the PREDIMED randomised controlled trial . Lancet Diabetes Endocrinol 2016 ;4 :666 –76 . 10.1016/S2213-8587(16)30085-7 27283479 36 Kearney J Food consumption trends and drivers . Philos Trans R Soc Lond, B, Biol Sci 2010 ;365 :2793 –807 . 10.1098/rstb.2010.0149 20713385 37 Centers for Disease Control and Prevention . Fruit and vegetable consumption among adults–United States, 2005 . MMWR Morb Mortal Wkly Rep 2007 ;56 :213 –7 .17363889 38 Blanck HM , Gillespie C , Kimmons JE Trends in fruit and vegetable consumption among U.S men and women, 1994–2005 . Prev Chronic Dis 2008 ;5 :A35 .18341771 39 Krebs-Smith SM , Cook A , Subar AF US adults’ fruit and vegetable intakes, 1989 to 1991: a revised baseline for the Healthy People 2000 objective . Am J Public Health 1995 ;85 : 1623 –9 . 10.2105/AJPH.85.12.1623 7503335 40 Serdula MK , Gillespie C , Kettel-Khan L Trends in fruit and vegetable consumption among adults in the United States: behavioral risk factor surveillance system, 1994–2000 . Am J Public Health 2004 ;94 :1014 –8 . 10.2105/AJPH.94.6.1014 15249308 41 Ludwig DS Always hungry? Conquer cravings, retrain your fat cells, and lose weight permanently . New York : Grand Central Publishing , 2016 . 42 Heymsfield SB , Wadden TA Mechanisms, pathophysiology, and management of obesity . N Engl J Med 2017 ;376 :254 –66 . 10.1056/NEJMra1514009 28099824 43 Knowler WC , Fowler SE , Hamman RF Diabetes Prevention Program Research Group . 10-year follow-up of diabetes incidence and weight loss in the Diabetes Prevention Program Outcomes Study . Lancet 2009 ;374 :1677 –86 . 10.1016/S0140-6736(09)61457-4 19878986 44 Shai I , Schwarzfuchs D , Henkin Y Weight loss with a low-carbohydrate, Mediterranean, or low-fat diet . N Engl J Med 2008 ;359 :229 –41 . 10.1056/NEJMoa0708681 18635428 45 North Carolina State Center for Health Statistics . Annual Survey Results: Behavioral Risk Factor Surveillance System (BRFSS). http://www.schs.state.nc.us/data/brfss/survey.htm. (accessed 7 Apr 2016 ). 46 Gustafson A , Khavjou O , Stearns SC Cost-effectiveness of a behavioral weight loss intervention for low-income women: the Weight-Wise Program . Prev Med 2009 ;49 :390 –5 . 10.1016/j.ypmed.2009.09.007 19747937

PMC005xxxxxx/PMC5372066.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01438710.1136/bmjopen-2016-014387Health EconomicsResearch150617011698Economic evaluation of aerobic exercise training in older adults with vascular cognitive impairment: PROMoTE trial Davis Jennifer C 12Hsiung Ging-Yuek Robin 3Bryan Stirling 4Best John R 12Eng Janice J 1Munkacsy Michelle 12Cheung Winnie 12Chiu Bryan 12Jacova Claudia 3Lee Philip 5Liu-Ambrose Teresa 12 1 Department of Physical Therapy, University of British Columbia, Vancouver, British Columbia, Canada 2 Centre for Hip Health and Mobility, Vancouver Coastal Research Institute, Vancouver, British Columbia, Canada 3 Department of Medicine, Division of Neurology, University of British Columbia, Vancouver, British Columbia, Canada 4 Centre for Clinical Epidemiology and Evaluation, University of British Columbia, Vancouver, British Columbia, Canada 5 Department of Medicine, Division of Geriatric Medicine, University of British Columbia, Vancouver, British Columbia, CanadaCorrespondence to Dr Jennifer C Davis; Jennifer.davis@ubc.ca2017 29 3 2017 7 3 e01438721 9 2016 9 2 2017 24 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Background/objectives Evidence suggests that aerobic exercise may slow the progression of subcortical ischaemic vascular cognitive impairment (SIVCI) by modifying cardiovascular risk factors. Yet the economic consequences relating to aerobic training (AT) remain unknown. Therefore, our primary objective was to estimate the incremental cost per quality-adjusted life years (QALYs) gained of a thrice weekly AT intervention compared with usual care. Design Cost–utility analysis alongside a randomised trial. Setting Vancouver, British Columbia, Canada. Participants 70 adults (mean age of 74 years, 51% women) who meet the diagnostic criteria for mild SIVCI. Intervention A 6-month, thrice weekly, progressive aerobic exercise training programme compared with usual care (CON; comparator) with a follow-up assessment 6 months after formal cessation of aerobic exercise training. Measurements Healthcare resource usage was estimated over the 6-month intervention and 6-month follow-up period. Health status (using the EQ-5D-3L) at baseline and trial completion and 6-month follow-up was used to calculate QALYs. The incremental cost–utility ratio (cost per QALY gained) was calculated. Results QALYs were both modestly greater, indicating a health gain. Total healthcare costs (ie, 1791±1369 {2015 $CAD} at 6 months) were greater, indicating a greater cost for the thrice weekly AT group compared with CON. From the Canadian healthcare system perspective, the incremental cost–utility ratios for thrice weekly AT were cost-effective compared with CON, when using a willingness to pay threshold of $CAD 20 000 per QALY gained or higher. Conclusions AT represents an attractive and potentially cost-effective strategy for older adults with mild SIVCI. Trial registration number NCT01027858. mild cognitive impairmentcost-utility analysiseconomic evaluationolder adultsexerciseaerobic training ==== Body Strengths and limitations of this study Our study is one of the first to investigate together the economic and health consequences relating to aerobic training. Very few randomised controlled trials with concurrent economic evaluations of exercise have been conducted in populations at risk for dementia such as those with vascular cognitive impairment. There was wide variability in the cost estimates and outliers due to a smaller sample size. Introduction Cerebrovascular disease is the second most common aetiology contributing to dementia in older adults1–4 and may be the most underdiagnosed and yet most treatable form of cognitive dysfunction in older adults.5 Vascular cognitive impairment (VCI)—defined as the loss of cognitive function due to vascular burden in the brain—is a prevalent condition that places a growing burden on the healthcare system.6 Cerebral small vessel disease plays a critical role in covert ischaemia and the development of subcortical ischaemic vascular cognitive impairment (SIVCI),7 the most common form of VCI_.8 SIVCI is defined by the presence of white matter lesions (WMLs) and lacunar infarcts and has the clinical consequence of increased dementia risk.8 9 Research has demonstrated that one-third of all dementias are attributable to VCI.10–12 More specifically, the proportion of vascular dementia attributable to small-vessel disease ranges from 36 to 67%.13 14 The worldwide economic burden of dementia is increasing at an unprecedented rate. In 2015, a 35% increase led to a worldwide annual estimate of 818 billion US dollars. The worldwide costs of dementia are expected to exceed 1 trillion US dollars by 2018.15 Notably, vascular dementia has among the highest annual direct costs and highest hospitalisation-related costs compared with other dementias such as Alzheimer’s disease.16 The average annual cost per patient with VCI was $33 740,6 compared with a variable range of $1500–$91 000 for Alzheimer's disease. The costs per VCI admission were ∼$9545 with the average number of admissions increasing through the progression of the disease.6 Epidemiological data suggest that modification of vascular risk factors may be beneficial in slowing the progression of VCI.17–20 Hence, aerobic-based exercise training is one promising approach to delay the progression of VCI by reducing key vascular risk factors associated with metabolic syndrome. What remains unknown is whether aerobic-based exercise training as an intervention strategy compared with ‘usual care’ for individuals with mild SIVCI is a cost-effective strategy. Until now, the simultaneous impact of healthcare costs and consequences remains unknown. It is an essential next step to provide an estimate of the costs and consequences (ie, health gains or losses) related to the aerobic training (AT) intervention, given that this type of intervention could be delivered at a population level and thus have an immense impact. Therefore, we conducted a concurrent economic evaluation with individual level data on cost and effectiveness outcomes collected during a proof-of-concept single-blinded randomised controlled trial—the Promotion Of the Mind Through Exercise (PROMoTE) trial.21 Our primary objective was to determine the incremental cost–utility ratio (incremental cost per incremental quality-adjusted life year (QALY) gained) of thrice weekly AT compared with usual care among individuals with mild SIVCI. Methods Overview of economic evaluation This cost–utility analysis was conducted concurrently with a 6-month proof-of-concept single-blinded randomised controlled trial with a 6-month follow-up study (ie, 6 months postintervention).21 22 The details of the PROMoTE trial were previously reported.21 22 Measurements were made at three times: baseline, end of the intervention period (6 months postrandomisation) and 6-month postintervention (ie, 12 months postrandomisation). Of 440 individuals screened for eligibility, 70 were deemed eligible for this economic evaluation. This economic evaluation used a Canadian healthcare system perspective, and a 6-month (ie, trial completion) and a 12-month (ie, 6-month postintervention) time horizon for the primary economic evaluation assessing the efficiency of the thrice weekly progressive AT and usual care plus education compared with the usual care plus education (CON; comparator) group. Participants in the CON group received usual care as well as monthly educational materials about VCI and healthy diet. However, no specific information regarding physical activity was provided. Briefly, usual care included whatever healthcare services a patient with mild SIVCI would usually receive in their clinical care. The main outcome for the cost–utility analysis was the incremental cost per QALY gained. We obtained approval for this study from the University of British Columbia Clinical Ethics Review Board (H13-00715). We previously described study design, participant recruitment, randomisation, demographics, methods and results of the PROMoTE trial.21 We recruited participants from the University of British Columbia Hospital Clinic for AD and related disorders, the Vancouver General Hospital Stroke Prevention Clinic and specialised geriatric clinics in Metro Vancouver, BC. Recruitment occurred between December 2009 and April 2014 with randomisation occurring on an ongoing basis. The assessors were blinded to the participants' group allocation. The primary outcome measures for the PROMoTE study were the Alzheimer's Disease Assessment Scale Cognitive Subscale (ADAS-Cog23), the Executive Interview (EXIT-2524) and the Alzheimer's Disease Co-operative Study—Activities of Daily Living (ADCS-ADL25). Secondary outcome measures included executive functions, cardiovascular capacity, physical activity level, physiological markers and health-related quality of life. We included 70 community dwelling older adults who were diagnosed with SIVCI,26 which requires the presence of cognitive syndrome21 and small vessel ischaemic disease.21 Other inclusion criteria included: (1) Montreal Cognitive Assessment (MoCA)27 score <26 at screening; (2) Mini-Mental State Examination (MMSE)28 score of ≥20 at screening; (3) community-dwelling; (4) live in Metro Vancouver; (5) had a caregiver, family member or friend who interacted with him/her on a weekly basis; (6) sufficient ability to read, write and speak English; (7) acceptable visual and auditory acuity to complete psychometric tests; (8) stable on a fixed dose of cognitive medications that is not expected to change during the 6-month intervention period; (9) provided a personally signed and dated informed consent document indicating that the individual (or a legally acceptable representative) has been informed of all pertinent aspects of the trial; (10) able to walk independently and (11) in sufficient health to participate in the study’s aerobic-based exercise training programme. Costs We tracked healthcare resource usage prospectively. Our primary method used cost diaries where participants were asked to fill out a monthly diary detailing any health resource usage. We also telephoned participants every 3 months using a health resource usage questionnaire. For individuals who did not fill out their calendars, the health resource usage questionnaire was the primary mode of healthcare resource usage data collection. For participants who missed the 3-month follow-up telephone call and who did not return their calendar, they were asked to recall their healthcare resource usage over the 6-month intervention period. We also collected healthcare resource usage for the 6-month follow-up period postintervention. We analyse these end points separately (ie, trial completion at 6 months and follow-up completion at 1 year). The healthcare resource usage questionnaire included the following categories: any visits to healthcare professionals (including general practitioners, specialists, physiotherapists, etc); all visits, admissions or procedures carried out in a hospital and laboratory and diagnostic tests. We calculated the costs of delivering the thrice weekly AT intervention and the CON group. Our base case analysis considered the costs of all healthcare resource use. Research protocol-driven costs were excluded from our analysis. A unit cost was assigned for each component of healthcare resource usage. Costs for admission to hospital were based on the fully allocated cost model of a tertiary care hospital, Vancouver General Hospital. We based costs on fee for service rates from the British Columbia Medical Services Plan 2013 price list for all healthcare professional-related costs. Unit costs for specialised services (ie, physiotherapy, chiropractic or naturopathic medicine) were taken from the BC Association website for each specialty. We inflated costs to 2015 Canadian dollars using the consumer price index reported by Statistics Canada. Given that our analytic time horizon was ≤12 months, discounting was not applied. Effectiveness outcome Briefly, we assessed health status using the EQ-5D-3L.29 The EQ-5D-3L is a short five-item multiple choice questionnaire that measures an individual's health-related quality of life (HRQoL) and health status according to the following five domains: mobility, self-care, usual activities, pain and anxiety/depression.29 Each domain has three possible options: no problems, some problems or severe problems. The EQ-5D-3L health state utility values (HSUVs) at each time point are bounded from −0.54 to 1.00 where a score of <0 is indicative of a health state worse than death. The HSUVs represent values that individuals within society assign––these are Canadian societal values for given health states.30 We administered the EQ-5D-3L at baseline, trial completion and at the 6-month follow-up period to patients and a patient proxy (ie, see below under ‘Caregiver’). From these data points, we calculated the total QALYs lost or gained at 6 (trial completion) and 12 (follow-up completion) months for the two experimental groups. We used multiple linear regression to calculate the incremental QALYs based on patient and proxy ratings for each participant adjusted for the baseline utility score. Baseline utility scores are often imbalanced between treatment arms. Given that a patient's utility score at baseline is most often highly correlated with that individual's QALYs over the study period, failure to control for this imbalance can lead to a misleading ICER. As such, we followed the recommendations of Manca and Palmer31 using multiple linear regression to control for imbalances in baseline utility scores between the two treatment groups. All statistical analyses were carried out using STATA V.10.0. Caregiver (proxy) The caregivers had to be able to read, write and speak English in which the questionnaires were provided with acceptable visual and auditory acuity. Caregivers completed the EQ-5D-3L from their own perspective of the participant (ie, proxy's own perspective). Adverse events and mortality Participants were advised to report any adverse effects due to the intervention. Our safety monitoring committee reviewed all adverse events on a monthly basis. Handling missing data In the PROMoTE study, 17% of participants had incomplete 6-month health resource usage data and 7% had incomplete 6-month EQ-5D-3L data at trial completion including dropouts. For the 6-month follow-up period, 19% of participants had incomplete 6-month follow-up health resource usage data and 19% had incomplete 6-month follow-up EQ-5D-3L data. The reasons for missing data included: drop out, participant burden and administration error. We calculated the cost and effectiveness estimates for available cases (dropping observations with missing values), complete case sets and an imputed data set. We examined the pattern of missing data using the STATA code: ‘mvpatterns’. Missing data appeared to be missing at random, and therefore, we imputed missing data using Bayesian analyses following recommendations,31–34 35 in which all baseline study variables (including treatment assignment) were used to create 40 imputed data sets; parameter estimates and SEs were pooled across the 40 data sets. For multiple imputation, we used the ‘mi imput mvn’ procedure in STATA. The imputed data are reported as our base case analysis. We report the results using deletion of missing data as our sensitivity analysis (ie, complete case analysis). Cost–utility analysis We calculated the incremental cost–utility ratio for thrice weekly AT compared with the CON group twice using the patient-rated EQ-5D-3L and the caregiver proxy-rated EQ-5D-3L. Briefly, the incremental cost–utility ratio provides an index of the cost per QALY gained at intervention completion (ie, 6 months) and cost per QALY gained at 6-month postintervention (ie, 12 months). The incremental cost–utility ratio is the ratio between the difference in total mean costs between the AT and the CON groups and the difference in the mean QALY gained between the AT and the CON groups. 1 Nested imputation and nonparametric bootstrapping were used to model uncertainty around the estimates for costs and effectiveness. For each of the 40 cycles, we imputed missing values and bootstrapped the complete data set. For each cycle of imputation and bootstrapping, we calculated the total healthcare resource use cost and QALYs according to group allocation. The results of each cycle of imputation for participants were averaged in each of the two participant groups. The contribution of each cost item in relation to the total healthcare resource use was estimated for each group. Plots of the cost-effectiveness plane and cost-effectiveness acceptability curves were generated based on 5000 iterations of nested imputation/bootstrapping using Fiellers' method to generate 95% confidence ellipses for the joint distribution of cost and effectiveness outcomes.36 The differences in mean costs and health outcomes in each group were expressed by reporting the incremental cost per QALY (ie, the incremental cost–utility ratio). The observed health benefit (ie, QALY) difference was close to zero; therefore, we used 5000 bootstrapped replications of mean cost and QALY differences.37 We used these values to generate cost–utility acceptability curves to estimate the probability that thrice weekly AT is considered cost-effective compared with CON over a select range of willingness to pay values.38 Sensitivity analysis For our sensitivity analysis, we restricted our data to a complete case analysis, thus including only participants for whom we had complete cost and effectiveness data. We applied multiple imputation, bootstrapped CI estimation, adjustment for imbalances in baseline utility and bootstrapped estimates of the incremental cost–utility ratios. Results Baseline characteristics and exercise compliance Seventy-one eligible participants were randomised to AT or CON. One participant was deemed ineligible due to the presence of mixed dementia detected after randomisation and was excluded from all analyses. As such, our analytic sample consisted of 70 participants. Table 1 provides the baseline descriptive characteristics separated by the study group. Average class attendance was 68% for the AT group. Table 1 Baseline characteristics of participants CON group n=35 AT group n=35 Variables at baseline Mean (SD) or n (%) or median (IQR) Mean (SD) or n (%) or median (IQR) Descriptive variables and covariates  Age, years 73.7 (8.3) 74.8 (8.4)  Gender, female 17 (49%) 19 (54%)  Education, >high school 27 (82%) 24 (69%)  Functional Comorbidity Index 2.8 (2.2) 2.8 (1.5)  Hypertensive, yes 20 (61%) 17 (49%)  Mini-mental state examination 26.4 (3.1) 26.3 (2.7)  Montreal cognitive assessment 21.7 (4.4) 20.7 (3.3)  Waist-to-hip ratio 0.93 (0.07) 0.88 (0.08)  Short physical performance battery 10.51 (1.20) 10.62 (1.86)  Time-up-and-go (s) 8.67 (2.26) 8.82 (2.36)  Physiological profile assessment 0.94 (1.42) 0.94 (1.39) Medications  Taking beta blockers, yes 7 (20%) 7 (20%)  Central-effecting medications, no 0.5 (1.0) 0.6 (0.9)  Total medications, no 4.2 (3.4) 3.5 (2.7) Primary clinical and economic outcome variables  Alzheimer's disease assessment scale, cognition 10.2 (5.4) 11.7 (5.5)  Executive interview 13.3 (6.4) 13.7 (4.7)  ADCS-ADL 46.5 (5.1) 46.1 (6.8)  EQ-5D-3L (patient rated) 0.797 (0.109)  0.817 (0.135) 0.822 (0.072)  0.826 (0.108)  EQ-5D-3L (caregiver rated) 0.799 (0.136)  0.826 (0.117) 0.829 (0.064)  0.843 (0.108) Secondary outcome variables  Stroop Test 3-2 (s) 57.12 (24.13) 67.82 (28.36)  Trail making test B-A (s) 75.18 (83.27) 59.70 (42.28)  Digit span forward—backward 3.8 (1.95) 3.37 (2.44)  6-minute walk (m) 486.9 (97.9) 502.8 (98.4)  Weight (kg) 72.39 (14.11) 70.05 (14.31)  Body mass index 26.54 (3.97) 25.26 (3.54)  Resting heart rate (bpm) 70.24 (15.10) 67.26 (12.38)  Resting systolic blood pressure (mm  Hg) 132.29 (18.66) 139.80 (17.73)  Resting diastolic blood pressure (mm  Hg) 76.71 (11.38) 80.26 (10.05)  Physical activity scale for the elderly 118.59 (55.41) 124.44 (73.47) AT, aerobic exercise training group; CON, nutrition education. Healthcare use and costs Complete healthcare resource usage data were provided by 58 (83%) participants at 6 months and 57 (81%) participants at 12 months. Response rates for healthcare usage data were comparable across the two participant groups. Unit costs for healthcare cost items are provided in table 2. In summary (table 2), the mean (SD) costs (2015 $CAD) for healthcare professional visits, admissions to hospital and laboratory tests/investigations at 6 months were as follows: 940 (1194), 187 (325) and 113 (128). The mean (SD) costs (2015 $CAD) for healthcare professional visits, admissions to hospital and laboratory tests/investigations at 6 months were as follows: 682 (465), 554 (1648) and 108 (132). The mean total healthcare resource usage costs for the control group (2015 $CAD) at 6 and 12 months were 1434 (1674) and 2964 (2947). The mean total healthcare resource usage costs for the AT group (2015 $CAD) at 6 and 12 months were as follows: 1434 (1674) and 2964 (2947). Table 2 Unit costs for each component of resource usage Item 6-month HRU 2015 CAN$ Mean (SD) Median (IQR) 12-month HRU 2015 CAN$ Mean (SD) Median (IQR) Unit Reference Cost of delivering control group 0 − Cost per person year Study records Cost of delivering thrice weekly aerobic training 576 − Cost per person year Study records Healthcare professional visit, mean (SD) 940 (1194) 586 (1097) 682 (465) 632 (726) Cost per person 2013 Medical services plan Admission to hospital 187 (325)    0 (277) 552 (1648)    0 (207) Cost per person 2005 Vancouver General Hospital fully allocated cost model* Emergency department presentations 42 Cost per hour 2005 Vancouver General Hospital fully allocated cost model* Laboratory procedures, mean (SD) 113 (128)   44 (204) 108 (132)   59 (129) Cost per person 2009 Medical services plan *Taken from the fully allocated cost model at Vancouver General Hospital. All costs were inflated to 2015 Canadian Dollars. Health outcomes Complete data for the EQ-5D-3L at baseline were provided by 69 (99%) patients and 63 (90%) caregivers. Complete data for the EQ-5D-3L at 6 months were provided by 65 (93%) patients and 54 (77%) caregivers. Complete data for the EQ-5D-3L at 12 months were provided by 57 (81%) patients and 49 (70%) caregivers. The response rates of patients or caregivers for dropouts were comparable between treatment groups. The mean EQ-5D-3L at 6 and 12 months and adjusted incremental QALYs for patients and caregivers are provided in table 3. Table 3 Results of imputed case analysis CON at 6 months Mean (SD) CON at 12 months Mean (SD) AT at 6 months Mean (SD) AT at 12 months Mean (SD) Cost of delivering programme per person (2015 CAN$) 0 (usual care) 0 (usual care) 730 730 Mean healthcare resource use cost (2015 CAN$) per person 1434 (1674) 2964 (2947) 956 (861) 2110 (1857) Adjusted incremental QALY based on  EQ-5D-3L patient* 0 (reference) 0 (reference) 0.804 (0.080) 0.800 (0.075)  EQ-5D-3L caregiver* 0 (reference) 0 (reference) 0.806 (0.096) 0.810 (0.078) Incremental cost (2015 $CAD)  EQ-5D-3L patient Reference Reference 1770 (1369) 3112 (2499)  EQ-5D-3L caregiver Reference Reference 1770 (1369) 3112 (2499) Incremental cost (2015 $CAD) per QALY based on†  EQ-5D-3L patient Reference Reference 2129 3761  EQ-5D-3L caregiver Reference Reference 2124 3715 *Incremental QALYs are adjusted for the baseline utility using a linear regression model. †ICER based on total HRU costs, fall related costs and cost of delivering programmes. Adjusting QALYs for imbalances in baseline utility Imputed case analysis After controlling for imbalances in baseline utility, the mean (SD) incremental QALY after 6 months calculated using the EQ-5D-3L was 0.82 (0.06) as rated by patients and 0.83 (0.06) as rated by the caregivers’ perspective for the patients in the AT group and 0.78 (0.09) as rated by patients and 0.79 (0.12) as rated by the caregivers’ perspective for the patients in the CON group (table 3). After controlling for imbalances in baseline utility, the mean (SD) incremental QALY after 12 months calculated using the EQ-5D-3L was 0.82 (0.06) as rated by patients and 0.83 (0.05) as rated by the caregivers’ perspective for the patients in the AT group and 0.78 (0.08) as rated by patients and 0.79 (0.10) as rated by the caregivers’ perspective for the patients in the CON group (table 3). Complete cases analysis After controlling for imbalances in baseline utility, the mean (SD) incremental QALY after 6 months calculated using the EQ-5D-3L was 0.82 (0.06) as rated by patients and 0.83 (0.05) as rated by the caregivers’ perspective for the patients in the AT group and 0.78 (0.09) as rated by patients and 0.78 (0.12) as rated by the caregivers’ perspective for the patients in the CON group (table 3). After controlling for imbalances in baseline utility, the mean (SD) incremental QALY after 12 months calculated using the EQ-5D-3L was 0.82 (0.03) as rated by patients and 0.82 (0.01) as rated by the caregivers’ perspective for the patients in the AT group and 0.78 (0.05) as rated by patients and 0.78 (0.03) as rated by the caregivers’ perspective for the patients in the CON group. Cost–utility analysis From the Canadian healthcare system perspective, the incremental cost–utility ratios for thrice weekly AT were cost-effective compared with the comparator group, when using a willingness to pay threshold of $CAD 20 000 per QALY gained or higher. Specifically, on the point estimates from our base case analysis, we found that AT is more effective and also more costly than the CON alternative. Figure 1A (using the patient’s own ratings of their health status) demonstrates that for three times weekly AT at 6 months (ie, intervention completion) compared with CON, most of the bootstrapped cycles (>80% of the 4000 cycles) were represented in the northeast quadrant. Figure 1B (using the caregiver’s ratings of the patient’s health status) demonstrates that for three times weekly AT at 6 months (ie, intervention completion) compared with CON, most of the bootstrapped cycles (>80% of the 4000 cycles) were represented in the northeast quadrant. Figure 1C, D (using the patient’s own ratings and caregiver's ratings of their health status, respectively) demonstrates that for three times weekly AT at 12 months (ie, intervention completion) compared with CON, most of the 4000 bootstrapped cycles were represented in the northeast quadrant. Figure 2A–D reports the cost-effectiveness acceptability curves highlighting the probability of the AT being cost-effective over different willingness to pay values. Figure 1 (A) Cost-effective plane (time horizon—6 months) depicting the 95% confidence ellipses of incremental cost and effectiveness (patient-rated health status) for comparison between thrice weekly aerobic training and usual care (control, comparator). (B) Cost-effective plane (time horizon—6 months) depicting the 95% confidence ellipses of incremental cost and effectiveness (caregiver (patient-proxy) rated health status) for comparison between thrice weekly aerobic training and usual care (control, comparator). (C) Cost-effective plane (time horizon—12 months) depicting the 95% confidence ellipses of incremental cost and effectiveness (patient-rated health status) for comparison between thrice weekly aerobic training and usual care (control, comparator). (D) Cost-effective plane (time horizon—12 months) depicting the 95% confidence ellipses of incremental cost and effectiveness (caregiver (patient-proxy) rated health status) for comparison between thrice weekly aerobic training and usual care (control, comparator). Figure 2 (A) Cost-effectiveness acceptability curve showing the probability that thrice aerobic training intervention is cost-effective compared to usual care over a range of values for the maximum acceptable ceiling ratio (λ—willingness to pay) in the PROMoTE trial (6-month time horizon, patient-rated health status). (B) Cost-effectiveness acceptability curve showing the probability that thrice aerobic training intervention is cost-effective compared to usual care over a range of values for the maximum acceptable ceiling ratio (λ—willingness to pay) in the PROMoTE trial (6-month time horizon, caregiver (patient-proxy)-rated health status). (C) Cost-effectiveness acceptability curve showing the probability that thrice aerobic training intervention is cost-effective compared to usual care over a range of values for the maximum acceptable ceiling ratio (λ—willingness to pay) in the PROMoTE trial (12-month time horizon, patient-rated health status). (D) Cost-effectiveness acceptability curve showing the probability that thrice aerobic training intervention is cost-effective compared to usual care over a range of values for the maximum acceptable ceiling ratio (λ—willingness to pay) in the PROMoTE trial (12-month time horizon, caregiver (patient-proxy) rated health status). Sensitivity analysis Our complete case analysis demonstrated the same trend with regard to a significant improvement in QALYs and an overall increase in health resource usage costs for the AT group. Discussion Among a population of individuals at high risk for future cognitive decline, this study demonstrated, using a Canadian healthcare system perspective, that the incremental cost per QALY gained by participating in thrice weekly AT was more effective and more costly than the usual care plus education group. We observed a trend towards improvement in the adjusted incremental QALYs determined from the EQ-5D-3L (by patients and proxies) for the AT group compared with the usual care plus education group at trial completion and 6-month follow-up. Importantly, AT is an alternative to resistance training and is accessible to older adults with mild SIVCI. Further, the delivery of a walking programme on an individual basis requires a low financial investment (ie, the cost of walking poles) by an individual. As such, the results of this economic evaluation represent a substantive contribution to the evidence base on how to efficiently minimise cognitive decline among those with mild SIVCI. The findings of this study build on previous research demonstrated that AT has significant and beneficial effects on overall health-related quality of life and quality of life more broadly.39 The overall incremental cost–utility ratios were not significantly different regardless of whether QALYs were ascertained from patient-reported or proxy-reported health status using the EQ-5D-3L suggesting that for this population, use of patient or proxy ratings should not alter healthcare decision-making. The cost-effectiveness acceptability curves confirm that AT is the preferred treatment option for a wide range of plausible willingness to pay thresholds. From both our sensitivity analyses, we found that all analyses supported the conclusions that AT resulted in clinically important gains in QALYs. However, our imputed case analysis demonstrated that the intervention was not cost-saving, while the complete case analysis demonstrated that the intervention was cost-saving. One potential explanation for this was that the complete case analysis may better reflect the per protocol findings (ie, those that had greater adherence to the trial). The time horizon of our study was limited to the duration of the intervention (ie, 6 months) and the follow-up period (ie, 12 months). The number of randomised controlled trials of exercise conducted in populations at high risk for dementia is accumulating.40 One study demonstrated that resistance training post a 6-month (frequency of two to three times weekly) intervention significantly improved global cognitive function while maintaining executive and global benefits for at least 18 months postintervention.41 42 However, the number of randomised controlled trials of exercise among individuals diagnosed with SIVCI remains low.22 Given that cardiovascular risk factors play a primary role in the onset and progression of VCI, examining the cost-effectiveness of AT is a logical starting place. In adults with mild VCI, 6 months of thrice weekly progressive AT improved cognitive function, relative to CON.22 Previous research that AT in older adults has longer term health benefits that we hypothesise would be applicable to adults with mild SIVCI benefit of the intervention may be ideally captured by a longer time horizon.43 Further, the sample size of our study was small. As such, there was wide variability in the cost estimates and outliers (ie, ±3 SDs from the mean) had a stronger impact than would be expected in a larger sample. However, we did not have any reason to remove any health resource usage outliers in our intention to treat analysis. The health resource usage questionnaire may be subject to recall bias, thus causing potential underestimation of costs. To minimise recall bias, participants were provided with a monthly diary to track and report their healthcare resource usage. Given that cost underestimation may have occurred in both groups, we do not estimate any impact on the incremental cost–utility ratio given that this was a randomised controlled trial. A key strength of our study is that it deals with a largely understudied yet important population. Importantly, this population actually may represent an ideal target population for intervention given that individuals have not yet crossed the dementia threshold. Hence, it is important to gain a better understanding of the effectiveness and efficiency of targeted interventions. Given that even mildly impaired cognition may impede an individual's ability to self-assess their HRQoL we also used a patient-proxy (ie, caregiver) assessment of the patient’s health status.44 45 In this study, we found that the use of the patient or the proxy did not significantly alter our findings. In all instances, we observed a significant increase in QALYs at 6 and 12 months regardless of the rater. This is a useful observation because it suggests that among individuals with VCI, the rater should not result in changes in healthcare decision-making. Finally, a highly relevant strength of this study is that the intervention is widely accessible and relatively easy to implement for any community dwelling older adult who is able to walk. The low cost required by an individual to start walking is also appealing from an implementation perspective. Our proof-of-concept findings suggest that this exercise (ie, AT) therapy delivered over a span of 6 months holds promise for improving cognitive function and health-related quality of life in older adults with mild VCI. While our findings suggest that this intervention is not cost-saving, it appears to be cost-effective depending on a decision maker’s willingness to pay. The authors thank the PROMoTE study participants. Contributors: TL-A and JCD had full access to all the data in the study, take responsibility for the integrity of the data and the accuracy of the data analysis and involved in study concept and design. TL-A, JCD, SB and JRB contributed to acquisition, analysis or interpretation of data. JCD, TL-A and SB drafted the manuscript. JCD, G-YRH, SB, JRB, JJE, MM, WC, BC, CJ, PL and TL-A contributed to critical revision of the manuscript for important intellectual content. JCD and JRB involved in statistical analysis. TL-A, CJ, G-YRH, JJE, PL and JCD obtained funding. WC, MM and BC provided administrative, technical, or material support. TL-A, JCD, MM and WC contributed to study supervision. Funding: This study is jointly funded by the Canadian Stroke Network and the Heart and Stroke Foundation of Canada (grant number: PG-09-0443). TL-A is a Canada Research Chair in Physical Activity, Mobility, and Cognitive Neuroscience, a Michael Smith Foundation for Health Research (MSFHR) Scholar, a Canadian Institutes of Health Research (CIHR) New Investigator and a Heart and Stroke Foundation of Canada's Henry JM Barnett's Scholarship recipient. These funding agencies did not play a role in study design. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Unpublished data are available on request. Please email Dr Teresa Liu-Ambrose: teresa.ambrose@ubc.ca. ==== Refs References 1 Rockwood K , Wentzel C , Hachinski V Prevalence and outcomes of vascular cognitive impairment. Vascular Cognitive Impairment Investigators of the Canadian Study of Health and Aging . Neurology 2000 ;54 :447 –51 . 10.1212/WNL.54.2.447 10668712 2 Desmond DW , Erkinjuntti T , Sano M The cognitive syndrome of vascular dementia: implications for clinical trials . Alzheimer Dis Assoc Disord 1999 ;13 (Suppl 3 ):S21 –9 .10609678 3 Erkinjuntti T , Bowler JV , DeCarli CS Imaging of static brain lesions in vascular dementia: implications for clinical trials . Alzheimer Dis Assoc Disord 1999 ;13 (Suppl 3 ):S81 –90 .10609686 4 Pantoni L , Leys D , Fazekas F Role of white matter lesions in cognitive impairment of vascular origin . Alzheimer Dis Assoc Disord 1999 ;13 (Suppl 3 ):S49 –54 .10609681 5 Román GC Vascular dementia may be the most common form of dementia in the elderly . J Neurol Sci 2002 ;203–204 :7 –10 . 10.1016/S0022-510X(02)00252-6 6 Ramos-Estebanez C , Moral-Arce I , Rojo F Vascular cognitive impairment and dementia expenditures: 7-year inpatient cost description in community dwellers . Postgrad Med 2012 ;124 : 91 –100 . 10.3810/pgm.2012.09.2597 7 O'Brien JT , Erkinjuntti T , Reisberg B Vascular cognitive impairment . Lancet Neurol 2003 ;2 :89 –98 . 10.1016/S1474-4422(03)00305-3 12849265 8 Román GC , Erkinjuntti T , Wallin A Subcortical ischaemic vascular dementia . Lancet Neurol 2002 ;1 :426 –36 . 10.1016/S1474-4422(02)00190-4 12849365 9 Vermeer SE , Longstreth WT Jr, Koudstaal PJ Silent brain infarcts: a systematic review . Lancet Neurol 2007 ;6 :611 –19 . 10.1016/S1474-4422(07)70170-9 17582361 10 de Bruijn RF , Bos MJ , Portegies ML The potential for prevention of dementia across two decades: the prospective, population-based Rotterdam Study . BMC Med 2015 ;13 :132 10.1186/s12916-015-0377-5 26195085 11 Barnes DE , Yaffe K The projected effect of risk factor reduction on Alzheimer's disease prevalence . Lancet Neurol 2011 ;10 : 819 –28 . 10.1016/S1474-4422(11)70072-2 21775213 12 Norton S , Matthews FE , Barnes DE Potential for primary prevention of Alzheimer's disease: an analysis of population-based data . Lancet Neurol 2014 ;13 :788 –94 . 10.1016/S1474-4422(14)70136-X 25030513 13 Jokinen H , Kalska H , Mäntylä R Cognitive profile of subcortical ischaemic vascular disease . J Neurol Neurosurg Psychiatr 2006 ;77 :28 –33 . 10.1136/jnnp.2005.069120 14 Kramer JH , Reed BR , Mungas D Executive dysfunction in subcortical ischaemic vascular disease . J Neurol Neurosurg Psychiatr 2002 ;72 :217 –20 . 10.1136/jnnp.72.2.217 15 Wimo A , Guerchet M , Ali GC The worldwide costs of dementia 2015 and comparisons with 2010 . Alzheimers Dement 2017 ;13 :1 –7 . 10.1016/j.jalz.2016.07.150 27583652 16 Rojas G , Bartoloni L , Dillon C Clinical and economic characteristics associated with direct costs of Alzheimer's, frontotemporal and vascular dementia in Argentina . Int Psychogeriatr 2011 ;23 :554 –61 . 10.1017/S1041610210002012 21044400 17 Abbott RD , Donahue RP , MacMahon SW Diabetes and the risk of stroke. The Honolulu Heart Program . JAMA 1987 ;257 :949 –52 . 10.1001/jama.1987.03390070069025 3806877 18 Knopman D , Boland LL , Mosley T Cardiovascular risk factors and cognitive decline in middle-aged adults . Neurology 2001 ;56 :42 –8 . 10.1212/WNL.56.1.42 11148234 19 Anstey KJ , Lipnicki DM , Low LF Cholesterol as a risk factor for dementia and cognitive decline: a systematic review of prospective studies with meta-analysis . Am J Geriatr Psychiatry 2008 ;16 :343 –54 . 10.1097/JGP.0b013e31816b72d4 18448847 20 Tzourio C , Anderson C , Chapman N Effects of blood pressure lowering with perindopril and indapamide therapy on dementia and cognitive decline in patients with cerebrovascular disease . Arch Intern Med 2003 ;163 :1069 –75 . 10.1001/archinte.163.9.1069 12742805 21 Liu-Ambrose T , Eng JJ , Boyd LA Promotion of the mind through exercise (PROMoTE): a proof-of-concept randomized controlled trial of aerobic exercise training in older adults with vascular cognitive impairment . BMC Neurol 2010 ;10 :14 10.1186/1471-2377-10-14 20158920 22 Liu-Ambrose T , Best JR , Davis JC Aerobic exercise and vascular cognitive impairment: a randomized controlled trial . Neurology 2016 ;87 :2082 –90 .27760869 23 Rosen WG , Mohs RC , Davis KL A new rating scale for Alzheimer's disease . Am J Psychiatry 1984 ;141 :1356 –64 . 10.1176/ajp.141.11.1356 6496779 24 Royall DR , Mahurin RK , Gray KF Bedside assessment of executive cognitive impairment: the executive interview . J Am Geriatr Soc 1992 ;40 :1221 –6 . 10.1111/j.1532-5415.1992.tb03646.x 1447438 25 Galasko D , Bennett D , Sano M An inventory to assess activities of daily living for clinical trials in Alzheimer's disease. The Alzheimer's disease Cooperative Study . Alzheimer Dis Assoc Disord 1997 ;11 (Suppl 2 ):S33 –9 . 10.1097/00002093-199700112-00005 9236950 26 Erkinjuntti T , Inzitari D , Pantoni L Research criteria for subcortical vascular dementia in clinical trials . J Neural Transm Suppl 2000 ;59 :23 –30 .10961414 27 Nasreddine ZS , Phillips NA , Bédirian V The montreal cognitive assessment, MoCA: a brief screening tool for mild cognitive impairment . J Am Geriatr Soc 2005 ;53 :695 –9 . 10.1111/j.1532-5415.2005.53221.x 15817019 28 Folstein MF , Folstein SE , McHugh PR “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician . J Psychiatr Res 1975 ;12 :189 –98 . doi:0022-3956(75)90026-61202204 29 Dolan P Modeling valuations for EuroQol health states . Med Care 1997 ;35 :1095 –108 . 10.1097/00005650-199711000-00002 9366889 30 Bansback N , Tsuchiya A , Brazier J Canadian valuation of EQ-5D health states: preliminary value set and considerations for future valuation studies . PLoS ONE 2012 ;7 :e31115 10.1371/journal.pone.0031115 22328929 31 Manca A , Palmer S Handling missing data in patient-level cost-effectiveness analysis alongside randomised clinical trials . Appl Health Econ Health Policy 2005 ;4 :65 –75 . 10.2165/00148365-200504020-00001 16162026 32 Briggs A , Clark T , Wolstenholme J Missing…presumed at random: cost-analysis of incomplete data . Health Econ 2003 ;12 :377 –92 . 10.1002/hec.766 12720255 33 Oostenbrink JB , Al MJ The analysis of incomplete cost data due to dropout . Health Econ 2005 ;14 :763 –76 . 10.1002/hec.966 15729743 34 Oostenbrink JB , Al MJ , Rutten-van Molken MP Methods to analyse cost data of patients who withdraw in a clinical trial setting . Pharmacoeconomics 2003 ;21 :1103 –12 . 10.2165/00019053-200321150-00004 14596629 35 Schafer JL Analysis of incomplete multivariate data . London, England : Chapman & Hall , 1997 . 36 Laska EM , Meisner M , Siegel C Statistical inference for cost-effectiveness ratios . Health Econ 1997 ;6 :229 –42 . 10.1002/(SICI)1099-1050(199705)6:3<229::AID-HEC268>3.0.CO;2-M 9226141 37 Briggs AH , Gray AM Handling uncertainty when performing economic evaluation of healthcare interventions . Health Technol Assess 1999 ;3 :1 –134 . 38 Fenwick E , Claxton K , Sculpher M Representing uncertainty: the role of cost-effectiveness acceptability curves . Health Econ 2001 ;10 :779 –87 . 10.1002/hec.635 11747057 39 Awick EA , Wójcicki TR , Olson EA Differential exercise effects on quality of life and health-related quality of life in older adults: a randomized controlled trial . Qual Life Res 2015 ;24 : 455 –62 . 10.1007/s11136-014-0762-0 25074734 40 Lautenschlager NT , Cox KL , Flicker L Effect of physical activity on cognitive function in older adults at risk for Alzheimer disease: a randomized trial . JAMA 2008 ;300 :1027 –37 . 10.1001/jama.300.9.1027 18768414 41 Fiatarone Singh MA , Gates N , Saigal N The Study of Mental and Resistance Training (SMART) study-resistance training and/or cognitive training in mild cognitive impairment: a randomized, double-blind, double-sham controlled trial . J Am Med Dir Assoc 2014 ;15 :873 –80 . 10.1016/j.jamda.2014.09.010 25444575 42 Lam LC , Chan WM , Kwok TC Effectiveness of Tai Chi in maintenance of cognitive and functional abilities in mild cognitive impairment: a randomised controlled trial . Hong Kong Med J 2014 ;20 (3 Suppl 3) :20 –3 . 43 McCartney N , Hicks AL , Martin J Long-term resistance training in the elderly: effects on dynamic strength, exercise capacity, muscle, and bone . J Gerontol A Biol Sci Med Sci 1995 ;50 :B97 –104 . 10.1093/gerona/50A.2.B97 7874586 44 Makai P , Beckebans F , van Exel J Quality of life of nursing home residents with dementia: validation of the German version of the ICECAP-O . PLoS ONE 2014 ;9 :e92016 10.1371/journal.pone.0092016 24632775 45 Bryan S , Hardyman W , Bentham P Proxy completion of EQ-5D in patients with dementia . Qual Life Res 2005 ;14 :107 –18 . 10.1007/s11136-004-1920-6 15789945

PMC005xxxxxx/PMC5372068.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01443110.1136/bmjopen-2016-014431Public HealthResearch150617241692Differences in predictors of permanent work disability between immigrants and natives: a cohort study of adults with sick leave due to common mental disorders Werlen Laura 1Helgesson Magnus 2Mittendorfer-Rutz Ellenor 2 1 Epidemiology, Biostatistics and Prevention Institute, University of Zurich, Zurich, Switzerland 2 Division of Insurance Medicine, Department of Clinical Neuroscience, Karolinska Institutet, SE-171 77 Stockholm, SwedenCorrespondence to Dr Magnus Helgesson; magnus.helgesson@ki.se2017 17 3 2017 7 3 e01443127 9 2016 13 2 2017 21 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives Immigrants with common mental disorders (CMDs) are reported to have a higher risk of disability pension (DP) compared with native residents; however, the reasons for this are not fully understood. This study aimed to investigate (1) differences in morbidity (3 measures) and socioeconomic status in native Swedes, ‘Western’ and ‘non-Western’ immigrants with CMDs and (2) interactions between morbidity and socioeconomic status and immigrant status regarding subsequent DP. Design The study was a prospective population-based cohort study using national register data. Crude and multivariate HRs with 95% CIs were calculated using the Cox regression (2007–2010). Participants All individuals aged 18–59 with an incident sick-leave spell due to CMDs during 2006 were included in the study (N=66 097). The study population was divided into 3 groups based on country of birth: (1) Sweden, (2) immigrants from ‘Western’ countries (EU25, Norway, Iceland, North America and Oceania) and (3) immigrants from ‘non-Western’ countries (east Europe, Africa, Asia and South America). Results Particularly, immigrants born in non-Western countries had higher levels of morbidity and lower socioeconomic status than natives (p>0.001). No significant differences in the associations between specialised psychiatric and somatic care with regard to subsequent DP were found between immigrants and native Swedes. Being prescribed more than 1 type of psychiatric medication was associated with higher HRs for DP in immigrants from Western (HR 3.34; CI 2.3 to 4.9) and non-Western countries (3.6; 1.9 to 6.4) than in native Swedes (2.55; 2.3 to 2.8) (pinteraction=0.003). Low education was a marginally stronger predictor for DP in non-Western immigrants than in native Swedes and Western immigrants (pinteraction=0.03). Conclusions Morbidity measured by medication, but not by specialised healthcare, was a stronger predictor for DP in immigrants than in native Swedes, warranting scrutiny of differences in care and treatment in immigrants and native Swedes with CMDs. Common Mental DisordersSick LeaveDisability, InsurancePsychotropic DrugsEmigrants and ImmigrantsForskningsrådet om Hälsa, Arbetsliv och Välfärdhttp://dx.doi.org/10.13039/5011000066362015-00742 ==== Body Strengths and limitations of this study The study is a prospective cohort study using a large study population with a relatively long follow-up period of 4 years. The Swedish national registers used here are of good quality with almost no dropout. The validity of the diagnoses of sick-leave spells has often been discussed, but a previous related study reported good validity. Findings of this study cannot be generalised to a population with depression and anxiety symptoms that did not come to the attention of the healthcare system. Grouping immigrants according to regions of country of birth could mask critical differences, including country of origin and reason for emigration. Introduction Common mental disorders (CMDs) are a significant and growing cause of morbidity in Sweden and across the globe. By 2030, depressive disorders are predicted to be the leading diagnoses resulting in loss of disability-adjusted life years in high-income countries.1 Moreover, CMDs have a strong impact on work capacity,2 3 putting individuals with these disorders at risk for temporal and permanent work disability (ie, sickness absence (SA) and disability pension (DP)).4–6 To date, immigrants form a considerable proportion of the population in Sweden, and this proportion is estimated to increase in the years to come.7 8 Previous studies have reported immigrants to Sweden and Denmark to be at higher risk of psychiatric morbidity compared with the native population9–11 and that there are differences between immigrants and the native population regarding the work-related consequences of having a CMD. For example, immigrants from non-Western countries have been found to be at higher risk for temporal and permanent work disability compared with the native population;6 12 however, the reasons for this are not yet fully understood. Previously reported risk factors for an increased risk of DP in individuals with a CMD are the severity of the underlying mental disorder, a comorbid somatic disorder and low socioeconomic status.5 6 However, it is not known whether these patterns differ between immigrants and the native population. Adding to the complexity of this question, migration is not a homogenous phenomenon, so the pathways to permanent work disability might vary for different immigrant groups with respect to country of birth. Especially, immigrants from non-Western countries have been shown to have a higher prevalence of CMDs and work disability.6 12 It is therefore important to consider different immigrant groups in studies on this topic. In this study, immigrants were divided into Western immigrants, a group mostly consisting of voluntary immigrants, for example, labour migrants and students, and non-Western immigrants, a group largely comprised refugees and relatives seeking reunification with refugees.13 The pathways to DP are not only strongly affected by a number of other sociodemographic factors, for example, age, sex and type of living area in Sweden, but also by previous spells of unemployment and/or sick leave.14 Therefore, further consideration of these factors in related studies is crucial. Moreover, due to the reported differences in access to healthcare as well as in diagnostics and treatment in immigrants compared with the native population, different measures of morbidity should be used in studies that include immigrants.9 Aim This study aimed to describe: (1) differences in morbidity and/or socioeconomic status with regard to immigrant status in adults with temporary work disability due to CMDs and (2) interactions between morbidity and socioeconomic status and immigrant status regarding subsequent permanent work disability among individuals with temporary work disability due to a CMD. Methods Study population This study is a prospective, population-based cohort study including all non-pensioned individuals aged 18–59 years and living in Sweden on 31 December 2005 who had a new, incident sick-leave spell due to a CMD during 2006 (N=66 097). Individuals were followed regarding DP from 1 January 2007 until 31 December 2010. Using the unique deidentified personal identification number for all Swedish residents, data on individuals were linked at the individual level. These data were obtained from registers from the following three agencies: (1) Statistics Sweden: the Longitudinal Integration Database for Health Insurance and Labour Market Studies (LISA) containing information on age, sex, country of birth, educational level, type of living area, family composition and length of unemployment. (2) The Social Insurance Agency: the Micro Data for Analyses of Social Insurance (MiDAS) register including information on the date, grade and diagnoses of SA and DP. (3) The National Board of Health and Welfare: the National Patient Register (NPR) comprising data on the date and cause of inpatient and specialised outpatient healthcare from 1973 and from 2001, respectively; the Cause of Death Register (CDR) with information on the date of death from 1960 and onwards; and the Prescribed Drug Register (PDR) with information on the prescription of dispensed psychiatric medication (date of dispensing and type), from July 2005 and onward. The Swedish social insurance system Swedish sickness insurance covers all people above the age of 16 who are living in Sweden and have at least a minimum annual income from work.15 Compensation can be provided for individuals with reduced working capacity of at least 25% due to a disease or injury by either the employer or the Social Insurance Agency (SIA). Employees receive sick pay from days 2 to 14 of a sick-leave spell from the employer (the first day being a qualifying day). From day 15, employees get sickness benefit from SIA. A certificate from a physician is required from the eighth day of a sick-leave spell. Unemployed individuals and individuals on parental leave can be granted sickness benefit from SIA from the second day of a sick-leave spell, whereas self-employed individuals receive sick pay from SIA according to which insurance coverage they had chosen. In Sweden, DP can be granted to all individuals living in Sweden whose work capacity has been reduced due to a disease or injury. Individuals between 19 and 29 years can be granted time-restricted DP due to impaired work capacity and if they have not completed their compulsory education at 19 years of age. Variables The study population was divided into three groups based on country of birth: (1) Sweden (natives), (2) immigrants from ‘Western’ countries consisting of EU25 plus Norway and Iceland, North America and Oceania and (3) immigrants from all remaining countries, labelled as ‘non-Western’ countries (Europe outside EU25 and Nordic countries, Africa, Asia and South America). The outcome measure was defined as having been granted all-cause DP in the period from 2007 through 2010. Morbidity was conceptualised using three different measures, namely: (1) specialised inpatient and outpatient healthcare due to mental disorders (2001–2006), (2) specialised inpatient and outpatient healthcare due to somatic disorders (2001–2006) and (3) psychiatric medication in 2006 grouped as (I) no medication, (II) antidepressants only, (III) anxiolytics/hypnotics/sedatives only and (IV) more than one of these types of psychiatric medication in any combination. Moreover, socioeconomic status was measured as educational level in 2005 divided into (I) low education level (through elementary school; ≤9 years), (II) medium education level (high school, 10–12 years) and (III) high education level (university; >12 years). Missing data regarding education were treated as low educational level. Covariates included age, sex, type of living area, family composition as well as previous unemployment and previous SA, which were measured for the year preceding baseline and coded as indicated in table 1. Table 1 Descriptive statistics of the 66 097 women and men, aged 18–59 years on 31 December 2005 with at least one new sick-leave spell due to a common mental disorder (CMD) in 2006 in Sweden and the share with disability pension between 2007 and 2010   All DP N Per cent N Per cent Study population, all 66 097 100 4610 7.0 Country of birth*  Sweden 57 011 86.3 3696 6.5  Western countries 3520 5.3 302 8.6  Non-Western countries 5566 8.4 612 11.0 Sex  Male 20 373 30.8 1540 7.6  Female 45 724 69.2 3070 6.7 Age in 2005  18–24 4420 6.7 295 6.7  25–34 17 029 25.8 731 4.3  35–44 21 228 32.1 1151 5.4  45–54 16 516 25.0 1363 8.3  55–59 6 904 10.4 1070 15.5 Type of living area†  Big cities 25 450 38.5 1610 6.3  Medium-sized cities 23 186 35.1 1585 6.8  Small cities/villages 17 461 26.4 1415 8.1 Family composition  Married/cohabiting with children at home 26 976 40.8 1434 5.3  Married/cohabiting without children at home 6792 10.3 690 10.2  Single‡ with children at home 9579 14.5 628 6.6  Single‡ without children at home 22 750 34.4 1858 8.2 Unemployment in 2005  No days 54 637 82.7 3276 6.0  1–90 days 5435 8.2 500 9.2  >90 days 6025 9.1 834 13.8 Sick leave in 2005  No days 48 749 73.8 2769 5.7  1–90 days 13 944 21.1 1309 9.4  >90 days 3404 5.2 532 15.6 Sick leave diagnosis in 2006  Depressive 24 419 36.9 2296 9.4  Anxiety 7827 11.8 806 10.3  Stress-related 33 851 51.2 1508 4.5 *Country of birth categorised according to: (1) Sweden, (2) Western countries include the Nordic countries (Denmark, Finland, Iceland and Norway), EU25 (all countries included in the European Union in 2006 without Sweden), North America and Oceania, (3) non-Western countries comprise all remaining countries. †Type of living area: big cities, Stockholm, Gothenburg and Malmo; medium-sized cities, cities with >90 000 inhabitants within 30 km distance from the centre of the city; small cities/villages, all remaining cities/villages. ‡Single means living without partner and also includes divorced, separated or widowed. Diagnostics and codes All diagnoses were coded according to the International Classification of Diseases V.10 (ICD 10) codes. Here, CMDs were categorised into three diagnostic groups: (1) depressive disorders (F32: depressive episode and F33: recurrent depressive disorder), (2) anxiety disorders (F40: phobic anxiety disorders, F41: other anxiety disorders and F42: obsessive–compulsive disorder) and (3) stress-related disorders (F43: reaction to severe stress and adjustment disorders). Mental (F00–99) and somatic (all other ICD codes) diagnoses were used to measure inpatient and specialised outpatient healthcare. Information on psychiatric medication was categorised according to the Anatomical Therapeutic Chemical (ATC) classification system: antidepressants (ATC: N06A); anxiolytics, sedatives and hypnotics (ATC: N05B or N05C). Statistical methods Pearson's χ2 tests were performed in order to examine differences between immigrant groups and native Swedes regarding measures of morbidity and socioeconomic status. Crude and multivariate HRs and CIs regarding the association between morbidity and socioeconomic factors and subsequent granting of DP were calculated by the Cox proportional hazard regression after assuring that the proportional hazard assumption was met. These analyses were stratified by immigrant status, that is, native Swedes and immigrants born in Western and non-Western countries. Censoring in the regression analyses was due to emigration, death or end of follow-up, whichever came first. Multivariate analyses were adjusted for all variables, that is, measures of morbidity, socioeconomic status, sociodemographics, previous unemployment, SA and diagnoses of SA in 2006 (all variables mentioned in tables 1 and 2). Partial likelihood ratio tests were performed in order to test potential interactions between the different measures of morbidity and socioeconomic status in addition to immigrant status and subsequent granting of DP. Analyses were conducted using IBM SPSS Statistics V.22.0. Table 2 Descriptive statistics of 66 097 women and men, aged 18–59 years on 31 December 2005 with at least one new sick-leave spell due to a common mental disorder in 2006 in Sweden with regard to morbidity and socioeconomic status, stratified by migration status   Sweden Western countries* Non-Western countries† Significance test‡ N (%) N (%) N (%) Specialised healthcare 2001–2006 Mental diagnosis  None 44 612 (78.3) 2748 (78.1) 4073 (73.2) χ2=75.8  Any 12 399 (21.7) 772 (21.9) 1493 (26.8) p<0.001 Somatic diagnosis  None 9509 (16.7) 569 (16.2) 690 (12.4) χ2=68.2  Any 47 502 (83.3) 2951 (83.8) 4876 (87.6) p<0.001 Psychiatric medication§ 2006  None 21 049 (36.9) 1185 (33.7) 1800 (32.3) χ2=145.6  Antidepr only 10 918 (19.2) 618 (17.6) 903 (16.2) p<0.001  Hyp/Sed/Anx only 6413 (11.2) 416 (11.8) 684 (12.3)  Mixed medications 18 631 (32.7) 1301 (37.0) 2179 (39.1) Socioeconomic status Education, 2005  Low 6982 (12.2) 546 (15.5) 1071 (19.2) χ2=263.8  Medium 29 960 (52.6) 1645 (46.7) 2697 (48.5) p<0.001  High 20 069 (35.2) 1329 (37.8) 1798 (32.3) *Western countries include the Nordic countries (Denmark, Finland, Iceland and Norway), EU25 (all countries included in the European Union in 2006 without Sweden), North America and Oceania. †Non-Western countries comprise all remaining countries. ‡χ2 test. §Antidepr, antidepressants; Hyp, hypnotics; Sed, sedatives; Anx, anxiolytics. Results During the 4 years of follow-up, 7% of individuals with SA due to CMDs were granted DP (table 1). There was an obvious gradient showing that immigrants from non-Western countries (11.0%) had the highest rate of DP followed by immigrants from Western countries (8.6%) and native Swedes (6.5%). More than two-thirds of individuals with at least one new sick-leave spell due to a CMD in 2006 were women (table 1). Still, a slightly higher proportion of men (7.6%) was granted DP during follow-up when compared with women (6.7%). Age was related to subsequent DP in a J-shaped fashion (table 1). Moreover, DP was more common among persons living in small cities/villages, married/cohabiting without children living at home and those with long earlier periods of unemployment and/or SA. The proportion of individuals who were subsequently granted DP was highest among those with anxiety disorders followed by depressive disorders and stress-related disorders (table 1). There were significant differences in the proportion of all markers of morbidity and socioeconomic status across immigrant groups and native Swedes. Immigrants from non-Western countries had higher morbidity expressed as a higher prevalence of mental (p<0.001) and somatic (p<0.001) specialised healthcare compared with native Swedes and immigrants from Western countries (table 2). Moreover, a higher proportion of immigrants from outside the Western world had a higher prescription rate of more than one type of psychiatric medication compared with native Swedes and lower rates of being prescribed only antidepressants (p<0.001). Immigrants from non-Western countries also had a significantly higher prevalence of low educational level compared with native Swedes (p<0.001). There were fewer differences between native Swedes and immigrants from the Western world (table 2). The univariate and multivariate models show that the HRs of being granted DP during follow-up after having received specialised healthcare due to mental diagnoses were higher for all groups, while the HRs of DP related to previous specialised healthcare due to somatic diagnoses were increased only among native Swedes (table 3). Table 3 Univariate and multivariate* HRs, with 95% CIs, for being granted disability pension (2007–2010) of the 66 097 women and men, aged 18–59 years on 31 December 2005 with at least one sickness absence spell due to a common mental disorder in 2006, stratified by migration status   Sweden Western countries† Non-Western countries‡ % DP Univariate (CI 95%) Multivariate (CI 95%) % DP Univariate (CI 95%) Multivariate (CI 95%) % DP Univariate (CI 95%) Multivariate (CI 95%) Morbidity Specialised healthcare 2001–2006 Mental diagnosis§  None 4.7 1 1 6.6 1 1 8.1 1 1  Any 12.8 2.84 (2.66 to 3.03) 2.03 (1.89 to 2.18) 15.5 2.47 (1.96 to 3.11) 1.71 (1.33 to 2.19) 19.0 2.53 (2.16 to 2.97) 1.96 (1.66 to 2.32) Somatic diagnosis¶  None 4.3 1 1 7.0 1 1 7.2 1 1  Any 6.9 1.64 (1.48 to 1.82) 1.33 (1.20 to 1.48) 8.9 1.28 (0.92 to 1.78) 0.97 (0.69 to 1.37) 11.5 1.61 (1.21 to 2.15) 1.32 (0.98 to 1.78) Psychiatric medication** 2006  None 3.1 1 1 2.9 1 1 3.9 1 1  Antidepr only 7.6 2.53 (2.28 to 2.81) 1.82 (1.63 to 2.03) 10.4 3.74 (2.46 to 5.66) 3.02 (1.96 to 4.65) 13.3 3.55 (2.65 to 4.76) 2.41 (1.78 to 3.27)  Hyp/Sed/Anx only 4.5 1.49 (1.29 to 1.71) 1.22 (1.06 to 1.40) 5.5 1.96 (1.15 to 3.32) 1.72 (1.01 to 2.93) 5.3 1.35 (0.90 to 2.01) 1.21 (0.81 to 1.81)  Mixed medications†† 10.4 3.53 (3.23 to 3.86) 2.07 (1.88 to 2.28) 13.4 5.12 (3.55 to 7.39) 3.34 (2.27 to 4.94) 17.7 4.84 (3.76 to 6.24) 3.24 (2.49 to 4.22) Socioeconomic status Education‡‡2005  Low 9.4 1.75 (1.59 to 1.92) 1.30 (1.18 to 1.43) 11.5 1.75 (1.27 to 2.41) 1.31 (0.94 to 1.82) 14.8 1.79 (1.43 to 2.23) 1.91 (1.52 to 2.40)  Medium 6.5 1.18 (1.10 to 1.27) 1.06 (0.98 to 1.14) 9.0 1.32 (1.02 to 1.72) 1.16 (0.89 to 1.51) 11.2 1.33 (1.09 to 1.61) 1.56 (1.28 to 1.90)  High 5.5 1 1 6.8 1 1 8.5 1 1 *All variables were mutually adjusted for all other variables in the table and additionally for sex, age, type of living area, family composition, unemployment in 2005, sick leave in 2005 and sick leave diagnose in 2006. †Western countries include the Nordic countries (Denmark, Finland, Iceland and Norway), EU25 (all countries included in the European Union in 2006 without Sweden), North America and Oceania. ‡Non-Western countries comprise all remaining countries. §Inpatient and/or outpatient specialised care for a mental diagnosis; p value for interaction=0.402. ¶Inpatient and/or outpatient specialised care for a somatic diagnosis; p value for interaction=0.198. **Antidepr, antidepressants; Hyp, hypnotics; Sed, sedatives; Anx, anxiolytics; p value for interaction=0.003. ††Any combination of antidepressants, hypnotics, sedatives and anxiolytics. ‡‡p Value for interaction=0.026. The risk for DP in the event of psychiatric healthcare was around twofold higher than the risk without such care in all groups. No significant interaction was observed between diagnosis-specific healthcare measures and immigrant group and subsequent granting of DP. The HRs of DP were higher among individuals in all groups when prescribed only antidepressants or more than one type of psychiatric medication compared with no prescribed psychiatric medication (table 3). Native Swedes and immigrants from the Western world also had higher HRs for DP when they had been prescribed hypnotics/sedatives/anxiolytics only. There was a significant interaction between psychiatric medication and immigrant status with regard to subsequent granting of DP (p=0.003). In the multivariate analyses, the HRs of DP related to having more than one type of psychiatric medication were twofold and threefold higher in native Swedes and immigrants, respectively (table 3). Individuals in all groups with low educational level had higher HRs for DP in the univariate models (table 3). Interactions of education and immigrant status were significant (p=0.026). In the multivariate model, low educational level was associated with a 30% and 90% increased risk of DP for natives and immigrants from non-Western countries, respectively. Discussion Main findings In a population of individuals with SA due to CMDs, higher proportions of immigrants from Western (8.6%) and non-Western (11.0%) countries were subsequently granted DP during follow-up compared with native Swedes (6.5%). Immigrants from non-Western countries had higher levels of morbidity and lower socioeconomic status than natives and immigrants from Western countries. Morbidity measured by being prescribed more than one type of psychiatric medication, but not as specialised psychiatric healthcare, was a stronger predictor for DP in immigrants than in natives. Compared with native Swedes, a higher proportion of immigrants, particularly those born in non-Western countries, was granted DP during follow-up in individuals with sick leave due to CMDs. This finding is in line with previous studies on psychiatric outpatients with depressive disorders and people with SA due to stress-related disorders.12 16 This study also found significant differences in measures of morbidity with regard to immigrant status in individuals with SA due to CMDs. Immigrants from non-Western countries in particular had higher rates of morbidity than native Swedes. This is also in line with previous studies showing more psychiatric and somatic morbidity in immigrants from outside Europe.17 In this study, morbidity was conceptualised in three different ways, that is, by specialised psychiatric and somatic healthcare as well as having been prescribed specific psychiatric medication. These measures can be regarded as reflecting the varying medical severity of the underlying mental disorders. Higher rates of specialised healthcare and being prescribed more than one type of psychiatric medication observed in immigrants from countries other than the Western world may thus be seen as an indicator of more severe morbidity, possibly due to a high proportion of refugees in this group.5 Moreover, this study found significant interactions between morbidity and immigrant status with regard to subsequent granting of DP in individuals with SA due to CMDs. Being prescribed more than one type of psychiatric medication was associated with higher risk estimates in immigrants than in native Swedes. Such interactions were not found when morbidity was measured as specialised psychiatric and somatic healthcare. These discrepancies may not only be a reflection of differences in the severity of the underlying morbidity, but may also reflect differences in access and acceptance of specialised healthcare, culturally sensitive diagnostics, reporting of mental symptoms and treatment in terms of psychiatric medication.17–21 Preconceptions on the part of healthcare professionals could also contribute to differential treatment.19 22 Future studies are warranted to elucidate potential differences in immigrant groups and native Swedes with SA due to CMDs regarding the type of treatment provided and the role of healthcare in preventing transition from temporal to permanent work disability. In line with previous studies, this study also shows that immigrants, particularly from non-Western countries, had lower levels of education compared with native Swedes.17 We could now also report that low educational level was a stronger predictor for subsequent DP in immigrants from non-Western countries than for natives. Potential explanations include differences in access to labour market and social insurance measures among immigrants with a culture that differs from the new country, but may also be explained by discrimination.23–25 The share of individuals with a high level of education is generally around the same among non-Western immigrants as among native Swedes.26 The most striking discrepancy in these two populations is that the share of individuals with a low level of education is much higher among non-Western immigrants, and that many of them, especially women, have difficulties with reading and writing.26 This might be one explanation for why low educational level among immigrants more often leads to permanent work disability. There is also literature suggesting that the educational level of immigrants in Sweden is under-reported to some extent.26 Strengths and limitations Strengths of this study include the population-based and prospective cohort design using a large study population. In addition, the follow-up period of 4 years was relatively long. The Swedish national registers used here are of good quality with almost no dropout, and missing data are also rare.27 28 The validity of the diagnoses of sick-leave spells has often been discussed, but a previous related study reported good validity.29 This study was also subject to some limitations. Findings of this study cannot be generalised to a population with depression and anxiety symptoms that did not come to the attention of the healthcare system. Previous studies have found barriers to seeking and accessing mental healthcare among immigrants, including cultural views on mental disorders and discrimination in the host country.19 24 25 This may lead to only the most severe cases coming to the attention of the healthcare system; in this case, immigrants' risk of DP in the case of specialised healthcare would be underestimated compared with that of native Swedes. Finally, immigrant groups to Sweden are very diverse. Owing to restrictions in our database, we unfortunately do not have data on specific country of birth or reason for immigration. Grouping immigrants by region of country of birth, that is, into Western and non-Western immigrants, could mask critical differences such as reason for migration.9 Conclusion Prescribed psychiatric medication, but not specialised healthcare, was a stronger predictor for permanent work disability in immigrants than in native Swedes. This warrants scrutiny of differences in care and treatment of immigrants and native Swedes with CMDs in the disabling process finally resulting in permanent work disability. Contributors: LW, MH and EM-R conceived and designed the study, interpreted the data and wrote and edited the manuscript. EM-R was the guarantor of this work. LW, MH and EM-R reviewed and approved the final version submitted for publication. Funding: This study was funded by the Swedish Research Council for Health, Working Life and Welfare, grant number 2015-00742. Competing interests: None declared. Ethics approval: Regional ethical review board in Stockholm, Sweden. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: We use data from three Swedish governmental authorities, Statistics Sweden, The Swedish Insurance Agency and the Swedish National Board of Health and Welfare. Data are available to research projects that have ethical approval from a regional ethics review board. ==== Refs References 1 Whiteford HA , Degenhardt L , Rehm J Global burden of disease attributable to mental and substance use disorders: findings from the Global Burden of Disease Study 2010 . Lancet 2013 ;382 :1575 –86 . 10.1016/S0140-6736(13)61611-6 23993280 2 Lerner D , Henke RM What does research tell us about depression, job performance, and work productivity? J Occup Environ Med 2008 ;50 :401 –10 . 10.1097/JOM.0b013e31816bae50 18404013 3 Simon GE , Barber C , Birnbaum HG Depression and work productivity: the comparative costs of treatment versus nontreatment . J Occup Environ Med 2001 ;43 :2 –9 . 10.1097/00043764-200101000-00002 11201765 4 Bültmann U , Christensen KB , Burr H Severe depressive symptoms as predictor of disability pension: a 10-year follow-up study in Denmark . Eur J Public Health 2008 ;18 :232 –4 . 10.1093/eurpub/ckm132 18202084 5 Ishtiak-Ahmed K , Perski A , Mittendorfer-Rutz E Risk markers of all-cause and diagnosis-specific disability pension—a prospective cohort study of individuals sickness absent due to stress-related mental disorders . BMC Public Health 2014 ;14 :805 10.1186/1471-2458-14-805 25099303 6 Mittendorfer-Rutz E , Härkänen T , Tiihonen J Association of socio-demographic factors, sick-leave and health care patterns with the risk of being granted a disability pension among psychiatric outpatients with depression . PLoS ONE 2014 ;9 :e99869 10.1371/journal.pone.0099869 24963812 7 Arslan C A New Profile of Migrants in the Aftermath of the Recent Economic Crisis. OECD Social, Employment and Migration Working Papers , OECD Publishing , 2014 ;No. 160 . 8 Migrationsverket . Verksamhets- och utgiftsprognos. (The Swedish Migration Agency: Operating and expenditure forecast) [In Swedish] Februari 2015 . 9 Gilliver SC , Sundquist J , Li X Recent research on the mental health of immigrants to Sweden: a literature review . Eur J Public Health 2014 ;24 :72 –9 . 10.1093/eurpub/cku101 25108001 10 Tinghög P , Al-Saffar S , Carstensen J The association of immigrant- and non-immigrant-specific factors with mental ill health among immigrants in Sweden . Int J Soc Psychiatry 2010 ;56 :74 –93 . 10.1177/0020764008096163 19592449 11 Barghadouch A , Carlsson J , Norredam M Psychiatric disorders and predictors hereof among refugee children in early adulthood: a register-based cohort study . J Nerv Ment Dis 2016 [Epub ahead of print 1 Aug 2016] 10.1097/NMD.0000000000000576 12 Ishtiak-Ahmed K , Perski A , Mittendorfer-Rutz E Predictors of suicidal behaviour in 36,304 individuals sickness absent due to stress-related mental disorders—a Swedish register linkage cohort study . BMC Public Health 2013 ;13 :492 10.1186/1471-2458-13-492 23687984 13 Migration and Asylum 2005–2015; Residence permits granted 1980–2015; Residence permits granted 1980–2014—refugees; Residence permits granted 1986–2014—family reunification; Residence permits granted 1986–2014—labour market; Residence permits granted 1986–2014—students; Overview and time series. Swedish Migration agency (Migrationsverket). https://www.migrationsverket.se/English/About-the-Migration-Agency/Facts-and-statistics-/Statistics/Overview-and-time-series.html 14 Helgesson M , Johansson B , Nordqvist T Sickness absence at a young age and later sickness absence, disability pension, death, unemployment and income in native Swedes and immigrants . Eur J Public Health 2015 ;25 :688 –92 . 10.1093/eurpub/cku250 25634955 15 Social Insurance Agency (SIA) . Social insurance in figures. Stockholm 2015. ISBN 978-91-7500-386-6 . 16 Mittendorfer-Rutz E , Kjeldgard L , Runeson B Sickness absence due to specific mental diagnoses and all-cause and cause-specific mortality: a cohort study of 4.9 million inhabitants of Sweden . PLoS ONE 2012 ;7 :e45788 .23049861 17 Hjern A Migration and public health Health in Sweden: The National Public Health Report 2012. Chapter 13 . Scand J Public Health 2012 ;40 :255 –67 . 10.1177/1403494812459610 23238411 18 Kirmayer LJ , Narasiah L , Munoz M Common mental health problems in immigrants and refugees: general approach in primary care . CMAJ 2011 ;183 :E959 –67 . 10.1503/cmaj.090292 20603342 19 Baarnhielm S , Javo C , Mosko MO Opening up mental health service delivery to cultural diversity: current situation, development and examples from three northern European countries . Adv Psychosom Med 2013 ;33 :40 –55 . 10.1159/000348730 23816862 20 Klinthäll M Ethnic background, labour market attachment and severe morbidity: hospitalisation among immigrants in Sweden 1990–2001 . Int Migr Integr 2008 ;9 :45 –61 . 10.1007/s12134-008-0043-2 21 Meershoek A , Krumeich A Multiculturalism and the construction of ethnic identities in labour and health practices: avoiding the culturalistic fallacy in applied research . Health Care Anal 2009 ;17 :173 –97 . 10.1007/s10728-008-0097-5 18758963 22 Hjern A High use of sedatives and hypnotics in ethnic minorities in Sweden . Ethn Health 2001 ;6 :5 –11 . 10.1080/13557850124836 11388086 23 Claussen B , Smeby L , Bruusgaard D Disability pension rates among immigrants in Norway . J Immigr Minor Health 2012 ;14 :259 –63 . 10.1007/s10903-010-9430-9 21188530 24 Pascoe EA , Smart Richman L Perceived discrimination and health: a meta-analytic review . Psychol Bull 2009 ;135 :531 –54 . 10.1037/a0016059 19586161 25 Hatch SL , Gazard B , Williams DR Discrimination and common mental disorder among migrant and ethnic groups: findings from a South East London Community sample . Soc Psychiatry Psychiatr Epidemiol 2016 ;51 :689 –701 . 10.1007/s00127-016-1191-x 26875153 26 OECD . Finding the way: a discussion of the Swedish migrant integration system. OECD International Migration Division, Directorate for Employment, Labour and Social Affairs. July 2014 . 27 Ludvigsson JF , Andersson E , Ekbom A External review and validation of the Swedish national inpatient register . BMC Public Health 2011 ;11 :450 10.1186/1471-2458-11-450 21658213 28 Statistiska Centralbyrån (SCB). Årsredovisning 2014 (Statistics Sweden – Annual Report 2014) [In Swedish] Statistics Sweden 2015. 29 Ljungdahl LO , Bjurulf P The accordance of diagnoses in a computerized sick-leave register with doctor's certificates and medical records . Scand J Public Health 1991 ;19 :148 –53 .

PMC005xxxxxx/PMC5372069.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01345510.1136/bmjopen-2016-013455EpidemiologyResearch1506169217241683Residential proximity to major roads, exposure to fine particulate matter and aortic calcium: the Framingham Heart Study, a cohort study http://orcid.org/0000-0002-3182-5633Dorans Kirsten S 12Wilker Elissa H 123Li Wenyuan 12Rice Mary B 24Ljungman Petter L 25Schwartz Joel 13Coull Brent A 36Kloog Itai 7Koutrakis Petros 3D'Agostino Ralph B 89Massaro Joseph M 810Hoffmann Udo 11O'Donnell Christopher J 81213Mittleman Murray A 12 1 Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, Massachusetts, USA 2 Cardiovascular Epidemiology Research Unit, Department of Cardiology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts, USA 3 Department of Environmental Health, Harvard T.H. Chan School of Public Health, Boston, Massachusetts, USA 4 Division of Pulmonary, Critical Care and Sleep Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts, USA 5 Unit of Environmental Epidemiology, Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden 6 Department of Biostatistics, Harvard T.H. Chan School of Public Health, Boston, Massachusetts, USA 7 Department of Geography and Environmental Development, Ben-Gurion University of the Negev, Beer Sheva, Israel 8 National Heart, Lung, and Blood Institute's and Boston University's Framingham Heart Study, Framingham, Massachusetts, USA 9 Department of Mathematics and Statistics, Boston University, Boston, Massachusetts, USA 10 Department of Biostatistics, Boston University School of Public Health, Boston, Massachusetts, USA 11 Cardiac MR PET CT Program, Department of Radiology, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, USA 12 Cardiology Section, Department of Medicine, Boston Veteran's Administration Health System, Boston, Massachusetts, USA 13 Cardiovascular Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, USACorrespondence to Murray A Mittleman; mmittlem@hsph.harvard.edu2017 16 3 2017 7 3 e01345513 7 2016 28 10 2016 22 12 2016 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives Traffic and ambient air pollution exposure are positively associated with cardiovascular disease, potentially through atherosclerosis promotion. Few studies have assessed associations of these exposures with thoracic aortic calcium Agatston score (TAC) or abdominal aortic calcium Agatston score (AAC), systemic atherosclerosis correlates. We assessed whether living close to a major road and residential fine particulate matter (PM2.5) exposure were associated with TAC and AAC in a Northeastern US cohort. Design Cohort study. Setting Framingham Offspring and Third Generation participants residing in the Northeastern USA. Participants and outcome measures Among 3506 participants, mean age was 55.8 years; 50% female. TAC was measured from 2002 to 2005 and AAC up to two times (2002–2005; 2008–2011) among participants from the Framingham Offspring or Third Generation cohorts. We first assessed associations with detectable TAC (logistic regression) and AAC (generalised estimating equation regression, logit link). As aortic calcium scores were right skewed, we used linear regression models and mixed-effects models to assess associations with natural log-transformed TAC and AAC, respectively, among participants with detectable aortic calcium. We also assessed associations with AAC progression. Models were adjusted for demographic variables, socioeconomic position indicators and time. Results There were no consistent associations of major roadway proximity or PM2.5 with the presence or extent of TAC or AAC, or with AAC progression. Some estimates were in the opposite direction than expected. Conclusions In this cohort from a region with relatively low levels of and variation in PM2.5, there were no strong associations of proximity to a major road or PM2.5 with the presence or extent of aortic calcification, or with AAC progression. EPIDEMIOLOGYPUBLIC HEALTHNational Heart, Lung, and Blood Institutehttp://dx.doi.org/10.13039/100000050HHSN268201500001IN01-HC-25195T32HL007575National Institute of Environmental Health Scienceshttp://dx.doi.org/10.13039/100000066F32ES023352R00 ES022243Environmental Protection Agencyhttp://dx.doi.org/10.13039/501100001589RD-83479801 ==== Body Strengths and limitations of this study In this study, we used two measures of aortic calcification, abdominal aortic calcium score (AAC) and thoracic aortic calcium score (TAC), which are markers of systemic atherosclerosis. We assessed the associations of two different exposures, residential proximity to a major roadway and residential exposure to fine particulate matter (PM2.5), with the presence and extent of TAC and AAC and with AAC progression. Many participants did not have detectable calcification and this study was carried out in the Northeastern USA, a region with relatively low levels of and variation in PM2.5, which might have limited our ability to detect associations with aortic calcification and may limit generalisability of the results. Because we did not randomly allocate where people live, there is the possibility for residual or unmeasured confounding, though we did adjust for many potential confounders. Exposure measurement error and the relatively short time period between AAC measurements (median 6.2 years) may have limited power to detect associations with AAC progression. Introduction Many studies have found that long-term particulate matter air pollution exposure is associated with cardiovascular disease (CVD) morbidity and mortality.1–5 Particulate matter air pollution may lead to cardiovascular disease by initiating or accelerating atherosclerosis. In susceptible animal models, exposure to particulate matter has been shown to lead to atherosclerosis progression.6–8 Epidemiological studies have also suggested ambient air pollution and traffic are positively associated with markers of atherosclerosis.9–16 Possible mechanisms through which particulate matter could lead to atherosclerosis promotion include imbalance of the autonomic nervous system, a systemic oxidative stress and inflammatory response and, possibly, the transportation of particles directly into arterial blood circulation.1 17–19 Even in regions with relatively low levels of fine particulate matter (aerodynamic diameter ≤2.5 µm, PM2.5), studies have reported positive associations between PM2.5 and cardiovascular or all-cause mortality.20 21 For instance, PM2.5 has been positively associated with mortality in New England.22 In Massachusetts, traffic exposure has been linked to adverse cardiovascular outcomes23–25 and PM2.5 to acute myocardial infarction.26 Among elderly men living in the Boston area, carotid intima-media thickness (CIMT), a marker of subclinical atherosclerosis, was associated with black carbon, a traffic correlate.27 Though many studies have found positive associations between particulate matter air pollution and CIMT,14 15 27 only two have explored associations of these exposures with aortic calcification,13 28 a marker of systemic atherosclerosis and of a later stage of atherosclerotic disease than CIMT. Calcification in the thoracic aorta or abdominal aorta is correlated with other markers of subclinical atherosclerosis29–33 and predicts incident cardiovascular disease.34–41 In a study in Germany, PM2.5 and night-time traffic noise were associated with thoracic aortic calcium Agatston score (TAC).13 Living closer to a highly trafficked road was associated with TAC, though this was attenuated after adjusting for night-time traffic noise.13 In a multicity US study, residential PM2.5 exposure was weakly associated with higher risk of detectable abdominal aortic calcium Agatston score (AAC);28 there was no association with the quantity of AAC.28 There was no association of residential proximity to a major roadway with AAC. We built on prior work and assessed associations of residential distance to a major roadway and residential exposure to spatially resolved average PM2.5 with TAC, AAC and AAC progression among Framingham Offspring or Third Generation cohort participants living in the Northeastern USA. TAC and AAC were measured from 2002 to 2005 and AAC was measured again from 2008 to 2011. During this period, PM2.5 levels in this Northeastern US study region were generally lower than levels in German-based13 and US-based28 studies that assessed associations with TAC and AAC. This work adds to studies that have assessed associations of particulate matter or traffic with other markers of atherosclerosis in Boston27 and the Northeastern USA.42 Methods Study participants The study population consists of participants from the Framingham Offspring and Third Generation cohorts, which have previously been described.43 44 A subset of the participants underwent the first round of multidetector CT (MDCT) scans from 2002 to 2005.29 45 For this substudy, exclusion criteria included weighing ≥350 pounds, age<35 years (men) or <40 years (women) and pregnancy. From 2008 to 2011, some participants scanned from 2002 to 2005 and additional Framingham Heart Study (FHS) participants underwent MDCT scans. Overall, 3530 participants (Offspring: 1419; Third Generation: 2111) had aortic calcium measured from 2002 to 2005 and 2749 participants (Offspring: 1320; Third Generation: 1429) had AAC measured from 2008 to 2011. Age of participants ranged from early adulthood to older age (many participants were middle aged), approximately half were women and participants were largely free of clinically apparent CVD. For these analyses, we included participants who had at least one aortic calcium score from the first or second MDCT period, lived in the Northeastern USA and were not missing exposure or covariate information. For scans carried out during the first MDCT round, we included participants who attended either Offspring Examination 7 (1998–2001) or Third Generation Examination 1 (2002–2005). For scans carried out during the second round, we included participants who attended either Offspring Examination 8 (2005–2008) or Third Generation Examination 2 (2008–2011). All participants provided written consent for FHS examinations and the MDCT scans. Participant characteristics At examination visits, each participant underwent a physical examination, physician interview and laboratory tests. Details of visits44 46 and descriptions47 of blood pressure measurement, definitions of diabetes and use of antihypertensive and lipid-lowering medications have previously been described. Following published FHS criteria,44 46 a panel of three investigators determined history of clinically apparent CVD (coronary heart disease, intermittent leg claudication, heart failure, stroke or transient ischaemic attack). Using each study participant's primary address, we used US Census 2000 data to assign the median value of owner-occupied housing units at the census-tract level, an area-level marker of socioeconomic position. Exposure assessment Distance to roadway: ArcGIS was used to geocode primary addresses of participants (Environmental Systems Research Institute (Esri), Redlands, CA). Using address reported at Offspring Examination 7 or Third Generation Examination 1, we evaluated distance to nearest major roadway, defined as A1, A2 or A3 road (US Census Features Class). Research has shown that particulate matter mass concentration and elemental carbon approach urban background levels within ∼100 to 400 m from major roadways.48 In an effort to reflect the decay of traffic-related pollution levels, we categorised distance to major roadway as <50, 50 to <200, 200 to <400 and 400 to <1000 m. We also assessed the association between the natural logarithm of distance to a major roadway and aortic calcium, based on our prior findings of log-linear associations between distance to a major roadway and cardiovascular health outcomes.24 47 49 For natural log of distance to nearest major roadway, we present results scaled to the IQR of distance to a major roadway, contrasting participants living 417.9 m to those living 58.0 m from the nearest major roadway. Participants residing 1000 m or further from a major roadway are likely to reside in rural areas, in which distance to a major road is unlikely to be an indicator of traffic-related pollution exposure. We therefore excluded observations from participants living ≥1000 m from a major road (572 observations, 11%) in all analyses of distance to roadway. Spatially resolved average PM2.5: We used estimates of daily PM2.5 levels at participant residential address, estimated by a spatiotemporal model to calculate each participant's average residential PM2.5 exposure. Kloog et al50 predicted daily PM2.5 at a scale of 1×1 km across the Northeastern USA (New England, New York and New Jersey) using satellite-derived measurements of aerosol optical depth (AOD; a quantitative measure of particle abundance in the atmospheric column), PM2.5 measurements from monitoring stations and meteorological and land-use regression terms. There was excellent out-of-sample 10-fold cross-validated R2 (mean out of sample R2=0.88; year-to-year variation 0.82–0.90) and a cross-validated slope of observed versus predicted PM2.5 concentrations of 0.99 (year-to-year variation 0.98–1.01) among 1×1 km grid cells that had a PM2.5 monitor measurement and AOD measurement.50 For each PM2.5 monitoring site, daily PM2.5 residuals (predicted vs measured) were regressed against local spatial and temporal variables. Using the fit of this model, we estimated daily predictions at home address for each residential address that represent deviations from grid predictions. To estimate total daily PM2.5, we summed the daily grid prediction and daily local residual PM2.5 prediction corresponding to each address. We then averaged these daily total PM2.5 predictions over the year to estimate annual average residential PM2.5. As has been previously done,47 51–53 we used the same index year for all observations. This approach aims to preserve PM2.5 spatial variation among residential locations, while minimising PM2.5 secular trend influences. For the main analyses, we chose the year 2003. To assess sensitivity of the results to choice of PM2.5 index year, we also ran analyses using PM2.5 averaged from 2003 to 2009. We scaled results to the IQR of 2003 PM2.5 (1.4 µg/m3). Aortic artery calcium For the first round of scans, eight-slice MDCT technology was used to carry out scans of the chest and abdomen (LightSpeed Ultra, General Electric, Milwaukee, Wisconsin, USA).29 45 For the second round, 64-slice MDCT scanners (General Electric Discovery VCT 64-slice PET/CT scanner, GE Healthcare) were used. For the chest, 48 2.5 mm slices were acquired from the carina to the diaphragm (120 kVp, 500 ms gantry rotation time, tube current 320/400 mA (<100 and ≥100 kg of body weight, respectively29). In the abdomen, using the top of the S1 vertebrae as the anatomic landmark of the lower field, slices were obtained for a total coverage of 15 cm in the Z-direction.29 Abdominal imaging parameters were: 120 kVp, 400 mA and gantry rotation time of 500 ms.29 A calcified lesion was defined as an area of ≥3 connected pixels with an attenuation >130 Hounsfield units.45 54 A modified Agatston score was calculated by multiplying the area of each lesion by a weighted attenuation score based on maximal lesion attenuation. TAC was measured twice in each participant (sequential scans); we use average TAC from these two scans. Statistical analyses Owing to the zero-inflated and heavily right-skewed distribution of TAC, we used a two-step approach to separately assess associations of exposure with detectable TAC and then with the amount of TAC among participants with TAC>0. For the binary outcome of detectable TAC, we used logistic regression. We used linear regression models to examine associations with the amount of TAC among participants with TAC>0. Owing to its skewed distribution, we modelled natural log-transformed TAC. We estimated 95% CIs by bootstrapping 1000 samples and reported percentile CIs, as we did not expect residuals to be normally distributed.55 We used a similar approach for AAC. To account for repeated AAC measures among 51% of participants, we used generalised estimating equation logistic regression, robust SEs and an unstructured correlation structure for the binary outcome of detectable AAC. To assess associations with the amount of natural log-transformed AAC among participants with AAC>0, we used linear mixed-effects models (95% CIs block bootstrapped). We also assessed associations with AAC incidence and with annual change in AAC. We defined AAC incidence as progressing from AAC=0 to detectable AAC (AAC>0). Prior work in FHS has shown that CVD risk factors were associated with incident AAC.56 We used logistic regression to evaluate associations with odds of incident AAC. Next, we used a linear mixed-effects modelling approach that simultaneously assessed associations of exposure with a participant's first measured AAC and change in AAC per year (the latter assessed by including interaction terms of time between scans with exposure; block bootstrapped CIs).21 54 55 We used all observations for this second approach, including participants scanned at one or both rounds of MDCT scans. We adjusted for covariates considered a priori to be potential confounders of the air pollution and atherosclerosis relationship. In all models, we adjusted for age and age2 at scan, sex, body mass index, smoking status (current, former, never), pack-years, individual-level educational attainment (high school or less, some college, college graduate), median census-tract value of owner-occupied housing units (quartiles), cohort (Offspring or Third Generation) and time. We focused on parameter estimation and the overall pattern of results, rather than results of specific hypothesis tests, and so do not report p values. We present all point estimates with 95% CIs. We used SAS V.9.3 (SAS Institute, Cary, North Carolina, USA) and Stata V.12 (StataCorp, College Station, Texas, USA) for analyses. Sensitivity analyses To assess robustness of our findings, we ran sensitivity analyses. We assessed whether results were similar when we only adjusted for age and sex. We also assessed whether results changed when we added other potential confounders. First, we additionally adjusted for potential confounders or predictors of aortic calcium: physical activity (tertiles of physical activity index57), alcohol (0, 0 to 7 and ≥7 drinks per week) and menopausal status in women (periods stopped for 1 year or more before examination). We then further adjusted for variables that could be confounders, outcome predictors or mediators: diabetes, systolic blood pressure, diastolic blood pressure, total cholesterol/high-density lipoprotein (HDL) ratio, triglycerides, antihypertensive medication and lipid-lowering medication. We tested for heterogeneity of associations between distance to major roadway and PM2.5 with the presence or extent of aortic calcium by: sex, age (≤65, >65), cohort (Offspring, Third Generation) and American College of Cardiology/American Heart Association's 10-year predicted risk of atherosclerotic CVD (tertiles: 0.1–1.6%, 1.6–5.3%, 5.3–81.3%).58 To run these analyses, we included interaction terms between the exposure and covariate of interest. Owing to sparse data, we did not assess for heterogeneity by 10-year atherosclerotic CVD risk for the outcomes of detectable TAC or loge (TAC) or by age for the outcome of detectable AAC. We also assessed whether results differed when we restricted to participants without clinically apparent CVD (4891 observations, 92%). We assessed associations of aortic calcium with residential distance to nearest A1 or A2 road (rather than nearest A1, A2 or A3 road), as A1 and A2 roads are larger than A3 roads. In an effort to take into account long-term secular PM2.5 trends, we adjusted for MDCT scan year as a categorical variable. We also ran AAC analyses separately for the first and second rounds of MDCT scans, to assess whether results differed by period. We used restricted cubic splines to assess linearity of the exposure–outcome relationships for natural log distance to roadway and PM2.559 and created plots using the POSTRCSPLINE package in Stata.60 To assess the robustness of the generalised estimating equation regression results, we reran the binary AAC>0 analyses using mixed-effects logistic regression. In line with prior work,13 we used linear regression to assess associations with loge(TAC+1) and mixed-effects linear regression to assess associations with loge(AAC+1). As in prior FHS analyses,61 we also assessed associations with TAC and AAC greater than the 75th or 90th age-specific and sex-specific centiles among a healthy referent subpopulation, using logistic regression for TAC and generalised estimating equation regression (logit link, unstructured covariance and robust SEs) for AAC. Results Study participants Participant characteristics are listed in table 1 and online supplementary table S1. Of the 6279 aortic calcium scores measured during either the first or second MDCT round, these analyses included 5328 observations from 3506 participants. Mean age of participants during the first MDCT round was 52.6 years, and for the second round was 59.6 years. In total, women contributed 50% of observations. In the second round of MDCT scans, participants were more likely to report being on hypertension medication (37% vs 20%) or lipid medication (40% vs 15%) and were less likely to be current smokers (8% vs 13%) than participants from the first round of MDCT scans. Table 1 Characteristics of study participants Characteristics mean±SD or n (%) CT 2002–2005 (N=2945) CT 2008–2011 (N=2383) Age at CT scan (years) 52.6±11.8 59.6±12.0 Male sex, % 1506 (51) 1165 (49) Offspring, % 1201 (41) 1155 (48) Education (years)  Some college 954 (32) 771 (32)  College graduate 1301 (44) 1063 (45) Median census value owner-occupied housing, $ 221 976±100 340 222 381±102 327  Current smokers, % 381 (13) 179 (8)  Former smokers, % 1109 (38) 1027 (43) Pack-years  Current smokers 30.5±14.4 33.5±15.7  Former smokers 17.5±17.9 17.5±17.1 Alcohol (average drinks per week) 4.9±7.3 4.6±7.3 Physical Activity Index 37.6±7.4 36.1±6.3 Menstrual periods stopped*, % 761 (53) 936 (77) Diabetes history, % 168 (6) 223 (9) Clinically apparent CVD at CT scan, % 216 (7) 221 (9) Body mass index (kg/m2) 27.8±5.3 28.5±5.3 Hypertension medications, % 590 (20) 891 (37) Systolic blood pressure (mm  Hg) 122±16 123±16 Diastolic blood pressure (mm  Hg) 76±9 74±9 Lipid medications, % 438 (15) 957 (40) Triglycerides (mg/dL) 129±92 118±79 Total cholesterol/HDL 4.0±1.3 3.5±1.1 10-year predicted risk of CVD†, % 0.05±0.07 0.07±0.09 MDCT scan results TAC>0, % 716 (25) N/A  TAC, among those with TAC>0‡ 156.6 (712.8) N/A AAC>0, % 1608 (55) 1604 (67)  AAC, among those with AAC>0‡ 464.6 (2066.8) 745.5 (3107.0) *Among women. †American College of Cardiology/American Heart Association 2013 10-year predicted risk of atherosclerotic CVD; Median, IQR.58 ‡Median, IQR. Data calculated from 5328 observations, from 3506 participants with a TAC measurement or at least one AAC measurement. Number of missing observations: TAC: 45; AAC: 16; alcohol: 4; physical activity index: 47; diastolic blood pressure: 4; triglycerides: 10; total cholesterol/HDL: 15; 10-year predicted CVD risk: 1223. AAC, abdominal aortic calcium Agatston score; CVD, cardiovascular disease; HDL, high-density lipoprotein; MDCT, multidetector CT; TAC, thoracic aortic calcium Agatston score. 10.1136/bmjopen-2016-013455.supp1supplementary tables Exposure distributions Exposure distributions are summarised in table 2. Twenty-three per cent of observations came from participants who lived within 50 m of a major roadway; the median distance to the closest major roadway was 201 m. In 2003, median residential PM2.5 concentration was 10.7 µg/m3. The current annual PM2.5 US Environmental Protection Agency (EPA) annual National Air Quality Standard (NAAQS) is 12 µg/m3. Table 2 Distributions of distance to roadway, PM2.5 Exposure Median (IQR) or n (%) Range (min, max) Range (5th–95th) Proximity to major roadway (m)* 201 (360) 0.01–999.7 6.9–816.1 Total PM2.5 (µg/m3), 2003† 10.7 (1.4) 2.9–26.7 8.2–12.6 Total PM2.5 (µg/m3), 2003–2009† 9.8 (1.1) 2.6–17.2 7.3–11.1 Residential proximity in categories  <50 m 1104 (23%)  50 to <200 m 1265 (27%)  200 to <400 m 1114 (23%)  400 to <1000 m 1273 (27%) *Proximity to a major roadway analyses restricted to individuals living <1000 m from a major road; 4756 observations (572 observations, 11% excluded). †PM2.5 2003 calculated from 5328 observations. PM2.5 2003–2009 calculated from 5326 observations. PM2.5, fine particulate matter. Associations with odds of detectable aortic calcium and continuous aortic calcium There were no consistent associations of residential proximity to a major roadway or PM2.5 with the presence or extent of TAC or AAC (table 3). Some of the associations were in the opposite direction than expected. Table 3 Associations of distance to a major roadway and PM2.5 with TAC and AAC* TAC† TAC>0 Linear regression (among those with TAC>0) OR 95% CI Per cent difference 95% CI <50 m 0.85 (0.59 to 1.22) −0.6 (−30.9 to 52.0) 50 to<200 m 0.70 (0.49 to 1.00) −0.4 (−35.4 to 51.0) 200 to<400 m 0.91 (0.64 to 1.29) −16.1 (−43.5 to 28.7) 400 to<1000 m Log of distance to major road‡ 1.06 (0.90 to 1.24) −2.6 (−19.1 to 14.4) 2003 PM2.5 (µg/m3)§ 0.89 (0.78 to 1.02) 2.4 (−13.0 to 20.8) 2003–2009 PM2.5 (µg/m3)§ 0.88 (0.76 to 1.01) 19.6 (−1.4 to 45.9) AAC¶ AAC>0 Linear mixed effects (among those with AAC>0) OR 95% CI Per cent difference 95% CI <50 m 0.85 (0.65 to 1.11) 3.2 (−12.9 to 22.8) 50 to <200 m 0.79 (0.62 to 1.00) −11.8 (−23.1 to 5.8) 200 to <400 m 0.96 (0.75 to 1.23) −3.8 (−17.4 to 13.5) 400 to <1000 m 1.00 Log of distance to major road‡ 1.09 (0.96 to 1.23) −0.4 (−8.6 to 7.9) 2003 PM2.5 (µg/m3)§ 0.99 (0.90 to 1.08) 2.2 (−3.2 to 9.9) 2003–2009 PM2.5 (µg/m3)§ 0.97 (0.89 to 1.07) 5.8 (−0.1 to 14.5) *Adjusted for age, age2 body mass index, sex, cohort (Offspring, Third Generation), smoking status (current, former, never), pack-years, individual-level education (high school or less, some college, college graduate), median census-tract value of owner-occupied housing (quartiles), date of CT scan, number of days between CT scan and examination at which individual-level covariates reported. For AAC, also adjusted for CT scan (first or second round). †For TAC>0 logistic regression, distance to roadway includes 2576 observations (647 TAC>0). PM2.5 2003 includes 2900 observations (716 TAC>0); PM2.5 2003–2009 includes 2899 observations (716 with TAC>0). For linear regression, distance to roadway includes 647 observations and PM2.5 716 observations; percentile bootstrapped CIs (n=1000 bootstrap samples). ‡Natural log of proximity to a major road scaled to the difference between living at the 75th (417.9 m) versus the 25th (58.0 m) centile from a major road. §PM2.5 scaled to the IQR for the 2003 average (1.4 µg/m3). ¶For AAC>0 GEE model, distance to roadway includes 4743 observations (2902 AAC>0). PM2.5 2003 includes 5312 observations (3212 AAC>0). PM2.5 2003–2009 includes 5310 observations (3212 AAC>0). For linear mixed-effects regression, distance to roadway includes 2902 observations and PM2.5 3212 observations. Percentile bootstrap CIs (n=1000 clustered bootstrap samples). AAC, abdominal aortic calcium Agatston score; PM2.5, fine particulate matter; TAC, thoracic aortic calcium Agatston score; Associations with change in AAC There were 1811 participants with AAC measured during MDCT round one and round two; 818 participants had no detectable AAC during MDCT round one. Of these, 30% had detectable AAC during round two. We found no association of residential distance to major roadway or PM2.5 with incident AAC (table 4, figure 1). Distance to nearest major road and PM2.5 were also not associated with annual change in AAC (table 4, figure 1). Table 4 Associations of distance to a major roadway, PM2.5 with incidence of newly detectable AAC and change in AAC Incident AAC* Annual change in AAC† OR 95% CI Change 95% CI <50 m 0.74 (0.45 to 1.23) 23.4 (−6.7 to 55.9) 50 to <200 m 0.66 (0.41 to 1.07) −28.9 (−59.4 to −3.1) 200 to <400 m 0.74 (0.44 to 1.23) −9.1 (−31.7 to 17.4) 400 to <1000 m 1.00 Log of distance to a major road‡ 1.15 (0.91 to 1.46) −5.8 (−21.7 to 7.4) 2003 PM2.5 (µg/m3)§ 0.93 (0.79 to 1.10) 1.1 (−6.7 to 9.7) 2003–2009 PM2.5 (µg/m3)§ 0.95 (0.79 to 1.14) 7.5 (−2.6 to 16.3) *Adjusted for age at MDCT1 (age, age2) and covariates reported at Offspring Examination 7 or Generation 3 Examination 1: body mass index, sex, cohort, smoking status (current, former, never), pack-years, individual-level education (high school or less, some college, college graduate), median census-tract value of owner-occupied housing (quartiles), date of 1st scan, number of days between scan and examination at which individual-level covariates reported, time between MDCT1 and MDCT2. †Adjusted for age at 1st scan (age, age2, body mass index, sex, cohort, scan (first or second round), smoking status (current, former, never), pack-years, individual-level education (high school or less, some college, college graduate), median census-tract value of owner-occupied housing (quartiles), date of scan, number of days between scan and examination at which individual-level covariates reported, time since 1st scan. Includes the following interaction terms with time since first scan: age at first scan (age, age2, sex, cohort. Percentile bootstrap CIs (n=1000 clustered bootstrap samples). ‡Natural log of proximity to a major road scaled to the difference between living at 75th (417.9 m) versus 25th (58.0 m) centile from a major road. For AAC incidence, 716 observations (215 with incident AAC). For change in AAC mixed model, all 4743 observations are included (1625 with two AAC measurements). §PM2.5 scaled to the IQR for the 2003 average (1.4 µg/m3). For 2003 PM2.5 AAC incidence, 818 observations (248 with incident AAC). For 2003–2009 PM2.5 AAC incidence, 817 observations (248 with incident AAC). For 2003 PM2.5 change in AAC mixed model, all 5312 observations are included (1811 with two AAC measurements). For 2003–2009 PM2.5 change in AAC mixed model, all 5310 observations are included (1810 with two AAC measurements). AAC, abdominal aortic calcium Agatston score; PM2.5, fine particulate matter. Figure 1 Associations of distance to a major roadway, PM2.5 with incidence of newly detectable AAC and change in AAC. (A) adjusted for age at MDCT1 (age, age2) and covariates reported at Offspring Examination 7 or Generation 3 Examination 1: body mass index, sex, cohort, smoking status (current, former, never), pack-years, individual-level education (high school or less, some college, college graduate), median census-tract value of owner-occupied housing (quartiles), date of 1st scan, number of days between scan and examination at which individual-level covariates reported, time between MDCT1 and MDCT2. (B) Adjusted for age at 1st scan (age, age2, body mass index, sex, cohort, scan (first or second round), smoking status (current, former, never), pack-years, individual-level education (high school or less, some college, college graduate), median census-tract value of owner-occupied housing (quartiles), date of scan, number of days between scan and examination at which individual-level covariates reported, time since 1st scan. Includes the following interaction terms with time since first scan: age at first scan (age, age2, sex, cohort. Percentile bootstrap CIs (n=1000 clustered bootstrap samples). Natural log of proximity to a major road scaled to the difference between living at 75th (417.9 m) versus 25th (58.0 m) centile from a major road. For AAC incidence, 716 observations (215 with incident AAC). For change in AAC mixed model, all 4743 observations are included (1625 with two AAC measurements). PM2.5 scaled to the IQR for the 2003 average (1.4 µg/m3). For 2003 PM2.5 AAC incidence, 818 observations (248 with incident AAC). For 2003–2009 PM2.5 AAC incidence, 817 observations (248 with incident AAC). For 2003 PM2.5 change in AAC mixed model, all 5312 observations are included (1811 with two AAC measurements). For 2003–2009 PM2.5 change in AAC mixed model, all 5310 observations are included (1810 with two AAC measurements). AAC, abdominal aortic calcium Agatston score; MDCT, multidetector CT; PM2.5, fine particulate matter. Sensitivity analyses We did not find evidence of non-linearity of the associations between any of the exposures with the presence or extent of aortic calcium, or with the incidence of AAC. When we only adjusted for age and sex, results were not changed substantially. We observed similar results when we also adjusted for physical activity, alcohol, menopausal status, diabetes, systolic blood pressure, diastolic blood pressure, total cholesterol/HDL ratio, triglycerides, antihypertensive medication and lipid-lowering medication or when we adjusted for year of CT scan as categorical variable. We did not observe consistent patterns of heterogeneity of associations with the presence or extent of aortic calcium by age, sex (men vs women), cohort or predicted 10-year risk of atherosclerotic CVD (see online supplemental table S2). Of note, there was sparse data in some subgroups. When we restricted to participants without clinically apparent CVD (92%), results were similar. When we used a different PM2.5 averaging period (2003–2009 instead of 2003), some point estimates were stronger, though CIs were wide (tables 3 and 4). We did not observe consistent associations of distance to nearest A1 or A2 road with the presence or extent of aortic calcium or with the incidence of AAC, and some of the associations were in the opposite direction than expected. Results were similar when we ran the AAC analyses separately by MDCT round. When we ran the AAC>0 analysis with a mixed logistic regression, instead of the generalised estimating equation regression, we obtained similar results, though with wider CIs. Distance to a major roadway and PM2.5 were not strongly associated with loge(TAC+1) or with TAC greater than the 75th or 90th age-specific and sex-specific healthy referent cut points. Similarly, we did not see associations with loge (AAC+1) or with AAC greater than the 75th or 90th age-specific and sex-specific healthy cut points. Discussion We assessed two residential exposures: distance to a major roadway, which is a surrogate of local traffic exposure, and PM2.5, which is emitted by a variety of local and regional pollution sources, including traffic. We found no consistent or strong associations between these exposures and the presence or extent TAC or AAC or with AAC progression. These findings were robust across a range of sensitivity analyses. This study has several limitations. Though there is potential for residual or unmeasured confounding, we adjusted for many potential confounders, including individual and area-level markers of socioeconomic position. TAC and AAC are measured with error and are correlates, not direct estimates of systemic atherosclerosis. However, TAC and AAC predict incident CVD34–39 and are correlated with other markers of atherosclerosis.29–33 We previously reported no strong associations of these exposures with coronary artery calcium Agatston scores (CAC), another marker of atherosclerosis, in FHS.42 However, our study would have been stronger if we had assessed associations with other atherosclerosis markers, such as CIMT. We did not assess associations with other markers of traffic, such as nitrogen dioxide. We do not have information on time activity patterns or time spent at home and our exposures are measured with error. However, we aimed to use residential location to estimate order of exposure levels. Importantly, we do not expect exposure measurement error to be related to TAC or AAC. We did not model year-to-year variability in exposures. However, results were similar when we adjusted for year of CT scan as a categorical variable in an effort to adjust for long-term time trends. Additionally, results were generally robust to choice of PM2.5 index period. We did not differentiate between type of major roadway (A1, A2, and A3); however, results were generally similar when we restricted to distance to A1 or A2 roadway only. Finally, the study population is predominantly of European ancestry and resides in the Northeastern USA. Though the study population included older adults, many participants were middle aged or younger. Results might not be generalisable to other populations. There are also several strengths of this study. Unlike prior studies, we assess two measures of aortic calcification—in the thoracic and abdominal aorta—in the same study population. We assessed associations with incident AAC and annual change in AAC. We used a validated, spatially and temporally resolved PM2.5 model based on satellite and land use data to estimate individual exposure, rather than relying on community monitors. Additionally, we explored different exposures: distance to major roadway as a correlate of local traffic, and PM2.5, which has local and regional sources. The two prior studies that assessed associations of distance to a major roadway or PM2.5 with aortic calcium score supported the presence of cross-sectional associations between proximity to major roadway and TAC and between PM2.5 and TAC and AAC.13 28 In the Heinz Nixdorf Recall Study, carried out in an industrial region of Germany, residential PM2.5 was associated with the extent of TAC, even after adjusting for night-time traffic noise.13 Living close to a highly trafficked road was also associated with TAC, though this was attenuated when adjusted for night-time noise.13 In the Multiethnic Study of Atherosclerosis (MESA) study, based in the USA, higher residential PM2.5 exposure was weakly associated with a higher risk of detectable AAC.28 Associations were slightly stronger among participants whom the authors expected to have less exposure measurement error—those with long-term residence near a PM2.5 monitor and those who did not recently work outside of the home. There were no associations of PM2.5 with the extent of AAC or of distance to a major roadway with the presence or extent of AAC. Unlike prior studies, we assessed associations of these exposures with TAC and AAC in the same study region and population. We were also able to assess associations with AAC progression. We did not observe associations of these outcomes with residential distance to a major roadway or residential estimates of PM2.5. In analyses of associations with the presence of detectable TAC and AAC, we observed some estimates that were in the opposite direction than expected. For instance, higher PM2.5 exposure was associated with lower odds of detectable TAC, though CIs were wide. There is no clear biological rationale for this direction of association, as we would expect that higher PM2.5 would be associated with higher odds of detectable TAC. Though we adjusted for many potential confounders, these results could possibly be due to chance or residual confounding. The lack of association with AAC progression may have been due to the relatively short time period between AAC scans (median 6.2 years). Importantly, there was little variation in average PM2.5 levels in our study, which reduced power to observe associations between PM2.5 and aortic calcification. Additionally, PM2.5 levels were lower in our study compared with the two existing studies that have studied air pollution and TAC or AAC. For instance, in our cohort, 2003 annual average PM2.5 predicted at residential address had a median of 10.7 µg/m3, while median 365-day PM2.5 average was 16.6 µg/m3 in the Heinz Nixdorf Recall Study.13 Additionally, use of diesel is much more common in Germany than in the USA and diesel exhaust may be more harmful than gasoline exhaust. Unlike our study, the Heinz Nixdorf Recall Study found an association between PM2.5 and the extent of TAC.13 As in MESA, we did not observe a consistent association between distance to a major roadway or PM2.5 with the extent of AAC.28 However, in contrast to our study, Allen and colleagues found a suggested weak positive association of PM2.5 with detectable AAC among MESA participants.28 In the MESA study region, mean 2000–2002 average PM2.5 ranged from 10.9 µg/m3 (St. Paul) to 22.8 µg/m3 (Los Angeles). Studies have also assessed associations of traffic and particulate matter air pollution with CAC, which provides an estimate of total coronary atheroma62 and predicts coronary heart disease.63 We did not find evidence of strong associations of residential proximity to a major road or PM2.5 with the presence, extent or progression of CAC.42 In the Heinz Nixdorf Recall Study, residing closer to a major road (mean daily vehicle counts of 10 000–130 000) was associated with CAC; PM2.5 was only associated among study participants who had not recently been employed full time.9 In line with our findings in FHS,42 in cross-sectional analyses, PM2.5 was not consistently associated with CAC in MESA,10 though there was an association of PM2.5 and nitrogen oxides with CAC progression in MESA.16 Importantly, the lack of associations of these exposures with artery calcification in our study population does not mean that ambient air pollution does not cause atherosclerosis. When considering policy implications, these results should be interpreted within the broader context of air pollution and cardiovascular disease literature. Studies in susceptible animals have found particulate matter exposure can lead to atherosclerosis progression.6–8 Additionally, controlled human and animal studies have found evidence that short-term ambient air pollution exposure can lead to proinflammatory and oxidative stress responses and autonomic nervous system imbalance, which could potentially lead to atherosclerosis.1 17–19 Age is strongly related to the presence and extent of detectable artery calcification. Though participant ages ranged from early to late adulthood, many were middle aged or younger and had little or no detectable calcification, thereby reducing power to detect associations in this study. As noted above, the Heinz Nixdorf Recall Study and MESA found evidence of positive associations of these exposures with artery calcification. Additionally, many studies have reported positive associations between particulate matter air pollution and CIMT.14 15 27 Importantly, CIMT represents an earlier stage in vascular injury64 than artery calcification, which only directly measures calcified elements. Perhaps relatively recent particulate matter pollution exposure contributes more to the development of soft plaque than to progression to arterial calcification, which may be more influenced by longer term exposures. Potential future work could explore whether, in this population, PM2.5 and residential proximity to roadway are associated with measures of soft plaques. In summary, in a cohort residing in a region with relatively low levels of and variation in PM2.5, we observed no strong associations between residential distance to a major roadway or PM2.5 with the presence, extent or progression of aortic calcification. These findings add to the prior studies exploring the association of traffic and PM2.5 with aortic calcification in humans. We thank the Framingham Offspring and Third Generation Study participants. We also thank the editors at BMJ Open and the reviewers, Dr Mark Miller and Dr Jess Lambrechtsen, for their thoughtful feedback. Contributors: All authors have contributed to the conception or design of the study and acquisition of the data. KSD analysed the data and EHW and WL reviewed the analysis. KSD wrote the first draft of the manuscript and all authors critically revised the manuscript. All authors have approved the manuscript. Funding: This work was supported by the National Heart, Lung, and Blood Institute (grant number T32HL007575) and the National Institute of Environmental Health Sciences (grant numbers R00 ES022243, F32ES023352) of the National Institutes of Health. This publication was made possible by US Environmental Protection Agency (USEPA) grant number RD-83479801. From the Framingham Heart Study of the National Heart, Lung, and Blood Institute of the National Institutes of Health and Boston University School of Medicine; this work was supported by the National Heart, Lung, and Blood Institute's Framingham Heart Study (contract numbers N01-HC-25195 and HHSN268201500001I). Disclaimer: The contents of this study are solely the responsibility of the grantee and do not necessarily represent the official views of the USEPA. Further, USEPA does not endorse the purchase of any commercial products or services mentioned in the publication. The views expressed in this manuscript are those of the authors and do not necessarily represent the views of the National Heart, Lung, and Blood Institute; the National Institutes of Health or the US Department of Health and Human Services. Competing interests: KSD reports grants from National Institutes of Health during the conduct of this study and outside the submitted work. EHW reports grants from National Institutes of Health-National Institute of Environmental Health Sciences, during the conduct of the study; and non-financial support from Servier Labs, outside the submitted work. JS reports grants from USEPA, during the conduct of the study; and serving as Expert Consultant for The US Department of Justice in a lawsuit against Volkswagen for violating emissions limits, Expert Consultant for Natural Resources Defense Fund in a lawsuit against coal-burning power plant for violating emissions limits. BAC reports grants from National Institutes of Health, grants from USEPA, outside the submitted work. UH reports grants from National Heart, Lung, and Blood Institute-Framingham Heart Study, during the conduct of the study; grants from Kowa Company, grants from American College of Radiology Imaging Network, personal fees from the American Heart Association, outside the submitted work. MAM reports grants from USEPA, grants from National Institutes of Health, during the conduct of the study; grants from USEPA, grants from National Institutes of Health, outside the submitted work. Ethics approval: Institutional Review Boards of Boston University Medical Center, Beth Israel Deaconess Medical Center and the Massachusetts General Hospital. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Brook RD , Rajagopalan S , Pope CA III Particulate matter air pollution and cardiovascular disease: An update to the scientific statement from the American Heart Association . Circulation 2010 ;121 :2331 –78 . 10.1161/CIR.0b013e3181dbece1 20458016 2 Pope CA III, Dockery DW Health effects of fine particulate air pollution: lines that connect . J Air Waste Manag Assoc 2006 ;56 :709 –42 . 10.1080/10473289.2006.10464485 16805397 3 Hoek G , Krishnan RM , Beelen R Long-term air pollution exposure and cardio- respiratory mortality: a review . Environ Health 2013 ;12 :43 10.1186/1476-069X-12-43 23714370 4 Dockery DW , Pope CA III, Xu X An association between air pollution and mortality in six U.S . cities. N Engl J Med 1993 ;329 :1753 –9 . 10.1056/NEJM199312093292401 8179653 5 Cesaroni G , Forastiere F , Stafoggia M Long term exposure to ambient air pollution and incidence of acute coronary events: prospective cohort study and meta-analysis in 11 European cohorts from the ESCAPE Project . BMJ 2014 ;348 :f7412 10.1136/bmj.f7412 24452269 6 Suwa T , Hogg JC , Quinlan KB Particulate air pollution induces progression of atherosclerosis . J Am Coll Cardiol 2002 ;39 :935 –42 . 10.1016/S0735-1097(02)01715-1 11897432 7 Sun Q , Wang A , Jin X Long-term air pollution exposure and acceleration of atherosclerosis and vascular inflammation in an animal model . JAMA 2005 ;294 :3003 –10 . 10.1001/jama.294.23.3003 16414948 8 Miller MR , McLean SG , Duffin R Diesel exhaust particulate increases the size and complexity of lesions in atherosclerotic mice . Part Fibre Toxicol 2013 ;10 :61 10.1186/1743-8977-10-61 24330719 9 Hoffmann B , Moebus S , Möhlenkamp S Residential exposure to traffic is associated with coronary atherosclerosis . Circulation 2007 ;116 :489 –96 . 10.1161/CIRCULATIONAHA.107.693622 17638927 10 Diez Roux AV , Auchincloss AH , Franklin TG Long-term exposure to ambient particulate matter and prevalence of subclinical atherosclerosis in the Multi-Ethnic Study of Atherosclerosis . Am J Epidemiol 2008 ;167 :667 –75 . 10.1093/aje/kwm359 18227099 11 Wilker EH , Mittleman MA , Coull BA Long-term exposure to black carbon and carotid intima-media thickness: the Normative Aging Study . Environ Health Perspect 2013 ;121 :1061 –7 .23820848 12 Rivera M , Basagaña X , Aguilera I Association between long-term exposure to traffic-related air pollution and subclinical atherosclerosis: the REGICOR study . Environ Health Perspect 2013 ;121 :223 –30 . 10.1289/ehp.1205146 23384708 13 Kälsch H , Hennig F , Moebus S Are air pollution and traffic noise independently associated with atherosclerosis: the Heinz Nixdorf Recall Study . Eur Heart J 2013 ;35 :853 –60 . 10.1093/eurheartj/eht426 24194529 14 Provost EB , Madhloum N , Int Panis L Carotid intima-media thickness, a marker of subclinical atherosclerosis, and particulate air pollution exposure: the meta-analytical evidence . PLoS ONE 2015 ;10 :e0127014 10.1371/journal.pone.0127014 25970426 15 Liu X , Lian H , Ruan Y Association of exposure to particular matter and carotid intima-media thickness: a systematic review and meta-analysis . Int J Environ Res Public Health 2015 ;12 :12924 –40 . 10.3390/ijerph121012924 26501300 16 Kaufman JD , Adar SD , Barr RG Association between air pollution and coronary artery calcification within six metropolitan areas in the USA (the Multi-Ethnic Study of Atherosclerosis and Air Pollution): a longitudinal cohort study . Lancet 2016 ;338 :696 –704 . 10.1016/S0140-6736(16)00378-0 17 Brook RD , Rajagopalan S Particulate matter air pollution and atherosclerosis . Curr Atheroscler Rep 2010 ;12 :291 –300 . 10.1007/s11883-010-0122-7 20617466 18 Campen MJ , Lund A , Rosenfeld M Mechanisms linking traffic-related air pollution and atherosclerosis . Curr Opin Pulm Med 2012 ;18 :155 –60 . 10.1097/MCP.0b013e32834f210a 22189455 19 Miller MR The role of oxidative stress in the cardiovascular actions of particulate air pollution . Biochem Soc Trans 2014 ;42 :1006 –11 . 10.1042/BST20140090 25109994 20 Crouse DL , Peters PA , van Donkelaar A Risk of nonaccidental and cardiovascular mortality in relation to long-term exposure to low concentrations of fine particulate matter: a Canadian national-level cohort study . Environ Health Perspect 2012 ;120 :708 –14 . 10.1289/ehp.1104049 22313724 21 Zeger SL , Dominici F , McDermott A Mortality in the Medicare population and chronic exposure to fine particulate air pollution in urban centers (2000–2005) . Environ Health Perspect 2008 ;116 :1614 –19 . 10.1289/ehp.11449 19079710 22 Shi L , Zanobetti A , Kloog I Low-Concentration PM2.5 and mortality: estimating acute and chronic effects in a population-based study . Environ Health Perspect 2016 ;124 :46 –52 . 10.1289/ehp.1409111 26038801 23 Tonne C , Melly S , Mittleman M A case-control analysis of exposure to traffic and acute myocardial infarction . Environ Health Perspect 2007 ;115 :53 –7 . 10.1289/ehp.9587 17366819 24 Wilker EH , Mostofsky E , Lue SH Residential Proximity to High-Traffic Roadways and Poststroke Mortality . J Stroke Cerebrovasc Dis 2013 ;22 :e366 –e72 . 10.1016/j.jstrokecerebrovasdis.2013.03.034 23721619 25 Lue SH , Wellenius GA , Wilker EH Residential proximity to major roadways and renal function . J Epidemiol Community Health 2013 ;67 :629 –34 . 10.1136/jech-2012-202307 23669275 26 Madrigano J , Kloog I , Goldberg R Long-term exposure to PM2.5 and incidence of acute myocardial infarction . Environ Health Perspect 2013 ;121 :192 –6 . 10.1289/ehp.1205284 23204289 27 Wilker EH , Mittleman MA , Coull BA Long-term exposure to black carbon and carotid intima-media thickness: the normative aging study . Environ Health Perspect 2013 ;121 :1061 –7 . 10.1289/ehp.1104845 23820848 28 Allen RW , Criqui MH , Diez Roux AV Fine particulate matter air pollution, proximity to traffic, and aortic atherosclerosis . Epidemiology 2009 ;20 :254 –64 . 10.1097/EDE.0b013e31819644cc 19129730 29 Chuang ML , Massaro JM , Levitzky YS Prevalence and distribution of abdominal aortic calcium by gender and age group in a community-based cohort (from the Framingham Heart Study) . Am J Cardiol 2012 ;110 :891 –6 . 10.1016/j.amjcard.2012.05.020 22727181 30 Allison MA , Criqui MH , Wright CM Patterns and risk factors for systemic calcified atherosclerosis . Arterioscler Thromb Vasc Biol 2004 ;24 :331 –6 . 10.1161/01.ATV.0000110786.02097.0c 14656730 31 Wong ND , Lopez VA , Allison M Abdominal aortic calcium and multi-site atherosclerosis: the Multiethnic Study of Atherosclerosis . Atherosclerosis 2011 ;214 :436 –41 . 10.1016/j.atherosclerosis.2010.09.011 21035803 32 Rivera JJ , Nasir K , Katz R Relationship of thoracic aortic calcium to coronary calcium and its progression (from the Multi-Ethnic Study of Atherosclerosis [MESA]) . Am J Cardiol 2009 ;103 :1562 –7 . 10.1016/j.amjcard.2009.02.004 19463516 33 Takasu J , Budoff MJ , O'Brien KD Relationship between coronary artery and descending thoracic aortic calcification as detected by computed tomography: the Multi-Ethnic Study of Atherosclerosis . Atherosclerosis 2009 ;204 :440 –6 . 10.1016/j.atherosclerosis.2008.09.041 19027115 34 Gonçalves FB , Voûte MT , Hoeks SE Calcification of the abdominal aorta as an independent predictor of cardiovascular events: a meta-analysis . Heart 2012 ;98 :988 –94 . 10.1136/heartjnl-2011-301464 22668866 35 Levitzky YS , Cupples LA , Murabito JM Prediction of intermittent claudication, ischemic stroke, and other cardiovascular disease by detection of abdominal aortic calcific deposits by plain lumbar radiographs . Am J Cardiol 2008 ;101 :326 –31 . 10.1016/j.amjcard.2007.08.032 18237594 36 Wilson PW , Kauppila LI , O'Donnell CJ Abdominal aortic calcific deposits are an important predictor of vascular morbidity and mortality . Circulation 2001 ;103 :1529 –34 . 10.1161/01.CIR.103.11.1529 11257080 37 Witteman JC , Kannel WB , Wolf PA Aortic calcified plaques and cardiovascular disease (the Framingham Study) . Am J Cardiol 1990 ;66 :1060 –4 . 10.1016/0002-9149(90)90505-U 2220632 38 Witteman JC , Kok FJ , van Saase JL Aortic calcification as a predictor of cardiovascular mortality . Lancet 1986 ;2 :1120 –2 . 10.1016/S0140-6736(86)90530-1 2877272 39 Hermann DM , Lehmann N , Gronewold J Thoracic aortic calcification is associated with incident stroke in the general population in addition to established risk factors . Eur Heart J Cardiovasc Imaging 2015 ;16 :684 –90 . 10.1093/ehjci/jeu293 25550362 40 Budoff MJ , Nasir K , Katz R Thoracic aortic calcification and coronary heart disease events: the multi-ethnic study of atherosclerosis (MESA) . Atherosclerosis 2011 ;215 :196 –202 . 10.1016/j.atherosclerosis.2010.11.017 21227418 41 Eisen A , Tenenbaum A , Koren-Morag N Calcification of the thoracic aorta as detected by spiral computed tomography among stable angina pectoris patients: association with cardiovascular events and death . Circulation 2008 ;118 :1328 –34 . 10.1161/CIRCULATIONAHA.107.712141 18779448 42 Dorans KS , Wilker EH , Li W Residential proximity to major roads, exposure to fine particulate matter, and coronary artery calcium: the Framingham Heart Study . Arterioscler Thromb Vasc Biol 2016 ;36 :1679 –85 . 10.1161/ATVBAHA.116.307141 27312220 43 Kannel WB , Feinleib M , McNamara PM An investigation of coronary heart disease in families. The Framingham Offspring Study . Am J Epidemiol 1979 ;110 :281 –90 . 10.1093/oxfordjournals.aje.a112813 474565 44 Splansky GL , Corey D , Yang Q The Third Generation cohort of The National Heart, Lung, and Blood Institute's Framingham Heart Study: design, recruitment, and initial examination . Am J Epidemiol 2007 ;165 :1328 –35 . 10.1093/aje/kwm021 17372189 45 Hoffmann U , Massaro JM , Fox CS Defining normal distributions of coronary artery calcium in women and men (from the Framingham Heart Study) . Am J Cardiol 2008 ;102 :1136 –41 , 41 e1 10.1016/j.amjcard.2008.06.038 18940279 46 Cupples LA DAR , Kiely D Section 34. Some risk factors related to the annual incidence of cardiovascular disease and death using pooled repeated Biennial measurements . In: Kannel WW , Garrison RJ eds. The Framingham Study. An Epidemiological Investigation of Cardiovascular Disease . Bethesda, MD : National Heart, Lung, and Blood Institute , 1987 :1 –26 . 47 Wilker EH , Ljungman PL , Rice MB Relation of long-term exposure to air pollution to brachial artery flow-mediated dilation and reactive hyperemia . Am J Cardiol 2014 ;113 :2057 –63 . 10.1016/j.amjcard.2014.03.048 24793676 48 Zhou Y , Levy JI Factors influencing the spatial extent of mobile source air pollution impacts: a meta-analysis . BMC Public Health 2007 ;7 :89 10.1186/1471-2458-7-89 17519039 49 Rosenbloom JI , Wilker EH , Mukamal KJ Residential proximity to major roadway and 10-year all-cause mortality after myocardial infarction . Circulation 2012 ;125 :2197 –203 . 10.1161/CIRCULATIONAHA.111.085811 22566348 50 Kloog I , Chudnovsky AA , Just AC A new hybrid spatio-temporal model for estimating daily multi-year PM2.5 concentrations across northeastern USA using high resolution aerosol optical depth data . Atmospheric Environment 2014 ;95 :581 –90 . 10.1016/j.atmosenv.2014.07.014 28966552 51 Rice MB , Ljungman PL , Wilker EH Long-term exposure to traffic emissions and fine particulate matter and lung function decline in the Framingham Heart Study . Am J Respir Crit Care Med 2015 ;191 :656 –64 . 10.1164/rccm.201410-1875OC 25590631 52 Miller KA , Siscovick DS , Sheppard L Long-term exposure to air pollution and incidence of cardiovascular events in women . N Engl J Med 2007 ;356 :447 –58 . 10.1056/NEJMoa054409 17267905 53 Wilker EH , Preis SR , Beiser AS Long-term exposure to fine particulate matter, residential proximity to major roads and measures of brain structure . Stroke 2015 ;46 :1161 –6 . 10.1161/STROKEAHA.114.008348 25908455 54 Agatston AS , Janowitz WR , Hildner FJ Quantification of coronary artery calcium using ultrafast computed tomography . J Am Coll Cardiol 1990 ;15 :827 –32 . 10.1016/0735-1097(90)90282-T 2407762 55 Efron B Nonparametric Standard Errors and Confidence Intervals . Canadian J Stat 1981 ;9 :139 –58 . 10.2307/3314608 56 Onuma OK , Pencina KM , Massaro JM Abstract P325: association of risk factors and extra-coronary Atherosclerosis with Progression of Coronary Artery Calcification in the Framingham Heart Study . Circulation 2014 (129: AP325 ). 57 Kannel WB , Sorlie P Some health benefits of physical activity. The Framingham Study . Arch Intern Med 1979 ;139 :857 –61 . 10.1001/archinte.1979.03630450011006 464698 58 Goff DC Jr, Lloyd-Jones DM , Bennett G 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines . Circulation 2014 ;129 (Suppl 2 ):S49 –73 . 10.1161/01.cir.0000437741.48606.98 24222018 59 Harrell FE Regression modeling strategies : with applications to linear models, logistic regression, and survival analysis . New York : Springer , 2001 . 60 Buis M POSTRCSPLINE: Stata module containining post-estimation commands for models using a restricted cubic spline , 2009 . 61 Tsao CW , Pencina KM , Massaro JM Cross-sectional relations of arterial stiffness, pressure pulsatility, wave reflection, and arterial calcification . Arterioscler Thromb Vasc Biol 2014 ;34 :2495 –500 . 10.1161/ATVBAHA.114.303916 25169933 62 Budoff MJ , Achenbach S , Blumenthal RS Assessment of coronary artery disease by cardiac computed tomography a scientific statement from the American Heart Association Committee on Cardiovascular Imaging and Intervention, Council on Cardiovascular Radiology and Intervention, and Committee on Cardiac Imaging, Council on Clinical Cardiology . Circulation 2006 ;114 :1761 –91 . 10.1161/CIRCULATIONAHA.106.178458 17015792 63 Pletcher MJ , Tice JA , Pignone M Using the coronary artery calcium score to predict coronary heart disease events: a systematic review and meta-analysis . Arch Intern Med 2004 ;164 :1285 –92 . 10.1001/archinte.164.12.1285 15226161 64 Stein JH , Korcarz CE , Hurst RT Use of carotid ultrasound to identify subclinical vascular disease and evaluate cardiovascular disease risk: a consensus statement from the American Society of Echocardiography Carotid Intima-Media Thickness Task Force . Endorsed by the Society for Vascular Medicine. J Am Soc Echocardiogr 2008 ;21 :93 –111 . quiz 89–90 10.1016/j.echo.2007.11.011 18261694

PMC005xxxxxx/PMC5372071.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01552610.1136/bmjopen-2016-015526Health Services ResearchResearch150617041704‘We both just wanted to be normal parents’: a qualitative study of the experience of maternity care for women with learning disability Malouf Reem 1McLeish Jenny 2Ryan Sara 3Gray Ron 1Redshaw Maggie 4 1 National Perinatal Epidemiology Unit, University of Oxford, Oxford, UK 2 Policy Research Unit in Maternal Health and Care, National Perinatal Epidemiology Unit, University of Oxford, Oxford, UK 3 Department of Primary Care, University of Oxford, Oxford, UK 4 Nuffield Department of Population Health, NPEU, Oxford, UKCorrespondence to Maggie Redshaw; maggie.redshaw@npeu.ox.ac.uk2017 24 3 2017 7 3 e01552614 12 2016 1 2 2017 23 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Background More women with learning disability (LD) are becoming mothers. Women with LD have rights to equal access to maternity care that meets their needs, however, many have poor pregnancy and birth outcomes compared to other women in the UK. Research is limited in this area. Objectives The aim of the study was to explore the lived experiences of pregnancy, childbirth, prenatal and postnatal care and services received by this group of women in the UK, including their expressed information and support needs relating to maternity care. Methods A qualitative study in which data were generated using in-depth semistructured interviews with learning disabled women who were pregnant or had given birth within the last 3 years in the UK; data were analysed using interpretative phenomenological analysis. Results 9 women with varying levels of cognitive impairment took part. 4 super-ordinate themes were identified: ‘I hate being treated differently’, ‘I find it harder to understand than other people’, ‘We've had to prove ourselves’ and ‘Make sure you've got very good support around you’. Subthemes included: ‘Negative attitudes and denial of choice’, ‘Understanding of normal care’, ‘Written information’ and ‘Being judged by professionals’. Conclusions With support from family and services, learning disabled women can become confident and successful parents. Maternity services should make reasonable adjustments when providing care to this group, including adapting to their individual communication and learning needs: allowing sufficient time in appointments, offering clear explanations of each aspect of care and sensitive support for autonomy and fully informed choice. Mothers who will be subject to a social care assessment of their parenting skills need clear information about the process, their choices and the level of skill they must demonstrate, as well as access to sufficient antenatal and postnatal support to give them the best possible chance of passing the assessment. learning disabilitymaternity caredisabilitypregnancypostnatal care ==== Body Strengths and limitations of this study In-depth qualitative interviews were undertaken with a group of people who generally do not often take part in research. The study has the potential to inform health professionals in maternity care and those specifically working with learning disabled women and their families. Easy to read participant information materials were specifically designed and used with this group. Limitations of the study include difficulties in accessing this population and unequal geographical distribution of participants. Background People diagnosed with learning disabilities (or intellectual impairment) are commonly described as having difficulties in understanding information and communication,1 or deficits in social and practical skills and self-care.2 Learning disability (LD) is a condition that starts before adulthood which reduces the ability to learn complex or new information and to cope independently.1 In 2010, there were an estimated 1 037 400 adults in the UK with LD3 though many people may remain undiagnosed. In the past, sterilisation and abortion were used to deny women with LD the possibility of becoming parents4–6 but the equal rights of learning disabled people to have children and to be supported in bringing them up are now enshrined in policy and legislation.7–11 However, women with LD are often faced with strong social and professional disapproval when they disclose their pregnancy8 12–15 and are more likely than other mothers to have their children removed and permanently placed outside the family home due to child protection or welfare concerns.16–18 People with LD experience poorer health outcomes and shorter life expectancy than the wider population,3 and mothers with LD and their babies are at higher risk of adverse outcomes during pregnancy and afterwards.8 19–23 For example, research in Australia found that women with LD are more likely to have pre-eclampsia and to give birth to low birthweight babies who were admitted more frequently to neonatal intensive care unit.8 Moreover, women with LD may face barriers to accessing antenatal care including difficulties with public transport, with remembering appointments, and with understanding the purpose of the appointment.24 Women with LD have been found to be less satisfied with their maternity care compared with non-disabled women, particularly the interpersonal aspect of care as reflected in interactions with staff.20 The limited research on their maternity experiences describes unmet needs for accessible pregnancy information, consistent care and control over decisions relevant to their care. Study objectives The main objective of our study was to better understand the individual experience of maternity care for women with LD, with a particular focus on access to maternity care, making choices, understanding information and feeling in control. In this way we would be able to reflect on the key support and services women with LD felt they needed during their pregnancy, birth and afterwards. We also aimed to increase the diversity of women whose views of UK maternity care are heard and for extracts from these interviews to be published at healthtalk.org, a freely available web resource of health experiences. Methods The qualitative methodology was similar to that adopted by a number of studies involving people with LD.14 25 26 Women's experiences were elicited using face-to-face, in-depth, semistructured interviews which allowed participants to express their views freely. The data were analysed using interpretative phenomenological analysis (IPA), a method suited to the in-depth analysis of the experiences and understanding of small group of participants27 28 and aiming to draw out the unique and shared aspects of experience.29 The ‘double hermeneutic’ of IPA, in which the researcher tries to make sense of the participant trying to make sense of their world,27 enabled a multidisciplinary group of researchers to explore participants' lived experiences by combining the women's ‘insider’ perspectives with the researchers' ‘outsider’ perspectives.28 30 A formal evaluation of intellectual ability was not undertaken in this study. The National Research Ethic Services (NRES) Committee for South Central—Berkshire ethical committee approved the study. The Research Ethics Committee Reference (REC) number: 12/SC/0495. Data collection Women were eligible to participate in the study if they had a LD and were pregnant or had given birth within the last 3 years in the UK. Recruitment of a convenience sample involved contacting advocacy groups, support workers, midwives and health visitors with an outline of the project and copies of the information leaflet (RM). Potential participants were approached by their care workers or workers in the support organisations, and the study was described. If a woman agreed to take part, her consent was sought to provide her contact details to the researchers who would visit and conduct the interview. A time was then arranged for the interviews and women were sent the accessible format illustrated ‘easy-read’ study information sheet, which was read to them when needed. A freephone number was also given to the women and their support workers for any further discussion about the study. At the time of interviews, carried out by a researcher experienced in the field (SR), information was again provided about the study and what participation involved, including assurance that participants could withdraw from the study at any point. Questions were asked carefully and clearly with language like that used in the ‘easy-read’ information sheet. Written consent was obtained before the interview using an ‘easy-read’ consent form. Contextual information was collected on age, family circumstances and support. Interviews were video and audio recorded with participants' permission and before each interview photos and example clips were played using illustrative scenes of non-disabled women talking about pregnancy from healthtalk.org. Six interviews were conducted at the participants' homes, two in community centres and one in a hotel close to the participant's home. A topic guide was used with a broad opening question asking women to talk about their experiences of being pregnant and giving birth, and subsequent questions exploring specific aspects of their experiences such as communication and interpersonal care. The participants also were asked in a general way about any changes they would like to make to their maternity care. All women were interviewed once, with interviews ranging from 40 to 100 min. All interviews were fully transcribed, and copies sent to the interviewees and their care workers for comment. Minor changes were made by eight participants to some of the details provided. Data analysis The interview data were analysed thematically using IPA as described by Smith 2009.28 This allows a very detailed analysis of a limited number of participant accounts and understandings as was appropriate for this kind of study.27 Each transcript was first analysed as an individual case before considering common themes. As each transcript was read and re-read, emergent themes were identified inductively and recorded using NVIVO software, while a manual log of researcher comments and reactions was also made. Emergent themes were clustered into groups under superordinate themes. This process was repeated for each transcript with previous transcripts reconsidered iteratively in the light of subsequently emergent themes. Finally, the themes from each transcript were combined. Results Characteristics of the study participants Nine women with LD participated in this study, eight of whom were reported by their support workers to have been diagnosed with mild-to-moderate LD. One participant had severe LD as reported by a family member. Three women also had a long-term health condition or physical disability and one had a serious mental health disorder. All participants were able to talk about their experiences, however, one was only able to give very short answers and one participant came to the interview with a set of notes as a prompt. Participants were White and aged between 25 and 39 years (mean and median 30 years). Four women were living with their partners and five were no longer in a relationship with the father of their baby. One woman lived with her mother who cared for her child with her. At the time of the interview, one woman was pregnant (20 weeks gestation), four women had one child only, one woman had two children, one had three children, one had five children and one had one child and had also experienced a stillbirth. All children were reported to be non-disabled apart from one child who was diagnosed with LD. Two women had had their child removed from their care. Two women had some administrative work in their advocacy groups, and one woman worked as a waitress locally. All were receiving social services support, two women reported receiving mental health services support and four reported receiving support from their advocacy groups. Six said they had support from their family during their pregnancy and afterwards. Potentially identifying information has been removed to protect the anonymity of the participants, and pseudonyms are used. Emerging themes Four superordinate themes emerged from the analysis: ‘I hate being treated differently’, ‘I find it harder to understand than other people’, ‘We've had to prove ourselves’ and ‘Make sure you've got…very good support around you’. The themes and subthemes are listed in table 1. Table 1 Themes and subthemes identified in the analysis of interviews with LD women about their experience of care during pregnancy, childbirth and afterwards Key themes Subthemes ‘I hate being treated differently’ Positive attitudes and support for choice Negative attitudes and denial of choice Understanding of normal care Suggestions for improvement ‘I find it harder to understand than other people’ Written information Other sources of information Verbal communication with healthcare professionals ‘We've had to prove ourselves’ Family reactions Being judged by professionals Experiencing safeguarding as discrimination ‘Make sure you've got very good support around you’ Family support Professional support LD, learning disability. ‘I hate being treated differently’ The mothers expressed a great longing for ‘normality’, contrasting themselves with “normal parents” (Lisa) while believing that “women who have LD are not classed as normal” (Ellie). This theme considers how the mothers felt treated during their maternity care, and whether they felt this was ‘normal’. Positive attitudes and support for choice Most of the mothers had enjoyed some or all of their antenatal care and spoke warmly about maternity staff who had supported them well during pregnancy and birth: “very good” (Dawn), “a great bunch” (Morgan), “very polite, very helpful” (Laura), “very well treated” (Ellie), “a lot of support” (Rachel) and “really quite supportive… very encouraging” (Jo). A friendly attitude was important to Sasha, who experienced it: “They can have a laugh. They're not all like stuck up,” and to Katie, who did not: “[The midwife]'s grumpy. She doesn't smile or anything.” Other mothers were pleased with continuity of care: “I had the same midwife all the time so that was quite good” (Jo); or with proactive and personalised support: “[The midwife] phoned me up like, ‘How's it going?’…And sometimes she'll come out to me if I can't make it” (Rachel). Morgan felt well supported by her community midwife who had some personal experience: “She had two children who had disabilities, so she understood what it was all about.” Six of the mothers gave examples where they had successfully made choices about their maternity care, including to have an elective caesarean (Ellie and Laura); to decide on pain relief during labour (Rachel, Jo and Sasha); to have the father cut the umbilical cord (Morgan); to discharge themselves from hospital against medical advice (Jo and Rachel) and to request a change of health visitor (Jo). None reported any difficulty with attending antenatal appointments. Negative attitudes and denial of choice There were also negative experiences related to autonomy and staff attitudes. Lisa in particular felt that she was discriminated against by maternity staff, because “[they] think you're stupid…treat you as though you don't exist.” She described how staff more than once discussed her care in front of her without including her in the conversation “like I wasn't even in the room,” and how they told her that she would have to give birth by caesarean section because of a medical condition:I would ask why was that and they were like, ‘Oh, it's our choice. It's our decision.’ And just felt like we didn't have a say in in how…we could have our son…felt like we were invisible really…no need for us to even be there because they'd already made a decision. As well as feeling disempowered by her exclusion from decision-making, Lisa was ignored when she tried to find out where her newborn baby had been taken minutes after birth:We kept asking, ‘Where's he, where has he gone?’…Three hours after I had him, they brought him back to the ward and then they told us that he was in the special care ward because he needed his lungs looking at. And I asked them why they couldn't just tell us that's where they took [him] and they just walked away. Morgan described how she was not able to make an informed choice about postnatal contraception before leaving hospital because she felt under pressure to make a decision: “I felt…they wanted me to give answers there and then without even giving me a chance to think about it. I felt pressurised into like getting the [long-acting reversible contraceptive] injection when I didn't really want it. Cos if I said no I don't know what they would have said.” Rachel experienced poor care when she was left unattended for a long time during one of her labours and was spoken to disrespectfully by a midwife who addressed her as ‘Speedy’ following a quick labour: “They treat you…like a piece of dirt really. Like they don't really care.” However, Rachel did not attribute poor care to her LD but to the fact that the staff were overworked, as did Jo: “They just haven't got enough staff really, it's not just purely because of having a learning disability…They're just rushed off their feet…So it's just a normal aspect of the NHS really.” This contrasted with Lisa who observed how the midwives interacted with others to check whether she was being singled out for different treatments: “I hate being treated differently so like all through my appointments and on the ward I'd been watching how they respond to other people, and how they'd respond to other people was complete different to how they responded to me.” Understanding of normal care The mothers' sense of themselves as outside the category of ‘normal’ may have led some of them to interpret what could have been ordinary aspects of maternity care as abnormal and discriminatory. For example, Katie disliked her midwife perhaps because she asked her the type of questions that all mothers are asked at their booking or first midwifery appointment: “She just wants to ask too—loads of questions. Just wants to be nosey in my private life.” Some mothers were offered antenatal classes and others were not, consistent with the provision of antenatal education in some hospitals and not in others.31 Lisa assumed that only being offered antenatal classes starting when she was 8 months pregnant represented a deliberate exclusion: “I think they just thought, ‘Oh, it's a pair of disabled people. It doesn't matter if they know what's going on or not.’” She also believed she was denied the conversations which “normal” women would have with their midwife during appointments, although similar concerns about brief and functional midwifery appointments have been expressed by many mothers without disabilities:31 32I just felt I was being treated like a disabled person…because most women, they can go in and have a talk with the midwife and that, but even though the midwife was really nice and supportive…she didn't really talk much to me about things. It's like when like she checks to see if the baby is okay, she just goes, ‘Yep, the baby is fine’. And then that's it. Suggestions for improvement Asked whether there was anything they would change about their maternity care, three mothers said there was nothing they would change. Three talked about staff attitudes and gaining information: “Just be more un-bossy, un-nosey. Just talk to me more about pregnancy. And actually smile” (Katie); and three talked about more staff training: “I'd like people to be more aware and learn and train about what a learning disability is so we don't get mistreated and misjudged” (Lisa). Rachel had made two formal complaints about her care and believed this had resulted in better treatment in her third pregnancy. ‘I find it harder to understand than other people’ This theme considers a range of difficulties which the mothers described with understanding written and oral information about their maternity care and looking after a baby. Written information Ellie, Rachel and Sasha said they were able to understand the ‘normal’ information they were given. Morgan and Katie had been given ‘easy read’ information, designed for parents with learning disabilities, by their support worker or children's centre, and Katie could text her midwife with any questions. Laura asked a friend to help her with words she did not understand and Katie asked her mother. Lisa and Jo said they were not given any information: “I went through the whole pregnancy not really knowing what to do because I wasn't getting information off the doctors or the midwives” (Lisa), although Jo also said that someone came to her home to give her “the official labour talk” after the local antenatal classes were cancelled. Other sources of information Most of the mothers said that they had found the internet to be a good source of information on pregnancy and birth, especially the NHS Choices website;33 some had found this site accessible and others had used it with the support of their partner or family members: “There's bits of it that were quite confusing…I wasn't sure what I had to click onto” (Jo). They had found videos particularly useful. Those who had gone to antenatal or postnatal classes had found them “very useful” (Ellie), and Laura particularly valued being “showed how” to carry out baby care tasks such as bathing, nappy changing and bottle feeding in hospital. Katie had been able to practise with a ‘Real Care Baby’: “It just cries, and then you've just got to use this little tag thing near its belly button and you've just got to try and see if it needs rocking, feeding, or changing.” Verbal communication with health professionals The mothers' experiences of understanding health professionals at their appointments were also mixed. Some did not have any difficulties, with Rachel describing how “[Midwives] were sort of like, help you out and tell you what you can and can't do, what you can and can't eat.” Dawn was supported to communicate by her mother and Jo's partner helped her to understand at a scan. Lisa and Jo described situations where they had not understood something, but when they asked the health professionals to explain, their requests were ignored: “I said, ‘I don't understand what you're talking about’. ‘Oh, we haven't got the time for that, I'm just going to prod you and poke you and I'm going to go’” (Jo). Jo explained that she did not challenge this poor communication because “I know that they don't have any specific training for people like me, so I just left it at that.” In contrast, when Morgan could not follow what was being said at a scan, she felt able to ask the radiographer for more help: “I couldn't really see that screen up at the top…So if they said, ‘This is the…’, I'm saying, ‘Well, where? Cos I can't see it’. When they showed me on their computer bit, then I could see it or her…So they were really clear.” Lisa described how she signed a consent form for a caesarean without properly understanding it: “They just went through a consent form really quickly and then I signed that but I didn't understand what I was signing…And they didn't explain…the caesarean procedure or anything so I was just clueless.” The consequence of this lack of adequate explanation was that her baby's birth was a very frightening experience: “Before the caesarean and all the way through the caesarean I was just shaking really badly because I was that scared because I just didn't know what was happening.” For Lisa, this incident was part of a repeated pattern in which she felt that staff were deliberately withholding information from her: doctors “wouldn't explain why” they felt it was safer for her baby to be born by caesarean, a physiotherapist or doctor “wouldn't advise me on what else I could do to stop the pain,” staff “wouldn't explain to us what jaundice was,” midwives took her baby to the special care unit and “wouldn't tell us where they took him.” This was in contrast to Sasha's experience of midwives explaining placental abruption in terms she could understand: “They said that the placenta eruption means that your placenta moves away from your body whilst your baby's still inside. Instead of waiting for baby to come out, and then come out as well,” and Dawn's understanding of having an anti-D injection because of her rhesus status: “[My baby]'s blood is different than mine, that's why I had a drug to stop it killing the baby off.” Katie also said that her midwife made an effort to explain things that she did not understand: “She does try and explain to me. But I do find it hard to understand it. Because I've got a learning disability, so I find it harder to understand than other people.” Jo had a health visitor who “explained things to me, so she was fine.” There were also some conversations recounted by mothers where it was apparent that the midwives had not communicated clearly, although the mother had not realised that she had not understood them. For example, midwives had advised Morgan against mixing breastfeeding and bottle feeding at birth, in line with the usual advice to enable breastfeeding to become well established before a bottle is introduced. Morgan had, however, understood this to mean that she should exclusively bottle feed: “I said I wanted to do both…I got told that it would mess up the baby's head really…So I just went on bottle-feeding her.” ‘We've had to prove ourselves’ This theme describes how almost all of the mothers felt that becoming a mother had been overshadowed by the requirement to demonstrate to others their ability to be good-enough parents: “For the last 8 months we've had to prove ourselves really. And we should n't have had to” (Morgan). Family reactions For Laura, this meant succeeding in the teeth of her family's expectations: “Most members of my family thought I'm not going to cope with a child. Even my Dad said that I wouldn't do it. But I proved him wrong” (Laura). Others were pleasantly surprised at the positive reaction and support from their family, for example Katie: “Not as I would have expected. My Grandad still doesn't like the idea…but everybody else seems to be fine with it.” Being judged by professionals A bigger issue was the need to prove themselves to social workers concerned about safeguarding the welfare of their children. All the mothers had some involvement with social services, either for themselves as vulnerable adults, or for the children. Owing to their learning disabilities, two of the mothers had had to undergo an assessment of their parenting capabilities when their babies were born and were not allowed to care for their children independently as a result (Ellie had failed the assessment and Lisa had declined to go to a specialist residential unit for assessment). Rachel and Laura had social workers involved for reasons unrelated to their learning disabilities, and Morgan's baby had been referred to children's social services after an incident in hospital had triggered a safeguarding concern, but this had been poorly explained to her: “We would have understood from the beginning if they told us the reasons why. And they did n't.” A recurrent theme in many of the mothers' accounts was the strain and inequity of becoming a parent under the judgemental gaze of the health and social care professionals: “They were watching over my shoulder” (Rachel). Jo said that professionals had expected her to fail as a parent: “A lot of people within midwives and health visitors, they just look at you and they make assumptions”; but Rachel had received affirmation from the midwives that exceeded her self-perception: “[They] said that I was doing well. Which, I don't basically think that I was doing well. But I done well. They could see it. But I couldn't see it.” Jo had decided that in the context of professionals looking for deficits, it was not safe to be honest about her postnatal emotions because “I thought if you said something how you's is exactly feeling, and if you was feeling a bit down that particular day, that they would use that against you.” This was echoed by Sasha who accepted social workers' advice in order to avoid problems: “I just go along with them. Keep the peace. Otherwise it's never going to end.” For some mothers, this scrutiny was an unanticipated contrast to being treated ‘normally’ during pregnancy. Morgan described how “Afterwards, things started to change. Me and my partner were treated differently,” and she found it particularly hard being judged on carrying out babycare tasks correctly when different professionals gave contradictory advice about what to do: “A lot of the things that the community healthcare assistant taught us was completely different to how we were shown in the hospital.” Two mothers who did not have care of their children vividly described how stressful it had been to be constantly watched with their newborn child: “I couldn't cope and I felt under pressure…I was actually being monitored and watched every day, every time, everywhere in the assessment unit. And I didn't feel like I was at peace” (Ellie). Lisa found being scrutinised immediately after birth had undermined her ability to behave naturally:We weren't allowed to do anything unless the midwife was supervising us…I felt too nervous to enjoy it…like when you bottle-feed and you just have this odd glance at your husband, or someone who knows it's a proud moment and I just found I had to fix my eyes on my son…I didn't feel I could look somewhere else in case I got marked down for it. Experiencing safeguarding as discrimination As well as the emotional pain of not being allowed to care for their children, these two mothers felt “guilty” (Lisa) and experienced distress at the disappointment of their dreams of parenthood: “This was something I've always been looking forward to. To being a mum” (Ellie). They both interpreted the decision that their babies should be looked after by the grandparents or the father as discrimination: “I felt like my rights were not heard as a mother with LD…they're always picking on us” (Ellie); “People aren't seeing past the learning disability…There's no reason for you to judge me on whether I can understand how to look after a baby or not” (Lisa). Ellie's self-assessment was that “I was a fantastic mother” and she blamed the court which had awarded care to the father, her partner at the time, for not “thinking about me and obviously, not just the child.” She had not been expecting this outcome: “I thought that instead of going to court that social services would work with me to support me and encourage me to keep my little girl.” Lisa had also not understood the postbirth child welfare system. Having decided not to go to a mother and baby unit for support and assessment, she was nonetheless taken by surprise when social services said that she and her husband (who also had learning disabilities) would not be able to take their baby home from hospital. With hindsight she thought that this outcome was the inevitable result of her decision: “From the moment social services were involved, they always planned according to my husband's parent's care…that's where [the baby] was going because I refused to go in the mother and baby unit.” Likewise Morgan was “not sure” whether social workers had told her that she might be monitored postbirth. ‘Make sure you've got…very good support around you’ This theme considers how the mothers were aware that to succeed as parents they needed to have good support, and how several of the mothers who had been well supported were flourishing. Family support Some had this support from their partner or family members, and Morgan described how this was a formal part of her discharge plan from hospital: “It was agreed that I would go and stay with [my partner's] mum to start off with.” Morgan's advice to other mothers was “Make sure basically you've got a good team of people around you and very good support around you.” Ellie did not have family support and believed this had contributed to her losing care of her child: “At the time I pushed my family away because I fell out with them…Which was a mistake I made and regret.” Laura had successfully managed a situation where her parents in law were “trying to take over” and had developed the confidence to make her own parenting choices: “I thought to myself, ‘I need to do this for myself’…I tried their advice. Sometimes it worked, and sometimes it didn't. But I chose which option that I stuck to.” Professional support Most of the mothers also had substantial support in the community from social workers, health visitors, support workers, children's centres and voluntary sector agencies. Rachel had fought to keep her social worker when she moved to a different area: “I still wanted that support,” and Morgan had argued for her support to be increased postnatally: “I also managed to get an extra day of support as well.” Sasha described her social worker a source of support rather than scrutiny: “She just comes over, trying to get a bigger house…they're good, when you need them”; and she described the qualities in a social worker that she liked: “She's calm… She don't just take in what she wants to hear and then don't do nothing about it, she actually listens.” She identified the dilemma that the support, which she wanted, came with unsolicited advice, which could be useful but could also feel intrusive: “I don't mind them being around, because they're quite helpful. But it does bug me a bit, because I know that I can look after [the children]. But they give you good tips, like on what to do, and what advice to go for…I like the support from them.” Jo and her partner had exercised autonomy as parents when they decided not to follow the health visitor's advice about waking their baby up to feed her, instead devising their own successful solution to her falling weight; she noted that the health visitor did not like this, but “we still did it.” She had likewise made up her own mind to disregard the baby care advice from epilepsy specialists, after she had a fit shortly after birth: “I observed the situation for a while and then just went back to normal.” Jo was one of several mothers who described how they had become confident and successful parents: “[Health professionals] said, ‘Oh, you're not going to do well’, so I didn't have a lot of encouragement at all. And so I proved them all wrong.” Lisa and Ellie said that they had not been given enough support to learn to be a mother before being assessed on their parenting; Ellie felt that “Social services could have…put me on to parenting courses and give me the support that I need and encouragement to manage to look after my daughter as a mother.” Despite separation from their children, both were working hard to maintain their identities as mothers: “I've been writing letters every day to my daughter, so that when she's older she can read them and understand and know that…if I fight and I don't win that I've tried my best for her” (Ellie). This had been challenging for them; Ellie said she had forgotten her child's birthday and Lisa said that “I kept forgetting I was a mum because I couldn't look after my son.” Discussion The mothers who took part in this study had varying levels of LD and diverse experiences within the maternity services. Most were positive about their antenatal and intrapartum care and the attitude of their midwives, consistent with Redshaw et al20 finding that 93% of mothers with learning disabilities rated their antenatal care as good or better. Although one mother had the clear impression that healthcare professionals were expecting her to fail (which she did not), none had had their right to be a parent challenged by maternity practitioners. This is in line with the findings of a small qualitative study of English midwives' attitudes and experiences34 but contrary to a survey of Swedish midwives, a third of whom partly or wholly agreed that women with learning disabilities should not have children.15 Research and national guidance has repeatedly emphasised the importance of assessing a learning disabled parent's support needs, providing accessible information in a variety of easy read or audio-visual formats, offering longer appointments and checking that information given has been understood.8 11 17 35 Despite the clear duty imposed on services under the Equality Rights Act (2010)10 to make ‘reasonable adjustments’ to meet the needs of disabled people, several mothers had experienced rushed appointments where their communication needs were not met and they were left feeling confused and marginalised. Two said they had been denied explanations when they specifically asked for them. None had been given easy-read pregnancy and birth information by maternity practitioners (although some had received this from other sources). This failure to adapt standard care to meet mothers' needs reflects earlier research findings.20 24 36–38 Mothers' suggestions that maternity professionals need more training are echoed by midwives themselves, who feel out of their depth when trying to communicate with and appropriately support parents with learning disabilities.34 36 38 39 In situations where a midwife had not checked whether the mothers had understood what was said to them, some mothers were left to draw potentially mistaken inferences from midwives' words or behaviour. Particularly in some cases where mothers felt highly conscious that they were not seen as ‘normal’ women, their LD appeared to become a lens through which every encounter was understood, and what may have been ordinary aspects of maternity care were therefore interpreted as discriminatory. This highlights the importance of careful explanations of all aspects of maternity care, including explaining what the usual offer of care is and any reasons for different treatments. Some mothers had attended antenatal classes and found the hands-on learning very helpful.36 Three more mothers had been keen to attend classes, in contrast to a mother in Porter's et al study,37 who was embarrassed to attend a group. Without accessible written information or explanations from midwives, some mothers described how they relied instead on family members, as in Quartermaine,40 and the internet to find information. Although six of the mothers recounted situations in which their right to make choices about their care had been respected (including mode of birth), two described how professionals either made a decision for them (for birth by elective caesarean) or did not ensure that they gave genuine informed consent (to caesarean birth or a contraceptive injection). Unlike earlier studies, none of the mothers had experienced maternity practitioners inappropriately seeking consent from a companion.36 Two mothers had encountered disrespectful attitudes from midwives. Lisa in particular had a range of frustrating and disempowering encounters with maternity staff that had left her feeling marginalised and ‘invisible’, lacking information she needed about pregnancy, birth and parenting, feeling denied aspects of care she believed were given to ‘normal’ mothers and believing that a large number of professionals had withheld information from her. Cox et al41 describe how specialist doulas work with expectant mothers with learning disabilities during pregnancy, birth and up to 6 months afterwards, supporting them to understand their pregnancy and to learn parenting skills, and acting as a ‘translator’ with professionals. This type of personalised, empowering support could have transformed Lisa's maternity experience and her understanding of it. This might also allow a safe space in which mothers could feel able to express any feelings of depression or low mood that they may not be able to disclose to health professionals. Consistent with many other studies of learning disabled parents,25 36 40 42 43 many of the mothers in this study felt that they could not take motherhood for granted: they had to prove to the authorities and some family members that they were able to be suitable parents. Being supervised and judged by maternity professionals and social workers was stressful and disempowering, particularly for the three mothers who had been assessed at or shortly after birth. They were unclear what standards of behaviour they were being judged against, a process described by Booth and Booth16 as “like playing a game without being told the rules.” They were also disadvantaged by receiving contradictory advice from professionals.17 For the mothers about whom there were child protection or welfare concerns, it was a shock to experience the difference in their treatment between pregnancy (when, as a vulnerable adult, their needs were central to professionals' planning and support), and after birth (when their child's needs displaced their needs as the focus of professionals' duties). It has been argued that parents with learning difficulties “are disproportionately represented in care proceedings around the world,”16 but this has been challenged by a review of case files from 10 local authorities in England which found that there “was no evidence to suggest that parental learning disability in itself was the reason children were removed from their parents’ care” and that removing children was used as a “last resort” after “a substantial input of services.”44 The two mothers in our study who formally did not have care of their children understood their loss straightforwardly as discrimination by social services and the court. They were comfortable with the language of their rights as a parent with learning disabilities but did not show any awareness that the best interests of their child might not always be identical with their rights as a mother; or that where there were child welfare concerns, professionals had a specific duty to consider “how capable each of [the child]'s parents…is of meeting [the child]'s needs” (Children Act 1989, section 145); or that a father has an equal right with a mother to be the primary carer following relationship breakdown. Instead they felt failed by services that had not sufficiently taught and supported them to be parents. Neither had been expecting the outcome and their distress was compounded by not having understood how the child welfare system might work in their situations. This system lacks transparency, given that the courts themselves continue to develop case law on the fundamental rights of learning disabled parents to “parent with support,”46 while accepting that there will be individual situations where no amount of support will enable the parents to provide their children with adequate care.47 All the mothers acknowledged that they needed support to gain confidence and succeed as parents, either from their families, or from social workers and other support workers.17 Some accepted that the price of this support was parenting advice, which might sometimes be useful and other times intrusive. A sense of power imbalance37 did not affect all mothers equally: some showed a sophisticated ability to negotiate the health and social care systems and their own family networks to get the support they wanted while ignoring unsolicited advice. They might do this covertly, to “keep the peace,” or overtly, exercising their own responsibility as parents. The confidence with which some of these mothers managed their parental role, their choices and their relationships with service providers is very different from the experiences of mothers with learning disabilities reported in earlier research, which has tended to emphasise negative and disempowering experiences.48 It seems that the LD advocacy movement has achieved a great deal in normalising parenthood as a choice for people with learning disabilities, in the 20 years since Brown challenged services to see that women with learning disabilities are “women first” and “mind about the things other women mind about,” including motherhood.49 Strengths and limitations It was a strength of this study that it involved in-depth qualitative interviews with a group of women who do not often take part in research, following a careful and lengthy process of engagement with LD support and advocacy groups. The study used easy read participant information materials specifically designed for women with LD. The chosen analytic method, IPA, enabled consideration of mothers' subjective experiences with the meanings the experiences had for them, and also a critical analysis of what may have lain behind some of those experiences and meanings. There were also some limitations. A convenience sample was used as gaining access to this group of participants was not easy and support organisations and family members acted as gatekeepers. One reason this group of women were reluctant to be approached directly appeared to be the assumption that researchers would assess their capacity and question their parenting ability. Others whose children had been removed may have felt reluctant to recall their maternity experiences. Conclusions Maternity professionals should be aware that the diagnosis of ‘learning disability’ masks a wide range of abilities, expectations and experiences, and that with support from family and services, women with LD can become confident and successful parents. In order for mothers with LD to have a ‘normal’ experience of the transition to motherhood, maternity services must make reasonable adjustments, including adapting to individual communication and learning needs, and taking the time to check that they have fully understood. It is important that each aspect of maternity care is clearly explained, so that the mother can understand what ‘normal’ care is like and is supported to make fully informed choices. Mothers who will be subject to a social care assessment of their parenting skills need clear information about the process, their choices and the level of skill they must demonstrate, as well as access to sufficient antenatal and postnatal support to give them the best possible chance of passing the assessment. The authors thank all the women who agreed to take part in the study. They also thank members of the advocacy groups, midwives and social workers who put them in touch with these women. Contributors: MR, RG and RM designed the study. RM contacted the advocacy groups, support workers and health professionals, made visits and gave talks about previous disability research and the planned study. RM and SR undertook the visits, and SR carried out the interviews. JM undertook the main analysis supported by MR and RM. JM, MR and RM drafted the manuscript, and all authors commented and approved the final version. Funding: The study was funded by the Policy Research Programme in the Department of Health. Disclaimer: The views expressed are not necessarily those of the Department of Health. Competing interests: None declared. Ethics approval: National Research Ethic Services (NRES) Committee for South Central—Berkshire Ethical Committee approved the study. The Research Ethics Committee Reference (REC) number: 12/SC/0495. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: The sensitive interview data collected will not be available for data-sharing. Selected recordings may be made available on the Health Experiences website. ==== Refs References 1 Department of Health . Valuing people: a new strategy for learning disability for the 21st Century . London : The Stationary Office https:// http://wwwgovuk/government/uploads/system/uploads/attachment_data/file/250877/5086pdf (accessed May 2016 ). 2 American Association of Psychiatric . Diagnostic and statistical manual of mental disorders . 5th edn Alexandria (VA ): American Counseling Association , 2003 . 3 Emerson E , Baines S Health inequalities & people with learning disabilities in the UK: 2010 . Learning Disabilities Observatory http://wwwimprovinghealthandlivesorguk/uploads/doc/vid_7479_IHaL2010-3HealthInequality2010pdf (accessed August 2016 ). 4 Hamilton C Sterilisation and intellectually disabled people in New Zealand—still on the agenda? Kōtuitui: NZ J Soc Sci Online 2012 ;7 :61 –71 . 5 Hamilton C Sterilization and intellectually disabled people in New Zealand: rights, responsibility and wise counsel needed . Kōtuitui: NZ J Soc Sci Online 2015 ;10 :36 –45 . 10.1080/1177083X.2014.953185 6 Roy A , Roy A , Roy M The human rights of women with intellectual disability . J R Soc Med 2012 ;105 :384 –9 . 10.1258/jrsm.2012.110303 22977048 7 Mitra M , Parish SL , Clements KM Pregnancy outcomes among women with intellectual and developmental disabilities . Am J Prev Med 2015 ;48 :300 –8 . 10.1016/j.amepre.2014.09.032 25547927 8 McConnell D , Mayes R , Llewellyn G Women with intellectual disability at risk of adverse pregnancy and birth outcomes . J Intellect Disabil Res 2008 ;52 :529 –35 . 10.1111/j.1365-2788.2008.01061.x 18422528 9 Department of Health . Valuing people now: a new three-year strategy for people with learning disabilities . London : The Stationery Office http://base-ukorg/sites/base-ukorg/files/[user-raw]/11-06/valuing_people_now_strategy_pdf (accessed October 2016 ). 10 Equality act London : HMSO , 2010 . 11 Department of Health & Department for Education & Skills . Good practice guidance on working with parents with a learning disability. http://wwwdmuacuk/documents/health-and-life-sciences-documents/nursing-and-midwifery/goodpracticeinparentingsupport2013pdf (accessed 10 Aug 2016 ). 12 Aunos M , Feldman M Attitudes towards sexuality, sterilization, and parenting rights of persons with intellectual disabilities . J Appl Res Intellect Disabil 2002 ;15 :28 –296 . 10.1046/j.1468-3148.2002.00135.x 13 Cuskelly M , Bryde R Attitudes towards the sexuality of adults with an intellectual disability: parents, support staff, and a community sample . J Intellect Dev Disabil 2004 ;29 :255 –64 . 10.1080/13668250412331285136 14 Mayes R , Llewellyn G , McConnell D Misconception: the experience of pregnancy for women with intellectual disabilities . Scand J Disabil Res 2006 ;8 :120 –31 . 10.1080/15017410600774178 15 Höglund B , Larsson M Struggling for motherhood with an intellectual disability—a qualitative study of women's experiences in Sweden . Midwifery 2013 ;29 :698 –704 . 10.1016/j.midw.2012.06.014 22981380 16 Booth T , Booth W Parents with learning difficulties in the child protection system experiences and perspectives . J Intellect Disabil 2005 ;9 :109 –29 . 10.1177/1744629505053922 15937050 17 Tarleton B , Ward L , Howarth J Finding the right support. A review of issues and positive practice in supporting parents with learning difficulties and their children . London : Baring Foundation , 2006 . 18 Elvish J , Hames A , English S Parents with learning disabilities: an audit of referrals made to a learning disability team . Learn Disabil Rev 2006 ;11 :26 –33 . 10.1108/13595474200600016 19 Parish SL , Mitra M , Son E Pregnancy outcomes among U.S. women with intellectual and developmental disabilities . Am J Intellect Dev Disabil 2015 ;120 :433 –43 . 10.1352/1944-7558-120.5.433 26322390 20 Redshaw M , Malouf R , Gao H Women with disability: the experience of maternity care during pregnancy, labour and birth and the postnatal period . BMC Pregnancy Childbirth 2013 ;13 :174 10.1186/1471-2393-13-174 24034425 21 Llewellyn G , McConnell D , Mayes R Health of mothers with intellectual limitations . Aust N Z J Public Health 2003 ;27 :17 –9 . 10.1111/j.1467-842X.2003.tb00374.x 14705262 22 Šumilo D , Kurinczuk JJ , Redshaw ME Prevalence and impact of disability in women who had recently given birth in the UK . BMC Pregnancy Childbirth 2012 ;12 :31 10.1186/1471-2393-12-31 22540347 23 Goldacre AD , Gray R , Goldacre MJ Childbirth in women with intellectual disability: characteristics of their pregnancies and outcomes in an archived epidemiological dataset . J Intellect Disabil Res 2015 ;59 :653 –63 . 10.1111/jir.12169 25331275 24 Gaudion A , Godfrey C , Cutts H , Homeyard C Mothers with learning disability: access, information provision and ongoing engagement with antenatal care. A pilot project developing antenatal recourses. The Polyanna Project. http://thepolyannaprojectorguk/resources/polyanna_project-mothers_with_a_learning_disabilitypdf (accessed August 2016 ). 25 Mayes R , Llewellyn G , McConnell D “That's who I choose to be”: the mother identity for women with intellectual disabilities . Womens Stud Int Forum 2011 ;34 :112 –20 . 10.1016/j.wsif.2010.11.001 26 Baum S , Burns J Mothers with learning disabilities: experiences and meanings of losing custody of their children . Tizard Learn Disabil Rev 2007 ;12 :3 –14 . 10.1108/13595474200700018 27 Smith JA Reflecting on the development of interpretative phenomenological analysis and its contribution to qualitative research in psychology . Qual Res Psychol 2004 ;1 :39 –54 . 28 Smith JA , Flower P , Larkin M Interpretative phenomenological analysis: theory, method and research . Qual Res Psychol 2009 ;6 :346 –7 . 10.1080/14780880903340091 29 Hood R Combining phenomenological and critical methodologies in qualitative research . Qual Soc Work 2016 ;15 :160 –74 . 10.1177/1473325015586248 30 Larkin M , Watts S , Clifton E Giving voice and making sense in interpretative phenomenological analysis . Qual Res Psychol 2006 ;3 :102 –20 . 10.1191/1478088706qp062oa 31 Redshaw M , Henderson J Safely delivered: a national survey of women's experience of maternity care 2014 . Oxford : National Perinatal Epidemiology Unit https://wwwnpeuoxacuk/downloads/files/reports/Safely%20delivered%20NMS%202014pdf 32 McLeish J , Redshaw M ‘We have beaten HIV a bit’: a qualitative study of experiences of peer support during pregnancy with an HIV Mentor Mother project in England . BMJ Open 2016 ;6 :e011499 10.1136/bmjopen-2016-011499 33 NHS . NHS choices website. http://www.nhs.uk/Conditions/pregnancy-and-baby/pages/pregnancy-and-baby-care.aspx (accessed September 2016 ). 34 Castell E , Kroese B Midwives' experiences of caring for women with learning disabilities—a qualitative study . Midwifery 2016 ;36 :35 –42 . 10.1016/j.midw.2016.02.001 27106942 35 National Collaborating Centre for Women's and Children's Health . Pregnancy and complex social factors: a model for service provision for pregnant women with complex social factors . London : NICE , 2010 https://www.nice.org.uk/guidance/cg110/resources/pregnancy-and-complex-social-factors-a-model-for-service-provision-for-pregnant-women-with-complex-social-factors-35109382718149 (accessed June 2016 ). 36 Begley C , Hames A , Lalor J The strengths and weaknesses of publicly-funded Irish Health Services Provided to Women with Disabilities in Relation to Pregnancy, Childbirth and Early Motherhood School of Nursing and Midwifery, Trinity College Dublin. https://nursing-midwiferytcdie/assets/publications/pdf/nda-report-final-2010pdf (accessed September 2016 ). 37 Porter E , Kidd G , Murray N Developing the pregnancy support pack for people who have a learning disability . Br J Lear Disabil 2012 ;40 :310 –17 . 10.1111/j.1468-3156.2011.00713.x 38 Höglund B , Lindgren P , Larsson M Midwives’ knowledge of, attitudes towards and experiences of caring for women with intellectual disability during pregnancy and childbirth: a cross-sectional study in Sweden . Midwifery 2013 ;29 :950 –5 . 10.1016/j.midw.2012.12.002 23415362 39 Beake S , Clark LL , Turner T A mixed methods study to develop and pilot a competency assessment tool to support midwifery care of women with intellectual disabilities . Nurse Educ Today 2013 ;33 :901 –6 . 10.1016/j.nedt.2012.08.007 22981517 40 Quartermaine D ed. Maternity services for women with learning difficulties: a report of a partnership of midwives, community nurses and parents . London : The Maternity Alliance , 1999 . 41 Cox R , Kroese BS , McGarry A Pregnancy, birth and postnatal support for women with intellectual disabilities: the Can Do Doula model . Int J Birth Parent Educ 2015 ;2 :19 –22 . 42 Booth T , Booth W Unto us a child is born: the trials and rewards of parenthood for people with learning difficulties . Aust N Z J Dev Disabil 1995 ;20 :25 –39 . 43 Sheerin FK , Keenan PM , Lawler D Mothers with intellectual disabilities: interactions with children and family services in Ireland . Br J Learn Disabil 2013 ;41 :189 –96 . 10.1111/bld.12034 44 Cleaver H , Nicholson D Parental learning disability and children's needs: family experiences and effective practice . London : Department for Children, Schools and Families http://webarchivenationalarchivesgovuk/20130401151715/https:/wwweducationgovuk/publications/eOrderingDownload/DCSF-00273-2008pdf (accessed June 2016 ). 45 Children Act. London: HMSO , 1989 . 46 Re G and A (care order: freeing order: parents with a learning disability) [2006] NI Fam 8. http://wwwcourtsnigovuk/en-gb/judicial%20decisions/publishedbyyear/documents/2006/2006%20nifam%208/j_j_gilc5486htm (accessed September 2016 ). 47 Re D. (A Child) (No 3) [2016] EWFC 1. https://wwwjudiciarygovuk/judgments/re-d-a-child-no-3/ (accessed September 2016 ). 48 Homeyard C , Montgomery E , Chinn D Current evidence on antenatal care provision for women with intellectual disabilities: a systematic review . Midwifery 2016 ;32 :45 –57 . 10.1016/j.midw.2015.10.002 26518177 49 Brown H Ordinary women: issues for women with learning disabilities . Br J Learn Disabil 1996 ;24 :47 –51 . 10.1111/j.1468-3156.1996.tb00201.x

PMC005xxxxxx/PMC5372072.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01365010.1136/bmjopen-2016-013650Cardiovascular MedicineResearch150616831694Global cardiovascular risk assessment in the primary prevention of cardiovascular disease in adults: systematic review of systematic reviews http://orcid.org/0000-0003-1330-3636Collins Dylan R J 1Tompson Alice C 1Onakpoya Igho J 1Roberts Nia 2Ward Alison M 1Heneghan Carl J 1 1 Nuffield Department of Primary Care Health Sciences, Centre for Evidence-Based Medicine, University of Oxford, Oxford, UK 2 Bodleian Health Care Libraries, University of Oxford, Oxford, UK▸ Additional material is available. To view please visit the journal online (http://dx.doi.org/10.1136/bmjopen-2016-013650). Correspondence to Dylan R J Collins; dylan.collins@phc.ox.ac.uk2017 24 3 2017 7 3 e01365027 7 2016 8 11 2016 15 12 2016 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective To identify, critically appraise and summarise existing systematic reviews on the impact of global cardiovascular risk assessment in the primary prevention of cardiovascular disease (CVD) in adults. Design Systematic review of systematic reviews published between January 2005 and October 2016 in The Cochrane Library, EMBASE, MEDLINE or CINAHL databases, and post hoc analysis of primary trials. Participants, interventions, outcomes Systematic reviews of interventions involving global cardiovascular risk assessment relative to no formal risk assessment in adults with no history of CVD. The primary outcomes of interest were CVD-related morbidity and mortality and all-cause mortality; secondary outcomes were systolic blood pressure (SBP), cholesterol and smoking. Results We identified six systematic reviews of variable but generally of low quality (mean Assessing the Methodological Quality of Systematic Reviews 4.2/11, range 0/11 to 7/11). No studies identified by the systematic reviews reported CVD-related morbidity or mortality or all-cause mortality. Meta-analysis of reported randomised controlled trials (RCTs) showed small reductions in SBP (mean difference (MD) −2.22 mm Hg (95% CI −3.49 to −0.95); I2=66%; n=9; GRADE: very low), total cholesterol (MD −0.11 mmol/L (95% CI −0.20 to −0.02); I2=72%; n=5; GRADE: very low), low-density lipoprotein cholesterol (MD −0.15 mmol/L (95% CI −0.26 to −0.05), I2=47%; n=4; GRADE: very low) and smoking cessation (RR 1.62 (95% CI 1.08 to 2.43); I2=17%; n=7; GRADE: low). The median follow-up time of reported RCTs was 12 months (range 2–36 months). Conclusions The quality of existing systematic reviews was generally poor and there is currently no evidence reported in these reviews that the prospective use of global cardiovascular risk assessment translates to reductions in CVD morbidity or mortality. There are reductions in SBP, cholesterol and smoking but they may not be clinically significant given their small effect size and short duration. Resources need to be directed to conduct high-quality systematic reviews focusing on hard patient outcomes, and likely further primary RCTs. Trial registration number CRD42015019821. total cardiovascular riskrisk scoreprimary preventioncardiovascular risk assessment ==== Body Strengths and limitations of this study This systematic review summarises evidence from six systematic reviews on the use of global cardiovascular disease (CVD) risk assessment for the primary prevention of CVD in adults and reports important patient outcomes. The quality of the systematic reviews was assessed using Assessing the Methodological Quality of Systematic Reviews and was generally poor, with inconsistencies in methods, outcomes, quality appraisal and reporting. Owing to the deficiencies in the reviews, we undertook post meta-analyses, which included further analysis of the identified primary studies within the reviews and quality appraisal using GRADE. Our systematic review was strengthened by its broad inclusion criteria, which reduced the chance we missed relevant systematic reviews. The findings of our post hoc meta-analyses should be interpreted with caution given the paucity of evidence and low to very low GRADE scores. Introduction Cardiovascular disease (CVD) is the leading cause of death worldwide.1 Contrary to popular belief, death and disability from CVD is also a major burden in low-resource settings2 3 and despite impressive global reductions in mortality over the last two decades, years-of-life-lost due to CVD is rising in low-income and middle-income countries.4 Prevention is therefore a worldwide priority. Global CVD risk assessment (also referred to as absolute risk assessment, total risk assessment or risk scoring) is an integrated approach to prevention that recognises the hazards of multiple risk factors to determine the absolute risk of experiencing a CVD event in a given time period. Almost all CVD guidelines recommend some form of risk scoring as a way to prioritise and plan primary prevention interventions.5–7 This practice is entrenched in most high-income countries,5 6 8 and endorsed by the WHO for low-income and middle-income countries.7 Even with an abundance of clinical guidance and effective interventions targeting modifiable risk factors, the majority of patients at risk do not achieve sufficient risk factor control.9–14 In jurisdictions where risk scoring is part of routine clinical practice, this may be due to implementation challenges, such as physician attitudes to and understanding of CVD risk assessment,14–16 challenges communicating risk to patients and patients' understanding of risk17 18 or lack of a true effect. Although calibration and discrimination studies of prognostic risk score models are important, trials on the prospective use of risk scores in practice are required to determine their impact on patient outcomes. While the development of global cardiovascular risk assessment has been the focus of much research,19–22 there is little certainty about its effectiveness in the prevention of CVD. We conducted a systematic review of systematic reviews to identify, critically appraise and summarise evidence on the impact of global cardiovascular risk assessment in the primary prevention of CVD. We chose to conduct a systematic review of systematic reviews because they can be used to synthesise evidence for a given intervention on a diversity of outcomes, in addition to identifying limitations in the methodology and quality of existing systematic reviews.23 Methods We prospectively registered our protocol on PROSPERO (registration CRD42015019821),24 and deviations from this plan, which were minor, are described in online supplementary S1 appendix 1. 10.1136/bmjopen-2016-013650.supp1supplementary S1 appendix 1 Search strategy We searched the Cochrane Library, EMBASE, MEDLINE and CINAHL databases from January 2005 to October 2016 (see online supplementary S2 appendix 2 for detailed search strategies). We searched from 2005 because we felt that more recent systematic reviews were likely to capture the same primary studies as older systematic reviews. No language restrictions were applied. Search strategies used a mix of Medical Subject Heading (MeSH) terms and keywords and were developed with an information specialist. We supplemented this strategy by handsearching the bibliographies of included reviews, forward citation searching on Google Scholar and Web of Science and by contacting experts. 10.1136/bmjopen-2016-013650.supp2supplementary S1 appendix 2 Inclusion criteria We included systematic reviews of studies of any design that included adults (18 years of age or older) with no history of CVD (including atrial fibrillation). Reviews were eligible for inclusion if they considered interventions involving global CVD risk assessment relative to no risk assessment, irrespective of who performed the risk assessment, how it was performed and to whom the information was directed. We excluded studies of perioperative CVD risk scores, analogue patients, patients with existing CVD, validation studies and accuracy studies of CVD risk scores. The primary outcomes of interest were CVD-related morbidity or mortality and all-cause mortality. We secondarily considered systolic blood pressure (SBP), cholesterol and smoking, because together these cause the majority of premature CVD and have effective interventions to treat them.25 26 Study selection, data extraction and quality assessment Study selection, data extraction and quality assessment was done in two phases. We first screened systematic reviews for inclusion, and then screened primary studies reported in the systematic reviews as required. Two reviewers screened titles and abstracts for full text review and disagreement was resolved by consensus. We screened full text articles for inclusion using the same method. Two reviewers independently extracted salient characteristics about the included reviews and their description of included studies, including the classification of interventions as either directed toward the patient, provider or the patient and provider. The same reviewers independently assessed methodological quality using the Assessing the Methodological Quality of Systematic Reviews (AMSTAR) checklist.27 We documented the method of quality assessment and extracted the quality of primary studies as reported by the review authors. If a primary study was reported by more than one review, we extracted the quality score from the review with the highest AMSTAR rating. We did not initially plan to extract data from primary studies; however, to address poor or insufficient reporting, or to verify intervention classifications, the second phase involved the same two reviewers extracting additional data from the primary studies as needed. We extracted data from primary studies as reported in systematic reviews and tabulated all results by outcome using Microsoft Excel 2010 and Microsoft Word 2010. In the event of a discrepancy, we used the data reported in the primary study rather than the systematic review. When outcomes were reported at multiple time points in a single study, we selected the longest follow-up time with complete data. Data synthesis We narratively reported systematic reviews and tabulated their salient characteristics, including review, author conclusions and AMSTAR ratings. For our post hoc analysis of primary studies reported by systematic reviews, we used RevMan V.5.3 to construct forest plots and inverse variance-weighted random effects meta-analyses.28 One author entered data into RevMan and these were independently verified by a second author. We only included randomised studies in meta-analyses; other study designs or randomised studies with insufficient data to meta-analyse were reported narratively. For continuous outcomes, we calculated mean differences and for dichotomous outcomes, risk ratios. We performed subgroup analyses by the intended target of the intervention (patient, provider or both). We imputed standard error (SE) when reasonable, using the average SE of studies reporting the same outcome and in the same subgroup, and in such cases a sensitivity analysis was performed. For each outcome meta-analysed, we used the GRADE methodology to rate the quality of evidence using GRADEpro GDT software.29 30 This was done by one author and checked by a second; disagreements were resolved by consensus. We used I2 values to estimate statistical heterogeneity. Results We screened 6877 studies for inclusion, 6727 of which were excluded as irrelevant based on title and abstract (figure 1). Following full-text review of 150 studies, six systematic reviews met the inclusion criteria. We excluded 144 articles, the majority (n=110) due to the study design. The six included systematic reviews together reported 122 unique primary studies, 16 of which were relevant based on our inclusion criteria, including adults with no history of CVD and global CVD risk assessment as an intervention relative to no risk assessment. Of these, none reported CVD-related morbidity or mortality or all-cause mortality, 12 reported SBP, 10 reported cholesterol and 9 reported smoking cessation (figure 1). Figure 1 Flow diagram of included systematic reviews, primary studies and the outcomes reported. Summary of included systematic reviews Characteristics of included systematic reviews are summarised in table 1. Reviews were generally broad in scope, published from 2006 to 2015, considered all study designs and collectively searched 11 databases. AMSTAR ratings ranged from 0/11 to 7/11 (mean 4.2/11); we did not exclude any reviews based on quality (see online supplementary S3 appendix 3). The conclusions of reviews were consistently of general uncertainty about the impact of cardiovascular risk assessment on hard patient outcomes. Table 1 Characteristics of included systematic reviews Systematic review ID Brindle 200631 Sheridan 200832 Sheridan 201033 Waldron 201134 van Dieren 201235 Usher-Smith 201536 Review objective(s) ‘To determine the accuracy of assessing CVD risk in the primary prevention of CVD and its impact on clinical outcomes’ ‘To assess whether global CHD risk scores result in clinical benefits or harms’ ‘To assess the effect of providing global CHD risk information to adults’ ‘To compare different interventions used to communicate cardiovascular risk and assess their impact on patient related outcomes’ ‘To review the primary prevention studies that focused on the development, validation and impact assessment of a cardiovascular risk model, scores or rules’ ‘To systematically review whether the provision of information on cardiovascular disease (CVD) risk to healthcare professionals and patients impacts their decision-making, behaviour and ultimately patient health.’ Population People ‘predominantly free from symptomatic CVD’ ‘Adults (>18) with no prior history of CVD’ ‘Adults with no history of CVD’ Adults (>18) ‘People with type 2 diabetes’ People ‘with no history of CVD’ Intervention(s) ‘Healthcare professional using a CVD risk score to aid primary prevention’ ‘Physician knowledge of a global CHD risk score’ ‘Global CHD risk presentation as the primary intervention or part of a multipart intervention’ ‘Communication interventions (of any format) for individualised CVD assessment’ ‘CVD predictions models that have been developed… or validated in a diabetes population’ ‘Intervention strategy consisted of provision of a CVD risk model estimate to either physicians or patients’ Comparison(s) Usual care ‘Either simple risk factor counting or no formal assessment of risk’ Not prespecified Control or usual care arm Not prespecified No ‘provision of a CVD risk model estimate’ Country of primary review author UK United States of America United States of America UK The Netherlands UK AMSTAR 4/11 4/11 7/11 5/11 0/11 5/11 Eligible study designs RCTs Any design Any design ‘Any quantitative design’ Not specified Randomised and non-randomised primary studies Databases searched CENTRAL, MEDLINE, EMBASE, CINAHL, PsycINFO, ZETOC, ISI Proceedings MEDLINE MEDLINE, PsycINFO, CINAHL, Cochrane Database ASSIA, CINAHL, EMBASE, MEDLINE, PsycINFO, Science Citation Index expanded MEDLINE MEDLINE, PubMed Date of database search 2004 2007 2008 2008 2011 2013 Meta-analysis? No No No No No Yes Review method of quality assessment No formal assessment. ‘Information on the methodological quality of the trials including the method of randomisation, concealment of allocation, baseline group comparisons and blind outcome assessment was collected’ Criteria adapted from the US Preventive Services Task Force Criteria adapted from the US Preventive Services Task Force Downs and Black checklist Quality not assessed Followed critical appraisal skills programme guidelines Review authors' conclusions ‘Evidence supporting the use of cardiovascular risk scores for primary prevention is scarce’ ‘We found surprisingly little evidence that physician knowledge of global CHD risk currently translates into improved clinical outcomes’ ‘Global CHD risk information seems to improve the accuracy of risk perception and may increase intent to initiate CHD prevention among individuals at moderate to high risk. The effect of global risk presentation on more distal outcomes is less clear and seems to be related to the intensity of accompanying interventions.’ ‘Better quality trials are needed that compare different risk presentation formats before conclusions can be drawn’ ‘The impact of applying these risk scores in clinical practice is almost completely unknown, but their use is recommended in various national guidelines.’ ‘There seems evidence that providing CVD risk model estimates to professionals and patients improves perceived CVD risk and medical prescribing, with little evidence of harm on psychological well- being.’ CHD, coronary heart disease. 10.1136/bmjopen-2016-013650.supp3supplementary S1 appendix 3 Brindle et al31 (AMSTAR 4/11) sought to determine the accuracy and impact of CVD risk assessment. To assess impact, they considered randomised controlled trials (RCTs) of healthcare professionals using a CVD risk score to aid in primary prevention, relative to usual care, in populations predominantly free from symptomatic CVD (table 1). Although the authors identified 31 primary studies, the majority of the studies assessed the accuracy of CVD risk assessment and therefore were beyond the scope of our review. The authors concluded that there was little evidence to support the use of risk scores for primary prevention. Sheridan and Crespo32 (AMSTAR 4/11) conducted a much more focused review, specifically aimed to assess the benefits and harms of physician knowledge of a global CVD risk score, relative to no formal risk assessment, in adults with no history of CVD. Although the authors only searched MEDLINE, they identified 11 primary studies, most of which were relevant to our review (table 1), and reached similar conclusions as Brindle 2006: ‘surprisingly little evidence that physician knowledge of global CHD risk translates into improved clinical outcomes’.32 Sheridan et al33 (AMSTAR 7/11) was distinct from Sheridan 2008 in that eligible interventions were targeted at patients rather than physicians. To be eligible, studies (of any design) had to be interventions of global CVD risk information, solely or as part of a complex intervention, directed at adults with no history of CVD to determine the impact of patient-targeted risk information on patient outcomes (table 1). In this review, Sheridan et al searched multiple databases and identified 18 primary studies; most were relevant to our review. They concluded that the effect of risk information on patients may improve proximal outcomes, such as accuracy or risk perception, in those who are high risk but the effect on distal outcomes was not clear. Waldron et al34 (AMSTAR 5/11) reviewed studies of any quantitative design that used communication interventions for individualised cardiovascular risk and compared them to a control or usual care and assessed their impact on patient-related outcomes (table 1). Although the authors performed a comprehensive search, only three of the 15 primary studies identified were in actual patients—the rest were of analogue patients and therefore were not relevant to our review. The authors were unable to draw strong conclusions citing the need for better quality trials. Van Dieren et al35 (AMSTAR 0/11) conducted a two-part review; the first part sought to identify risk scores that had been developed or validated for people with type two diabetes, and the second part sought to find studies aimed at assessing the impact on patient outcomes when these scores were implemented in practice (table 1). The authors therefore only searched for impact studies on the risk scores that had been identified in the former half of their review. Although the authors identified 45 primary studies, only three studies were identified from the latter half of the review and were relevant to our review. They concluded that ‘the impact of applying these risk scores in clinical practice is almost completely unknown’.35 Usher-Smith et al36 (AMSTAR 5/11) conducted the most recent and most comprehensive review, although the searches were conducted in 2013. The authors considered randomised or non-randomised studies of interventions consisting of the provision of a CVD risk estimate to patients, or their providers, in patients with no history of CVD (table 1). Searching MEDLINE and PubMED, they identified 17 primary studies, the vast majority of which had previously been reported by one of the aforementioned systematic reviews. However, this review was the first and only review to conduct meta-analyses, which helped draw stronger conclusions about impact. The authors concluded that global risk assessment could improve prescribing and perceived risk, but not smoking, and their meta-analyses showed no significant effect on SBP or pooled total cholesterol and low-density lipoprotein (LDL) cholesterol. Effects of interventions As the included reviews were of variable, but generally low quality and only one attempted meta-analysis, we conducted a post hoc analysis of the primary studies reported by included systematic reviews that met our inclusion criteria to gain further insight from the reported data. Furthermore, some of the reviews insufficiently described included primary studies. Therefore, in addition to meta-analysis, we tabulated the characteristics of included primary studies, including quality ratings for each study as reported by review authors, which are included in online S4 supplementary appendix 4. The median follow-up of all included primary studies reported by reviews was 12 months, but ranged from 2 to 36 months (see online supplementary S4 appendix 4). Most interventions used Framingham or a Framingham-derived risk score; these interventions and their controls are summarised in detail in online supplementary S4 appendix 4. None of the systematic reviews used GRADE to evaluate the quality of studies, and we therefore included GRADE tables for each outcome and subgroup in online supplementary S5 appendix 5. 10.1136/bmjopen-2016-013650.supp4supplementary S1 appendix 4 10.1136/bmjopen-2016-013650.supp5supplementary S1 appendix 5 Primary outcomes Mortality No reviews included studies reporting CVD-related or all-cause mortality (table 2). Table 2 Meta-analysis of the impact of global cardiovascular risk assessment on cardiovascular disease morbidity, mortality, systolic blood pressure, total cholesterol, low-density lipoprotein cholesterol and smoking cessation Outcome Patient directed Provider directed Patient and provider directed Overall effect estimate CVD morbidity Effect estimate − − − − I2 (%) − − − − Quality (GRADE) − − − − Mortality (CVD-related or all-cause) Effect estimate − − − − I2 (%) − − − − Quality (GRADE) − − − − Systolic blood pressure (mm Hg) Effect estimate (MD) (95% CI) −4.88 (−8.57 to −1.19) −1.14 (−2.09 to −0.19) −2.77 (−5.91 to 0.37) −2.22 (−3.49 to −0.95) I2 (%) 0 46 76 66 Quality (GRADE) very low (●○○○) low (●●○○) very low (●○○○) very low (●○○○) Total cholesterol (mmol/L) Effect estimate (MD) (95% CI) −0.07 (−0.13 to −0.02) −0.01 (−0.08 to 0.06) −0.26 (−0.38 to −0.15) −0.11 (−0.20 to −0.02) I2 (%) 0 single study 0 72 Quality (GRADE) very low (●○○○) very low (●○○○) very low (●○○○) very low (●○○○) LDL cholesterol (mmol/L) Effect estimate (MD) (95% CI) −0.15 (−0.27 to −0.03) − −0.23 (−0.47 to 0.01) −0.15 (−0.26 to −0.05) I2 (%) 58 − single study 47 Quality (GRADE) very low (●○○○) − very low (●○○○) very low (●○○○) Smoking cessation (risk of quitting) Effect estimate (RR) (95% CI) 1.53 (1.07 to 2.19) 1.90 (0.43 to 8.29) 0.30 (0.01 to 7.07) 1.62 (1.08 to 2.43)   I2 (%) 0 57 single study 17 Quality (GRADE) low (●●○○) very low (●○○○) moderate (●●●○) low (●●○○) Subgroup analyses were performed by interventions targeting the patient, the provider or both. LDL, low-density lipoprotein; MD, mean difference, the difference of the mean differences from baseline to follow-up of the intervention arm compared with the control arm; RR, risk ratio. CVD morbidity No reviews included studies reporting CVD-morbidity (table 2). Secondary outcomes Systolic blood pressure Five reviews reported a total of 12 unique primary studies (n=23 346) (see online supplementary S4 appendix 4). Of the 12 primary studies, 3 were cluster RCTs, 8 were RCTs and 1 was an observational follow-up of an RCT. Nine of the controlled trials (n=7537) were reported with sufficient data to meta-analyse, resulting in a mean difference of mean change of SBP of −2.22 mm Hg (95% CI −3.49 to −0.95; I2=66%; GRADE: very low) (figure 2; table 2). Subgroup analyses by studies of interventions targeting the patient (n=3), provider (n=3) or both (n=3), resulted in mean differences of −4.88 mm Hg (95% CI −8.57 to −1.19; I2=0%; GRADE: very low), −1.14 mm Hg (95% CI −2.09 to −0.19; I2=46%; GRADE: low) and −2.77 mm Hg (95% CI −5.91 to 0.37; I2=76%; GRADE: very low), respectively, in the intervention group relative to control (figure 2; table 2). We imputed the SE of Lovibond (1986); excluding this study from the meta-analysis did not significantly change the results of the subgroup analysis or overall estimate: −4.82 mm Hg (95% CI −9.36 to −0.27; I2=0%) and −2.12 mm Hg (95% CI −3.41 to −0.83; I2=68%), respectively. Figure 2 Forest plot and meta-analysis of the mean difference of mean change in systolic blood pressure (mm Hg) from baseline to follow-up between intervention and control groups. Two additional patient-directed studies were identified. The first, an RCT,37 evaluated the impact of health checks which comprised a Framingham-derived risk score and counselling about risk, relative to usual care, in middle-aged adults. The study found a difference in SBP of −2.5 mm Hg (95% CI −3.7 to −1.3) between the intervention and control group after 3 years; however, this was not adjusted to baseline values.33 The second, an observational follow-up of an RCT,38 found no significant difference between patients who received their risk score along with a decision analysis tool, relative to usual care, after 3 years: adjusted difference of SBP between intervention and control at follow-up was 0.94 mm Hg (95% CI −3.2 to 5.1, p value 0.65). Finally, one RCT, targeting patients and providers, trained nurses on risk factor measurement and counselling, and provided patients with a CVD risk score and two to six sessions of counselling.39 Compared with usual care after 1 year, unadjusted differences in SBP between intervention and control were −7.3 mm Hg (p value not reported) and −6.2 mm Hg (p value not reported), in men and women, respectively.33 Cholesterol Six reviews reported a total of nine RCTs and one cluster RCT (see online supplementary S4 appendix 4). We considered total cholesterol and LDL cholesterol, and analysed them separately. Total cholesterol Of the seven trials (n=23 406) reporting total cholesterol, five (n=7813) were reported with sufficient data to meta-analyse, resulting in a mean difference of −0.11 mmol/L (95% CI −0.20, −0.02; I2=72%; GRADE: very low) (figure 3; table 2). Subgroup analyses by studies of interventions targeting the patient (n=2), provider (n=1) or both (n=2), resulted in mean differences of −0.07 mmol/L (95% CI −0.13,−0.02; I2=0%; GRADE: very low), −0.01 mmol/L (95% CI −0.08 to 0.06; GRADE: very low) and −0.26 mmol/L (95% CI −0.38 to −0.15; I2=0%; GRADE: very low) (figure 3; table 2). Figure 3 Forest plot and meta-analysis of the mean difference of mean change in total cholesterol (mmol/L) from baseline to follow-up between intervention and control group. Two additional RCTs were reported, one targeting patients and one targeting patients and providers. The former,37 evaluated the impact of health checks which comprised a Framingham-derived risk score and counselling about risk, relative to usual care, in middle-aged adults. The study found a difference in total cholesterol of −0.18 mmol/L (95% CI −0.26 to −0.12) between the intervention and control group after 3 years; however this was not adjusted to baseline values.33 The latter,39 targeting patients and providers, trained nurses on risk factor measurement and counselling, and provided patients with a CVD risk score and two to six sessions of counselling.39 Compared with usual care after 1 year, unadjusted differences in total cholesterol between intervention and control were −0.13 mmol/L (p value not reported) and –0.09 mmol/L (p value not reported), in men and women, respectively.33 LDL cholesterol Five trials (n=4940) reported data on LDL cholesterol, four of which (n=4505) when meta-analysed yielded a greater reduction in LDL cholesterol in the intervention group compared with the control group: mean difference of −0.15 mmol/L (95% CI −0.26 to −0.05, I2=47%; GRADE: very low) (figure 4; table 2). Subgroup analyses by studies of interventions targeting the patient (n=3) and the patient and provider (n=1), resulted in mean differences of −0.15 mmol/L (95% CI −0.27 to −0.03; I2=58%; GRADE: very low) and −0.23 mmol/L (95% CI −0.47 to 0.01; GRADE: very low), respectively (figure 4; table 2). Figure 4 Forest plot and meta-analysis of the mean difference of mean change in LDL cholesterol (mmol/L) from baseline to follow-up between intervention and control group. LDL, low-density lipoprotein. One additional RCT targeting patients assessed the impact of mailing personalised CVD risk assessment and encouraging discussion with their primary care physician to usual care.40 The primary outcome was the occurrence of an LDL cholesterol measurement of at least 0.78 mmol/L (30 mg/dL) lower than the baseline measurement during a 9-month follow-up. The trial showed no significant difference between intervention and control, intention-to-treat analysis resulted in an OR of 0.99 (95% CI 0.56 to 1.74, p=0.96).40 Smoking Three reviews reported a total of eight RCTs and one cluster RCT and (n=19 873) (see online supplementary S4 appendix 4). Seven of these trials (n=4131) were reported with sufficient data to meta-analyse, resulting in a risk ratio for quitting smoking of 1.62 (95% CI 1.08 to 2.43; I2=17%; GRADE: low) (figure 5; table 2). Subgroup analyses by studies of interventions targeting the patient (n=4), provider (n=2) or patient and provider (n=1) resulted in risk ratios of 1.53 (95% CI 1.07 to 2.19; I2=0%; GRADE: low), 1.90 (95% CI 0.43 to 8.29; I2=57%; GRADE: very low) and 0.30 (95% CI 0.01 to 7.07; GRADE: moderate), respectively (figure 5; table 2). Figure 5 Forest plot and meta-analysis of the risk ratio of quitting smoking from baseline to follow-up between intervention and control group. Two additional RCTs were reported by included reviews, one targeted patients and one targeted patients and providers. Of the former, one RCT targeting patients assessed the impact of mailing personalised CVD risk assessment and encouraging discussion with their primary care physician to usual care.40 As a secondary outcome, the trial sought to record the number of smokers who quit after nine months, but failed to provide the actual data stating, ‘non-smoking (among smokers) was rare and did not vary by group’.40 The other, an RCT targeting patients and providers, trained nurses on risk factor measurement and counselling, and provided patients with a CVD risk score and two to six sessions of counselling.39 Compared with usual care after one year, unadjusted differences in smoking prevalence between intervention and control were −4.1% (p value not reported) and −2.9% (p value not reported), in men and women, respectively.33 Discussion To the best of our knowledge, this is the first systematic review of systematic reviews on the impact of global cardiovascular risk assessment in the primary prevention of CVD. Reviews on this topic vary in the databases searched, search strategies used and the specific populations and interventions of interest, and their quality was generally poor (mean AMSTAR 4.2/11, range 0/11 to 7/11). However, they have consistently reported uncertainty about the effectiveness of global CVD risk assessment. Our results show that to date, no reviews or their included primary studies have reported CVD-related morbidity or mortality or all-cause mortality. Post hoc analysis of data reported by included reviews illustrated very low-quality evidence for small reductions in SBP, total cholesterol, LDL cholesterol and low-quality evidence for small increases in smoking cessation. We conducted a post hoc analysis of data reported by included systematic reviews to compensate for the poor evidence synthesis of some of the included reviews and to reduce research waste. Several of the included reviews highlighted a need for better quality trials on this topic, and our quality appraisal using GRADE corroborates these conclusions. Our post hoc analysis illustrated that all effect estimates were of low GRADE or very low GRADE quality. Our meta-analyses of data reported by systematic reviews, although not the primary aim of our systematic review, advances our knowledge on the impact of total cardiovascular risk assessment beyond what was reported by individual systematic reviews. Given the disparate and poor quality of data reporting and quality appraisal of the included systematic reviews, we felt it was responsible to perform better quality evidence synthesis and quality appraisal. Our meta-analysis of SBP added five additional RCTs compared with the previously reported meta-analysis36 and revealed for the first time a statistically significant small reduction in SBP. We also report for the first time statistically significant, but small, reductions in total cholesterol, LDL cholesterol and smoking. These results might be explained by improvements in the prescription of cholesterol-lowering and blood pressure-lowering treatment,36 and increasing accuracy of patient risk perception.33 36 Reductions in SBP, LDL cholesterol and smoking tended to be greatest when cardiovascular risk information was directed at patients. The delineation between patient-directed and provider-directed interventions is not well resolved, since patients may discuss risk information with their provider and vice versa. This phenomenon may contaminate the subgroup analysis. A previous review33 concluded that greater effects were seen in patients with increased baseline CVD risk; adjusting for baseline CVD risk in our analyses did not explain the observed trend toward patient-directed interventions. Observed reductions in blood pressure, cholesterol and smoking could lead to reductions in overall risk and eventual reduction in CVD events. For this to be true, such changes must be sustained long term to have a meaningful impact on event rates in individual patients. However, the median follow-up of all eligible studies reported by included systematic reviews was only 12 months. Therefore, we can have little confidence from the systematic review literature in the prospective use of CVD risk scores translating into long-term clinical benefits. There seems to be a widespread acceptance, including among several guidelines, of total cardiovascular risk assessment despite no systematic review evidence for its effectiveness on long-term patient outcomes: neither the National Institute of Health and Care Excellence (NICE)8 or WHO41 currently acknowledge this evidence gap. However, the 2013 American College of Cardiology (ACC)/American Heart Association (AHA) Guideline on the Assessment of Cardiovascular risk acknowledges that ‘none of the risk assessment tools…have been formally evaluated in randomised controlled trials… with clinical events as outcomes’.42 NICE does go so far as to make recommendations for research, but overlooks the need to evaluate the global CVD risk approach.43 Further, global cardiovascular risk assessment is included in the Quality and Outcomes Framework in the UK as an incentive for general practitioners to be compliant with guidance for primary prevention.44 There is evidence, from studies retrospectively risk scoring patients, that the absolute reduction in CVD events from cholesterol-lowering and blood pressure-lowering treatment is associated with baseline CVD risk.45 46 It is important to distinguish, however, that these data do not illustrate that the prospective use of global CVD risk assessment improves patient outcomes, but suggest that targeting high-risk patients is an efficient way to allocate resources to reduce CVD events. Nor do studies of calibration and discrimination imply effectiveness. We speculate the paucity of data may partly be explained by the assumption that accuracy of risk scores is a proxy for effectiveness, therefore undermining rationale for further study. While mechanical prediction, such as the use of global cardiovascular risk assessment, is generally a superior method of data combination than clinical judgement;47 it is unknown whether this holds true for cardiovascular risk. Moreover, clinicians still rely on clinical judgement to integrate risk scores within the broader context of the patient—especially for risk factors that are not included in the risk assessment algorithm—therefore risk stratification is not strictly a practice that can be replaced with mechanical prediction. The uncertainty around the clinical utility of this practice has been highlighted by UK general practitioners, who expressed that broad implementation created considerable confusion, and emphasised the need for simplification and for guidance to be updated to reflect how risk scores are actually used.15 Strengths and limitations Our systematic review of systematic reviews was strengthened by its broad inclusion criteria, which reduced the chance we missed relevant literature and allowed us to include reviews whose primary objective may not have been assessing the impact of risk assessment. This method provides quality appraised systematic reviews and their primary studies and therefore can help inform policy and practice, in addition to providing new insight from rigorous data synthesis from a breadth of sources. It also highlights trends in the systematic review literature. While there was heterogeneity in the risk scores used, the exact nature of the interventions, comparisons, follow-up time and patient demographics, we feel that such heterogeneity approximates the reality of clinical practice. The results therefore provide a good approximation of impact. While we believe we have taken steps to ensure an exhaustive search, it is possible some systematic reviews were missed. Our post hoc analysis of primary studies reported by included systematic review should be interpreted with caution. Its purpose is to better appraise and synthesise the primary studies which are already available in the systematic review literature, and therefore it does not represent the conclusions from a systematic review of individual primary studies, and it is possible that further primary studies exist. Given that these data were not appraised and synthesised to a high standard in the included systematic reviews, we felt it necessary to do so. There are likely further primary studies that can add to our meta-analysis, and high-quality systematic reviews should be conducted and these analyses should be updated. We were unable to perform sensitivity analyses due to a paucity of data. We did not contact authors of primary studies, which may have allowed us to adjust unadjusted study results which we were forced to report narratively. However, as these studies were reported in systematic reviews, some of which contacted study authors, we believe the likelihood of further data becoming available to be minimal, and the findings of these studies did not contradict our overall findings of the meta-analyses. In addition, we did not assess publication bias and reporting bias, not least because of the paucity of evidence, which may lead to overestimations of the effective sizes. We imputed the SE of one study,48 which may underestimate heterogeneity; however sensitivity analysis revealed only a 2% absolute increase in I2 for the overall effect estimate when this study was excluded. The current evidence base is limited in that all patients were from high-income countries, many were hypertensive and most interventions used a Framingham-derived risk score. Implications for research and practice The findings of our post hoc meta-analyses should be interpreted for practice with caution given the paucity of evidence and its low quality. Policymakers worldwide should take care when implementing global cardiovascular risk approaches, especially in an era where prevention efforts are expanding worldwide, as there was no systematic review evidence to suggest that the prospective use of global cardiovascular risk assessment in clinical practice reduces CVD events. Although the most recent systematic review was published in 2015, its literature search was conducted in 2013. Researchers should therefore update and improve the quality of systematic reviews and focus on CVD-related morbidity and mortality and other hard patient outcomes. In some jurisdictions, such as many low-income and middle-income countries, high-quality primary studies may be needed and these should be done before the widespread implementation of CVD risk scoring or, if not possible, in concert. Conclusion There is currently no evidence from systematic reviews that the prospective use of global cardiovascular risk assessment translates to reductions in CVD events. Systematic reviews, which were generally of poor quality, have consistently reported uncertainty about the impact of this practice on meaningful patient outcomes. These reviews found no studies reporting CVD-related morbidity or mortality, and the reductions in SBP, cholesterol and smoking may not be clinically significant given their small effect size and short duration. Despite the widespread popularity of global cardiovascular risk assessment, resources need to be directed in the first instance to conduct high-quality systematic reviews, and further RCTs powered to measure CVD-related morbidity and mortality may be needed. Contributors: DRJC conceived of the study, screened articles, extracted data, analysed the data and wrote the manuscript. ACT screened articles, extracted data, contributed to analysis and manuscript writing. IJO contributed to GRADE appraisal, analysis and manuscript writing. NR reviewed and contributed to the database search strategies. AMW contributed to the analysis and manuscript writing. CJH contributed to the methodological approach, the analysis and manuscript writing. Funding: This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 World Health Organization . Cardiovascular diseases (CVDs) 2015 http://www.who.int/mediacentre/factsheets/fs317/en/ (accessed 4 Aug 2015 ). 2 WHO . Noncommunicable diseases, fact sheet 2015 http://www.who.int/mediacentre/factsheets/fs355/en/ 3 Mathers CD , Loncar D Projections of global mortality and burden of disease from 2002 to 2030 . PLoS Med 2006 ;3 :e442 10.1371/journal.pmed.0030442 17132052 4 Roth GA , Huffman MD , Moran AE Global and regional patterns in cardiovascular mortality from 1990 to 2013 . Circulation 2015 ;132 :1667 –78 . 10.1161/CIRCULATIONAHA.114.008720 26503749 5 Pearson TA , Blair SN , Daniels SR AHA guidelines for primary prevention of cardiovascular disease and stroke: 2002 update: consensus panel guide to comprehensive risk reduction for adult patients without coronary or other atherosclerotic vascular diseases. American Heart Association Science Advisory and Coordinating Committee . Circulation 2002 ;106 :388 –91 . 10.1161/01.CIR.0000020190.45892.75 12119259 6 National Institute for Health and Care Excellence . Lipid modification: cardiovascular risk assessment and the modification of blood lipids for the primary and secondary prevention of cardiovascular disease 2015 https://www.nice.org.uk/guidance/cg181/resources/guidance-lipid-modification-cardiovascular-risk-assessment-and-the-modification-of-blood-lipids-for-the-primary-and-secondary-prevention-of-cardiovascular-disease-pdf 7 WHO . Package of essential noncommunicable (PEN) disease interventions for primary health care in low-resource settings . Geneva, Switzerland 2010 http://apps.who.int/iris/handle/10665/44260 (accessed 27 Oct 2014 ). 8 NICE . Cardiovascular risk assessment and the modification of blood lipids for the primary and secondary prevention of cardiovascular disease: methods, evidence, and recommendations. Published online first 2014 http://www.nice.org.uk/guidance/cg181/evidence/lipid-modification-update-full-guideline-243786637 9 Tang L , Patao C , Chuang J Cardiovascular risk factor control and adherence to recommended lifestyle and medical therapies in persons with coronary heart disease (from the National Health and Nutrition Examination Survey 2007–2010) . Am J Cardiol 2013 ;112 :1126 –32 . 10.1016/j.amjcard.2013.05.064 23827404 10 McFarlane SI , Jacober SJ , Winer N Control of cardiovascular risk factors in patients with diabetes and hypertension at urban academic medical centers . Diabetes Care 2002 ;25 :718 –23 . 10.2337/diacare.25.4.718 11919131 11 Pearson TA , Laurora I , Chu H The lipid treatment assessment project (L-TAP): a multicenter survey to evaluate the percentages of dyslipidemic patients receiving lipid-lowering therapy and achieving low-density lipoprotein cholesterol goals . Arch Intern Med 2000 ;160 :459 –67 . 10.1001/archinte.160.4.459 10695686 12 Wolf-Maier K , Cooper RS , Kramer H Hypertension treatment and control in five European countries, Canada, and the United States . Hypertension 2004 ;43 :10 –17 . 10.1161/01.HYP.0000103630.72812.10 14638619 13 Hajjar I , Kotchen TA Trends in prevalence, awareness, treatment, and control of hypertension in the United States, 1988–2000 . JAMA 2003 ;290 :199 –206 . 10.1001/jama.290.2.199 12851274 14 Camisasca P , Avanzini F , Alli C Overall cardiovascular risk still ignored in general practice care of hypertension . J Cardiovasc Risk 2002 ;9 :147 –52 . 10.1097/00043798-200206000-00003 12202837 15 Liew SM , Blacklock C , Hislop J Cardiovascular risk scores: qualitative study of how primary care practitioners understand and use them . Br J Gen Pract 2013 ;63 :e401 –7 . 10.3399/bjgp13X668195 23735411 16 Williams CL , Bollella M , Wynder E Preventive cardiology in primary care . Atherosclerosis 1994 ;108 (Suppl):S117 –26 . 10.1016/0021-9150(94)90157-0 7802718 17 Goldman RE , Parker DR , Eaton CB Patients’ perceptions of cholesterol, cardiovascular disease risk, and risk communication strategies . Ann Fam Med 2006 ;4 :205 –12 . 10.1370/afm.534 16735521 18 van Steenkiste B , van der Weijden T , Timmermans D Patients’ ideas, fears and expectations of their coronary risk: barriers for primary prevention . Patient Educ Couns 2004 ;55 :301 –7 . 10.1016/j.pec.2003.11.005 15530768 19 Conroy RM , Pyörälä K , Fitzgerald AP Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project . Eur Heart J 2003 ;24 :987 –1003 . 10.1016/S0195-668X(03)00114-3 12788299 20 D'Agostino RBS , Grundy S , Sullivan LM Validation of the Framingham coronary heart disease prediction scores: results of a multiple ethnic groups investigation . JAMA 2001 ;286 :180 –7 . 10.1001/jama.286.2.180 11448281 21 Hippisley-Cox J , Coupland C , Vinogradova Y Derivation and validation of QRISK, a new cardiovascular disease risk score for the United Kingdom: prospective open cohort study . BMJ 2007 ;335 :136 10.1136/bmj.39261.471806.55 17615182 22 Thomsen TF , McGee D , Davidsen M A cross-validation of risk-scores for coronary heart disease mortality based on data from the Glostrup Population Studies and Framingham Heart Study . Int J Epidemiol 2002 ;31 :817 –22 . 10.1093/ije/31.4.817 12177028 23 Higgins JPT , Green S , eds. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011] . The Cochrance Collaboration , 2011 www.cochrane-handbook.org 24 Collins D , Tompson A , Roberts N Global cardiovascular risk assessment in the primary prevention of cardiovascular diseases in adults: an overview of systematic reviews. PROSPERO Int. Prospect. Regist. Syst. Rev 2015 http://www.crd.york.ac.uk/PROSPERO/display_record.asp?ID=CRD42015019821 (accessed 4 Aug 2015 ). 25 Emberson JR , Whincup PH , Morris RW Re-assessing the contribution of serum total cholesterol, blood pressure and cigarette smoking to the aetiology of coronary heart disease: impact of regression dilution bias . Eur Heart J 2003 ;24 :1719 –26 . 10.1016/S0195-668X(03)00471-8 14522566 26 Cooper A , Nherera L , Calvert N Clinical guidelines and evidence review for lipid modification: cardiovascular risk assessment and the primary and secondary prevention of cardiovascular disease . London: National Collaborating Centre for Primary Care and Royal College of General Practitioners 2008 ;1 –241 . 27 Shea BJ , Grimshaw JM , Wells GA Development of AMSTAR: a measurement tool to assess the methodological quality of systematic reviews . BMC Med Res Methodol 2007 ;7 :10 10.1186/1471-2288-7-10 17302989 28 Review Manager (RevMan) [Computer program]. Version 5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration , 2014 . 29 Guyatt GH , Oxman AD , Vist GE GRADE: an emerging consensus on rating quality of evidence and strength of recommendations . BMJ 2008 ;336 :924 –6 . 10.1136/bmj.39489.470347.AD 18436948 30 GRADEpro GDT . GRADEpro Guideline Development Tool 2015 http://www.gradepro.org . 31 Brindle P , Beswick A , Fahey T Accuracy and impact of risk assessment in the primary prevention of cardiovascular disease: a systematic review . Heart 2006 ;92 :1752 –9 . 10.1136/hrt.2006.087932 16621883 32 Sheridan SL , Crespo E Does the routine use of global coronary heart disease risk scores translate into clinical benefits or harms? A systematic review of the literature . BMC Health Serv Res 2008 ;8 :60 10.1186/1472-6963-8-60 18366711 33 Sheridan SL , Viera AJ , Krantz MJ The effect of giving global coronary risk information to adults . Arch Intern Med 2010 ;170 :230 –9 . 10.1001/archinternmed.2009.516 20142567 34 Waldron C-A , van der Weijden T , Ludt S What are effective strategies to communicate cardiovascular risk information to patients? A systematic review . Patient Educ Couns 2011 ;82 : 169 –81 . 10.1016/j.pec.2010.04.014 20471766 35 van Dieren S , Beulens JWJ , Kengne AP Prediction models for the risk of cardiovascular disease in patients with type 2 diabetes: a systematic review . Heart 2012 ;98 :360 –9 . 10.1136/heartjnl-2011-300734 22184101 36 Usher-Smith JA , Silarova B , Schuit E Impact of provision of cardiovascular disease risk estimates to healthcare professionals and patients: a systematic review . BMJ Open 2015 ;5 :e008717 10.1136/bmjopen-2015-008717 37 Effectiveness of health checks conducted by nurses in primary care: final results of the OXCHECK study. Imperial Cancer Research Fund OXCHECK Study Group . BMJ 1995 ;310 :1099 –104 . 10.1136/bmj.310.6987.1099 7742676 38 Emmett CL , Montgomery AA , Peters TJ Three-year follow-up of a factorial randomised controlled trial of two decision aids for newly diagnosed hypertensive patients . Br J Gen Pract 2005 ;55 :551 –3 .16004744 39 Randomised controlled trial evaluating cardiovascular screening and intervention in general practice: principal results of British family heart study. Family Heart Study Group . BMJ 1994 ;308 :313 –20 . 10.1136/bmj.308.6924.313 8124121 40 Persell SD , Lloyd-Jones DM , Friesema EM Electronic health record-based patient identification and individualized mailed outreach for primary cardiovascular disease prevention: a cluster randomized trial . J Gen Intern Med 2013 ;28 :554 –60 . 10.1007/s11606-012-2268-1 23143672 41 WHO . Prevention of cardiovascular disease: guidelines for the assessment and managment of cardiovascular risk 2007 http://www.who.int/cardiovascular_diseases/guidelines/Fulltext.pdf 42 Goff DCJ , Lloyd-Jones DM , Bennett G 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines . Circulation 2014 ;129 :S49 –73 . 10.1161/01.cir.0000437741.48606.98 24222018 43 NICE . Cardiovascular risk assessment and the modification of blood lipids for the primary and secondary prevention of cardiovascular disease: appendices. Published online first 2014 http://www.nice.org.uk/guidance/cg181/evidence/lipid-modification-update-appendices-243786638 44 NHS Employers . 2015/16 General Medical Services (GMS) contract Quality and Outcomes Framework (QOF) 2015 http://www.nhsemployers.org/~/media/Employers/Documents/Primary care contracts/QOF/2015—16/2015 16 QOF guidance for stakeholders.pdf 45 Sundstrom J , Arima H , Woodward M Blood pressure-lowering treatment based on cardiovascular risk: a meta-analysis of individual patient data . Lancet (London, England) 2014 ;384 :591 –8 . 10.1016/S0140-6736(14)61212-5 46 Mihaylova B , Emberson J , Blackwell L The effects of lowering LDL cholesterol with statin therapy in people at low risk of vascular disease: meta-analysis of individual data from 27 randomised trials . Lancet 2012 ;380 :581 –90 . 10.1016/S0140-6736(12)60367-5 22607822 47 Grove WM , Zald DH , Lebow BS Clinical versus mechanical prediction: a meta-analysis . Psychol Assess 2000 ;12 :19 –30 . 10.1037/1040-3590.12.1.19 10752360 48 Lovibond SH , Birrell PC , Langeluddecke P Changing coronary heart disease risk-factor status: the effects of three behavioral programs . J Behav Med 1986 ;9 :415 –37 . 10.1007/BF00845131 3540308

PMC005xxxxxx/PMC5372073.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01472610.1136/bmjopen-2016-014726Geriatric MedicineProtocol15061698171216941698The incidence of delirium after cardiac surgery in the elderly: protocol for a systematic review and meta-analysis Liao Yulin 1Flaherty Joseph H 2Yue Jirong 3Wang Yanyan 1Deng Chuanyao 1Chen Ling 1 1 Department of Geriatrics, West China Hospital, Sichuan University, Chengdu, Sichuan, China 2 Division of Geriatrics, Department of Internal Medicine, Saint Louis University, St Louis, Missouri, USA 3 Department of Geriatrics, West China Hospital, Sichuan University, The National Center for Geriatric Clinical Research in West China Hospital, Chengdu, Sichuan, ChinaCorrespondence to Dr Jirong Yue; yuejirong11@hotmail.com2017 29 3 2017 7 3 e01472613 10 2016 7 2 2017 28 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction Delirium is one of the most common complications after cardiac surgery in the elderly. Future studies aimed at preventing postoperative delirium will need an accurate estimate of incidence. However, there are no available systematic reviews on the incidence, and reports of incidence of postoperative delirium after a cardiac operation vary widely with significant heterogeneity. Therefore, we aim to perform a systematic review and meta-analysis to determine the most accurate incidence possible of postoperative delirium in individuals aged >65 years after cardiac surgery. Methods and analyses We will undertake a comprehensive literature search among PubMed, EMBASE, the Cochrane Library, PsycINFO and CINAHL, from their inception to January 2017. Prospective cohort and cross sectional studies that described the incidence of delirium will be eligible for inclusion. The primary outcome will be the incidence of delirium. Risk of bias and methodological quality for the included studies will be assessed using a risk of bias tool for prevalence studies and the Cochrane guidelines. Heterogeneity of the estimates across studies will be assessed. Incidence data will be pooled by selective or emergency surgery. This systematic review will be reported according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA). Ethics and dissemination This proposed systematic review and meta-analysis is based on published data, and thus there is no requirement for ethics approval. The study will provide an up to date and accurate incidence of postoperative delirium among the older population after cardiac surgery, which is necessary for future research in this area. The findings of this study will be presented at conferences and disseminated through publication in a peer reviewed journal. Trial registration number CRD42016047773. deliriumsystematic review and meta-analysisincidenceprotocolelderly ==== Body Strengths and limitations of this study This study will provide an up to date and accurate incidence of postoperative delirium in the elderly after cardiac surgery. A subgroup analysis of certain patient characteristics make it possible to determine the incidence for specific populations and screen the populations with higher than average risk of delirium. The findings of this study will help guide researchers as they develop intervention trials to prevent delirium in the older population. Non-English articles will not be included in our study due to language difficulties and this may cause publication bias to some extent. Introduction Postoperative delirium is one of the most common complications for elderly patients who undergo surgery, characterised by acute onset, fluctuating course, inattention and, at times, an abnormal level of consciousness. Recent studies have provided evidence that postoperative delirium in elderly patients is associated with poor outcomes. Postoperative delirium is associated with an increased risk of long term cognitive dysfunction and functional decline compared with patients without postoperative delirium.1 Moreover, patients who had experienced delirium were also at increased risk of death.2 Although systematic review and meta-analysis of incidence exist for acute stroke, emergency department visits, intensive care units and orthopaedic surgery, no such systematic review and meta-analysis exists for delirium after cardiac surgery.3–5 Such a systematic review and meta-analysis is needed for several reasons. Future intervention studies aimed at preventing delirium will need an accurate estimate of incidence. Currently, reports of incidence vary too widely (8–65%)6–11 to use for comparison. Furthermore, if researchers perform placebo controlled intervention trials, the incidence in the placebo group could be compared with that of the meta-analysis to ensure the placebo group had the expected incidence. The incidence could also be used to help researchers perform power calculations for number of subjects needed in trials.12 Finally, the incidence of postoperative delirium varies depending on the characteristics of patients and type of surgery.6–9 For example, delirium is more common among older patients with baseline cognitive impairment compared with patients with normal cognitive function.13 14 Type of cardiac surgery is another factor that affects the incidence of delirium.15 16 A systematic review and meta-analysis that analyses and identifies higher incidence of certain subgroups (such as baseline cognitive impairment) could help future researchers focus interventions on these higher than average risk populations. Additionally, characteristics of higher than average risk populations would be important to recognise for researchers who recruit participants into study protocols so that the researchers will measure all potential baseline variables that increase the risk of postoperative delirium in both the interventional and control groups. Thus we intend to conduct a systematic review and meta-analysis to ascertain the incidence of delirium among older people undergoing cardiac surgery. We will also ascertain incidence based on certain patient characteristics (if data are available) that might have a higher than average expected incidence. It has been confirmed that there are numerous risk factors or patient characteristics associated with delirium, but some are mentioned much more often and we will conduct a subgroup analysis based on these characteristics. These characteristics include, but are not limited to, age ≥65 years (including two subgroups of older patients: 65–79 years and ≥80 years), baseline cognitive impairment, diabetes mellitus, depression and cerebrovascular disease, and type of cardiac surgery (‘closed’ surgery, such as valve replacement and ‘open’ surgery, such as bypass surgery), all of which have been found in previous meta-analyses on risk factors to be associated with the occurrence of delirium.17 18 Methods Criteria for included studies We will conduct a comprehensive search on the databases MEDLINE, EMBASE, the Cochrane Library, PsycINFO and CINAHL from their inception to January 2017. References of eligible studies and relevant review articles or meta-analyses will be manually searched for additional studies. The MeSH word and free words used are as follow: ‘delirium,’ ‘acute confusion,’ ‘acute organic psycho syndrome,’ ‘toxic confusion,’ ‘delir$’, ‘surg$’, ‘operati$’, ‘perioperati$’, ‘postoperati$’, ‘prevalence’, ‘incidence’, ‘occurrence’ (see the Ovid search strategy in online supplementary appendix 1). 10.1136/bmjopen-2016-014726.supp1supplementary appendix We will include: (1) prospective studies that include preoperative assessment of cognition; (2) cross sectional studies and other cohort study designs as long as preoperative assessment of cognition was done; (3) publications in English; (4) studies in which a validated measurement tool was used to screen for delirium as well as those that utilised diagnostic criteria for delirium as described in the Diagnostic and Statistical Manual of Mental Disorders (DSM-V/DSM-IV-TR) or International Classification of Diseases ICD-10; and (5) studies that provide the data necessary to calculate incidence. We will exclude: (1) retrospective studies and interventional studies and (2) studies that only report prevalence of delirium because it is not clear in these studies if patients had delirium preoperatively. Incident delirium is preferred to investigate the actual condition of delirium after cardiac surgery. The DSM criteria are widely accepted as the gold standard in the diagnosis of delirium, and the ICD criteria are also applied widely by clinicians. Instruments such as the Confusion Assessment Method (CAM),19 Delirium Rating Scale (DRS),20 Delirium Symptom Interview (DRI),21 Intensive Care Delirium Screening Checklist (ICDSC),22 etc, have been validated against DSM or ICD criteria. Studies using any of these methods for diagnosis of delirium are acceptable. Studies using non-validated instruments, such as the Mini-Mental Status Examination (MMSE)23 and the Short Mental Status Test (SMST),24 will be excluded as they were developed to measure cognitive impairment rather than delirium.25 Studies will be limited to human subjects aged ≥60 years. The population of interest will be older hospitalised patients that undergo cardiac surgery, such as elective coronary artery surgery or heart valve replacement/repair, either with or without coronary bypass grafting. Outcome The primary outcome is the incidence of postoperative delirium. Postoperative delirium is defined as an episode of delirium occurring after surgery. Study selection and data extraction Two authors will independently screen the titles and abstracts of all citations identified by the searches for potentially eligible studies. Full text of potentially eligible studies will be obtained and assessed according to the aforementioned inclusion criteria. We will present the process of search and study selection using a flow process chart. A standardised data extraction form will be developed for data extraction. Two authors will independently perform data extraction. We will collect the following information from every included trial: (1) publication (title, first author, year of publication); (2) study design; (3) patient demographics (sample size, mean age, gender ratio, type of surgery, baseline cognitive status); (4) details of outcome measures; (5) details necessary to assess the risk of bias. Any disagreements will be resolved by consulting a third author or the original authors will be contacted for further information if necessary. Risk of bias (quality) assessment We will incorporate quality assessment into our analyses by evaluating sources of bias that may affect the overall estimations. We will assess the quality of included studies using the risk of bias tool for prevalence studies developed by Hoy, and the Cochrane guidelines.26 27 Risk of bias and quality scores will be presented in a table. In addition, a minimum sample size will be calculated to differentiate estimates with good precision. Studies with a sample size equal to or higher than the minimum sample size will be classified as ‘with good precision.’ Small study effect on the effect size will be explored by funnel plots, and symmetry will be tested using Egger's test. Statistical analysis Estimates for the incidence of postoperative delirium will be pooled into a meta-analysis and displayed with 95% CIs. We will derive SEs where studies have provided the corresponding numerator and denominator for delirium incidence estimates. The study specific estimates will be pooled to obtain an overall summary estimate of the incidence across studies using a random effects, after stabilising the variance of individual studies with the Freeman–Tukey double arc-sine transformation.28 Statistical heterogeneity across studies will be assessed by the Cochrane Q test and quantified by calculating I2. A value of I2 >50% indicates substantial heterogeneity. Where heterogeneity is statistically significant, we will conduct a subgroup analysis to investigate the possible sources of heterogeneity according to the following variables: age, type of surgery, methods used for delirium diagnosis or assessment, pre-existing cognitive impairment, diabetes mellitus, depression and cerebrovascular disease.17 As different diagnostic methods for delirium have various sensitivity and specificity29 and this may influence the incidence of delirium to some extent,30 we will also perform subgroup analyses based on diagnostic methods, if possible. Furthermore, a sensitivity analysis will be performed to find out how our results would change if only high quality studies were considered. Symmetry of the funnel plot will be used to assess publication bias across studies or selective reporting bias. Data will be analysed using Stata V.13.1 for Windows. Reporting of this review This systematic review will be reported following the guideline of Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA),31 and of meta-analyses of observational studies (MOOSE).27 Discussion This systematic review and meta-analysis will define the incidence data for delirium in the elderly after cardiac surgery. Determining the current burden of postoperative delirium in the elderly undergoing cardiac surgery based on certain baseline patient characteristics will be important for clinicians providing care to this potentially vulnerable population and will help guide researchers as they develop interventional trials to prevent this important syndrome in the older population. Furthermore, understanding baseline (preoperative) patient characteristics that increase postoperative delirium is critical for balanced randomisation in interventional trials to prevent postoperative delirium. A major possible limitation of this study may be the heterogeneity of the studies, as the study populations' baseline of each trial and the methods used to assess delirium are so heterogeneous. To explore the possible source of heterogeneity, we will conduct subgroup analysis based on patients' baseline characteristics, type of cardiac surgery and diagnostic methods. Due to the language barrier, only English articles will be included in our study. This may be another limitation of this study, for which we may lose relevant data from non-English spoken areas and may cause publication bias to some extent. Contributors: YL and JY conceived and designed the protocol, and YL drafted the protocol manuscript. JHF and JY critically revised the manuscript for methodological and intellectual content. YW participated in the development of the search strategy. CD and LC planned the data extraction. All authors approved the final version. Competing interests: None declared. Ethics approval: Given that this is a protocol for a systematic review and based on published data, there is no requirement for ethics approval. Provenance and peer review: Not commissioned; externally peer reviewed. ==== Refs References 1 Saczynski JS , Marcantonio ER , Quach L Cognitive trajectories after postoperative delirium . N Engl J Med 2012 ;367 :30 –9 . 10.1056/NEJMoa1112923 22762316 2 Witlox J , Eurelings LS , de Jonghe P Delirium in elderly patients and the risk of postdischarge mortality, institutionalization, and dementia: a meta-analysis . JAMA 2010 ;304 :443 –51 . 10.1001/jama.2010.1013 20664045 3 Shi Q , Presutti R , Selchen D Delirium in acute stroke: a systematic review and meta-analysis . Stroke 2012 ;43 :645 –9 . 10.1161/STROKEAHA.111.643726 22267831 4 Barron EA , Holmes J Delirium within the emergency care setting, occurrence and detection: a systematic review . Emerg Med J 2013 ;30 :263 –8 . 10.1136/emermed-2011-200586 22833596 5 Bruce AJ , Ritchie CW , Blizard R The incidence of delirium associated with orthopedic surgery: a meta-analytic review . Int Psychogeriatr 2007 ;19 :197 –214 . 10.1017/S104161020600425X 16973101 6 Whitlock EL , Vannucci A , Avidan MS Postoperative delirium . Minerva Anestesiol 2011 ;77 :448 –56 .21483389 7 Robinson TN , Eiseman B Postoperative delirium in the elderly: diagnosis and management . Clin Interv Aging 2008 ;3 :351 –5 .18686756 8 Inouye SK , Westendorp RGJ , Saczynski JS Delirium in elderly people . Lancet 2014 ;383 :911 –22 . 10.1016/S0140-6736(13)60688-1 23992774 9 Rudolph JL , Marcantonio ER Review articles: postoperative delirium: acute change with long-term implications . Anesth Analg 2011 ;112 :1202 –11 . 10.1213/ANE.0b013e3182147f6d 21474660 10 Koster S , Hensens AG , Schuurmans MJ Prediction of delirium after cardiac surgery and the use of a risk checklist . Eur J Cardiovasc Nurs 2013 ;12 :284 –92 . 10.1177/1474515112450244 22694810 11 Norkienė I , Ringaitienė D , Kuzminskaitė V Incidence and risk factors of early delirium after cardiac surgery . Biomed Res Int 2013 ;2013 :323491 10.1155/2013/323491 24102052 12 Mu JL , Lee A , Joynt GM Pharmacologic agents for the prevention and treatment of delirium in patients undergoing cardiac surgery: systematic review and metaanalysis . Crit Care Med 2015 ;43 :194 –204 . 10.1097/CCM.0000000000000673 25289932 13 Fong TG , Davis D , Growdon ME The interface between delirium and dementia in elderly adults . Lancet Neurol 2015 ;14 :823 –32 . 10.1016/S1474-4422(15)00101-5 26139023 14 Jankowski CJ , Trenerry MR , Cook DJ Cognitive and functional predictors and sequelae of postoperative delirium in elderly patients undergoing elective joint arthroplasty . Anesth Analg 2011 ;112 :1186 –93 . 10.1213/ANE.0b013e318211501b 21415433 15 Schenning KJ , Deiner SG Postoperative delirium in the geriatric patient . Anesthesiol Clin 2015 ;33 :505 –16 . 10.1016/j.anclin.2015.05.007 26315635 16 Vasilevskis EE , Han JH , Hughes CG Epidemiology and risk factors for delirium across hospital settings . Best Pract Res Clin Anaesthesiol 2012 ;26 :277 –87 . 10.1016/j.bpa.2012.07.003 23040281 17 Gosselt AN , Slooter AJ , Boere PR Risk factors for delirium after on-pump cardiac surgery: a systematic review . Crit Care 2015 ;19 :346 10.1186/s13054-015-1060-0 26395253 18 Hollinger A , Siegemund M , Goettel N Postoperative delirium in cardiac surgery: an unavoidable menace? J Cardiothor Vasc An 2015 ;29 :1677 –87 . 10.1053/j.jvca.2014.08.021 19 Inouye SK , van Dyck CH , Alessi CA Clarifying confusion: the confusion assessment method: a new method for detection of delirium . Ann Intern Med 1990 ;113 :941 –8 . 10.7326/0003-4819-113-12-941 2240918 20 Trzepacz PT , Baker RW , Greenhouse J A symptom rating scale for delirium . Psychiat Res 1998 ;23 :89 –97 . 10.1016/0165-1781(88)90037-6 21 Albert MS , Levkoff SE , Reilly C The delirium symptom interview: An interview for the detection of delirium symptoms in hospitalized patients . Geriatr Psychol Neu 1992 ;5 :14 –21 . 10.1177/002383099200500103 22 Bergeron N , Dubois MJ , Dumont M Intensive care delirium screening checklist: evaluation of a new screening tool . Intens Care Med 2001 ;27 :859 –64 . 10.1007/s001340100909 23 Folstein MF , Folstein SE , McHugh PR ‘Mini mental state’. A practical method for grading the cognitive state of patients for the clinician . J Psychiatr Res 1975 ;12 :189 –98 .1202204 24 Kokmen E , Naessens JM , Offord KP A short test of mental status: Description and preliminary results . Mayo Clin Proc 1987 ;62 :281 –8 . 10.1016/S0025-6196(12)61905-3 3561043 25 Smith MJ , Breitbart WS , Platt MM A critique of instruments and methods to detect, diagnose, and rate delirium . J Pain Symptom Manag 1995 ;10 :35 –77 . 10.1016/0885-3924(94)00066-T 26 Hoy D , Brooks P , Woolf A Assessing risk of bias in prevalence studies: modification of an existing tool and evidence of interrater agreement . J Clin Epidemiol 2012 ;65 :934 –9 . 10.1016/j.jclinepi.2011.11.014 22742910 27 Stroup DF , Berlin JA , Morton SC Meta-analysis of observational studies in epidemiology: a proposal for reporting. JAMA 2000 ;283 :2008 –12 .10789670 28 Barendregt JJ , Doi SA , Lee YY Meta-analysis of prevalence . J Epidemiol Community Health 2013 ;67 : 974 –8 . 10.1136/jech-2013-203104 23963506 29 Wong CL , Holroyd-Leduc J , Simel DL Does this patient have delirium? Value of bedside instruments . JAMA 2010 ;304 :779 –86 . 10.1001/jama.2010.1182 20716741 30 Laurila JV , Pitkala KH , Strandberg TE The impact of different diagnostic criteria on prevalence rates for delirium . Dement Geriatr Cogn 2003 ;16 :156 –62 . 10.1159/000071004 31 Moher D , Shamseer L , Clarke M Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement . Syst Rev 2015 ;4 :1 10.1186/2046-4053-4-1 25554246

PMC005xxxxxx/PMC5372074.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01148210.1136/bmjopen-2016-011482Renal MedicineResearch15061728169217231728Trajectories of CKD-MBD biochemical parameters over a 2-year period following diagnosis of secondary hyperparathyroidism: a pharmacoepidemiological study Filipozzi Pierre 12Ayav Carole 23Ngueyon Sime Willy 3Laurain Emmanuelle 12Kessler Michèle 12Brunaud Laurent 4Frimat Luc 125 1 Department of Nephrology, University Hospital, Vandœuvre-lès-Nancy, France 2 Néphrolor Network of Care, Vandœuvre-lès-Nancy, France 3 Pôle S2R, Epidemiology and Clinical Evaluation, University Hospital, Vandœuvre-lès-Nancy, France 4 Department of Hepatobiliary, Digestive and Endocrine Surgery, University Hospital, Vandœuvre-lès-Nancy, France 5 Lorraine University, Paris Descartes University, Nancy, FranceCorrespondence to Pr Luc Frimat; l.frimat@chu-nancy.fr2017 27 3 2017 7 3 e01148211 2 2016 20 6 2016 31 8 2016 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives To define groups of patients according to the changes of biochemical parameters, that is, serum calcium, phosphate and parathyroid hormone (PTH), over a 2-year follow-up period using group-based multi-trajectory modeling (GBMM) among a cohort of dialysis patients with newly diagnosed secondary hyperparathyroidism (SHPT) (ie, PTH≥500 ng/L for the first time) and to compare their patient characteristics and treatments. Design Pharmacoepidemiological study. Setting In the 12 dialysis units located in the French region of Lorraine. Participants A total of 269 dialysis patients with newly diagnosed SHPT were prospectively included from December 2009 to May 2012 and followed-up for 2 years. Results We identified four distinct trajectory groups: ‘rapid PTH drop’ experiencing a rapid and sharp decrease (over weeks) in PTH level associated with decreasing phosphate level within normal range (n=34; 12.7%), ‘gradual PTH decrease’ experiencing a gradual and continuous decrease (over months) in PTH level and maintaining phosphate at a middle level throughout the study (n=98; 36.4%), ‘slow PTH decrease with high phosphate’ experiencing a slow decrease in PTH level associated with a relatively high phosphate level (n=105; 39.0%) and ‘uncontrolled SHPT’ with high levels of PTH and phosphate throughout the study (n=32; 11.9%). Patients in the ‘uncontrolled SHPT’ group were significantly (p<0.00001) younger than patients in other groups. Kidney Disease Improving Global Outcomes (KDIGO) targets for PTH, phosphate and calcium were reached simultaneously for 14.9% of patients at baseline and 16.7% at the end of the study. Patients were given cinacalcet more frequently at months 3 and 6 in the ‘rapid PTH drop’ and at month 24 in the ‘uncontrolled SHPT’ groups. Conclusions Over 2 years following a new SHPT diagnosis, a younger age and a higher rate of alkaline phosphatase were associated to a continuous uncontrolled SHPT. Patients with the lowest PTH at the end of the follow-up tended to receive more often cinacalcet. Trial registration number ClinicalTrials.gov number, NCT02888639, post results. ==== Body Strengths and limitations of this study The first study to apply group-based multi-trajectory modeling (GBMM) in chronic kidney mineral and bone disorder (CKD-MBD). The EPHEYL study included all dialysis patients reporting newly diagnosed secondary hyperparathyroidism (SHPT) in a French administrative country, whereas most studies on biochemical assessment in SHPT patients included prevalent SHPT patients for whom data on disease duration were unavailable. The small sample size was counterbalanced by the rigorous methodology. Background and objectives Clinical practice guidelines on the management of chronic kidney mineral and bone disorder (CKD-MBD) recommend target levels for parathyroid hormone (PTH) in the range of two to nine times the upper normal limit and treatment modifications on the basis of trends rather than on a single value.1 Until now studies on CKD-MBD management assessed secondary hyperparathyroidism (SHPT) evolution using a single measurement value for a single biochemical parameter, mostly serum PTH level which is highly variable. Group-based multi-trajectory modeling (GBMM), in contrast, is a statistical approach recently used in clinical research for identifying distinct patterns of change within a population and estimating the likelihood for each individual of being assigned to a trajectory group.2 3 A full and accurate description of illness trajectories may help clinicians plan and deliver appropriate care. To the best of our knowledge, trajectories of biochemical parameters of CKD-MBD in patients with newly diagnosed SHPT were never described with this statistical approach. This study aimed to define groups of patients according to the changes of biochemical parameters, that is, serum calcium, phosphate and PTH over a 2-year follow-up period using GBMM among a cohort of dialysis patients with newly diagnosed SHPT, and to compare their patient characteristics and treatments. Methods Settings The pharmacoepidemiological EPHEYL (Étude PHarmacoÉpidémiologique de l'hYperparathyroïdie secondaire en Lorraine) study is a 2-year, open-cohort, prospective, observational study on incident SHPT, that is, newly diagnosed, with a 2-year follow-up, set in the 12 dialysis units located in the French region of Lorraine (public or private). Participants From 1 December 2009 to 31 May 2012, all patients who were on dialysis (haemodialysis or peritoneal dialysis) for at least 3 months were identified through the REIN registry—Region of Lorraine.4 Among them, those who experienced PTH ≥500 ng/L for the first time were included in our study. The PTH cut-off value of 500 ng/L was chosen at the time of 2003 K-DOQI.5 Indeed when we initiated the study, the updated Kidney Disease Improving Global Outcomes (KDIGO) recommendations were not effective and the advocated PTH target range was 150–300 ng/L in CKD stage 5 patients.5 All patients included in our study were prospectively followed-up for 2 years. Data collection Data collection was described in-depth elsewhere.6 7 In brief, all sociodemographic and clinical data (biochemical parameters, outcomes, treatments) were retrieved from the REIN registry and prospectively collected from medical reports.4 A standardised form was used to collect data from medical records. A steering committee composed of an epidemiologist (CA) and a nephrologist (LF) reviewed all forms and medical records when collected biochemical data were out of international standards. The following biochemical parameters were collected at the time of inclusion and during the study: PTH, calcaemia, phosphate, vitamin D, alkaline phosphatase, albumin, haemoglobin and measured ionised calcium. All drugs acting on phosphocalcic metabolism were collected during the study and classified into five groups: native vitamin D, vitamin D receptor activators, calcium-containing phosphate binder (CCPB), calcium-free phosphate binder (CFPB) and cinacalcet. A Charlson index was also calculated for each patient.8 Statistical analysis Group-based multi-trajectory modeling GBMM was used to identify distinct trajectories of phosphataemia, calcaemia and PTH.9 10 This statistical method is based on identifying groups of individuals following markedly distinct trajectories within a population and estimating probability for an individual to be assigned to a trajectory group.9 GBMM is a variant of group-based trajectory modeling devised to model trajectories for several parameters and to assess inter-relationship of trajectories for these parameters without significantly increasing the total number of trajectory groups.11 12 Serum phosphate and calcium levels as well as PTH/assay upper limit ratio were modeled using the SAS (SAS Institute, http://www.sas.com) procedure, PROC TRAJ.9 11 This procedure is an extension of the statistical software SAS which fits a semiparametric mixture model to longitudinal data using the maximum likelihood method. The total number of measurements used was up to nine for each parameter (at inclusion and every 3 months over 2 years). The median was used when different values for a biochemical parameter had been collected for a patient in the same time interval. The Bayesian information criterion (BIC) was used to compare zero-inflated Poisson models with up to six trajectory groups. For each number of trajectories, we compared models defined by intercept with linear, quadratic or cubic models using the following formula: 2(ΔBIC)>2.11 Each patient was assigned to a trajectory group for which the estimated probability of assignment was greatest. An assignment probability of 0.7 was considered a poor fit, while a value of 0.9 or higher was considered an excellent fit. Comparison of characteristics of the trajectory groups Qualitative variables were described by frequency and percentage and quantitative variables by mean and SD if the normality hypotheses (Shapiro–Wilks test) were fulfilled. Comparative bivariate analysis of all patient characteristics was performed using χ2 test or Fischer's exact test for qualitative variables, and analysis of variance or Kruskal-Wallis non-parametric test for Gaussian quantitative variables. Multivariate analyses of variance (MANOVA) were performed to compare the biochemical parameters of trajectory groups not used in the GBMM approach. When a difference was found, bivariate comparisons were performed using Bonferroni test or χ2 test by adjusting the α value using the Bonferroni correction. A comparative survival analysis between trajectory groups was also performed using the log-rank test. Model robustness We carried out a random sample among two-thirds of the patients of our population (N=179) to check the robustness of the model and account for uncertainties associated to the distribution of trajectory groups. These randomly selected patients were assigned to a trajectory group by iterating on the GBMM model. Then model calibration was performed using χ2 test and the Hosmer-Lemeshow test. Distribution of trajectory groups and assignment probabilities were assessed as recommended by Nagin and Odgers.10 Sensitivity analysis An alternative model by iterating statistical analyses was devised by excluding patients for whom less than four biochemical values were collected at four time periods. We considered that these patients were at increased risk of being less likely to be assigned to a group. We calculated the Cohen's kappa coefficient by excluding patients with less than four biochemical values or not in order to assess if distribution of trajectory groups was consistent. Missing data Latent trajectory model may be used if any biochemical parameter is missing. When the first day of a treatment was missing data and this treatment had not been actually reported in the quarterly report, we considered that this treatment began at mid-quarter, that is, 45 days before first reporting. Likewise, when the last day of a treatment was missing data, its discontinuation was considered to happen 45 days after the last reporting. Statistical analyses were performed using SAS software, V.9.3 (SAS Institute, Cary, North Carolina, USA). A significance threshold of 0.05 was adopted for all analyses. Results Among the 305 patients enrolled in the EPHEYL cohort, 269 patients with a PTH level ≥500 ng/L were included in this study. The mean duration for follow-up was 620.0±195.3 days. A total of 83 patients (30.9%) left the study before the end with a length of follow-up of 371.2±184.5 days; among them, 54 died, 26 underwent kidney transplantation, 2 moved outside the study area and 1 had recovered from renal function and stopped dialysis treatment. Several measurements for serum levels of PTH, phosphate and calcium were reported per patient: 6.9±2.2 PTH values/patient (median of 8; 25th–75th centile: 6–9), 7.8±2.2 phosphate values/patient (median of 9; 25th–75th centile: 7–9) and 7.8±2.0 calcium values/patient (median of 9; 25th–75th centile). Characteristics of four distinct trajectory groups The GBMM analysis using the zero-inflated Poisson modeling option allowed the identification of four distinct trajectories over a 2-year period after SHPT diagnosis (figure 1). The first group, identified as the ‘rapid PTH drop’ group, included 12.7% of patients (N=34) experiencing a rapid and sharp decrease in serum PTH level associated with decreasing serum phosphate level within normal range. Serum calcium levels were lower at the start of the study in this group when compared to other groups. The second group, identified as the ‘gradual PTH decrease’ group, included 36.4% of patients (n=98) experiencing a gradual and continuous decrease in serum PTH level and maintaining serum phosphate at a middle level throughout the study. The third group, identified as the ‘slow PTH decrease with high phosphate’ group, included 39.0% of patients (N=105) experiencing a slow decrease in serum PTH level associated with a relatively high serum phosphate level. The fourth group, identified as the ‘uncontrolled SHPT’ group, included 11.9% of patients (N=32) with high serum levels of PTH and phosphate throughout the study. Patients of this group had a significantly higher PTH/assay upper limit ratio at inclusion in the study (see online supplementary table S1). Figure 1 Trajectories for parathyroid hormone, calcium and phosphate over the two years after diagnosis of a severe secondary hyperparathyroidism. The solid lines indicate the observed trajectories, and the dashed lines indicate the predicted trajectories with 95% CIs. Blue, ‘rapid PTH drop’ group; red: ‘gradual PTH decrease’ group; green, ‘slow PTH decrease with high phosphate’ group; black, ‘uncontrolled SHPT’ group. PTH, parathyroid hormone; SHPT, secondary hyperparathyroidism. 10.1136/bmjopen-2016-011482.supp1supplementary table Biochemical parameters used for determination of trajectory groups Comparison of other biochemical values collected during the 24-month follow-up between groups is presented in table 1. The group ‘uncontrolled SHPT’ had a significantly (p<0.0001) higher rate of alkaline phosphatase when compared with other groups using MANOVA. Table 2 highlights descriptive statistics of patient characteristics at inclusion by the trajectory group. There were no difference of survival between groups (p=0.24). The three KDIGO targets were simultaneously reached for calcium, phosphate and PTH by 14.9% (40/269) and 16.7% (32/192) of patients at inclusion and at 24 months, respectively (see online supplementary table 2). There was no significant difference between groups (see online supplementary table 2). In the statistical analyses using the χ2 or Fischer's exact test, the ‘uncontrolled SHPT’ group was less likely to reach KDIGO targets for PTH at inclusion and at 24 months than the ‘gradual PTH decrease’ group (p=0.003 at inclusion; p=0.0002 at 24 months) and the ‘slow PTH decrease with high phosphate’ group (p<0.0001 at inclusion; p<0.0001 at 24 months). 10.1136/bmjopen-2016-011482.supp2supplementary table Distribution of patients reaching the KDIGO targets for CKD-MBD according to trajectory groups Description of clinical practice The average weekly duration of dialysis treatment at the inclusion was 12.0±2.0 hours without any significant difference between the trajectory patterns. A total of 157 of patients (58.4%) received at least one prescription of cinacalcet throughout the follow-up. At the end of the follow-up, 38 patients of these 157 (24.2%) versus 9 of the 108 patients who never received cinacalcet had a PTH value under two times the upper normal limit of the assay (p<0.001). Vitamin D analogues, native vitamin D, CFPB and CCPB were prescribed at least once throughout the follow-up in 85 (31.6%), 210 (78.4%), 178 (66.2%) and 213 (79.2%) patients, respectively. The total length of the prescription period for each therapeutic class was not statistically different between trajectory groups (table 3). Table 1 Biochemical parameters during the follow-up according to trajectory groups ‘Rapid PTH drop’ group N=34 (12.7%) “Gradual PTH decrease” group N=98 (36.4%) ‘Slow PTH decrease with high phosphate’ group N=105 (39.0%) ‘Uncontrolled SHPT’ group N=32 (11.9%)   N Mean SD N Mean SD N Mean SD N Mean SD p Value* Serum 25-OH vitamin D level (ng/mL)  At baseline 31 17.2 11.8 67 18.7 12.2 68 20.1 16.2 22 22.5 13.2 0.53  At 3 months 18 27.2 17.2 30 20.4 12.0 36 23.4 15.1 14 24.5 13.4 0.45  At 6 months 15 31.3 21.0 37 21.9 11.4 38 22.1 15.0 9 27.6 15.5 0.15  At 9 months 15 30.0 14.4 34 22.7 15.2 33 24.5 15.3 13 21.4 8.1 0.36  At 12 months 15 30.5 14.4 37 26.8 14.5 37 27.7 34.6 15 23.9 13.8 0.89  At 15 months 15 23.0 11.6 34 24.4 13.5 26 28.1 15.9 10 17.7 8.7 0.22  At 18 months 15 30.5 16.7 26 27.0 18.2 36 28.3 14.2 11 26.3 16.0 0.90  At 21 months 11 32.8 11.1 33 27.1 16.6 29 28.2 14.8 15 21.1 10.6 0.23  At 24 months 10 36.9 17.1 31 26.5 18.0 24 25.8 13.9 13 24.1 12.0 0.22 Serum alkaline phosphatase level (UI/L)  At baseline 34 94.5 36.1 96 102.0 46.1 104 98.3 44.5 31 162.1 199.2 <0.001  At 3 months 31 101.1 48.6 85 98.0 41.7 96 101.3 48.0 28 176.8 185.4 <0.0001  At 6 months 30 89.1 35.4 82 97.6 36.0 97 103.1 48.8 26 184.5 149.0 <0.0001  At 9 months 29 79.4 32.0 76 97.9 41.0 94 100.4 40.3 25 200.2 169.1 <0.0001  At 12 months 27 78.1 27.6 76 96.9 50.1 86 102.9 43.1 29 168.3   <0.0001  At 15 months 25 85.2 39.9 65 101.7 74.0 79 107.7 58.6 29 165.7 134.3 0.0001  At 18 months 24 93.3 44.3 67 96.2 38.9 69 103.5 53.9 25 158.3 117.7 <0.001  At 21 months 21 82.0 33.8 62 94.9 36.4 68 103.3 48.2 25 151.1 81.3 <0.0001  At 24 months 20 86.8 47.1 60 92.7 35.0 58 105.9 49.7 22 184.5 182.0 <0.0001 Serum albumin level (g/L)  At baseline 33 35.4 5.6 92 36.4 4.3 102 35.8 4.5 31 37.4 4.4 0.28  At 3 months 28 35.4 3.7 80 35.5 5.0 88 35.9 4.8 27 37.1 4.9 0.47  At 6 months 31 35.8 5.0 74 35.6 5.3 88 36.9 4.8 25 37.2 5.4 0.28  At 9 months 26 34.5 5.3 70 35.2 5.0 85 36.5 4.7 29 38.2 4.7 0.01  At 12 months 26 33.8 5.5 68 34.9 5.4 77 35.4 5.0 25 37.4 5.4 0.10  At 15 months 23 33.4 6.5 61 35.0 6.0 73 35.6 4.8 27 37.6 5.5 0.06  At 18 months 24 32.9 6.9 61 35.4 5.7 70 35.8 5.7 25 36.5 5.5 0.14  At 21 months 19 33.5 6.0 59 35.6 4.7 70 36.1 4.7 27 37.6 5.0 0.04  At 24 months 17 35.1 3.6 54 35.6 5.1 65 36.2 5.4 21 38.1 3.3 0.18 Serum bicarbonate level (mmol/L)  At baseline 34 22.4 3.3 98 22.2 3.6 103 22.1 3.3 32 22.1 4.3 0.97  At 3 months 33 23.3 3.2 91 23.2 3.4 100 22.4 3.4 29 21.1 3.8 0.02  At 6 months 32 23.1 2.6 87 23.4 3.7 97 22.6 3.9 30 22.1 2.7 0.26  At 9 months 29 23.9 3.1 84 23.3 3.4 96 22.8 3.5 27 22.8 2.6 0.36  At 12 months 27 23.7 3.1 80 22.8 3.1 88 23.1 4.0 25 22.0 3.2 0.31  At 15 months 26 23.7 3.4 72 23.4 3.8 79 22.9 3.6 27 22.8 3.2 0.66  At 18 months 26 24.2 3.1 72 23.3 3.7 80 23.1 4.0 27 22.0 3.1 0.16  At 21 months 23 24.1 3.2 68 23.9 3.6 76 23.2 3.2 27 22.2 3.2 0.10  At 24 months 21 24.1 2.6 63 24.3 3.8 73 23.5 3.5 26 22.6 3.0 0.15 Bold indicates statistical significance, p<0.05. *Analysis of variance or Wilcoxon's test used to compare continuous variables. PTH, parathyroid hormone; SHPT, secondary hyperparathyroidism. Table 2 Baseline characteristics of patients according to trajectory groups   ‘Rapid PTH drop’ group N=34 (12.7%) ‘Gradual PTH decrease’ group N=98 (36.4%) ‘Slow PTH decrease with high phosphate’ group N=105 (39.0%) ‘Uncontrolled SHPT’ group N=32 (11.9%) N %/ Mean SD N %/ Mean SD N %/ Mean SD N %/ Mean SD p Value* Sociodemographic data Age at baseline 34 68.8 14.6 98 68.5 13.6 105 67.2 14.9 32 54.1 19.4 <0.0001 Gender male 22 64.7 59 60.2 56 53.3 20 62.5 0.57 BMI 34 28.3 6.9 98 28.0 7.2 105 27.4 6.3 32 26.5 6.3 0.66 BMI>30 kg/m2 14 41.2 33 33.7 33 31.4 6 18.8 0.26 Former or active tobacco consumption 8 23.5 26 26.5 33 31.4 10 31.3 0.76 Professional status 0.004  Active 2 6.5 10 11.8 16 15.8 6 20.7  Retired 23 74.2 61 71.8 66 65.3 9 31.0  Other inactive 6 19.4 14 16.5 19 18.8 14 48.3 Medical data Comorbidities potentially related to SHPT 18 52.9 59 60.2 59 56.2 12 37.5 0.16  History of fracture 5 14.7 22 22.4 24 22.9 2 6.3 0.14 Other comorbidities  Diabetes 6 17.6 20 20.4 31 29.5 9 28.1 0.33  High blood pressure 15 44.1 40 40.8 50 47.6 19 59.4 0.32  Existence of a disability 5 14.7 11 11.2 14 13.3 3 9.4 0.89  Cardiovascular disease(s) 18 52.9 53 54.1 52 49.5 13 40.6 0.60 Charlson comorbidity index 34 5.7 2.2 98 5.9 2.1 105 5.6 1.9 32 4.4 2.3 0.005 History of kidney transplant 2 5.9 6 6.1 9 8.6 3 9.4 0.88 Dialysis data 0.52 Technique of dialysis  Haemodialysis or hemodiafiltration 33 97.1 90 91.8 93 88.6 29 90.6  Peritoneal dialysis 1 2.9 8 8.2 12 11.4 3 9.4 Dialysis vintage (months) 34 41.2 73.2 98 46.3 67.8 105 39.7 68.0 32 50.8 67.5 0.45 Dialysate calcium concentration 0.77  1.25 mmol/L 3 8.8 7 7.7 14 13.6 2 6.5  1.50 mmol/L 25 73.5 69 75.8 71 68.9 24 77.4  1.60 mmol/L 1 2.9 1 1.1 0 0.0 0 0.0  1.75 mmol/L 5 14.7 14 15.4 18 17.5 5 16.1 Use of ESA 27 81.8 74 75.5 73 69.5 20 64.5 0.34 Bold indicates statistical significance, p<0.05 *χ2 test or Fisher's exact test used to compare categorical variables. Analysis of variance or Wilcoxon's test used to compare continuous variables. BMI, body mass index; ESA, erythropoiesis stimulating agent; PTH, parathyroid hormone; SHPT, secondary hyperparathyroidism. Table 3 Total length of treatments for SHPT according to trajectory groups ‘Rapid PTH drop’ group N=34 (12.7%) ‘Gradual PTH drop’ group N=98 (36.4%) ‘Slow PTH decrease with high phosphate’ group N=105 (39.0%) ‘Uncontrolled SHPT’ group N=32 (11.9%) Length of treatments for SHPT N Mean (SD) N Mean (SD) N Mean (SD) N Mean (SD) p Value* Mean duration on the whole population (day/patient)  Vitamin D receptor activators 34 96.9 (198.6) 98 116.3 (223.7) 105 123.3 (213.1) 32 143.0 (226.0) 0.85  Native vitamin D supplementation 34 323.0 (283.4) 98 272.4 (253.1) 105 311.8 (252.3) 32 344.0 (264.5) 0.47  Cinacalcet 34 268.2 (256.0) 98 219.8 (257.0) 105 198.5 (241.5) 32 276.4 (246.3) 0.31  Calcium-free phosphate binders 34 249.4 (280.9) 98 301.3 (293.7) 105 317.9 (289.2) 32 410.5 (295.8) 0.15  Calcium-containing phosphate binders 34 468.1 (242.6) 98 343.0 (277.8) 105 371.4 (288.1) 32 446.8 (291.9) 0.07 Mean duration in patients receiving the treatment (day/patient)  Vitamin D receptor activators 9 365.9 (227.8) 28 406.9 (239.0) 38 340.6 (227.3) 14 326.8 (239.1) 0.64  Native vitamin D supplementation 26 422.3 (249.7) 76 351.2 (234.1) 85 385.2 (224.1) 27 407.7 (237.6) 0.49  Cinacalcet 22 414.4 (198.7) 58 371.3 (234.7) 63 330.9 (230.9) 27 327.6 (234.4) 0.42  Calcium-free phosphate binders 19 446.3 (227.8) 63 468.7 (235.0) 73 457.2 (237.1) 25 525.4 (223.9) 0.61  Calcium-containing phosphate binders 31 513.4 (201.8) 73 460.4 (221.6) 85 458.8 (249.3) 26 549.8 (216.2) 0.23 PTH, parathyroid hormone; SHPT, secondary hyperparathyroidism. A total of 157 patients (58.4%) received at least one dose of cinacalcet. Among them, 59 patients (21.9%) had to stop taking cinacalcet treatment at least once. Regarding therapy with cinacalcet, the ‘rapid PTH drop’ group was most often treated at the inclusion and at 3 and 6 months whereas the ‘uncontrolled SHPT’ group was most often treated at 24 months (figure 2). Figure 2 Cinacalcet prescription according to trajectory groups. Blue, ‘rapid PTH drop’ group; red: ‘gradual PTH decrease’ group; green, ‘slow PTH decrease with high phosphate’ group; black, ‘uncontrolled SHPT’ group. PTH, parathyroid hormone; SHPT, secondary hyperparathyroidism. Biological hypocalcaemia developed in 79.6% of cinacalcet-treated patients and nausea occurred in 12.1% of them. All other adverse events occurred in <1.0% of them. Regarding adverse events of cinacalcet, no significant difference was found between trajectory groups. A total of eight patients (3.0%) underwent a parathyroidectomy: four patients assigned in the ‘uncontrolled SHPT’ group after 464.5±175.3 days of follow-up; two patients in the ‘slow PTH decrease with high phosphate’ group after a 368-day and 447-day follow-up (407.5±55.86 days); one patient in the ‘gradual PTH decrease’ group after a 181-day follow-up and one patient in the ‘rapid PTH drop’ after a 202-day follow-up. Validity of data Probability assignments were 0.89±0.17 for the ‘rapid PTH drop’ group, 0.84±0.18 for the ‘gradual PTH decrease’ group, 0.84±0.19 for the ‘slow PTH decrease with high phosphate’ group and 0.99±0.23 for the ‘uncontrolled SHPT’ group. The percentage of patients with a probability assignment to a group under 0.7 was 17.6% for the ‘rapid PTH drop’ group, 22.4% for the ‘gradual PTH decrease’ group, 27.6% for the ‘slow PTH decrease with high phosphate’ group and 6.3% for the ‘uncontrolled SHPT’ group. Model calibration was satisfactory for all trajectory groups as the Hosmer-Lemeshow test was not significant. In figure 1, there were no significant differences between observed values and those predicted by the statistical model irrespective of the length of follow-up and of the groups, except for PTH level in ‘rapid PTH drop’ and ‘gradual PTH decrease’ groups at 3 months. After excluding 18 patients with less than four biochemical measurements throughout the follow-up, results remained significantly unchanged in the GBMM analysis (see online supplementary table S3). The distribution of patients into groups was preserved and the shape of trajectories was similar: 12.7% of patients were assigned in the ‘rapid PTH drop’ group, 35.5% in the ‘gradual PTH decrease’ group, 39.4% in the ‘slow PTH decrease with high phosphate’ group and 12.4% in the ‘uncontrolled SHPT’ group. The concordance between groups before and after excluding those patients was with a kappa coefficient of 0.99. 10.1136/bmjopen-2016-011482.supp3supplementary table Table of compliance between groups defined by GBTM according to the exclusion or not of patients with less than four available biological measurements Discussion In the EPHEYL study, we previously demonstrated: first, the benefit–risk balance of cinacalcet was presumably unfavourable in patients with a low PTH level;6 second, cinacalcet prescription was not associated with a better evolution of health-related quality of life after 1 year of treatment;7 third, cinacalcet was used in two-thirds of the patients, a weak proportion of them reaching KDIGO targets for PTH levels when compared to patients not treated with calcimimetics.13 Here we identify four distinct trajectories that mainly differed on the changes in serum PTH and phosphate levels over the first 2 years after diagnosis for SHPT in dialysis patients. These distinct progressive profiles actually reflect routine clinical situations.14 Patients in the ‘rapid PTH drop’ group experienced a drop of serum PTH level at the beginning of the study, that is, shortly after SHPT diagnosis, and were more likely to be overtreated because the PTH/assay upper limit ratio was close to two from 9-month treatment to the end of the follow-up. Patients in the ‘gradual PTH decrease’ group and those in the ‘slow PTH decrease with high phosphate’ group experienced a more moderate decrease of PTH. Phosphataemia of the latter group, however, was more poorly controlled. Finally, the patients in the ‘uncontrolled SHPT’ group did not experience major changes in serum PTH and phosphate levels; such a clinical condition could account for a resistance pattern. In addition, the significant high level of serum alkaline phosphatase throughout the follow-up in the ‘uncontrolled SHPT’ group highlights a more severe SHPT.15 Calcaemia was not found as a discriminatory variable in GBMM. This variable seemed less likely to vary widely within our population, whatever the group, as corroborated by a low SD for calcaemia in our follow-up.16 This could be explained by the fact that calcium level is a closely regulated physiological variable.17 Regarding baseline characteristics, differences between the ‘uncontrolled SHPT’ group and the other groups were observed and mainly concerned age, Charlson comorbidity index, professional status and current structure for dialysis unit. Age difference is likely to explain all differences regarding professional status and Charlson comorbidity index. Our results are consistent with those recently published by Parfrey et al.18 They showed that the more pronounced effect of cinacalcet in older patients than in younger patients may be explained by the association of bone demineralisation and vascular calcification with older age. In addition, serum alkaline phosphatase levels suggest that few patients immediately treated with cinacalcet had a high bone turnover. The difference for survival did not differ between groups, but this study did not aim to compare survival and was underpowered to assess any difference. Some explanations may be suggested for the reasons underlying heterogeneous biological progression among these patients. The association between serum levels of PTH and age is well-known and has been previously described in the literature.19 Nutritional status is also a well-known risk factor for low PTH.20 In our study, albumin levels did not differ in MANOVA but there was a trend for high rates in the ‘uncontrolled SHPT’ group with statistically significant differences at 9 and 21 months of follow-up. Even if there was no statistically significant difference for total lengths of drug prescriptions, there were differences between groups for times of prescription. Patients of the ‘rapid PTH drop’ group were more likely to receive CCPB and cinacalcet at the beginning of the study and their PTH dropped simultaneously to a level close to two times the upper normal limit of the assay as recommended by the KDIGO guidelines.5 These data suggest that patients from this group were overtreated.20 Patient profiles or trajectories observed in real-life are different from those identified in the first efficacy studies of cinacalcet, reflecting heterogeneity of patients in real-life whereas profiles of patients included in clinical trials are more homogenous due to a careful selection. In the EVOLVE trial, patients were rather young with a median age of 55 years, whereas patients in our study had a mean age of 66.3 years and a median age of 69 years, which was closer to the median age of the population in dialysis.21 The median serum calcium level was 2.175 mmol/L in our study, whereas it was 2.45 mmol/L in the EVOLVE trial. In addition, the EVOLVE population had a selected population profile excluding all patients with hypocalcaemia (serum calcium<8.4 mg/dL). Thus, pharmacoepidemiology studies are complementary to randomised controlled trials and provide a broad spectrum of information from real-life data. Clinicians adhere to the K/DOQI guidelines, and their main objective of prescribing is to reach advocated targets. In the EPHEYL study, frequencies of reaching simultaneously the three biochemical targets as recommended by the 2003 KDIGO guidelines were found to be low throughout the follow-up of CKD-MBD patients. Our results supported previous findings regarding the difficulty of reaching targets, particularly for phosphate.22 The proportion of deaths that occurred among the EPHEYL cohort by the end of the follow-up was consistent with the mortality rate of ∼120 deaths for 1000 people/year obtained for dialysis patients in the age group (60–70) through the national REIN registry in 2012.23 In addition, our study showed 3.0% of patients undergoing parathyroidectomy, a low percentage compared to 6.6% and 7.1% of patients found in the COSMOS and Photo-Graph cohorts, respectively.22 24 This difference with previous cohorts should be seen in the light of the proportion of patients who received cinacalcet that accounted for 58.4%. Comments should be pointed out regarding the most frequent use of cinacalcet instead of parathyroidectomy. First, we found a rather high frequency of adverse events in patients receiving cinacalcet. A higher frequency (>90% of patients) was found in the EVOLVE randomised clinical trial, but reported adverse events were mainly related to digestive disorders whereas ours were mainly hypocalcaemia.21 Second, in most studies evaluating the efficacy of cinacalcet, parathyroidectomy was therefore reported as a treatment failure or an unwanted event.4 25 26 No superiority of cinacalcet over parathyroidectomy has not been demonstrated yet, as well as no efficacy of cinacalcet on mortality from cardiovascular disease and all causes.24 25 In a medicoeconomic study, parathyroidectomy was more cost-effective than cinacalcet, except in patients with contraindication for surgery.27 Finally, previous findings demonstrated efficacy of parathyroidectomy on cardiovascular mortality.28 Lowest and highest values of PTH levels are probably associated with adynamic or hyperparathyroid bone disease, respectively. However, the majority of PTH levels will match with a grey zone where biochemical and analytical variability will hamper in-depth evaluation of CKD-MBD consequences.29 The major strength of this investigation lies in the use of GBMM which is a valuable and useful statistical tool for identifying groups of patients and for analysing large amounts of longitudinal data. This method avoids using subjective and objective criteria to identify groups. Another advantage of this method was to combine numerous biochemical measurements in our 2-year study whereas only two measurements at baseline and 12 months of follow-up were assessed in the study conducted by Block et al.30 As highlighted by Stevens et al and advocated in the KDIGO guidelines, a single value for a phosphocalcic parameter should no more be considered, and a tool assessing biological trends in markers of mineral-bone disorders should be preferred.31 32 GBMM meets therefore both objectives. To the best of our knowledge, this is the first study in which GBMM was applied in CKD-MBD. Another strength was the exhaustive and consecutive data collection among all dialysis patients reporting a SHPT after dialysis initiation. The EPHEYL study included therefore incident SHPT patients who were more homogeneous than prevalent patients regarding onset of the disease and history of a shortened treatment regimen. Most studies on biochemical assessment in SHPT patients, in contrast, included prevalent SHPT patients for whom data on disease duration were unavailable.19 26 33 The authors acknowledge several limitations. A number of nausea related to cinacalcet was likely not to be reported due to data collection on medical history and adverse events of cinacalcet from medical records. In addition, 500 ng/L was chosen as the cut-off PTH value before the publication of KDIGO clinical guidelines. It is now recognised that this value should not be interpreted as indicative of a confirmed diagnosis for SHPT.34 In our cohort, patients were not likely to have adynamic bone disease given mean values of PTH levels. In conclusion, our study highlighted four groups of patients with different trajectories for phosphocalcic parameters using a modern and original statistical methodology. There were significant differences in age and alkaline phosphatase between groups. GBMM is a relevant and appropriate methodology for longitudinal data analysis among SHPT patients. Further studies are necessary on larger pools of patients using other biochemical parameters or combining GBMM with propensity score. The authors are extremely grateful to the patients who contributed to this study. They thank all nephrologists and hospital directors who participated in the study, and particularly the study nurse, Catherine Campagnac, and Brigitte Bourdillat for the manuscript editing. Contributors: PF, CA, LB and LF conceived the idea of the study. PF, CA, EL, MK, LB and LF designed the study. CA, LB and LF were responsible for data collection. CA and LF reviewed all forms and medical records when collected biochemical data were out of international standards. PF, CA, WNS and LF were responsible for undertaking data analysis and contributed to the interpretation of the results. PF produced the tables and graphs. The first draft of the manuscript was prepared by PF and was then circulated repeatedly among all authors for critical revision. PF, CA, WNS, EL, MK, LB and LF gave their final approval regarding submission for publication. PF, CA, WNS, EL, MK, LB and LF read and approved the final manuscript. Funding: This work was supported by a public funding through a Hospital Clinical Research Program (‘PHRC inter-regional 2009’) of the French Health Ministry and ‘Agence de la Biomedicine’ (2008). Competing interests: None declared. Ethics approval: This study was conducted in compliance with French regulations concerning pharmacoepidemiological studies.35 Approvals from the French data protection agency (CNIL: no. 904163) and from the Advisory Committee on information processing research in the field of health located in the region of Lorraine (CCTIRS: no. 0428) were obtained through the national REIN registry. An information sheet was displayed in all dialysis units, and each patient was given an individual written information sheet at the initiation of dialysis. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Kidney Disease: Improving Global Outcomes (KDIGO) CKD-MBD Work Group . KDIGO clinical practice guideline for the diagnosis, evaluation, prevention, and treatment of Chronic Kidney Disease-Mineral and Bone Disorder (CKD-MBD) . Kidney Int 2009 ;113 :S1 –30 . 2 O'Hare AM , Batten A , Burrows NR Trajectories of kidney function decline in the 2 years before initiation of long-term dialysis . Am J Kidney Dis 2012 ;59 :513 –22 .22305760 3 Allen NB , Siddique J , Wilkins JT Blood pressure trajectories in early adulthood and subclinical atherosclerosis in middle age . JAMA 2014 ;311 :490 –7 . 10.1001/jama.2013.285122 24496536 4 Network and epidemiologic information in nephrology: kidney report 2010[In-ternet]. [quoted 2017 Mar 07]. Available: https://www.agence-biomedecine.fr/IMG/pdf/2012_rapport_annuel_rein.pdf . 5 Moe SM , Chertow GM , Coburn JW Achieving NKF-K/DOQI bone metabolism and disease treatment goals with cinacalcet HCl . Kidney Int 2005 ;67 :760 –71 . 10.1111/j.1523-1755.2005.67139.x 15673327 6 Laurain E , Ayav C , Erpelding ML Targets for parathyroid hormone in secondary hyperparathyroidism: is a “one-size-fits-all” approach appropriate? A prospective incident cohort study . BMC Nephrol 2014 ;15 :132 –43 . 10.1186/1471-2369-15-132 25123022 7 Filipozzi P , Ayav C , Erpelding ML Influence on quality of life from an early cinacalcet prescription for secondary hyperparathyroidism in dialysis . Pharmacoepidemiol Drug Saf 2015 ;24 :187 –96 . 10.1002/pds.3683 25111639 8 Charlson ME , Pompei P , Ales KL A new method of classifying prognostic comorbidity in longitudinal studies: development and validation . J Chronic Dis 1987 ;40 :373 –83 .3558716 9 Jones BL , Nagin DS , Roeder K A SAS procedure based on mixture models for estimating developmental trajectories . Sociol Methods Res 2001 ;29 :374 –93 . 10 Nagin DS , Odgers CL Group-based trajectory modeling in clinical research . Annu Rev Clin Psychol 2010 ;6 :109 –38 . 10.1146/annurev.clinpsy.121208.131413 20192788 11 Jones BL , Nagin DS Advances in group-based trajectory modeling and an SAS procedure for estimating them . Soc Methods Res 2007 ;35 :542 –71 . 12 Nagin DS , Odgers CL Group-based trajectory modeling (nearly) two decades later . J Quant Criminol 2010 ;26 :445 –53 . 10.1007/s10940-010-9113-7 21132047 13 Brunaud L , Ngueyon Sime W , Filipozzi P Minimal impact of calcimimetics on the management of hyperparathyroidism in chronic dialysis . Surgery 2016 ;159 :183 –92 . 10.1016/j.surg.2015.06.058 26476826 14 Jean G , Souberbielle JC , Lorriaux C Clinical and biological forms of secondary hyperparathyroidism in dialysis patients . Nephrol Ther 2012 ;8 :35 –40 . 10.1016/j.nephro.2011.05.002 21741337 15 Ureña P , Hruby M , Ferreira A Plasma total versus bone alkaline phosphatase as markers of bone turnover in hemodialysis patients . J Am Soc Nephrol 1996 ;7 :506 –12 .8704118 16 Levitt H , Smith KG , Rosner MH Variability in calcium, phosphorus, and parathyroid hormone in patients on hemodialysis . Hemodial Int 2009 ;13 :518 –25 . 10.1111/j.1542-4758.2009.00393.x 19758299 17 Courbebaisse M , Souberbielle JC Phosphocalcic metabolism: regulation and explorations . Nephrol Ther 2011 ;7 :118 –38 . 10.1016/j.nephro.2010.12.004 21273150 18 Parfrey PS , Drüeke TB , Block GA Evaluation of Cinacalcet HCl Therapy to Lower Cardiovascular Events (EVOLVE) Trial investigators. The effects of cinacalcet in older and younger patients on hemodialysis: the Evaluation of Cinacalcet HCl Therapy to Lower Cardiovascular Events (EVOLVE) trial . Clin J Am Soc Nephrol 2015 ;10 :791 –9 .25710802 19 Floege J , Kim J , Ireland E Serum iPTH, calcium and phosphate, and the risk of mortality in a European haemodialysis population . Nephrol Dial Transplant 2011 ;26 :1948 –55 . 10.1093/ndt/gfq219 20466670 20 Andress DL Adynamic bone in patients with chronic kidney disease . Kidney Int 2008 ;73 :1345 –54 . 10.1038/ki.2008.60 18337716 21 Chertow GM , Block GA , Correa-Rotter R Effect of cinacalcet on cardiovascular disease in patients undergoing dialysis . N Engl J Med 2012 ;367 :2482 –94 . 10.1056/NEJMoa1205624 23121374 22 Fernández-Martín JL , Carrero JJ , Benedik M COSMOS: the dialysis scenario of CKD-MBD in Europe . Nephrol Dial Transplant 2013 ;28 :1922 –35 .23166310 23 Agence de la Biomédecine . Registre français des traitements de suppléance de l'insuffisance rénale chronique. Réseau, Epidémiologie, Information, Néphrologie. Rapport annuel 2012. http://www.soc-nephrologie.org/PDF/enephro/registres/rapport_2012/rapport_2012.pdf 24 Pelletier S , Roth H , Bouchet JL Mineral and bone disease pattern in elderly haemodialysis patients . Nephrol Dial Transplant 2010 ;25 :3062 –70 . 10.1093/ndt/gfq128 20237058 25 Block GA , Martin KJ , de Francisco AL Cinacalcet for secondary hyperparathyroidism in patients receiving hemodialysis . N Engl J Med 2004 ;350 :1516 –25 . 10.1056/NEJMoa031633 15071126 26 Ureña P , Jacobson SH , Zitt E Cinacalcet and achievement of the NKF/K-DOQI recommended target values for bone and mineral metabolism in real-world clinical practice-the ECHO observational study . Nephrol Dial Transplant 2009 ;24 : 2852 –9 .19369690 27 Komaba H , Moriwaki K , Goto S Cost-effectiveness of cinacalcet hydrochloride for hemodialysis patients with severe secondary hyperparathyroidism in Japan . Am J Kidney Dis 2012 ;60 :262 –71 . 10.1053/j.ajkd.2011.12.034 22445709 28 Goldenstein PT , Elias RM , Pires de Freitas do Carmo L Parathyroidectomy improves survival in patients with severe hyperparathyroidism: a comparative study . PLoS One 2013 ;8 :e68870 10.1371/journal.pone.0068870 23940515 29 Garrett G , Sardiwal S , Lamb EJ PTH—a particularly tricky hormone: why measure it at all in kidney patients? Clin J Am Soc Nephrol 2013 ;8 :299 –12 . 10.2215/CJN.09580911 22403273 30 Block G , Do TP , Collins AJ Co-trending of parathyroid hormone and phosphate in patients receiving hemodialysis . Clin Nephrol 2016 ;85 :142 –51 . 10.5414/CN108629 26784912 31 Stevens LA , Djurdjev O , Cardew S Calcium, phosphate, and parathyroid hormone levels in combination and as a function of dialysis duration predict mortality: evidence for the complexity of the association between mineral metabolism and outcomes . J Am Soc Nephrol 2004 ;15 :770 –9 .14978180 32 Moe SM , Drueke TB Kidney disease: Improving Global Outcomes (KDIGO) CKD-MBD Work Group. KDIGO clinical practice guideline for the diagnosis, evaluation, prevention, and treatment of chronic kidney disease-mineral and bone disorder (CKD-MBD) . Kidney Int 2009 ;76 :S1 –130 . 33 Block GA , Klassen PS , Lazarus JM Mineral metabolism, mortality, and morbidity in maintenance hemodialysis . J Am Soc Nephrol 2004 ;15 :2208 –18 . 10.1097/01.ASN.0000133041.27682.A2 15284307 34 Barreto FC , Barreto DV , Moysés RM K/DOQI-recommended intact PTH levels do not prevent low-turnover bone disease in hemodialysis patients . Kidney Int 2008 ;73 :771 –7 . 10.1038/sj.ki.5002769 18185506 35 Claudot F , Alla F , Fresson J Ethics and observational studies in medical research: various rules in a common framework . Int J Epidemiol 2009 ;38 :1104 –8 . 10.1093/ije/dyp164 19336436

PMC005xxxxxx/PMC5372075.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01373410.1136/bmjopen-2016-013734Cardiovascular MedicineResearch1506168317151704Patient-reported outcomes and associations with pleural effusion in outpatients with heart failure: an observational cohort study Gundersen Guri H 1Norekvål Tone M 23Graven Torbjørn 1Haug Hilde H 1Skjetne Kyrre 1Kleinau Jens O 1Gustad Lise T 145Dalen Håvard 167 1 Department of Medicine, Levanger Hospital, Nord-Trøndelag Hospital Trust, Levanger, Norway 2 Department of Heart Disease, Haukeland University Hospital, Bergen, Norway 3 Department of Clinical Science, Faculty of Medicine and Dentistry, University of Bergen, Bergen, Norway 4 Department of Neuromedicine (INM), Norwegian University of Science and Technology (NTNU), Trondheim, Norway 5 Department of Circulation and Medical Imaging, Faculty of Medicine, Norwegian University of Science and Technology (NTNU), Trondheim, Norway 6 Department of Circulation and Medical Imaging, Faculty of Medicine, K.G. Jebsen Centre of Exercise in Medicine, Norwegian University of Science and Technology (NTNU), Trondheim, Norway 7 Department of Cardiology, St. Olavs University Hospital, Trondheim, NorwayCorrespondence to Dr Guri H. Gundersen; guri@ntebb.no2017 20 3 2017 7 3 e0137344 8 2016 9 12 2016 14 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives We aimed to study whether patient-reported outcomes, measured by quality of life (QoL) and functional class, are sensitive to pleural effusion (PLE) in patients with heart failure (HF), and to study changes in QoL and functional class during follow-up of PLE. Methods A cohort of 62 patients from an outpatient HF clinic was included. The amount of PLE was quantified using a pocket-sized ultrasound imaging device. Self-reports of QoL and functional class were collected using the Minnesota Living with Heart Failure Questionnaire (MLHFQ) and the New York Heart Association (NYHA) functional classification. Results At baseline, 26 (42%) patients had PLE of which 19 (31%) patients had moderate to severe amounts of PLE. Patients with no to mild PLE had a lower MLHFQ score (mean 42, SD 21) compared with patients with a moderate to severe amount of PLE (mean 55, SD 24), p=0.03. For 28 patients (45%) with follow-up data, we observed a linear improvement of the MLHFQ-score (3.2, 95% CI 1.2 to 5.1) with each centimetre reduction of PLE. Correspondingly, patient-reported NYHA-class followed the same pattern as the MLHFQ-score. Conclusions Our study indicates that patient-reported outcome measures as MLHFQ may be sensitive tools to identify patients with HF at highest risk of symptomatic PLE and that treatment targeting reduction of PLE during follow-up is essential to improvement of QoL and functional capacity of outpatients with HF. Trial registration number NCT01794715; Results quality of lifepleural effusionpoint of care ultrasound ==== Body Strengths and limitations of this study Patient-reported outcome measures of quality of life (QoL) and objective measurements of pleural effusion (PLE) by point-of-care ultrasound was followed up repeatedly in a heart failure (HF) outpatient clinic. QoL is increasingly emphasised in treatment strategies and point-of-care ultrasound is recently recommended as a tool to identify PLE in European Society of Cardiology guidelines. The study was conducted at one single HF outpatient clinic and the sample size was small. However, the presented population is previously shown to be comparable with studies from other HF clinics with regard to HF severity, age, sex, New York Heart Association class and comorbidity. Introduction Pleural effusion (PLE) is a common sign associated with congestion and worsening of heart failure (HF)1 and is often followed by dyspnoea.2 However, the degree of dyspnoea may not correlate with the amount of PLE3 4 and screening for PLE is often restricted to patients with HF with clinical signs of PLE. It is well known that patients with HF experience significant reductions in quality of life (QoL).5 6 Thus, initiatives to find treatment strategies that improve QoL have become an important objective in the follow-up of patients with HF.7 In cancer care, initiatives to reduce PLE seem to be an essential palliative step, which has proven to improve QoL and decrease symptom burden without particular side effects.8 9 However, poor self-reported QoL or functional capacity are not conventionally used indicators for screening and treatment of PLE in patients with HF. This may be due to the fact that it is largely unknown if QoL or functional class is associated with the presence of PLE in patients with HF. Further, to the best of our knowledge, no previous studies have investigated whether QoL and functional class improves in patients with HF when PLE is successfully treated. Thus, we aimed to study whether patient-reported outcomes (QoL and functional class) are sensitive to PLE, and to study changes in QoL and functional class during follow-up of PLE. Methods Study population and design This was a prospective follow-up study with inclusion of 62 patients admitted to the outpatient HF clinic at the non-university Levanger Hospital, Norway between 15 April and 21 June 2013.10 11 The sample size was determined in order to validate the feasibility and reliability of point-of-care ultrasound examination.10 11 The HF diagnosis was confirmed according to the European Society of Cardiology (ESC) guidelines by clinical examination,12 medical history and echocardiographic examination by one of four experienced cardiologists.10 11 All patients provided their written informed consent to participate. The exclusion criteria were inability or unwillingness to consent or HF worsening requiring hospital admission at entry. The study observed patient-reported QoL in the follow-up of a clinical trial regarding feasibility and reliability of point-of-care ultrasound in order to determine PLE (ClinicalTrials.gov: ID: NCT01794715). Data regarding feasibility and influence of point-of-care ultrasound have been recently published.10 11 The observational part of the study was conducted according to the Second Declaration of Helsinki. Follow-up The follow-up of patients at the outpatient HF clinic was managed by two nurses specialised in cardiovascular and intensive care, working in close cooperation with four cardiologists. Briefly, the visits included patient education, self-management counselling and optimisation of the treatment for HF according to the 2012 ESC HF guidelines12 in order to reduce the amount of PLE and improve the functional capacity, QoL and prognosis of the patients. None of the patients with HF needed therapeutic drainage of PLE. The follow-up schedule of the outpatient HF clinic was individualised depending on the condition of the patients. The number of visits ranged from one to four visits during the study period. The final follow-up was defined as the visit when the condition of the patient was satisfactory and stable. Twenty-eight (45%) patients had a final follow-up visit in the study period (figure 1). Figure 1 Shows the flow chart of study recruitment and follow-up. The follow-up schedule of the outpatient HF clinic was individualised depending on the condition of the patients. The number of visits ranged from one to four visits during the study period. The final follow-up was defined as the visit when the condition of the patient was satisfactory and stable. Twenty-eight (45%) patients had a final follow-up visit in the study period. HF, heart failure. Measurements Measurement of PLE The nurses received a short, but dedicated training programme in focused point-of-care ultrasound examinations of the pleural cavities. The high feasibility and excellent reliability for quantification of PLE by point-of-care ultrasound performed by the nurses are comprehensively described in a recent publication.10 Ultrasound examinations were performed using the Vscan (GE Ultrasound, Horten, Norway) at each visit. With the transducer placed in the intercostal space, the liver and spleen were used as reference points to identify the diaphragm of the right and left hemithoraces, respectively. During quiet breathing, the posterior chest was scanned along the paravertebral, scapular, posterior and medial axillary lines, continuously focusing on the diaphragm as a reference point. PLE was assessed in the costodiaphragmatic angle by assessing the dimension between the diaphragm and the lung surface (measured in the middle between the transducer and the mediastinum). The amount of PLE was categorised in four groups as (1) no PLE, (2) insignificant when present in the costodiaphragmatic angle only, (3) small to moderate when the measurement as described above was <3 cm and (4) significant when the measurement was ≥ 3 cm. We also dichotomised the PLE measures as no to mild (groups (1) and (2)) and moderate to severe (groups (3) and (4)). Patient-reported outcome measures The patients completed the questionnaires before the clinical examination at the baseline visit and at the final follow-up visit as recommended for patient-reported outcomes.13 The functional class was self-reported according to the New York Heart Association (NYHA) classification14 on the validated Norwegian version of this instrument.15 QoL was reported by the validated Minnesota Living with Heart Failure Questionnaire (MLHFQ), which quantifies how patients with HF perceive their own QoL associated with the HF symptoms and treatment the preceding month. The questionnaire consists of 21 items where the patients are asked to rate how each item prevents them from living the life they desire using a six-point Likert scale from 0 to 5 (0= not at all, 5= all the time). The MLHFQ items include: physical HF symptoms (dyspnoea, fatigue, peripheral oedema and sleeping difficulties); psychological HF symptoms (anxiety and depression) and social/functional impairment due to HF (walking, climbing stairs, work, household or labour, need to rest, going places away from home, doing things with family and friends, eating, concentration, memory, loss of self-control, being a burden to others and sexual activity). The total score range from 0 to105, where the higher the score, the poorer the QoL. A reduction of 5 units or more in the MLHFQ-score by treatment is considered as clinically meaningful.16 We defined a MLHFQ-score >40 to be an indicator for poor QoL as this cut-off includes the patients that scores in the upper quartile of the moderate QoL category.17 Clinical examination, blood tests and medications The specialised nurses examined the patients at all visits. The clinical examinations included ECG (sinus rhythm or not), measurement of the blood pressure (mm Hg), heart rate (beats/min) and weight (to nearest 0.5 kg).10 11 Blood samples were collected at each visit and analysed at the hospital's International Electrotechnical Commision (IEC) 17025-accredited laboratory. Creatinine was analysed by an enzymatic method developed by Abbott and N-terminal pro-brain natriuretic peptide (NT-proBNP) by electrochemiluminescence immunoassay (ECLIA). Estimated glomerulus filtration rate (e-GFR) was calculated by the chronic kidney disease epidemiology collaboration (CKD-EPI) formula.18 The use of β-blockers was coded as 0 (no use) or 1 (use of) and the same coding was applied for the prescription of ACE inhibitors and angiotensin receptor blockers (ARBs). For diuretics, we recalculated the doses of furosemide, hydrochlorthiazide and bendrofumethiazide to equivalent doses of bumetanide. In the analyses, 40 mg furosemide, 12.5 mg hydrochlorthiazide and 2.5 mg bendrofumethiazide were each recalculated to 1 mg bumetanide.11 Statistical analysis Comparisons of the baseline characteristics between the participants with no to mild PLE and the participants with moderate to severe PLE and between those participants that self-reported a MLHFQ-score ≤40 compared with those who scored higher were performed by t-test for continuous variables and χ2-test for categorical data. One-way analyses of variance were used in order to assess the differences between MLHFQ-scores in the different categories of PLE and NYHA-classes. For analyses of trend, we treated the PLE categories and NYHA-class as continuous variables in a logistic regression model. Pearson correlation was used to investigate the correlation between patient-reported NYHA-class and MLHFQ-score. Sensitivity, specificity, positive predictive and negative predictive values of the MLHFQ cut-off to detect PLE were calculated at baseline. Generalised estimating equation with robust SEs and exchangeable correlation structure and inbuilt correction for age was used to model the change of the MLHFQ-score associated with change of PLE over time. For NYHA-class, we used the log link function of generalised estimating equation. In our second model, we adjusted for sex. In the third to seventh model, we added potential confounders (β-blockers, bumetanide-equivalent diuretic doses and ACE inhibitors/ARBs, change in weight and change of systolic blood pressure). These variables were added one by one to model 1, as the sample size was limited. We used a two-way graph based on linear regression to model the change in MLHFQ-score by change in centimetre (cm) PLE. We investigated the potential effect modification by sex and defined the critical p value to be <0.10 for the interaction term. For all other analyses we considered p <0.05 as significant. Statistical analyses were performed in Stata SE/13.1 for windows (© Stata Corp LP, Collage Station, Texas, USA). Results Table 1 shows the baseline characteristics of the patients (n=62) by PLE amount at baseline. As expected, the group with no or mild PLE had lower NT-proBNP at baseline than the group with moderate to severe PLE. There was a higher proportion of patients with sinus rhythm (49%) in the group with no to mild PLE compared with the group with moderate to severe PLE (21%). However, the two groups were comparable in age, sex, blood pressure, heart rate and the use of HF medication (all p>0.1). Self-reported MLHFQ-score >40 at baseline was associated with higher NYHA-class compared with MLHFQ-score ≤40 (p<0.001) (data not shown). Table 1 Baseline characteristics by the amount of pleural effusion No or mild pleural effusion (n=43) Moderate to severe pleural effusion (N=19) p between groups Variables mean ±SD (range) mean ±SD (range) Age, years 74±12 (35–91) 75±11 (49–92) 0.69 Systolic blood pressure (mm Hg) 125±24 (85–171) 115±18 (80–150) 0.12 Diastolic blood pressure (mm Hg) 73±14 (50–107) 68±13 (50–99) 0.23 Heart rate (bpm) 75±18 (51–121) 88±26 (56–140) 0.02 NT-proBNP (ng/L) 3252±2907 (90–9999) 5130±3320 (93–9999) 0.04 e-GFR (ml/min) 51±20 (16–88) 53±23 (26–99) 0.73 Bumetanide-equivalent (mg) 2.2±2.5 (0–13) 2.1±1.5 (0–6) 0.80 Weight 85±19 (56–141) 75±18 (49–120) 0.05 MLHFQ-score 42±21 (0–92) 55±24 (11–91) 0.03 n (%) n (%) Women 20 (47) 10 (53) 0.66 Sinus rhythm (yes) 21 (49)  4 (21) 0.04 Medication  Beta-blockers (yes) 35 (81) 14 (74) 0.49  ACEI or ARB (yes) 26 (61) 11 (58) 0.85 Patient-reported NYHA-class 0.23  NYHA-class 1 3 (7) 1 (5)  NYHA-class 2 24 (56) 6 (32)  NYHA-class 3 9 (21) 5 (26)  NYHA-class 4 7 (16) 7 (37) ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blockers; eGFR, estimated glomerulus filtration rate; MLHFQ-score, Minnesota Living with Heart Failure Questionnaire score; NT-proBNP, N-terminal pro-brain natriuretic peptide; NYHA-class, New York Heart Association functional classification. As shown in figure 2, the patient-reported MLHFQ-score at baseline depended on the PLE quantity (p=0.02). Approximately half of the patients with no PLE and most patients with moderate-to severe PLE had a MLHFQ-score >40. The mean MLHFQ-score in the patients with no or mild PLE was 42 (SD 21) compared with 55 (SD 24) in the patients with moderate to severe PLE (p=0.03). Figure 2 Shows the MLHFQ-score by amount of PLE in four categories at baseline. MLHFQ-score, Minnesota Living with Heart Failure Questionnaire score; PLE, pleural effusion. As shown in figure 3, self-reported NYHA-class and MLHFQ-score correlated well at baseline (r=0.63, p<0.001). MLHFQ-score was higher with more severe symptoms (NYHA-class 1 to 4) with p=0.001 for trend. Figure 3 Shows MLHFQ-score by self-reported NYHA-class at baseline. MLHFQ-score, Minnesota Living with Heart Failure Questionnaire score; NYHA-class, New York Heart Association functional classification. A MLHFQ-score >40 was observed for seven of nine patients with severe PLE amounts and for 15 of 19 patients with moderate to severe PLE (sensitivity 78% and specificity 49%). The corresponding positive and negative predictive values were 40% and 84% for the detection of moderate to severe PLE. In total, the mean reduction of PLE during follow-up was 0.8 cm (SD 2.1 cm; range from −9.7 to +0.7 cm). For obvious reasons, the reduction of PLE was greatest among those with moderate to severe amount (−2.6 cm, SD 3.5, range from −9.7 to +0.7 cm) compared with those with only mild or no PLE (−0.07 cm, SD 0.2, range from −1.0 to 0 cm). During follow-up, the patients improved their MLHFQ-score by an average of 17 points (SD 27, range from −79 to +42). A total of 17 patients (61%) had a clinically significant improvement in MLHFQ-score of >5 points reduction and 25 (86%) patients reported an improvement in NYHA-class. There was no difference in the proportion of patients who experienced clinically meaningful improvement in MLHFQ-score by dichotomised PLE groups (p=0.2). The improvement in MLHFQ-score correlated with improvement of NYHA-class (r=0.44, p=0.02). We observed a linear improvement in the MLHFQ-score (3.2 points, 95% CI 1.2 to 5.1) with each cm reduction in PLE. This improvement in MLHFQ was not explained by any of the adjustments for age, sex, β-blockers, bumetanide-equivalent, ACEI/ARB, change in weight or change in blood pressure over time (see table 2). Figure 4 shows the linear two-way association of MLHFQ-score with the reduction of PLE over time. The odds for reduction in NYHA-class per cm reduction in PLE was 1.06 (95% CI 1.04 to 1.08, p<0.001) and none of the seven adjustments models explained the association (data not shown). Table 2 Change in patient-reported Minnesota Living with Heart Failure Questionnaire (MLHFQ)-score by decrease in pleural effusion (cm) Model Adjustments ß-coefficient (95% CI) Model 1 Age 3.1 (1.1 to 5.0) Model 2 Age + sex 3.1 (1.0 to 5.2) Model 3 Model 1+ beta-blockers 3.2 (1.3 to 5.2) Model 4 Model 1+ bumetanide equivalent (in mg) 3.1 (1.1 to 5.0) Model 5 Model 1+ weight (change over time) 3.2 (0.8 to 5.5) Model 6 Model 1+ACEI or ARB 3.1 (0.9 to 5.2) Model 7 Model 1+ systolic BP (change over time) 3.0 (0.6 to 5.5) ß indicates the decrease in MLHFQ-score per cm decrease of PLE (95% CI) over time. ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; BP, blood pressure (mm Hg). Figure 4 Shows change in MLHFQ-score by change in PLE graphed by linear regression approach. MLHFQ-score, Minnesota Living with Heart Failure Questionnaire score; PLE, pleural effusion. Discussion Our study of outpatients with HF demonstrates that patient-reported QoL, by MLHFQ, is a sensitive tool for identifying patients with HF with severe amount of PLE. Furthermore, we found an improvement in MLHFQ-score and NYHA-class associated with reduction of PLE. A reduction of 1 cm of PLE corresponded to a reduction of 3.2 points in MLHFQ-score and 1.06 higher odds of improvement in functional class. In the models used, it seems that the change in PLE over time explains the improvement in MLHFQ-score and NYHA-class. Thus, paying attention to and treating PLE in follow-up of patients with HF may be important also with respect to improved symptoms and QoL. To the best of our knowledge, no study has assessed the sensitivity of MLHFQ to detect PLE among outpatients with HF. We find that the MLHFQ is a sensitive tool for detecting severe PLE in HF follow-up, and that a MLHFQ-score >40 should warrant further examination to search for PLE. The MLHFQ-score cut-off at 40 ensured that the impact on QoL for most of the patients with moderate–severe PLE was observed. However, half of the patients without PLE had MLHFQ-score >40 and treatment of PLE based solely on QoL is not advisable. However, as early detection of PLE is crucial for optimal follow-up of patients with HF, a moderate or worsened QoL should drive further diagnostic approach, and thereafter treatment. The MLHFQ-score is an easy and cheap method to raise the suspicion of PLE in patients with HF, and thus, should be routinely applied. Our study is also the first to describe the associations of reduced amount of PLE with improved self-reported MLHFQ-score and NYHA-class among patients with HF. The relation between HF symptoms and patient-reported outcomes is an area outlined in need of future research.19 The improvement in MLHFQ-score and NYHA-class were intercorrelated and in line with a previous study that showed improvement of MLHFQ-score during follow-up in a multidisciplinary HF programme.20 Point-of-care ultrasound of the pleural cavities is a sensitive and specific test for detection of PLE,1 21 and is previously shown to be superior to chest X-ray for diagnostics of PLE.22 The current study indicates that caring for improvement of patient-reported outcomes is an additional reason for systematic use of ultrasound to assess PLE in patients with HF,21 as excessive volume status could be an explanation for the poor QoL. However, if the availability of ultrasound is limited, a MLHFQ-score >40 identifies ∼80% of those with moderate to severe PLE and these should be referred to ultrasound. A determinant for the improvement in MLHFQ-score in an earlier study was found to be correlated with increased systolic blood pressure.20 In our study, adjustment for change in systolic blood pressure did not alter the improvement in MLHFQ-score associated with PLE reduction. However, the blood pressure was considered before titrating HF medication in order to prevent hypotension. Respiratory symptoms have been reported to reduce QoL in patients with cancer23 and PLE is associated with respiratory symptoms in patients with HF.2 4 The observed improvement in QoL and NYHA-class may be influenced by the expected improvement in respiratory symptoms through other mechanisms causing dyspnoea in HF and not only by the PLE reduction per se. However, in palliative care, intermittent pleural drainage is often associated with superior palliation, improvement of QoL and decreased morbidity compared with no treatment of the PLE,9 24 and we observed a reduction of PLE to be copresent with improvement in symptoms following targeted treatment. Nevertheless, the study also has limitations. The study was conducted at a single HF outpatient clinic and the sample size was small with limited number of patients with severe HF and large PLE. However, the presented population is previously shown to be comparable with studies from other HF clinics with regard to HF severity, age, sex, NYHA-class and comorbidity and thus, it is plausible that our results are generalisable to other HF clinics.11 Assessment of HF symptoms is difficult because of the subjective nature of the symptoms. Incorporation of patient-reported outcome measures in clinical practice may be important to improve management and care for patients with HF.13 25 26 Poor patient-reported MLHFQ-score and NYHA-classification are previously shown to be a strong predictor for early death in patients with HF.5 20 25 27 Improvement of MLHFQ-score during follow-up of HF is previously shown to predict event-free survival.20 28 29 The low number of participants do not allow for survival analyses. However, self-report to detect PLE is useful as patients may prefer a better QoL over prolonged life.30 As patient-reported QoL is increasingly emphasised in treatment strategies, further studies should be performed in order to evaluate the need for more systematic identification and earlier intervention aiming to reduce PLE in follow-up of patients with HF. Conclusions The MLHFQ-score seems sensitive to detect severe PLE. Patient-reported outcome measures like the MLHFQ-score and NYHA-class were associated with the amount of PLE. Further, the long-term improvement in MLHFQ-score and NYHA-class were associated with reduced amount of PLE. Thus, reducing PLE may be important to improve the QoL in patients with HF. Routinely including patient-reported outcomes in follow-up of outpatients with HF may identify those at highest risk of having symptomatic amount of PLE and may allow for further improvement in the care for these patients. Further studies are needed to draw definitive conclusions regarding the interrelationship between PLE, QoL assessment and outcomes. The Department of Internal Medicine, Levanger Hospital, at Nord-Trøndelag Hospital Trust, gave invaluable support in order to conduct and write up the study. Contributors: HD designed the study. All authors contributed to acquisition, analyses or interpretation of data. GHG and LTG drafted the manuscript, and all authors critically reviewed and approved submission of the final version of the manuscript. Funding: This study was funded by The Department for Research and Development at Nord-Trøndelag Hospital Trust. Competing interests: HD held a position at the Medical Imaging Laboratory, Norwegian University of Science and Technology (NTNU), a centre of research-based innovation that is funded by the Research Council of Norway and industry. One of the industry partners was GE Vingmed Ultrasound, which contributed to the total budget with 7 million NOK (6%) for the 8-year period 2007–2014. Ethics approval: The Committee for Medical and Health Research Ethics (REK 2013/257). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Porcel JM , Azzopardi M , Koegelenberg CF The diagnosis of pleural effusions . Expert Rev Respir Med 2015 ;9 :801 –15 . 10.1586/17476348.2015.1098535 26449328 2 Freeman RK Treatment options for patients with recurrent, symptomatic pleural effusions secondary to heart failure . Curr Opin Pulm Med 2015 ;21 :363 –7 . 10.1097/MCP.0000000000000176 26016580 3 Kataoka H Ultrasound pleural effusion sign as a useful marker for identifying heart failure worsening in established heart failure patients during follow-up . Congest Heart Fail 2012 ;18 :272 –7 . 10.1111/j.1751-7133.2012.00285.x 22994441 4 Porcel JM , Light RW Pleural effusions . Dis Mon 2013 ;59 :29 –57 . 10.1016/j.disamonth.2012.11.002 23374395 5 Kato N , Kinugawa K , Seki S Quality of life as an independent predictor for cardiac events and death in patients with heart failure . Circ J 2011 ;75 :1661 –9 .21532181 6 Nieminen MS , Dickstein K , Fonseca C The patient perspective: quality of life in advanced heart failure with frequent hospitalisations . Int J Cardiol 2015 ;191 :256 –64 . 10.1016/j.ijcard.2015.04.235 25981363 7 Dunderdale K , Thompson DR , Miles JN Quality-of-life measurement in chronic heart failure: do we take account of the patient perspective? Eur J Heart Fail 2005 ;7 :572 –82 . 10.1016/j.ejheart.2004.06.006 15921797 8 Lenker A , Mayer DK , Bernard SA Interventions to treat malignant pleural effusions . Clin J Oncol Nurs 2015 ;19 :501 –4 . 10.1188/15.CJON.501-504 26414568 9 Lorenzo MJ , Modesto M , Perez J Quality-of-life assessment in malignant pleural effusion treated with indwelling pleural catheter: a prospective study . Palliat Med 2014 ;28 :326 –34 . 10.1177/0269216314521851 24523284 10 Dalen H , Gundersen GH , Skjetne K Feasibility and reliability of pocket-size ultrasound examinations of the pleural cavities and vena cava inferior performed by nurses in an outpatient heart failure clinic . Eur J Cardiovasc Nurs 2015 ;14 :286 –93 . 10.1177/1474515114547651 25122616 11 Gundersen GH , Norekval TM , Haug HH Adding point-of-care ultrasound to assess volume status in heart failure patients in a nurse-led outpatient clinic. A randomised study . Heart 2016 ;102 :29 –34 . 10.1136/heartjnl-2015-307798 26438785 12 McMurray JJ , Adamopoulos S , Anker SD ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure 2012: the Task Force for the Diagnosis and Treatment of Acute and Chronic Heart Failure 2012 of the European Society of Cardiology . Eur Heart J 2012 ;33 :1787 –847 . 10.1093/eurheartj/ehs104 22611136 13 Anker SD , Agewall S , Borggrefe M The importance of patient-reported outcomes: a call for their comprehensive integration in cardiovascular clinical trials . Eur Heart J 2014 ;35 :2001 –9 . 10.1093/eurheartj/ehu205 24904027 14 Holland R , Rechel B , Stepien K Patients’ self-assessed functional status in heart failure by New York Heart Association class: a prognostic predictor of hospitalizations, quality of life and death . J Card Fail 2010 ;16 :150 –6 . 10.1016/j.cardfail.2009.08.010 20142027 15 Morken IM , Isaksen K , Karlsen B Shock anxiety among implantable cardioverter defibrillator recipients with recent tachyarrhythmia . Pacing Clin Electrophysiology 2012 ;35 : 1369 –76 . 16 Rector TS , Kubo SH , Cohn JN Patients’ self-assessment of their congestive heart failure. Content, reliability and validity of a new measure, The Minnesota Living with Heart Failure Questionnaire . Heart Fail 1987 :198 –209 . 17 Behlouli H , Feldman DE , Ducharme A Identifying relative cut-off scores with neural networks for interpretation of the Minnesota Living with Heart Failure questionnaire . Conf Proc IEEE Eng Med Biol Soc 2009 ;2009 :6242 –6 . 10.1109/IEMBS.2009.5334659 19965089 18 Matsushita K , Mahmoodi BK , Woodward M Comparison of risk prediction using the CKD-EPI equation and the MDRD study equation for estimated glomerular filtration rate . JAMA 2012 ;307 :1941 –51 . 10.1001/jama.2012.3954 22570462 19 Zannad F , Garcia AA , Anker SD Clinical outcome endpoints in heart failure trials: a European Society of Cardiology Heart Failure Association consensus document . Eur J Heart Fail 2013 ;15 : 1082 –94 . 10.1093/eurjhf/hft095 23787718 20 Hole T , Grundtvig M , Gullestad L Improved quality of life in Norwegian heart failure patients after follow-up in outpatient heart failure clinics: results from the Norwegian Heart Failure Registry . Eur J Heart Fail 2010 ;12 :1247 –52 . 10.1093/eurjhf/hfq156 20847014 21 Ponikowski P , Voors AA , Anker SD 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: the task force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC) . Eur J Heart Fail 2016 ;18 :891 –975 . 10.1002/ejhf.592 27207191 22 Graven T , Wahba A , Hammer AM Focused ultrasound of the pleural cavities and the pericardium by nurses after cardiac surgery . Scand Cardiovasc J 2015 ;49 :56 –63 . 10.3109/14017431.2015.1009383 25611808 23 Neumann K , Sundset A , Espinoza A Changes in quality of life, dyspnea scores, and lung function in lung cancer patients with airway obstruction after a therapeutic bronchoscopy . J Bronchology Interv Pulmonol 2013 ;20 :134 –9 . 10.1097/LBR.0b013e3182917280 23609247 24 Sabur NF , Chee A , Stather DR The impact of tunneled pleural catheters on the quality of life of patients with malignant pleural effusions . Respiration 2013 ;85 :36 –42 . 10.1159/000342343 23154202 25 Mastenbroek MH , Versteeg H , Zijlstra W Disease-specific health status as a predictor of mortality in patients with heart failure: a systematic literature review and meta-analysis of prospective cohort studies . Eur J Heart Fail 2014 ;16 :384 –93 . 10.1002/ejhf.55 24515393 26 Norekval TM , Falun N , Fridlund B Patient-reported outcomes on the agenda in cardiovascular clinical practice . Eur J Cardiovasc Nurs 2016 ;15 :108 –11 . 10.1177/1474515115614133 26512075 27 Hoekstra T , Jaarsma T , van Veldhuisen DJ Quality of life and survival in patients with heart failure . Eur J Heart Fail 2013 ;15 :94 –102 . 10.1093/eurjhf/hfs148 22989869 28 Moser DK , Yamokoski L , Sun JL Improvement in health-related quality of life after hospitalization predicts event-free survival in patients with advanced heart failure . J Card Fail 2009 ;15 :763 –9 . 10.1016/j.cardfail.2009.05.003 19879462 29 Soriano N , Ribera A , Marsal JR Improvements in health-related quality of life of patients admitted for heart failure. The HF-QoL study . Rev Esp Cardiol 2010 ;63 :668 –76 .20515624 30 Lewis EF , Johnson PA , Johnson W Preferences for quality of life or survival expressed by patients with heart failure . J Heart Lung Transplant 2001 ;20 :1016 –24 .11557198

PMC005xxxxxx/PMC5372076.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01275910.1136/bmjopen-2016-012759Geriatric MedicineResearch1506169817131712Systematic review of systematic reviews of non-pharmacological interventions to treat behavioural disturbances in older patients with dementia. The SENATOR-OnTop series http://orcid.org/0000-0002-5440-775XAbraha Iosief 1Rimland Joseph M 1Trotta Fabiana Mirella 1Dell'Aquila Giuseppina 1Cruz-Jentoft Alfonso 2Petrovic Mirko 3Gudmundsson Adalsteinn 4Soiza Roy 5O'Mahony Denis 6Guaita Antonio 7Cherubini Antonio 1 1 Geriatrics and Geriatric Emergency Care, Italian National Research Center on Aging (IRCCS-INRCA), Ancona, Italy 2 Servicio de Geriatría, Hospital Universitario Ramón y Cajal, Madrid, Spain 3 Department of Internal Medicine (Geriatrics), Ghent University, Ghent, Belgium 4 Landspitali University Hospital Reykjavik, Reykjavik, Iceland 5 Department of Medicine for the Elderly, Woodend Hospital, Aberdeen, UK 6 Department of Medicine, University College Cork, Cork, Ireland 7 ‘Golgi Cenci’ Foundation, Milan, ItalyCorrespondence to Dr Iosief Abraha; iosief_a@yahoo.it2017 16 3 2017 7 3 e01275923 5 2016 5 9 2016 4 10 2016 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective To provide an overview of non-pharmacological interventions for behavioural and psychological symptoms in dementia (BPSD). Design Systematic overview of reviews. Data sources PubMed, EMBASE, Cochrane Database of Systematic Reviews, CINAHL and PsycINFO (2009–March 2015). Eligibility criteria Systematic reviews (SRs) that included at least one comparative study evaluating any non-pharmacological intervention, to treat BPSD. Data extraction Eligible studies were selected and data extracted independently by 2 reviewers. The AMSTAR checklist was used to assess the quality of the SRs. Data analysis Extracted data were synthesised using a narrative approach. Results 38 SRs and 142 primary studies were identified, comprising the following categories of non-pharmacological interventions: (1) sensory stimulation interventions (12 SRs, 27 primary studies) that encompassed: acupressure, aromatherapy, massage/touch therapy, light therapy and sensory garden; (2) cognitive/emotion-oriented interventions (33 SRs; 70 primary studies) that included cognitive stimulation, music/dance therapy, dance therapy, snoezelen, transcutaneous electrical nerve stimulation, reminiscence therapy, validation therapy, simulated presence therapy; (3) behaviour management techniques (6 SRs; 32 primary studies) and (4) other therapies (5 SRs, 12 primary studies) comprising exercise therapy, animal-assisted therapy, special care unit and dining room environment-based interventions. Music therapy was effective in reducing agitation (SMD, −0.49; 95% CI −0.82 to −0.17; p=0.003), and anxiety (SMD, −0.64; 95% CI −1.05 to −0.24; p=0.002). Home-based behavioural management techniques, caregiver-based interventions or staff training in communication skills, person-centred care or dementia care mapping with supervision during implementation were found to be effective for symptomatic and severe agitation. Conclusions A large number of non-pharmacological interventions for BPSD were identified. The majority of the studies had great variation in how the same type of intervention was defined and applied, the follow-up duration, the type of outcome measured, usually with modest sample size. Overall, music therapy and behavioural management techniques were effective for reducing BPSD. Non-pharmacological interventionAlzheimer's diseaseBehavioral and psychological symptoms in dementiaBPSD ==== Body Strengths and limitations of this study Non-pharmacological interventions have gained increasing attention in recent years as an alternative first-line approach to treat behavioural and psychological symptoms in dementia (BPSD). The strength of this review is its extensive, comprehensive systematic search of studies that investigated non-pharmacological interventions for BPSD. It provides a compendium of the types of non-pharmacological interventions, including the component of each single intervention, the dosage (when available) and the duration of the treatment. Primary studies were generally of limited sample size; there was substantial variation in the characteristics of the intervention and the authors of primary studies reported different conceptual frameworks, and sometimes broad, and quite generic descriptions, of the interventions. Introduction Dementia is a neuropsychiatric syndrome characterised by cognitive decline and progressive deterioration of daily function, often associated with behavioural disturbances. The prevalence of dementia in older participants is reported to be ∼6% worldwide1 and, with global population ageing, it is expected to rise, although some recent studies have suggested declining trends in dementia frequency.2 Dementia presents a considerable burden to families and caregivers and is becoming a major challenge for all healthcare systems, as well as for society at large.3 4 Alzheimer's disease (AD) is the most common form of dementia in older people, accounting for 60% of cases. Approximately five out of every six patients with dementia, including those living at home, will develop behavioural and psychological symptoms during the course of the disease.5–8 Behavioural and psychological symptoms in dementia (BPSD) are defined as signs and symptoms of disturbed behaviour, mood, thought or perception.9 These disturbances, namely agitation, depression, elation, delusions and hallucinations, are strongly correlated with each other.10 11 Twenty per cent of those initially without symptoms will manifest them within 2 years of dementia diagnosis,12 whereas 50–80% of those with clinically important symptoms remain agitated for several months.13 In addition, at least 50% of patients with dementia present with significant BPSD on a monthly basis.14 Agitation, together with depression, hinder activities and relationships, cause feelings of helplessness and distress in families and formal caregivers15 and are strong predictors for poor quality of life,16 as well as nursing home admission.17 Currently, options for treating BPSD include pharmacological and non-pharmacological therapies.18 19 Psychotropic medications are often used to reduce the frequency and severity of BPSD, but in the majority of patients, they provide only modest symptom control.20–22 A recent trial reported that the addition of citalopram to psychosocial support significantly reduced agitation and caregiver distress.23 However, their adverse effects are common and problematic, in particular the increased risk of falls and fractures,24 stroke and even mortality.25 In addition, there is some evidence that the use of benzodiazepines to treat agitation in patients with dementia may increase cognitive decline24 and may expose patients to an immediate risk of injurious falls.26 Finally, memantine and cholinesterase inhibitors are considered to be of very limited value to improve agitation in participants with AD.27 28 In general, non-pharmacological interventions are considered a preferable alternative to psychotropic pharmacotherapy for treating BPSD.29 However, there is conflicting evidence concerning the efficacy and practicality of non-pharmacological interventions to improve BPSD, particularly agitation.9 30 The purpose of the present overview is to assess the evidence supporting these non-pharmacological interventions with a view to providing a working compendium for the non-drug management of BPSD. The present overview updates the evidence on the same theme gathered by a previous systematic overview published in 2011.31 Methods This work is part of the Optimal Evidence-Based Non-drug Therapies in Older People (ONTOP) project, a work package of a European Union funded FP7 research named SENATOR (Software ENgine for the Assessment & Optimization of drug and non-drug Therapy in Older peRsons). The ONTOP aim is to undertake a literature search of systematic reviews (SRs) and provide clinical recommendations concerning evidence-based non-pharmacological treatments of several prevalent medical conditions affecting older people, including delirium,32 33 pressure ulcers,33–35 falls,36 37 stroke and heart failure. A protocol that describes the search strategy, screening and inclusion criteria, has been previously published.38 Briefly, to obtain the evidence regarding the non-pharmacological interventions, we first identified published SRs using a systematic search across several databases. After processing eligible SRs, we identified and obtained primary studies from these SRs to generate the compendium of non-pharmacological interventions. In a subsequent work will present the assessment of the body of evidence and provide recommendations according to the GRADE approach.38 Search strategy and inclusion criteria for systematic reviews The search sources included the Cochrane Database of Systematic Reviews, PubMed, PsycINFO and CINAHL (see online supplementary appendix 1). Two criteria were considered for further evaluation of an abstract: (1) a paper defined as a review or a meta-analysis; (2) the use of any non-pharmacological intervention to treat behavioural disturbances in patients with dementia. The publication years ranged from 2009 to March 2015. 10.1136/bmjopen-2016-012759.supp1supplementary appendix Subsequently, full-texts of relevant abstracts were obtained and screened to identify SRs of interest based on (1) the use of at least one medical literature database; (2) the inclusion of at least one primary study and (3) the use of at least one non-pharmacological intervention to treat behavioural disturbances in people aged 60+years. We assessed the methodological quality of each SR using the AMSTAR (A Measurement Tool to Assess Reviews) instrument that contains 11 items.39 Final grading of the methodological quality of each SR was based on the overall score and reported as either ‘high’ (score≥8), ‘medium’ (score 4–7) or ‘low’ (score≤3). Two reviewers independently assessed the quality of the SRs, and disagreements were resolved by consensus. Data extraction and management From each SR, the following data were collected: the publication year, the databases searched, the study population, the non-pharmacological interventions, the number of primary studies included, the outcome measures and the AMSTAR score. Pairs of reviewers independently screened titles, abstracts and full texts of articles. Disagreements were resolved by discussion or, where necessary, by consulting another author. Outcome measures We focused on reviews that considered BPSD, as a primary outcome, measured by (1) multidomain scales (eg, Neuropsychiatric Inventory (NPI), Brief Psychiatric Rating Scale, BPRS), (2) scales specific to agitation (eg, Cohen-Mansfield Agitation Inventory, CMAI) and (3) scales specific to depression or anxiety (eg, Cornell Scale for Depression in Dementia, CSDD). Inclusion criteria for primary studies and assessment From the included SRs, we obtained any experimental comparative study, either randomised or non-randomised, that investigated any non-pharmacological intervention to treat BPSD in older patients. Observational studies or before–after studies, with historical controls, were excluded. As outlined in our protocol, we extracted data from primary studies to perform meta-analyses and heterogeneity was addressed using the Cochrane Collaboration approach.38 Risk of bias assessment and grading the quality of evidence We used the Cochrane Collaboration method to evaluate the risk of bias. The domains considered were random sequence generation, allocation concealment,40 blinding of participants, personnel, or outcome assessor,40 incomplete outcome data,41 selective reporting42 and other potential biases (eg, balance in baseline characteristics). The overall quality of evidence was assessed using the GRADE (Grading of Recommendations, Assessment, Development and Evaluation) methodology that takes into account the risk of bias, consistency of results across the studies, precision of the results, directness and likelihood of publication bias.43 Results regarding the risk of bias assessment, and grading the quality of evidence, will be provided in a companion paper. Results Our search strategy identified 4392 abstracts of which 2549 were duplicates and were subsequently removed. After abstract screening, 67 records were identified for full-text assessment. Of these, 38 reviews were included in this overview. From these SRs, we obtained 142 primary studies from which we abstracted details of the non-pharmacological interventions. Figure 1 shows the study screening process. Table 1 depicts the basic characteristics of the included SRs. The characteristics of relevant primary studies are reported in online supplementary appendix 2 as electronic tables (etable). The AMSTAR evaluation are summarised in online supplementary appendix 3. Table 1 Characteristics of included systematic reviews/meta-analyses Author (year) Databases searched Population Non-pharmacological intervention Primary studies Outcome Aguirre 2013 MEDLINE, Embase, Cinahl, PsycINFO, the Cochrane Library, Lilacs, trial registers, grey literature Participants who had a diagnosis of dementia (Alzheimer's disease, vascular dementia mixed Alzheimer's and vascular dementia, other types of dementia), including all levels of cognitive impairment Cognitive stimulation 15 RCTs Mood, quality of life, well-being, ADL, communication, behaviour, neuropsychiatric symptoms and social interaction Alves 2013 MEDLINE, PsycINFO, Cochrane Library, EMBASE, metaRegister of Clinical Trials, OVID all, EBM Reviews Patients diagnosed with Alzheimer's disease (without mild cognitive impairment, mixed Alzheimer's disease, vascular dementia, and other types of dementia such as frontotemporal dementia or dementia with Lewy bodies) Memory-training program; attention-stimulating activities; computerized ‘cognitive training’ 4 studies Cognitive functioning; ADL; memory Symptoms; finger tapping test; depressive symptoms; QoL; reaction time; screening of mental status; neuropsychiatric symptoms Bernabei 2013 MEDLINE, Embase, PsycINFO Elderly patients affected by dementia or psychiatric disorders Animal-assisted interventions 10 studies on BPSD (3 case–control and 7 repeated measures design) Any psychiatric disorder Blake 2013 PubMed, Science Direct, the Cochrane Library and Web of Knowledge Adults diagnosed with dementia who have depressive symptoms Reminiscence group therapy 4 studies Change in level of depressive symptoms Carrion 2013 MEDLINE, EMBASE, PASCAL, the Cochrane Library, National Guidelines Clearinghouse, Trip database, HEALTHSTAR, CINHAL and PsycINFO Older people diagnosed as having Alzheimer's disease or probable Alzheimer's disease Cognition-oriented care approaches: 1.Reality orientation; 2. Skills training Reality orientation: 9 RCTs;Skills training: 8 RCTs Cognitive function; behavioural symptoms and mood Chaudhury 2013 MEDLINE, CINAHL, Ageline, Web of Science, and Simon Fraser University library catalogue Long-term facility residents with dementia Supportive dining environment 21 studies included: light therapy (1 study); music therapy (3 studies) Physiological and sociopsychological aspects of dining, including caloric intake, enjoyment in eating and social interaction Collet 2010 MEDLINE, PsycINFO and PubMed Nursing home patients suffering from either somatic illness or dementia combined with psychiatric disorders or severe behavioural problems Psychiatric care and nursing home care combination 8 RCTs Psychosis and depression, increase in global functioning, behaviour disorders, cognition and ADL Eggenberger 2013 MEDLINE, AMED, EMBASE, PsycINFO, CINAHL, The Cochrane Library, Gerolit, Web of Science People with dementia; professional and family caregivers Communication skills training by means of face-to-face interaction 12 RCTs QoL, social interactions Forbes 2014 MEDLINE, EMBASE, the Cochrane Library, CINAHL, PsycINFO, LILACS+several Registries, proceedings+other sites People with dementia Light therapy 5 studies met the inclusion criteria—only 3 were included in the analyses because of inappropriate reported Sleep, behaviour, mood, and cognitive disturbances associated with dementia Forrester 2014 MEDLINE, EMBASE, the Cochrane Library, CINAHL, PsycINFO, LILACS+several Registries, proceedings+other sites People with dementia Aromatherapy 2 RCTs Agitation, behavioural symptoms, quality of life and adverse effects Fung 2012 MEDLINE, CINAHL, Cochrane Library, PsycINFO, Social Sciences Citation Index, SCOPUS Participants with dementia Aromatherapy 11 studies (5 RCTs; 6 controlled trials) Behavioural problems Gonzalez 2014 MEDLINE, AMED, CINAHL, ISI Web of Knowledge, Embase and Scopus People with dementia Sensory gardens and horticultural activities 2 RCTs Agitation levels; cognitive status Guzman-Garcia 2013 MEDLINE, AgeInfo, EBM Reviews EBSCO-CINAHL, EMBASE, ISI Web, LILACS, SCOPUS ZETOC; reference lists; EthOS-Beta; ACER; Google;+ People with dementia living in long-term care homes Dance movement therapy; dance therapy; Psychomotor dance-based; Social dancing 10 studies (1 RCT) Not specified Kiepe 2012 MEDLINE and PsycINFO Patients with mental illness Dance therapy 1 study (RCT) for dementia Any physical and mental outcomes Kim 2012 MEDLINE, CINAHL, ProQuest Medical Library, and Cochrane and OT Persons with dementia Occupational therapy 9 studies Behavioural problems and depression Kverno 2009 MEDLINE, CINAHL, PsycINFO, EMBASE Individuals diagnosed with advanced dementia Any non-pharmacological intervention 460 primary studies Neuropsychiatric symptoms Lai 2009 MEDLINE, The Cochrane Library, EMBASE, PsycINFO and CINAHL Patients with a confirmed diagnosis of dementia or Alzheimer's disease or related disorders Special care units 8 non-randomised studies (0 RCT) Behavioural problems, mood, use of restraints and psychotropic medication Livingston 2014 MEDLINE; Web of Knowledge; EMBASE; British Nursing Index; the Health Technology Assessment programme database; PsycINFO; NHS Evidence; System for Information on Grey Literature Participants with dementia Sensory, psychological and behavioural interventions 160 primary studies Agitation McDermott 2013 MEDLINE, EMBASE, PsycINFO, CINAHL, Cochrane Library, Web of Science, J Music Therapy, and Nordic Journal of Music Therapy Participants with dementia Music therapy 15 studies (6 RCTs; 4 non-randomised trials; 5 before–after studies Behavioural, psychological aspects, hormonal and physiological changes, social and relational aspects of music therapy Moniz-Cook 2012 MEDLINE, EMBASE, CINAHL, PsycInfo and LILACS;+ People with dementia, irrespective of its cause or diagnostic subtype, with reported BPSD or ‘behaviours that challenge’, receiving support or treatment from mental health workers, care staff or family or other informal caregivers Formulation-led individualised interventions targeting reduction in the person's distress and/or resolution of the caregivers’ management difficulties 18 trials Challenging behaviours (eg, verbal and physical aggression, restlessness) and mood (depression).Changes in caregiver self-report of reaction to challenging behaviours. Moyle 2013 MEDLINE, CINHAL, PsycINFO, Cochrane Library, Scopus, Web of Science, Health Reference Center Academic Older people with dementia: Massage therapy Of 13 studies identified only 1 satisfied the quality of the inclusion criteria Agitated behaviour O'Neill 2011 MEDLINE, the Cochrane Library, PsycINFO Adults with mild, moderate, or severe dementia Non-pharmacological treatments 28 systematic reviews Behavioural symptoms of dementia Olazaran 2010 MEDLINE, PsycINFO, CINAHL, Embase, Lilacs and the Cochrane Dementia and Cognitive Improvement Group Specialized Register People with Alzheimer's disease and related disorders Any non-pharmacological intervention 213 Cognition; institutionalization; ADL; behaviour; mood; QoL; psychological well-being Padilla 2011 MEDLINE, the Cochrane Library, AgeLine, CINAHL, PsycINFO, EMBASE, and HealthSTAR, OT Seeker, and Allied and Complementary Medicine+reference list People with Alzheimer's disease and related dementias Environment-based interventions; multisensory approaches; other interventions 1 cross-overall trial (environmental-based intervention) Performance, affect and behaviour Potter 2011 MEDLINE, EMBASE, CINAHL, PsycINFO, AMED, the Cochrane Library, the UK National Research Register, Current Controlled Trials Older people with dementia Strength and flexibility; strength and balance Tai Chi classes sitting and standing; walking; stretching; seated exercises; balance training; endurance; aerobic training 13 RCTs Physical functioning, quality of life anddepression Robinson 2011 MEDLINE; EBM reviews; AMED; BNI; CINAHL; EMBASE Not specified Acupressure 1 RCT (of 71 acupressure studies) Any outcome Salami 2011 MEDLINE, EMBASE and the Cochrane Central Register People with Alzheimer's disease Any treatment option for sleep disturbance not attributable to other clinical conditions 9 RCTs Sleep disturbance Seitz 2012 MEDLINE, EMBASE, PsycINFO, the Cochrane Library and Google Scholar People with dementia Any non-pharmacological intervention 40 studies Neuropsychiatric symptoms Subramaniam 2012 MEDLINE, PsycINFO, CINAHL, the Cochrane Library, EMBASE and Web of Knowledge People with dementia Reminiscence therapy 5 RCTs Psychosocial benefits Thune-Boyle 2012 MEDLINE, EMBASE, PsycINFO People with dementia Exercise therapy 2 RCTs Behavioural and psychological symptoms Ueda 2013 MEDLINE, CINAHL, PsycINFO People with dementia Music therapy 18 of 20 studies considered agitation or anxiety (9 CCT; 9 RCTs) Behavioral and psychological symptoms; ADL; cognitive function Vasionytė 2013 JSTOR, EBSCO, ERIC, SCIRUS, MEDLINE, PsycINFO, Cochrane Library and ProQuest, the journal databases SAGE PUB and Cambridge journals Patients with dementia Music therapy 3 RCTs Affective, behavioural, cognitive and physiological outcomes Vasse 2010 PubMed, PsycINFO, Web of Science and the Cochrane library People with dementia A walking program combined with conversation, group validation therapy, life review programs, cognitive stimulation therapy, activity therapy and staff education 9 RCTs Communication between residents with dementia and care staff; neuropsychiatric symptoms of residents with dementia. Wall 2010 MEDLINE, CINAHL, PsycINFO Older people with dementia Music therapy 4 RCTs Behaviour and well-being Whear 2014 MEDLINE, PsycINFO, Embase, HMIC, AMED; Cochrane Library; CINAHL; British Nursing Index; ASSIA; Social Science Citation Index; EThOS; Social Care Online and OpenGrey November 2012. Elderly residents with dementia Mealtime interventions categorized into four types: music, changes to food service, dining environment alteration, and group conversation 11 studies (7 time series repeated measures; 3 pre-post study design; 1 controlled clinical trial) Behavioural symptoms (anxiety, agitation, aggression) Woods 2012 MEDLINE, the Cochrane Library, EMBASE, PsycINFO, CINAHL and LILACS, ongoing trial databases and grey literature sources People with dementia and their caregivers Cognitive stimulation 15 RCTs Cognitive functioning; mood; QoL; ADL; behavioural symptoms; neuropsychiatric symptoms Yu 2009 MEDLINE (PubMed), CINAHL, PsycINFO, and the Cochrane Library People with early-stage Alzheimer's disease and dementia Cognitive training 7 RCTs Any dementia symptoms in early-stage Alzheimer's disease and related dementia. Zimmerman 2013 MEDLINE, EMBASE, the Cochrane Library, the Cumulative Index to Nursing and Allied Health Literature, AgeLine, and PsycINFO People with dementia in nursing homes and other residential long-term care settings Effective characteristics of residential long-term care 14 studies: 4 prospective cohort studies, 9 RCTs, 1 non-randomized controlled trial Health and psychosocial outcomes ADL, activities of daily living; QoL, quality of life; RCT, randomised controlled trial. Figure 1 Study screening process. 10.1136/bmjopen-2016-012759.supp2supplementary appendix 10.1136/bmjopen-2016-012759.supp3supplementary appendix The interventions in this overview were classified according to the following categories: (1) sensory stimulation interventions that encompass acupuncture, aromatherapy, massage therapy, light therapy, sensory garden intervention, cognitive stimulation, music/singing and dance therapy, snoezelen and transcutaneous electrical nerve stimulation (TENS) therapy; (2) cognitive/emotion-oriented interventions that include reminiscence therapy, validation therapy, simulated presence therapy (SPT); (3) behavioural management technique and (4) other interventions, such as exercise therapy, pet-therapy and special care unit. Sensory stimulation interventions Shiatsu and acupressure Only one SR was identified. Robinson 201144 (AMSTAR=7) investigated the evidence available for shiatsu and acupressure in BPSD. Shiatsu is a form of complementary medicine primarily developed in Japan, which employs gentle manipulations, stretches and pressure with the fingers, elbows, knees and feet. Acupressure is similar, but exerts pressure for longer on specific meridian points according to traditional Chinese medicine or acupoints of the human body in order to ‘balance energy fields’. The authors identified 40 RCTs, 8 controlled clinical trials, 5 crossover trials, 6 within-participants studies, 1 observational study, 10 uncontrolled studies and 1 prospective study. Only one randomised trial (n=133 participants) using acupressure in dementia participants was relevant for our assessment.45 The authors reported that agitation, aggression and physically non-aggressive behaviour all declined significantly in demented participants. Aromatherapy Aromatherapy is proposed as a complementary intervention, to treat a wide-range of health problems, including lack of sleep and behavioural symptoms for people with dementia.46 Aromatherapy is based on the use of plant products or aromatic plant oils to produce essential oils and blends of aromatic compounds. Aromatherapy can be delivered through massage or topical application, inhalation and water immersion. Our systematic search identified three SRs that considered aromatherapy as an intervention to treat agitated behaviours and other outcomes in patients with dementia. The AMSTAR scores ranged from 6 to 8 across the reviews. The range of included primary studies varied from 4 to 13.29 47 48 The most recent SR was a Cochrane review,48 which had the highest AMSTAR quality score (8). The review included only randomised trials and launched its last search strategy in January 2013. Seven studies with 428 participants were identified. The types of interventions included lavender-based (four studies49–52), Melissa-based (two studies53 54) and lemon balm oil (1 study) aromatherapy. However, only two of these had usable data for pooling. The first study (n=71) reported a favourable treatment effect on measures of agitation (MD −11.1, 95% CI −19.9 to −2.2) and behavioural symptoms (MD −15.8, 95% CI −24.4 to −7.2), whereas the second trial (n=63) did not detect any difference in agitation (MD 0.00, 95% CI −1.36 to 1.36) or behavioural symptoms (n=63, MD 2.80, 95% CI −5.84 to 11.44). The review authors remarked that the published studies used different scales to assess the behavioural symptoms and were limited in sample size and methodological quality, particularly because of selective reporting bias. The second review by Seitz et al29 consisted of any non-pharmacological interventions, including aromatherapy, to treat outcomes relevant to patients with dementia. The review reported data in a narrative way and cited only one study of aromatherapy,53 which was also included in the Cochrane review above.48 The review received an AMSTAR score of 6. The third study was a review by Fung et al,47 which considered only aromatherapy as a non-pharmacological intervention. The review was judged to have moderate methodological quality (AMSTAR score=6). After performing a comprehensive search in several electronic databases, 11 studies were identified, with a total of 405 patients in different settings, including long-term care (LTC) homes, clinical centres and general and old age psychiatry. In addition to the trials included in the above cited Cochrane review, the review by Fung et al47 included one randomised trial55 which was excluded in the Cochrane review because the route of administration was not reported and there was no mention of the type of the aromatherapy, in addition to five controlled clinical trials.56–60 Moreover, the Fung et al review47 did not include the two trials49 51 61 that were evaluated in the Cochrane review. The controlled clinical trials could not be included in a meta-analysis because of heterogeneity. The review highlighted the methodological limitations of the studies and reported promising results of aromatherapy. Online supplementary etable 1 describes the type of interventions, the outcomes and the results of the primary studies included in the aromatherapy reviews. Massage therapy Massage and touch therapy have been proposed as non-pharmacological interventions to be used in dementia to offset manifestations of cognitive decline and behavioural disturbances, including related psychological problems, such as depression and anxiety, and to improve quality of life.62 Two reviews were identified. The first was a Cochrane review63 that was included in the review by O'Neil et al.31 This review assessed the efficacy of massage and touch therapy for the treatment of BPSD. Its last search strategy was launched in 2006. The aim of the overview was to evaluate the effects of a range of massage and touch therapies on conditions associated with dementia, such as anxiety, agitated behaviour and depression, to identify any adverse effects and to provide recommendations for future trials. The review considered only randomised trials. The primary outcome measures were changes in the frequency and severity of various types of agitated behaviour, as observed by staff or investigators (short-term and long-term using any rating method), and the emotional well-being and the quality of life of the patients (rated by staff, investigators and/or patients themselves using any method). Remington (2002)64 assessed the effect of music and massage in 68 nursing home residents with dementia (AD, multi-infarct dementia or senile dementia). The participants were randomly allocated into four groups: calming music, hand massage, simultaneous calming music and hand massage and no intervention. The intervention lasted 10 min and was given to each patient once. The efficacy of treatment on ‘agitation level’ was evaluated with a modified version of the CMAI administered by trained research assistants who were blinded to treatment allocation when possible. The method of randomisation was unclear and to conceal allocation, sealed envelopes, without further explanation, were used. However, patients could have been excluded after allocation (if they had a CMAI score of 0 at baseline) and consequently the study was considered to have high risk of selection bias. The trial found that agitated behaviour decreased, more so in the group receiving hand massage than in the group receiving no treatment. This treatment effect was consistently found, compared to baseline, for measurements taken during treatment, immediately after treatment and 1-hour after treatment, and it was practically identical among the three groups receiving treatment (hand massage, calming music or both). The mean agitation score was in favour of massage therapy immediately after treatment (MD 7.83 (4.30 to 11.36)) and 1-hour after treatment (MD 12.12 (6.58 to 17.66)). The second review by Moyle et al65 conducted a search in 10 databases in October 2011. The authors identified 13 studies that evaluated massage therapy for the treatment of behavioural disturbances in patients with dementia, but only one study with a high methodological score, using the Validity Rating Tool, was identified. The included study, performed by Holliday-Welsh,66 was a prospective before–after study in which 52 participants (39 women and 13 men; mean age 90 years) from two skilled nursing facilities in Northeastern Minnesota, USA, were enrolled. Patients were cognitively impaired and had a history of agitated behaviour confirmed by the facility staff. The intervention consisted of a 10-min to 15-min massage of the upper extremities (including the head, shoulders and hands), undertaken by a physical therapy assistant, during a 1-hour period identified by caregivers as the time the participant was usually most agitated (individualised for each participant). The outcomes of interest were assessed with a scale that used the five behavioural symptoms from the minimum data set; (1) wandering; (2) verbally abusive behavioural symptoms; (3) physically abusive behavioural symptoms; (4) socially inappropriate/disruptive behaviour and (5) resistance to care. Methodologically, the study was considered at high risk of selection and performance bias given the study design and the nature of the intervention. In addition, it was unclear whether the outcome assessor was blinded. Massage therapy was significantly associated with improvement for four of the five outcomes examined, including wandering (0.38 vs 0.16, p<0.001), verbally agitated behavioural symptoms (0.59 vs 0.49, p=0.002), physically agitated behavioural symptoms (0.82 vs 0.40, p<0.001) and resistance to care (0.10 vs 0.09, p=0.022). Online supplementary etable 2 describes the type of interventions, the outcomes and the results of the primary studies included in the massage therapy reviews. Light therapy Rest-activity and sleep-wake cycles are controlled by the endogenous circadian rhythm generated by the suprachiasmatic nucleus (SCN) of the hypothalamus. Degenerative changes in the SCN appear to be a biological cause of circadian rhythm disturbances in people with dementia. In addition to the internal regulatory loss, older people (especially those with dementia) experience a reduction in sensory input, due to less visual sensitivity to light and less exposure to bright environmental light. Evidence suggests that circadian rhythm disturbances may be reversed by stimulation of the SCN with light.67 Four reviews considered the use of bright light therapy to treat behavioural problems in patients with dementia. The first was a Cochrane review67 (AMSTAR=10) with the aim of evaluating the effectiveness of light therapy to improve cognition, activities of daily living (ADLs), sleep, challenging behaviour and psychiatric disturbances associated with dementia. The search strategy was launched in January 2014. The included studies were randomised trials that compared any bright light therapy, including dim red light or dim, low-frequency blinking light <300 lux, to usual care. The primary outcome measures included cognition (global or single domain, eg, memory), ADLs, sleep-wake disturbances, challenging behaviour (eg, agitation), psychiatric disturbances (eg, depression) and adverse effects. Secondary outcomes were rates of institutionalisation and overall cost of care. The authors identified 11 studies, but stated that three of the studies could not be included in the analyses either because the data were insufficient or could not be retrieved from the trial authors. Only four of the included studies considered challenging behaviour as an outcome, but the sample sizes were limited and the outcome measures were not the same across the studies.68–71 A meta-analysis of challenging behaviour, however, was performed and no substantial heterogeneity was found, although the results were not statistically in favour of bright light therapy. The second review aimed to identify which non-pharmacological interventions were most effective for BPSD in LTC.29 Only two studies68 69 were included in the review (which were already included in the Forbes review67), but were not assessed in detail. The review received four points in the AMSTAR rating system. The third review72 aimed to assess the role of physical environment in supporting person-centred dining in LTC. Only one study that evaluated the effect of ambient bright light in activity and dining areas among institutionalised people with dementia was identified.73 This study was not included in the previous two reviews. The fourth review74 that addressed the effectiveness of environment-based interventions for people with AD or dementia identified a cluster-unit crossover trial.75 The trial was conducted in two geriatric units in a state-operated psychiatric hospital and in a dementia-specific residential care facility in Oregon, USA, and enrolled 66 older adults with dementia to evaluate the effectiveness of ambient bright light therapy, delivered through a high-intensity, low-glare lighting system installed in the public areas of study units at both sites, at reducing depressive symptoms. Each lighting condition was provided for multiple 3-week periods in a predetermined sequence. The CSDD was used to assess depressive symptoms. Results did not support the use of ambient bright light therapy as a treatment for depressive symptoms in people with dementia.75 Online supplementary etable 3 describes the type of interventions, the outcomes and the results of the primary studies included in the light therapy reviews. Sensory garden and horticultural activities Whear 2014 76 (AMSTAR=7) investigated the impact of gardens and horticultural therapy on the mental and physical well-being of residents with dementia, in nursing homes and specialised dementia care facilities. This approach uses either ‘sensory’ gardens to stimulate the five senses (sight, vision, hearing, smell and touch), or plants and plant-related activities to improve well-being (horticultural therapy or therapeutic horticulture). Eighteen studies were identified: ten were quantitative studies (two RCTs (n=34), six pre-post studies, one crossover study, one prospective cohort study), seven qualitative and one used mixed methods. In one of the RCTs,77 there was a non-statistically significant decline in verbal and physical aggression and non-verbal aggression, and total CMAI score (Online supplementary etable 4). Gonzalez et al78 (AMSTAR=3) examined the effects of sensory garden and horticultural activities in dementia care. Sixteen studies were identified, including 2 RCTs (n=149), one of which was cluster randomised, 11 pre-post studies, 2 case studies and 1 survey. In the smaller of the two RCTs,79 verbal agitation significantly decreased in the outdoor horticultural group compared to the indoor horticultural group, while in the larger trial, the effect of participants in the horticultural group did not differ from the traditional activity group. (Connell et al79 was included in both SRs.76 78) Online supplementary etable 4 describes the type of interventions, the outcomes and the results of the primary studies included in the sensory garden and horticultural activities reviews. Music and dance therapy Music therapy is the application of music and/or its elements (melody, rhythm, harmony, sound) by a qualified musical therapist, in order to support and stimulate various aspects of cognitive, emotional, social and physical needs, such as expression, communication, learning and forming relationships. Participants can passively listen to music or actively participate by singing, playing an instrument or moving. Dance therapy is a psychotherapeutic intervention that uses movement to ‘further the emotional, cognitive, physical and social integration of the individual’.80 Six SRs that evaluated music therapy,29 81–85 and one review that assessed live singing to people affected with dementia,80 were identified. The number of included primary studies in the reviews varied from 3 to 18, and the AMSTAR scores of the reviews ranged from 2 to 7. The review by Ueda 201382 received the highest score (AMSTAR=7) and included nine randomised trials and nine controlled clinical trials that evaluated one music-related experience or a combination of music-related experiences, such as singing, listening, performing, rhythmic exercising and improvising. Uncontrolled before-and-after studies and case studies were excluded. Participants were allocated to music therapy (mean of 36 min/day, 2–3 days/week for 10 weeks (range 1 day to 11 months)) or usual care for BPSD assessment. The music therapy comprised listening,86–93 moving/dancing,86 88 89 94–97 singing/playing a musical instrument86 88 89 92 93 95 96 98–102 and in some occasion was administered in combination with exercise103 and reminiscence therapy.89 99 101 Music therapy was effective in reducing behavioural symptoms (6 RCTs+5 CTs; 397 participants) (SMD=−0.49 (95% CI −0.82 to −0.17)), despite a moderate and statistically significant heterogeneity (I2=58%, p=0.009). The same intervention achieved a statistically significant reduction on depression (4 RCTs+5 CTs; 250 participants) (SMD=−0.32 (95% CI −0.68 to −0.04); I2=44%, p=0.08) and anxiety (SMD −0.64, 95% CI −1.05 to −0.24; I2=55%; eight studies; 258 participants). Whear et al85 investigated the effectiveness of mealtime interventions, including music, on BPSD in people with dementia in residential nursing homes or care homes. Eleven studies were identified: one controlled trial, three before/after studies and seven repeated measure time series studies. The results of the studies were described narratively. One before/after study with 22 participants found that music played at mealtime improved physical and verbal, aggressive and non-aggressive, behaviour using the CMAI. Seitz et al29 (AMSTAR=6) identified 40 RCTs of non-pharmacological interventions, of which 3 studies with 133 participants96 97 104 evaluated music therapy for BPSD of dementia in LTC facilities. Owing to the heterogeneity of the studies (study design, patient populations, interventions, treatment duration and outcomes measured), the authors did not perform a meta-analysis. The behavioural outcome was measured either with a modified CMAI, Behavioural Pathology in Alzheimer's Disease Rating Scale (BEHAVE-AD) or the NPI. In one study, the music therapy was performed with movement, in a group, for 30 min, twice/week for 4 weeks.97 In a second study, the music intervention lasted 30 min, 3 times/week for 6 weeks.104 And in a third trial, the duration and frequency of individual sessions were not specified, but the therapy lasted 14 weeks.96 Two of the three studies employing music found a statistically significant difference between treatment and control groups, but all three were at risk of randomisation bias and two had unclear bias of incomplete outcome data. All the studies were included in Ueda's review.82 The review by McDermott et al81 (AMSTAR=4) searched MEDLINE, EMBASE, PsycINFO, CINAHL, the Cochrane Library, Web of Science, Journal of Music Therapy and Nordic Journal of Music Therapy and identified 18 studies of which 6 were RCTs (the remaining were non-randomised controlled studies (n=4), before-and-after studies (n=5) and qualitative and mixed-method studies (n=3)). Two trials96 100 and the case–control study104 were already included in the reviews described above.29 82 Three RCTs (n=165), two of which were carried out by the same group, measured BPSD using either the NPI or BEHAVE-AD. In one trial, the music therapy (patients and music therapist play musical instruments to express emotions and interact) was performed for 30 min, 3 times/week for 1-month, followed by a 1-month interruption, over 6 months (Raglio 2010). In another study by the same group, the music therapy (singing and body movement with music to stimulate communication) was administered for 30 min, 30 times over 16 weeks.96 In the third trial, the therapy was executed for 30 min, 3 times/week for 6 weeks (Svansdottir 2006). McDermott et al concluded that evidence for reduction of behavioural disturbance was consistent, but there were no high-quality longitudinal studies that demonstrated long-term benefits of music therapy. Of note, five of the RCTs included in the review were not included in the review by Ueda et al.82 Unlike the previous review, Vasionyt˙e and Madison83 (AMSTAR=4) provided a meta-analysis of the effects of music interventions (median=8 weeks; range 2–53weeks) in patients with dementia, differentiating between different types of interventions (listening, active music therapy, recorded music, live music, selected music, individualised music, classical/relaxation music, popular/native music and group and individual interventions). This SR included 18 studies comprised of 6 RCTs,87 90 105–108 6 CCTs 88 91 98 102 109 and 6 pre–post-test studies. The outcomes evaluated were behaviour (measured with the CMAI, NPI-Q, Multidimensional Observation Scale for Elderly Participants (MOSES), an agitation checklist or a behavioural chart), affect, cognition and physiology. There was no statistically significant effect on behaviour (effect size (ES) 1.16, 95% CI −0.65 to 2.98; 8 studies, n=217) or affect (ES 0.38, 95% CI −0.56 to 1.32; 6 studies, n=109), while cognition (ES 1.56, 95% CI 1.11 to 2.01; 4 studies, n=63) and physiology (ES 0.72, 95% CI 0.36 to 1.08; 4 studies, n=88) were affected. Three of the RCTs, and four of the controlled trials, in this review, were also included in Ueda et al.82 The review by Wall and Duffy84 included 13 studies that were presented narratively. The review was of low quality (AMSTAR score 2). The review by Chatterton et al80 evaluated the efficacy of ‘live’ singing to people with dementia for cognitive, behavioural, physiological and social outcomes. The study received an AMSTAR score of 1. An additional SR72 that aimed to assess the role of the physical environment in supporting person-centred dining in LTC identified four non-randomised studies, with different designs, that evaluated the effect of music on the incidence of agitated behaviours during mealtimes, among older adults with dementia, residing in special care units (SCUs).110–113 The results of these studies showed that playing music during mealtime reduced the incidence of agitated behaviour. Online supplementary etable 5a describes the type of interventions, the outcomes and the results of the primary studies included in the music therapy reviews. Dance therapy Two reviews evaluated dance therapy in patients with dementia.114 115 The first review's objective was to evaluate the evidence concerning dancing interventions in physical and mental illnesses compared to other types of interventions or non-specific interventions.115 The review received 3 points in the AMSTAR scoring system and identified 13 small studies reporting results from 11 randomised trials of which only one considered patients with dementia. The trial that considered participants with dementia included 29 participants (mean age 79 years, SD 7.7; 75% women) in a nursing home and evaluated the efficacy of dance and movement therapy delivered in nine sessions, lasting 30 to 45 min each, once-a-week.116 The outcome measures included the word list savings score, the Clock drawing test (for visual spatial ability), the Cookie Theft picture description task from the Boston Diagnostic Aphasia test and the Nurses' Observation Scale for Geriatric Patients (NOSGER). The results did not show any important differences in favour of dance therapy. The second SR aimed to evaluate the effects of dance (movement) therapy and ballroom dancing, compared to usual care, for adults with physical and mental illnesses.114 The review received only one point on the AMSTAR Scale and identified only one study that investigated the intervention in a population affected by dementia116 and which was also included in the review above. Snoezelen multisensory stimulation therapy Snoezelen multisensory stimulation therapy (SMST) comprises multiple stimuli and is aimed at stimulating the primary senses of sight, hearing, touch, taste and smell. The intervention is provided in specially designed rooms, which provide diverse sensory-stimulating effects/material including music, aroma, bubble tubes, fibre optic sprays and moving shapes projected across walls. SMST was investigated by two reviews.29 31 The first was an overview of reviews,31 and its evidence for SMST was based on a Cochrane review that included three studies.117 The inclusion criterion was any randomised trial that assessed the efficacy of SMST and/or multisensory stimulation to treat people over 60 years of age suffering from dementia. The outcomes of interest included behaviour, mood, cognition, physiological indices and client–carer communication, as well as short-term effects measured during the sessions or postsession, and longer term benefits measured postintervention and at follow-up. The three included primary studies evaluated a total of 311 patients with dementia, aged 60 or older. The first was a randomised trial118 that compared eight standardised multisensory programmes with eight standardised activity sessions. Both programmes were implemented on a one-to-one basis, twice-a-week, with each session lasting 30 min. Fifty participants (25 women, mean age 78) with a diagnosis of AD (N=33), vascular dementia (N=7) or a mixed diagnosis (N=10) were enrolled. The objectives of the trial were the immediate effects of SMST on the behaviours of older people with dementia, the carryover effects of SMST on mood and behaviour to day-hospitals and home environments and the maintenance effects of SMST on mood, behaviour and cognition over time. The effects of SMST on behaviour were measured by INTERACT.119 The generalisation effects were measured by three outcome measures: the carryover effect of day-hospitals was measured with the General Behaviour and Community Skills subscales of REHAB (Baker 1988); the carryover effect to home, at midintervention and postintervention, was measured with the Behaviour and Mood Disturbance Scale (BMD) and the Behaviour Rating Scale (BRS) of the Clifton Assessment Procedures for the Elderly (CAPE). The maintenance effect (at the 1-month postintervention follow-up) on behaviours and cognition were measured by REHAB, BMD, the Cognitive Assessment Scale (CAS) of CAPE and Mini-Mental State Examination (MMSE). No significant effects on any scale of behavioural symptoms were found either immediately after intervention or at 1-month follow-up. The second study120 was a quasi-experimental pre-test and post-test design with cluster randomisation performed at a ward level, which compared a 15-month, 24-hour individualised care plan that was integrated with SMST, with 15-month usual care. The study included 136 participants diagnosed with AD, vascular or mixed dementia from three different countries (UK=94 day patients, the Dutch sample=26 inpatients. Swedish sample=16 inpatients). There was a significant group difference in the mean baseline MMSE scores (data from the UK and the Dutch only) between the SMST group (9.4) and the control group (6.7) (p=0.01). All participants attended eight, 30-min sessions on a one-to-one basis according to their group assignment. The sessions were conducted by the same key workers throughout the study period. The following outcomes measured the short-term effects of SMST on behaviours: (1) INTERACT (22-item) measured behaviours during the sessions; (2) INTERACT (12-item) measured behaviours 10 min before and 10 min after the sessions and (3) Behaviour Observation Scale for Intra-mural Psycho-Geriatrics (GIP) measured behaviours that were videotaped during the sessions in the Netherlands sample. The study showed significant effects on two behavioural items of INTERACT during sessions: enjoying oneself (MD=−0.74; 95% CI (−1.29 to −0.19); z=2.62, p=0.01) and bored/inactive (MD=−0.56; 95% CI (−1.11 to −0.01); z=1.99, p=0.05). There were no longer term treatment effects of the integrated SMST-care programme on behaviour. The third study121 122 assessed the effects of SMST when integrated into 24-hour daily care on nursing home residents with dementia. A total of 125 patients with moderate or severe dementia and care dependency were recruited from six old age psychiatry wards for pre-test. A cluster randomised design was used to assign the wards to either experimental (integrated SMST-care programme) or control (usual activity) conditions. Twelve old age psychiatry wards in six nursing homes (out of 19 homes) were recruited to the study. At baseline, 125 participants (woman 79%, mean age 84) were recruited and were assigned to experimental or control conditions according to the ward in which they stayed. For the experimental group, participants were given a stimulus-preference screening in 10 weekly one-hour sessions to identify their preferred sensory stimuli. Subsequently, individual SMST-care plans were developed for each participant based on their life history, stimulus preference and discussions from multidisciplinary conferences. Certified nursing assistants (CNAs) used multisensory stimuli in the 24-hour care of the experimental participants. Participants in the control group were provided with individual usual care. A minimum period of 3 months was used for experimental and control conditions. The short-term effects of the integrated SMST-care programme on behaviours were measured using a modified version of INTERACT, in which six items were deleted and eight new items were added during morning care sessions. The long-term effects of integrated SMST-care programmes on behaviours, mood and interaction were evaluated at the 18-month follow-up using the eight items of GIP for apathy, anxiety and disoriented behaviours, the Dutch version of CMAI for agitated behaviours, physically non-aggressive behaviour and verbally agitated behaviours and the Cornell Scale for Depression for depressive symptoms. In terms of behavioural disturbances, when compared to the control, the 24-hour integrated SMST-care programme122 showed a significant effect on two behavioural items of INTERACT during sessions: enjoying self (MD=−0.74; 95% CI (−1.29 to −0.19); z=2.62, p=0.01) and bored/inactive (MD=−0.56; 95% CI (−1.11 to −0.01); z=1.99, p=0.05). There were no longer term treatment effects of the integrated SMST-care programme on behaviour. In terms of mood, there were significant improvements in one mood item of INTREACT during sessions: the SMST group was happier and more content than the control group (MD=−0.84; 95% CI (−1.39 to −0.29); z=2.98, p=0.003). There were no significant effects of the 24-hour integrated SMST at postintervention. The fourth review scored 6 in the AMSTAR evaluation and investigated different non-pharmacological interventions including SMST for the treatment of BPSD.29 The review identified only one study that was included in the above cited review.122 Online supplementary etable 6 describes SMST-based interventions, outcomes and results of the primary studies included in the reviews. Transcutaneous electrical nerve stimulation TENS is a simple, non-invasive, non-pharmacological intervention commonly used for pain control123 and occasionally for neurological and psychiatric conditions such as drug/alcohol dependency, headaches and depression.31 TENS consists of attaching electrodes to the skin and applying an electrical current, whose frequency can vary from low (<10 Hz) to high (>50 Hz). Two reviews were identified. One review that evaluated current treatment options for sleep disturbance in AD scored 3 in the AMSTAR evaluation.124 Different non-pharmacological interventions were considered, including bright light therapy, behavioural and multifaceted interventions (combined increased daytime physical activity and exercise, decreased daytime in-bed time, daily sunlight exposure, structured bedtime routine and decreased night-time noise and light) and TENS. For the latter intervention, only one randomised trial of 19 nursing home residents was identified. The study did not evaluate behavioural outcomes. A Cochrane review that was included in O'Neil's review31 was also considered.125 The review was focused only on RCTs that enrolled inpatients and outpatients of any age (with or without caregivers), with a diagnosis of dementia. The outcomes of interest included visual and verbal short-term and long-term memory, semantic verbal fluency, circadian rest-activity rhythm, affect/depression, level of independent functioning, adverse effects and dropouts due to inefficacy. The review identified and included nine trials that were performed in Japan and the Netherlands. The Dutch studies were performed by the same group of authors.126–131 These studies were randomised placebo-controlled trials, and the participants were chosen from a group of 350–500 residents of a residential home for older people. The age range of the participants was ∼70 to mid-90 years and were mostly women (>80%). All participants met NINCDS-ADRDA criteria for the clinical diagnosis of probable AD; most participants had early AD, but some had moderate AD. Participants generally had scores of 17 or less on the Hamilton Depression Rating Scale. All included studies used a similar TENS protocol, except the most recent one published in 2002, which addressed cranial electrostimulation. The remaining three publications were performed by a group of authors from Japan and describe the results of the same study.132 The study design was a double-blind crossover and, in contrast to the Dutch studies, participants were thought to have multiinfarct dementia or AD and were selected on the basis of irregular sleep-wake patterns in conjunction with nocturnal behaviour disorders and/or dementia. Twenty-seven participants completed the study. The intervention used a HESS-10 stimulator with rectangular pulse waveforms at a frequency of 6–80 Hz, a pulse duration of 0.2 ms maximum, 256 µAmps and an amplitude of 6–8 V. The outcomes evaluated were sleep disorder, motivation, behaviour disorder, intelligence, emotion, language, neurological signs, subjective symptoms and activities of daily life. All of these were rated on a five-point scale: absence of the related symptom, 0; mildly disturbed, 1; moderately disturbed, 2; markedly disturbed, 3 and severely disturbed, 4. Of the nine studies, only three could be included in a meta-analysis for a combined total of 63 participants. Two of these studies were conducted in the Netherlands, and one was conducted in Japan. Results, however, were inconclusive. It should be noted that none of the other studies mentioned adverse effects, although it is unclear if adverse events were monitored. Online supplementary etable 7 describes TENS-based interventions, outcomes and results of the primary studies included in the reviews. Cognitive/emotion-oriented interventions Cognitive stimulation Cognitive stimulation involves a variety of pleasurable activities, such as word games, puzzles, music, cooking, gardening and discussing past and present events, and is usually carried out by trained personnel with small groups of four to five people. It lasts for 45 min, minimally 2 times/week. It is based on Reality Orientation, which was developed in the 1950s to counteract the confusion and disorientation of older people during hospitalisations. Seven reviews were identified.133–139 Woods 2012133 (AMSTAR score=10) was a Cochrane review that identified 15 RCTs that used cognitive stimulation for people with dementia. The authors stated that most of the studies were of low quality, but that generally, investigators had taken measures to protect against the risk of allocation concealment bias. In a meta-analysis of three trials140–142 (n=190 participants), the intervention had no effect on problem behaviours (SMD −0.14, 95% CI −0.44 to 0.17; I2=0%, p=0.57). The review by Aguirre et al in 2013134 (AMSTAR score=5) evaluated the effectiveness of cognitive stimulation in patients with dementia and identified nine RCTs. Three trials that considered behaviour-related outcomes were identified. These trials were already included in Woods's review133 and reached the same conclusion. Alves et al in 2013135 (AMSTAR score=4) identified four RCTs of cognitive interventions for AD patients. Only one trial that measured BPSD as an outcome was identified. The study population was composed of 32 patients with a score between 10 and 24 on the Mini Mental State Examination, no history of antidepressant medication and a total NPI score >5 points arising from at least 2 domains of behaviour. The cognitive stimulation intervention was administered individually and focused on a set of tasks requiring executive functions and working memory. The study found a statistically significant reduction of BPSD (MD −2.06; 95% CI −2.91 to −1.21). The study of Carrion et al in 2013136 (AMSTAR score=4) found 17 RCTs of cognition-oriented interventions (reality orientation and skills training) for dementia sufferers. Challenging behaviour was evaluated in only two trials (n=156 and n=44, respectively) that employed the two categories of cognitive interventions, using the NPI and the Revised Memory and Behaviour Problems Checklist. In both RCTs, the intervention group had a smaller increase in change from baseline compared to the control group. Owing to the heterogeneity among the studies, the authors decided a meta-analysis was inappropriate. Yu 2009137 (AMSTAR=3) included 15 studies (9 RCTs, 5 CCTs and 1 before–after study), in addition to 5 case studies and 3 undefined studies, all of which investigated different types of cognitive interventions for AD and dementia. The only study, a CCT (n=32 with early-stage AD), that evaluated the effect of cognitive stimulation on behavioural disturbances, showed larger improvement than the cognitive training group. Olazarán et al in 2010138 (AMSTAR=4) identified 179 RCTs of diverse types of non-pharmacological interventions for AD patients and examined problem behaviour, mood, QoL, cognition, ADLs, mechanical restraint and institutionalisation of patients and mood, psychological well-being and QoL of CGs. The authors performed a meta-analysis of three low-quality RCTs to determine the effect of cognitive stimulation on problem behaviour and mood. There was a non-statistically significant reduction in problem behaviour (group session cognitive stimulation (ES=0.61; 95% CI 0.09 to 1.12)). The primary study by Baines et al140 was included in the Woods133 review above, while the study by Robichaud et al143 was included in the review by Kim144 which examined behaviour management techniques described below. Thirty-three RCTs, employing cognitive interventions for cognitively impaired individuals (dementia and mild cognitive impairment), were identified in Kurz et al145 (AMSTAR score=2). Twelve of these trials examined behavioural disturbances, but only three studies found a significant effect of the intervention. Zientz et al139 (AMSTAR score=2) identified three studies (two RCTs and one RCT or CCT; n=124 participants) of caregiver-administered cognitive stimulation for individuals with AD. One of the randomised trials (n=16) found that individuals who received the intervention displayed fewer behavioural problems compared to those who had not been given the intervention. Online supplementary etable 8 describes cognitive stimulation-based interventions, outcomes and results of the primary studies included in the reviews. Reminiscence therapy Reminiscence therapy is a non-pharmacological intervention that involves the discussion of past experiences, events and activities with family members or other groups of people. The intervention uses materials such as photographs, books, old newspapers and familiar items from the past to inspire reminiscences and facilitate people to share and value their experiences. Three reviews assessed reminiscence therapy as a non-pharmacological intervention to treat agitated behaviour in patients with dementia.29 146 147 The first review29 received the highest score (AMSTAR score of 6) and considered all non-pharmacological interventions to treat relevant outcomes in patients with dementia. The review identified two small studies involving a total of 107 patients148 149 performed in care facilities. The NPI and the Clifton Assessment Procedures for the Elderly-Behavioural Rating Scale (CAPE-BRS) were used to measure BPSD. Seitz et al29 reported that this outcome was unaffected in one study,149 while the effect of the intervention was unclear in the other study.148 The second review147 was focused only on reminiscence therapy as a sole treatment of behavioural outcomes for patients with dementia. The review was of low methodological quality (AMSTAR score=3). The results were presented in a narrative synthesis. The review included five trials with a before–after design, containing 258 patients affected by dementia. The studies considered different interventions. Two studies (one with 31 participants (Haight 2006)150 and the other with 17 participants (Morgan 2010151)) assessed a life review or story approach and found significant improvements in depression, communication, positive mood and cognition. The third study (101 participants (Lai 2004152)) evaluated specific reminiscence, which produced a life-story book using personalised triggers for each person's life history. No significant differences were observed between groups except for outcomes such as well-being and social engagement. The remaining two trials (involving 73 participants153) and 36 participants154) evaluated individual reminiscence approaches. One study used six weekly sessions, which focused on a particular life phase, such as childhood or family life, while the other study used a basket of visual and auditory activities, based on five themes, such as musical instruments, designed to stimulate reminiscence. No significant differences were observed between the groups in terms of behavioural outcomes. The third review146 focused on whether reminiscence therapy could alleviate depressive symptoms in adults with dementia, but its methodological quality was extremely low (AMSTAR score=1). Four primary studies with a pre–post-test design were included and were described individually, three of which were randomised trials and one of which comprised a single group. Online supplementary etable 9 describes reminiscence therapy interventions, outcomes and results of the primary studies included in the reviews. Validation therapy Validation therapy is based on the general principle of the acceptance of the reality and personal truth of another person's experience and incorporates a range of specific techniques. Validation therapy is intended to give the individual an opportunity to resolve unfinished conflicts by encouraging and validating the expression of feelings. The specific interventions and techniques are based on a synthesis of behavioural and psychotherapeutic methods. The approach can be used as a structured therapeutic activity in a group setting, usually lasting several weeks, or it can be conducted individually as part of an ongoing approach to facilitate communication as a supplement to group work. The validation therapy techniques comprised non-threatening, simple concrete words; speaking in a clear, low and empathic tone of voice; rephrasing and paraphrasing unclear verbal communication; responding to meanings through explicit and implicit verbal and non-verbal communication and mirroring verbal and non-verbal communication. One Cochrane review that evaluated the effectiveness of validation therapy to reduce BPSD was identified (AMSTAR score=7).155 The review included only randomised trials of participants over 65 years of age, diagnosed with Alzheimer's disease, dementia or other forms of cognitive impairment, according to ICD 10, DSM IV or comparable criteria. The outcomes of interest were cognition, behaviour, emotional state and activities of daily living. The review, updated in 2005, included three randomised trials (n=155 participants).156–158 Another SR159 that evaluated the effective characteristics of residential LTC settings for people with dementia identified one trial158 that was included in the Cochrane review.155 Primary studies Among the primary studies, the first study156 (n=31) was performed in a nursing home and used an intervention (30 min once-per-week for 6 weeks) that included activities such as discussion of a previously agreed subject, singing and movement, followed by a closing ritual and refreshments. Behaviour was measured with the Behaviour Assessment Tool. The control groups consisted of reminiscence therapy, which followed the guidance of a reality orientation manual (cues such as flannel boards and calendars were used to promote orientation) and usual care. At 6 weeks, validation therapy was associated with a decrease of problem behaviours (MD=−5.97, 95% CI −9.43 to −2.51; p<0.001; based on an analysis of participants who completed the study). The second study157 enrolled 36 patients with moderate-to-severe disorientation of which 25 had a diagnosis of dementia. The study was performed in a LTC institution in the USA. The validation therapy was performed twice-a-week for 9 months; details of the validation therapy were not given. Agitation was measured using the Minimal Social Behaviour Scale (MSBS; Farina 1957) where a reduction in score indicated improvement. No effects on behaviour were detected. The last study158 was carried out in ‘skilled-care nursing homes’ in the USA. In this study, patients were included if they had at least a moderate level of dementia (assessed by the Short Portable Mental Status Questionnaire—SPMSQ—and the Validation Screening Instrument) and displayed problem behaviours, such as physical aggression. Validation therapy (four meetings lasting 30 min per week for 52 weeks) was composed of groups divided into four sessions of 5–10 min each. The first session included introductions, salutations and singing. The second session involved conversation regarding a subject of interest; recalling past events was promoted. The third session comprised an activity programme and singing or poetry. The fourth session involved refreshments and individual goodbyes. Agitation was measured with the CMAI,160 carried out as CMAI(N) nurse observed and CMAI(O) non-participant observed. The authors reported that depression (MOSES) decreased at 12 months (MD −4.01, 95% CI −7.74 to − 0.28; p=0.04, based on an analysis of participants (66 out of 88) who completed the study. Online supplementary etable 10 describes validation therapy interventions, outcomes and results of the primary studies included in the reviews. Simulated presence therapy SPT involves the use of video/audiotapes made by family members containing scripted ‘telephone conversations’ about cherished memories from earlier parts of a person's life, in an effort to stir remote memory, improve behavioural symptoms and enhance the quality of life among people with dementia.161 Two SRs were identified.30 162 The first review was written by only one reviewer and scored 3 on the AMSTAR Scale. The review was aimed at investigating the effectiveness of SPT for challenging behaviours in dementia. The review searched PubMed, PsycINFO and the Web of Science, conducted hand searches of relevant articles and considered for inclusion, studies that reported pre-test and post-test, or pre-test and during-test data for SPT for challenging behaviours. The SPT consisted of audio or videotapes prepared by a spouse, family members, the caregiver, a psychologist, a surrogate or researchers. Of the seven included primary studies, only the data from four could be pooled, showing an overall mean effect of 0.70, with a 95% CI of 0.38 to 1.02, but with statistically significant heterogeneity (I2=71%, p=0.02). The second review examined the efficacy of any non-pharmacological intervention (including SPT) to reduce BPSD in patients with dementia.162 After searching the databases MEDLINE, CINAHL, PsycINFO, EMBASE, Dissertations International and the Cochrane Database of Systematic Review, from 1974 to May 2008, the review identified only two studies that were included in the Zetteler review above.30 Online supplementary etable 11 describes SPT, outcomes and results of the primary studies included in the reviews. Behavioural management techniques There is a multitude of behavioural interventions that constitute behavioural management techniques, which include behavioural or cognitive–behavioural therapy, functional analysis of specific behaviour, individualised behavioural reinforcement strategies, communication training and other therapies such as habit training, progressive muscle relaxation and token economies. These behavioural interventions can be realised either with the patient or by training caregivers to perform the intervention with the patient. One overview of reviews and four SRs that considered behavioural interventions were identified. The overview of reviews by O'Neil 201131 identified three SRs, and after performing additional searches of primary studies, included nine randomised trials.163–171 The overview authors' conclusions were in support of behavioural management techniques as effective interventions for behavioural symptoms of dementia although they admitted there were mixed results. In addition, the authors highlighted some concerns regarding the variety of specific interventions and methodological limitations in many studies and advocated additional research with carefully assessed outcomes. A Health Technology Assessment (HTA)172 report that aimed to evaluate the clinical and cost-effectiveness of sensory, psychological and behavioural interventions to manage agitation in older adults with dementia, systematically searched and identified four randomised trials.166 173–175 The intervention in all four trials was caregiver-based. The HTA authors concluded that the evidence in favour of the behavioural management techniques was limited. A Cochrane review176 aimed to assess the effects of functional analysis-based interventions for people with dementia (and their caregivers) living in their own home or other settings and identified 18 randomised trials. The development of the intervention was driven by various approaches and theories, including knowledge and/or training approaches, the stress-coping model, the progressively lowered stress threshold model and problem-solving approaches. In addition, the time frame in which the intervention was delivered varied from 9 days to 18 months and the number of sessions used to deliver the intervention varied widely, from 1 to 2 sessions to more than 10 sessions. Of the 18 studies included,164–168 173 174 177–187 the authors were able to meta-analyse data from 4 trials,178 180 182 188 of which one contained unpublished data. There were no significant reductions in the incidence of challenging behaviours reported postintervention in four family care studies (SMD 0.02, 95% CI −0.13 to 0.17, p=0.80, N=722). Among 179 RCTs of diverse types of non-pharmacological interventions for AD patients, identified by Olazarán 2010138 (AMSTAR score=4), the authors performed a meta-analysis of three low quality RCTs of behavioural interventions (analysis and modification of antecedents and consequences of behaviour) and found a statistically significant reduction in problem behaviour (ES=0.57, 95% CI 0.21 to 0.92; 3 trials; n=167). The same authors carried out another meta-analysis of four low-quality RCTs of care staff training in behavioural management and found a reduction in problem behaviour (ES=0.22, 95% CI 0.02 to 0.43; 4 trials; n=370). Two primary studies examined emotion-oriented care. The first study189 was a RCT of NH residents (n=146 older residents with AD, mixed AD and vascular dementia and dementia syndrome; mean age 84). The intervention of emotion-oriented care was associated with less anxious behaviour in the group of residents who needed less assistance/care compared to similar residents in the usual care group. The second study190 was a cluster randomised study of residential care homes (n=16 homes; n=151 residents). The authors reported that there was no statistically significant effect of the intervention on any behavioural outcome, including behavioural problems. Teri 2000174 was included in the HTA;172 Gormley 2001173 and Teri 2005166 were included in the Brodaty 2012 review;191 Gonyea 2006183 was included in reviews in behaviour management techniques and McCallion 1999 and Teri 2005 were included in Eggenberger 2013.192 Eggenberger 2013192 (AMSTAR score=3) aimed to evaluate interventions that were designed to enhance communication or interaction in dementia care, in any setting. Review authors identified 12 studies (7 randomised trials, 2 controlled clinical trials and 3 before–after studies) that focused on communication training for staff in institutions and family caregivers at home. In institutional settings, the results on challenging behaviour, of residents with dementia, were not consistent. Four studies reported a significant reduction of challenging behaviour.171 193 194 McCallion et al,171 for instance, demonstrated a decrease of physically aggressive behaviour (15.16 (SD 9.81) to 12.21 (SD 8.31), p<0.001)) and a reduced mean occurrence of verbally aggressive behaviour in patients with dementia (16.22 (SD 10.31) to 12.88 (SD 8.39), p<0.001)). In addition, one trial demonstrated a significant decrease of residents' agitation during care routines (F(1.7=5.12, p<0.05)).194 Conversely, three studies reported no effect on challenging behaviour of people with dementia.195–197 Only one trial167 was included in the Brodaty 2012 review.191 Kim et al144 conducted a review to assess the effectiveness of occupational therapy on behavioural problems and depression in patients with dementia. MEDLINE, CINAHL, ProQuest and The Cochrane Library were searched up to the end of March 2011. The AMSTAR score was 7. The authors defined occupational therapy as an application of ‘activity analysis, caregiver training, sensory stimulation, behaviour control skill teaching, physical and social environmental modification, cognitive training, and purposeful activity’. The review identified nine randomised trials with a total of 751 participants. On the basis of the type of intervention, the authors categorised four studies118 120 143 198 as sensory stimulation, three studies198–200 as functional task activities and two studies200 201 as environmental modification. The authors performed a meta-analysis of the trials with occupational therapy-based sensory stimulation and found an ES of 0.32 (95% CI 0.04 to 0.59; 250 participants; no significant heterogeneity). No significant effect was detected for OT-based functional task activities (0.15, 95% CI −0.17 to 0.47; 203 participants) or environmental modification (0.13, 95% CI −0.09 to 0.36; 298 participants). Primary studies Overall 22 trials were evaluated in the 6 reviews that were included. Except for one study performed in Taiwan, all the studies were carried out in Europe, the USA and Australia. Thirteen studies were performed in family care settings.164 167 168 173 174 178 180 182–187 Three studies with a total of 740 residents were conducted in care homes.165 179 181 Finally, one study was located in an assisted living setting166 and the other in a hospital setting.177 Characteristics of the interventions varied greatly across the trials. Fifteen trials were focused on enhancing communication skills in family and formal caregivers. Eighteen trials focused on functional activity of which four were described as a behavioural management intervention. The intervention in one trial involved caregiver training on verbal or non-verbal communication focused on activities of daily living. Another trial was dedicated to teaching participants the basic technique for progressive muscle relaxation.170 Time delivery of the intervention also varied widely. However, as noted by Moniz-Cook, the intervention delivery was determined by setting: the interventions in care homes were provided weekly and lasted for 6 months.176 In one family care study, the intervention was provided in just 4 sessions over 8 weeks.173 Follow-up data varied from a few weeks to 24 months. Setting-based description Family care: In this setting, family caregivers assisted people with dementia at home, with or without support from formal caregivers, healthcare workers and adult day care centres. Thirteen trials were conducted in a family care setting.164 166 168 173 174 176 178 180 182–185 187 200 201 Six of these trials investigated an intervention that was focused on enhancing communication skills of the caregiver. The duration of the intervention ranged from 3 weeks184 to 12 months.178 The number of weekly sessions administered were, according to a classification proposed by Moniz-Cook 2012,176 high (>10 session) in three trials,164 178 180 185 moderate-high (6–10 sessions) in one trial,185 moderate (3 to 5 sessions) in one trial182 and minimal (1 to 2 sessions) in one trial.184 The participants who delivered the interventions varied from trial to trial: occupational therapists;182 trained nurses or social workers;180 professionals specialised in the REACH programme;178 healthcare professionals supervised by an old age psychologist;164 psychologists185 or trial investigator together with an experienced nurse.184 Of the 13 trials in the family care setting, 4 investigated a behavioural intervention that was focused on providing support to the caregiver. The interventions lasted from 5 weeks183 to 18 months,186 with the number of sessions that varied from 4186 to 8 sessions,187 with home visits176 187 and associated with or followed by telephone contacts.166 168 Overall, the intervention dosage was high for three trials,166 168 176 medium-high in one trial187 and moderate in one trial.183 The interventions were delivered by different healthcare experts: community mental health nurses;176 therapists;187 occupational therapists;168 community consultants trained by an old age psychologist.166 The remaining two trials evaluated behavioural management techniques. Teri 2000174 compared the intervention consisting of eight weekly and three biweekly sessions (high-intensity intervention) with pharmacological interventions or placebo. The intervention was provided by a therapist with a master's degree and 1-year clinical experience, but was not reported in detail. The postintervention evaluation started at 4 months, and the follow-up lasted beyond 12 months. The second study173 did not completely describe the intervention for behavioural management. The intervention was delivered in four sessions (moderate intensity) over 8 weeks by the trial investigator. In terms of results, no statistically significant change in the incidence of challenging behaviours was observed in any of the studies. Moniz-Cook 2012 meta-analysed data of four studies (N=722), but did not find any difference among the groups (SMD 0.02, 95% CI −0.13 to 0.17, p=0.80; I2=0%).168 178 180 182 At follow-up of 6 months, two studies did not show any significant effect of behavioural management techniques.168 180 When the frequency of challenging behaviours was examined, none of the studies detected a significant difference even when a meta-analysis, using the data from 10 studies, was performed (SMD −0.05, 95% CI −0.17 to 0.07). Assisted living: In this setting, people with dementia lived in a residence, did not require full-time nursing care, but needed assistance with some ADLs, such as bathing, dressing and eating. Family members could still act as intermittent caregivers during visits by providing different types of support for ADLs, instrumental ADLs (eg, laundry washing, room cleaning, transportation to a doctor's office), socioemotional support (eg, talking, reminiscing, socialising), monitoring care provision or advocating.202 One study evaluated a behavioural management intervention to improve caregiver training to manage residents with dementia.166 The intervention intensity was medium-high, delivered by a clinical psychologist and graduate nursing students who performed two half-day group workshops and four individualised sessions with a follow-up 2 months after the termination of the intervention. Results for residents showed a statistically significant effect, in intent-to-treat analyses, in favour of the STAR-caregivers (STAR-C) intervention, general behavioural disturbance (measured by the Revised Memory and Behaviour Problems Checklist (RMBPC), NPI and ABID) and depression. Residential care: This setting referred to assisted living residences and nursing homes. The latter included facilities for people with dementia who needed significant nursing care. Three cluster randomised trials were conducted in residential care with a total of 743 residents.165 179 181 In 15 residential care sites across metropolitan areas in Sydney (Australia), Chenoweth et al179 examined the efficacy of person-centred care versus usual care. The intervention was a high-intensity, person-centred care, based on the needs-driven model in which staff, selected by managers, administered training sessions to caregivers. The topics covered during the sessions were derived from Bradford University's training manual. The duration of the intervention was 4 months, and the overall follow-up was 8 months. The total number of residents enrolled was 289. During follow-up, the mean agitation score (measured with the CMAI) in the person-centred care group decreased significantly, from 47.5 (9.1) at baseline, to 37.2 (9.1) at 6 months (p=0.01), compared to usual care in which agitation increased from 50.3 (6.8) at baseline to 57.7 (6.8) at 6 months (p value not reported). In 12 residential homes, Fossey 2006181 allocated 346 residents to an intervention that consisted of training and support delivered to nursing home staff over 10 months, focusing on person-centred care and skill development for the management of agitated behaviour in dementia. The comparison intervention was usual care. The high-intensity intervention was delivered during the whole period of follow-up (12 months) by a psychologist, an occupational therapist or a nurse supervised weekly by the trial investigators. The study's main outcome measure was mean levels of agitated and disruptive behaviour measured with the CMAI, but no significant difference between the groups was detected. In 10 residential homes, Proctor 1999165 allocated 120 patients to a staff-based intervention or usual care. The intervention, of high-medium intensity, consisting of training on “psychosocial management of residents’” behavioural problems, was delivered through seven, 1-hour seminars by members of the hospital outreach team and psychiatric nurse during the whole period of follow-up (6 months). The seminars covered topics that the staff had identified to improve their knowledge and skills (eg, management of dementia, aggression, etc). The Crichton Royal Behavioural Rating Scale was used to assess behavioural characteristics of residents (0=no problems, 38=severe problems). In addition, the geriatric mental state schedule and the diagnostic algorithm AGECAT (Automatic Geriatric Examination for Computer-Assisted Taxonomy) were used to assess the effect of the intervention on residents' organic and depressive symptoms. Despite the control group having mean scores on the Crichton Scale higher than the intervention group at follow-up, this difference was not statistically significant (mean score −0.7 (−3.0 to 1.6)). Although the clustered trials reported different types of interventions, intensities, durations and follow-up times, Moniz-Cook et al176 attempted an analysis using two studies and found a significant reduction in behavioural disturbances (SMD, −0.21, 95% CI −0.39 to −0.03; p=0.02; I2=9%). Online supplementary etable 12 describes behavioural management technique-based interventions, outcomes and results of the primary studies included in the reviews. Multicomponent interventions Integrated interventions combining psychiatric and nursing home care Collet 2010203 (AMSTAR score=5) carried out a SR in MEDLINE, PsycINFO and PubMed to determine the efficacy of interventions that combined psychiatric and nursing home care in nursing home residents. The authors identified 4 RCTs (n=371 participants), 1 retrospective cohort study and 3 prospective case studies. All the studies used tailored treatment plans that combined psychosocial, nursing, medical and pharmacological interventions. The results of the RCTs were described narratively. Three out of the four randomised trials reported an improvement in behaviour and mood, while one trial found no difference among the groups (online supplementary etable 13). Combination of environmental sensory stimulation A SR204 that evaluated the effective characteristics of residential LTC settings for people with dementia identified one controlled clinical trial.205 The intervention in this trial was provided in five nursing homes and consisted of 15 agitated participants with dementia taking showers, 15 agitated participants with dementia taking walks in an environment where natural elements such as large bright pictures coordinated with audio, including bird songs, bird pictures, the sound of water flowing gently, as well as food (such as banana, pudding and soda). The control group consisted of 15 other agitated participants with dementia that received only usual care. Agitation was measured with a modified version of CMAI. The analysis showed a significant decline in agitation in the treatment group with respect to the comparison group. Combination of music and hand massage Another review72 that aimed to assess the role of physical environment in supporting person-centred dining in LTC identified another trial206 that was not included in the previous reviews. This trial applied an experimental 3×3 repeated measures design and included 41 residents with dementia living in three SCUs. Participants were mostly women (78.0%), with a mean age of 84.5 years (SD=6.0). Residents in the treatment group received each of three treatments (hand massage, favourite music and the combination of both) with each treatment lasting 10 min; the control group did not receive any treatment. The CMAI was used to measure agitation. The results showed that each single and combined treatment were effective in significantly decreasing agitation immediately following the intervention and one-hour postintervention. Online supplementary etable 13 describes multicomponent interventions, outcomes and results of the primary studies included in the reviews. Other interventions Exercise therapy The systematic search identified two reviews207 208 that evaluated the efficacy of only exercise as a therapeutic intervention. The review by Potter et al207 received 6 points in the AMSTAR assessment and identified 13 randomised trials that evaluated the effects of physical activity on physical functioning, quality of life and depression in older people with dementia. Only four of these trials investigated depression as an outcome using four different rating scales (Geriatric Depression Scale (GDS15); Montgomery-Asberg Depression Rating Scale (MADRS); a Dutch Evaluation scale for older patients (subscale used) and the CSDD) and two trials measured behavioural disturbances (NPI and Stockton Geriatric Rating Scale). The review authors stated that the methods of randomisation were clear and adequate in six of the trials with only three of these also providing methods of allocation concealment; eight of the trials reported information regarding losses to follow-up and six trials declared intention-to-treat analysis. The first study, Burgener 2008,209 was a small trial (n=43) carried out in community-dwelling older people with dementia. The intervention was multimodal comprising Tai Chi (sitting and standing; 60 min, 3 times-a-week for 40 weeks) and cognitive–behavioural therapies. Depressive symptoms were measured with the GDS15. The authors reported that at 20 weeks of observation, there were no statistical differences between the groups. The second study, Rolland 2007,210 was a larger trial (n=134) carried out in nursing homes. Participants performed exercises including stretching, walking, strength, flexibility and balance training for 60 min, 2 times/week for 40 weeks. Depression was evaluated using the MADRS. After 12 months of observation, the MADRS score (13.4±8.0) was higher in the intervention group than in the control group (14.8±7.2), but without any statistical difference. The third study103 was also a small study (n=25) conducted in a psychiatric hospital. The invention was composed of strength, balance and flexibility exercises with music, 30 min daily for 12 weeks. Depression was measured in older patients with the subscale Beoordelingsschaal voor Oudere Patienten. At 3 months follow-up, no significant difference in depressive behaviour was observed. The last study164 was a larger trial (n=153) that enrolled community-dwelling patients and their caregivers. The exercise intervention, for patients, comprised aerobic, endurance, strength, balance and flexibility training, 30 min twice weekly, reducing to twice monthly, for 23 weeks. Caregivers were given training in behavioural management techniques. The CSDD was used to assess depression. At a 2-year follow-up, the mean difference was 2.14 (95% CI 0.14 to 4.17) and statistically significant in favour of the intervention. The four trials used different types of interventions, outcome measures and follow-up times that hindered the possibility of performing meta-analyses. The two randomised trials103 210 that considered behavioural disturbances used the NPI and Stockton Geriatric Rating Scale, respectively. The second review by Thuné-Boyle et al208 received an AMSTAR score of 2 and included six studies comprising two small randomised trials (n=31), two prospective design and two repeated measures studies that examined the effect of exercise on BPSD. In the first trial (Hokkanen 2003), the exercise intervention consisted of 16 sessions of dance and rhythmic movement lasting 30–45 min, once-a-week. This trial was already discussed in the dance section. The second trial211 aimed to assess the efficacy of a home-based exercise intervention programme to improve the functional performance of patients with AD. The intervention consisted of a daily programme of aerobic, balance and flexibility and strength training, given to patients and caregivers. Depression and apathy were measured using NPI and the CSDD at 6 and 12 weeks. Online supplementary etable 14 describes exercise therapy, outcomes and results of the primary studies included in the reviews. Animal-assisted therapy One review212 performed a comprehensive literature search in PubMed, EMBASE and PsycINFO to identify pertinent studies that evaluated the efficacy of animal-assisted therapy (AAT) in older patients with dementia or other psychiatric disorders. The authors identified 23 eligible studies of which 18 recruited patients with dementia, but only 10 studies investigated the effect of AAT on BPSD. The design of the studies was as follows: 3 case–control and 7 repeated measures (eg, interrupted time series analysis) studies. Overall, the authors concluded that AAT may have positive influences on patients with dementia by reducing the degree of agitation and improving the amount and quality of social interaction. However, they advocated more research examining the issue of optimal AAI duration, frequency of sessions and suitable target group. Primary studies Churchill et al213 included 28 residents of three SCUs with dementia (25% women; mean age 83.8 years; dementia evaluated with Bourke Dementia Rating Scale). The authors administered pet-therapy visits during the difficult ‘sundown’ time to examine the effect on residents with a history of agitated ‘sundowning’ behaviour. The active group was exposed to 30-min interaction with an investigator and a dog, which ameliorated agitated/aggressive behaviour measured with the Agitated Behaviours Mapping Instrument Scale. However, the study did not report the p values. In addition, the variability in resident response over time after the departure of the dog was not explored. The effect of dog-based AAT was also evaluated in another special care unit. McCabe et al214 enrolled 22 participants with dementia (women 68%; mean age 83.7, range 68–96 years). The study introduced a resident dog and agitated behaviour was measured using the Nursing Home Behaviour Problem Scale. Data were collected 1 week before and for the first 4 weeks after introduction of the dog. The authors reported a significant reduction in daytime behavioural disturbances among residents, but not during evening shift. In a small pilot study, Richeson215 evaluated visiting therapy dogs in 15 residents with dementia (14 women; age range 63–99 years; dementia MMSE mean score: 3.9; 26% with depression). The session with visiting therapy dogs lasted 1 hour daily for 3 weeks. Agitated behaviour, measured with the CMAI, decreased significantly after 3 weeks and increased significantly after 2 weeks washout subsequent to the end of AAT. Libin and Cohen-Mansfield216 assessed the efficacy of a robotic cat (NeCoRo) and a soft toy cat in reducing agitated behaviour in nine women with moderate dementia in nursing homes. The intervention consisted of two, 10-min interactive sessions on different days. The robotic cat produced a significant increase in pleasure and interest, but did not reduce agitation. Conversely, the soft toy cat significantly reduced agitation. Motomura et al217 included 8 women (mean age 84.8 years) residing in a nursing home and evaluated the efficacy of AAT, consisting of two dogs visiting for 1 hour, over four consecutive days, to reduce apathy or irritability. The outcomes were measured using the Geriatric Depression Scale, Physical Self-Maintenance Scale and MMSE. The intervention did not show any significant change on any of the outcomes evaluated. Sellers et al218 included four residents with dementia to evaluate the efficacy of a visiting dog. Agitation was measured with the Agitated Behaviours Mapping Instrument and Social Behaviour Observation Checklist. The authors reported that the intervention reduced agitated behaviour during treatment and increased observed social behaviour, but data and p values were not reported. Dining room environment Two small (n=38) pre-post studies included in Whear's review85examined the effect of improved lighting and table-setting contrast in a dining room environment. One study219 (Brush 2002; n=25) found a positive effect on problem behaviours using the Meal Assistance Screening Tool, while the other study220 found a statistically significant reduction in daily agitation. Special care units In a Cochrane review, Lai 2009 (AMSTAR=8) examined SCUs for dementia individuals with behavioural problems. SCUs are characterised by trained staff, special care programmes, an altered physical environment and involvement of families. This SR included one quasi-experimental study and seven observational studies (six prospective cohort studies and one prospective case–controlled study). The absence of randomised trials is likely a consequence of important practical and ethical issues in applying this methodology in older participants with dementia and behavioural problems. Only one case-controlled study evaluated agitation and used NPI and CMAI to measure the outcome in 65 participants with dementia.221 The results showed no significant changes in outcomes at 3 months; however, there were small, but significant improvements in the NPI score in favour of the SCU group at 6 months (WMD −4.30 (95% CI −7.22 to −1.38), 12 months (WMD −4.30 (95% CI −7.22 to −1.38)) and 18 months (WMD −5.40 (95% CI −9.16 to −1.65)). The same study also evaluated the effect of SCU on mood at 3 months, and the results showed a small significant effect in favour of SCU (WMD −6.30 (95% CI −7.88 to −4.72)).221 Discussion Given the well-known negative side effects of commonly prescribed drugs to control behavioural disturbances (BPSD) in patients with dementia, non-pharmacological interventions have gained increasing attention in recent years as an alternative first-line approach to treat BPSD. This overview addresses the evidence supporting the efficacy of these interventions in community and residential care settings. We identified a number of SRs, which often focused on single interventions although, in several instances, multicomponent interventions were also examined. With the present study, using the primary studies included in the SRs, we have created a compendium of the types of non-pharmacological interventions, including the component of each single intervention, the dosage (when available) and the duration of the treatment. In the absence of a validated taxonomy, we categorised the interventions according to the following classification: sensory stimulation interventions; cognitive/emotion-oriented interventions; behaviour management techniques (further subdivided according to the recipient of the intervention, ie, the person with dementia, the caregiver or the staff); multicomponent interventions and other interventions, such as exercise and animal-assisted therapies. Among sensory simulation interventions, the only convincingly effective intervention for reducing behavioural symptoms (specifically agitation and aggressive behaviour) was music therapy. According to the most comprehensive review of music therapy, this treatment also reduced anxiety. However, the evidence supporting the effectiveness of music therapy was limited by moderate, but significant, heterogeneity, probably related to the variability of the intervention (eg, type of music, active involvement, such as singing/playing a musical instrument and dancing, or passive involvement, such as listening) and the heterogeneity of the patient population in terms of the severity of dementia and the type of dementia. The efficacy of aromatherapy and massage therapy, both associated with conflicting results, remains unknown. Light therapy and SMST therapy did not show any noteworthy effect for clinical practice. The body of evidence concerning cognitive/emotion-oriented interventions, which include reminiscence therapy, SPT and validation therapy, had important methodological limitations. The quality of the primary studies was low, as reported by the review authors, and the sample size of the studies was not powered to detect statistically significant effects. Even when it was possible to combine studies in a meta-analysis, for example, for SPT, the pooled estimated effect was not statistically significant. Added to these shortcomings was the variability in the length and type of the interventions and the multitude of outcomes measured. Overall, convincing evidence supporting the effectiveness of these psychological interventions was lacking. The most frequently assessed intervention in several trials was behavioural management techniques. The elements in this type of intervention included behavioural or cognitive–behavioural therapy, functional analysis of specific behaviour, individualised behavioural reinforcement strategies, communication training and other therapies, such as habit training, progressive muscle relaxation and token economies.31 The body of evidence supporting the effectiveness of behavioural management techniques includes positive and negative studies. Among the types of behavioural management techniques which aimed to enhance communication skills, formal caregiver training and dementia mapping provided in residential care were found to be effective at reducing agitation. The evidence was convincing when the intervention was supervised by healthcare professionals, with the effectiveness possibly persisting for 3–6 months. There is some evidence that multicomponent interventions that use a comprehensive, integrated multidisciplinary approach combining medical, psychiatric and nursing interventions can reduce severe behavioural problems in nursing home patients. Other interventions such as animal-assisted and exercise therapy did not show any convincing effect on any BPSD. Strengths of this overview The present overview represents a substantial update of a previous overview,31 using a search strategy launched in 2009, that provided a comprehensive synthesis of the evidence about non-pharmacological interventions on BPSD. We systematically searched reviews available in four electronic databases and systematically collected the evidence regarding non-pharmacological interventions for the treatment of behavioural disturbances in patients with dementia. To allow the identification of SRs of all potential non-pharmacological interventions, we used a highly sensitive search strategy by avoiding the inclusion of any specific name of non-pharmacological interventions. We also assessed the methodological quality of the reviews using the AMSTAR criteria. Another strength of the present overview was the adoption of a systematic and transparent method, and the use of duplicate, independent reviewers who performed the phases of study selection, data abstraction and data interpretation separately.38 Limitations of the interpretation of the results Overall, the SRs had a number of methodological limitations that could have affected the confidence in the reported results. First, the heterogeneity of the types and characteristics of the interventions, even within the same class of non-pharmacological interventions, was the most significant problem that emerged from the present study. One implication is that there are serious methodological issues that question the correctness, in our opinion, of combining studies in a meta-analysis, as some authors have previously performed. Moreover, in some studies, the description of the interventions is too vague to allow a complete understanding of what was actually performed. In addition, even in cases in which the intervention is well characterised, the dosage of the intervention, and the means used for its delivery, varied considerably. For example, in the case of music therapy, music interventions such as listening to music via headphones, based on participants' musical preferences,87 differed from listening, playing percussion instruments, singing, movement or dance86 and was observed across all nine trials combined in the meta-analysis. In the case of aromatherapy, there were several essential oils that were used in the primary studies, but in some instances, even when similar components were used (eg, Melissa essential oil), the mode of administration differed among trials. Similarly, there was great variation in the intensity (from 2500 to 10 000 lux), duration (1–9 hours), frequency of exposure (10 days to 10 weeks) and type of device used (Dawn-Dusk Simulator222), when light therapy was investigated for behavioural problems in dementia. The variation in the characteristics of the interventions was particularly pronounced in the trials ascribed to behavioural management techniques. The trials used different conceptual frameworks, and sometimes broad and quite generic descriptions, to describe the interventions that at times were difficult to interpret and which influenced the content and quality of evidence of the SRs. In this area, it is therefore difficult to produce a satisfactory classification, which implies that different SRs did not consider the same group of studies, even when they clearly investigated non-pharmacological interventions specifically designed to improve behavioural management. Finally, the arbitrary age cut-off of the patients (more than 60 years of age) and the exclusion of reviews published before 2009 constitute other limitations of the present overview. We did not evaluate the methodological quality of the primary studies included in the reviews, as this will be the scope of our next publication, in which we will apply the GRADE criteria.38 Conclusion This overview succeeded in providing a complete and up-to-date compendium of non-pharmacological interventions in older people with dementia, using recently published SRs and meta-analyses. The most promising treatments appeared to be music therapy and some behavioural management techniques, particularly those involving caregiver-oriented and staff-oriented interventions. Despite the considerable number of published articles included in this overview, the evidence supporting the efficacy of non-pharmacological interventions is limited due to methodological quality and sample size and to the presence of important variations in the taxonomy of the non-pharmacological interventions, the outcomes assessed and the tools used to evaluate the outcomes. Twitter: Follow Roy Soiza @AbdnGeriatrics Contributors: IA, JMR, AC, RS, AC-J and DO conceived and designed the study. The manuscript of this protocol was drafted by IA, JMR, AC, RS, AC-J, AdG and BHM and revised by MP, AnG, FMT and GDA. IA and JMR designed the search strategies; IA, JMR, FMT and GDA performed the search, screening and assessment independently. AC arbitrated disagreements during the review. All authors contributed to data analysis and critical revision of the paper; additionally every author approved the final version. Funding: The research leading to these results has received funding from the European Union Seventh Framework programme (FP7/2007-2013) under grant agreement no. 305930 (SENATOR). Disclaimer: The funders had no role in the study design, data collection and analysis, the decision to publish or the preparation of the manuscript. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Prince M , Bryce R , Albanese E The global prevalence of dementia: a systematic review and metaanalysis . Alzheimers Dement 2013 ;9 :63 –75. e2 . 10.1016/j.jalz.2012.11.007 23305823 2 Grasset L , Brayne C , Joly P Trends in dementia incidence: evolution over a 10-year period in France . Alzheimers Dement 2016 ;12 :272 –80 . 10.1016/j.jalz.2015.11.001 26693893 3 Rattinger GB , Fauth EB , Behrens S Closer caregiver and care-recipient relationships predict lower informal costs of dementia care: the Cache County Dementia Progression Study . Alzheimers Dement 2016 ;12 :917 –24 . 10.1016/j.jalz.2016.03.008 27103262 4 Wimo A , Jönsson L , Bond J The worldwide economic impact of dementia 2010 . Alzheimers Dement 2013 ;9 :1 –11.e3 . 10.1016/j.jalz.2012.11.006 23305821 5 Finkel SI , Costa e Silva J , Cohen G Behavioral and psychological signs and symptoms of dementia: a consensus statement on current knowledge and implications for research and treatment . Int Psychogeriatr 1996 ;8 (Suppl 3 ):497 –500 . 10.1017/S1041610297003943 9154615 6 Lyketsos CG , Carrillo MC , Ryan JM Neuropsychiatric symptoms in Alzheimer's disease . Alzheimers Dement 2011 ;7 :532 –9 . 10.1016/j.jalz.2011.05.2410 21889116 7 Geda YE , Schneider LS , Gitlin LN Neuropsychiatric symptoms in Alzheimer's disease: past progress and anticipation of the future . Alzheimers Dement 2013 ;9 :602 –8 . 10.1016/j.jalz.2012.12.001 23562430 8 Ismail Z , Smith EE , Geda Y Neuropsychiatric symptoms as early manifestations of emergent dementia: provisional diagnostic criteria for mild behavioral impairment . Alzheimers Dement 2016 ;12 :195 –202 . 10.1016/j.jalz.2015.05.017 26096665 9 Kales HC , Gitlin LN , Lyketsos CG Assessment and management of behavioral and psychological symptoms of dementia . BMJ 2015 ;350 :h369 10.1136/bmj.h369 25731881 10 Petrovic M , Hurt C , Collins D Clustering of behavioural and psychological symptoms in dementia (BPSD): a European Alzheimer's disease consortium (EADC) study . Acta Clin Belg 2007 ;62 :426 –32 . 10.1179/acb.2007.062 18351187 11 Pawłucka U , Brzyski P , Kubicz D The determinants of behavioral symptoms in long-term care facility residents . Eur Geriatr Med 2016 ;7 :157 –62 . 10.1016/j.eurger.2016.01.013 12 Savva GM , Zaccai J , Matthews FE Prevalence, correlates and course of behavioural and psychological symptoms of dementia in the population . Br J Psychiatry 2009 ;194 :212 –19 . 10.1192/bjp.bp.108.049619 19252147 13 Hendriks SA , Smalbrugge M , Galindo-Garre F From admission to death: prevalence and course of pain, agitation, and shortness of breath, and treatment of these symptoms in nursing home residents with dementia . J Am Med Dir Assoc 2015 ;16 :475 –81 . 10.1016/j.jamda.2014.12.016 25736822 14 Ryu SH , Katona C , Rive B Persistence of and changes in neuropsychiatric symptoms in Alzheimer disease over 6 months: the LASER-AD study . Am J Geriatr Psychiatry 2005 ;13 :976 –83 . 10.1176/appi.ajgp.13.11.976 16286441 15 Givens JL , Jones RN , Mazor KM Development and psychometric properties of the family distress in advanced dementia scale . J Am Med Dir Assoc 2015 ;16 :775 –80 . 10.1016/j.jamda.2015.03.028 25940236 16 Wetzels RB , Zuidema SU , de Jonghe JF Determinants of quality of life in nursing home residents with dementia . Dement Geriatr Cogn Disord 2010 ;29 :189 –97 . 10.1159/000280437 20215750 17 Gaugler JE , Yu F , Krichbaum K Predictors of nursing home admission for persons with dementia . Med Care 2009 ;47 : 191 –8 . 10.1097/MLR.0b013e31818457ce 19169120 18 Cabrera E , Sutcliffe C , Verbeek H Non-pharmacological interventions as a best practice strategy in people with dementia living in nursing homes. A systematic review . Eur Geriatr Med 2015 ;6 :134 –50 . 10.1016/j.eurger.2014.06.003 19 Cammisuli DM , Danti S , Bosinelli F Non-pharmacological interventions for people with Alzheimer's disease: a critical review of the scientific literature from the last ten years . Eur Geriatr Med 2016 ;7 :57 –64 . 10.1016/j.eurger.2016.01.002 20 Howard R , McShane R , Lindesay J Donepezil and memantine for moderate-to-severe Alzheimer's disease . N Engl J Med 2012 ;366 :893 –903 . 10.1056/NEJMoa1106668 22397651 21 Olin J , Schneider L Galantamine for Alzheimer's disease . Cochrane Database Syst Rev 2002 (3) :CD001747 10.1002/14651858.CD001747 12137632 22 Locca JF , Büla CJ , Zumbach S Pharmacological treatment of behavioral and psychological symptoms of dementia (BPSD) in nursing homes: development of practice recommendations in a Swiss canton . J Am Med Dir Assoc 2008 ;9 :439 –48 . 10.1016/j.jamda.2008.04.003 18585647 23 Porsteinsson AP , Drye LT , Pollock BG Effect of citalopram on agitation in Alzheimer disease: the CitAD randomized clinical trial . JAMA 2014 ;311 :682 –91 . 10.1001/jama.2014.93 24549548 24 Bierman EJ , Comijs HC , Gundy CM The effect of chronic benzodiazepine use on cognitive functioning in older persons: good, bad or indifferent? Int J Geriatr Psychiatry 2007 ;22 :1194 –200 . 10.1002/gps.1811 17407168 25 Maher AR , Maglione M , Bagley S Efficacy and comparative effectiveness of atypical antipsychotic medications for off-label uses in adults: a systematic review and meta-analysis . JAMA 2011 ;306 :1359 –69 . 10.1001/jama.2011.1360 21954480 26 Berry SD , Placide SG , Mostofsky E Antipsychotic and benzodiazepine drug changes affect acute falls risk differently in the nursing home . J Gerontol A Biol Sci Med Sci 2016 ;71 :273 –8 . 10.1093/gerona/glv091 26248560 27 Howard RJ , Juszczak E , Ballard CG Donepezil for the treatment of agitation in Alzheimer's disease . N Engl J Med 2007 ;357 :1382 –92 . 10.1056/NEJMoa066583 17914039 28 Matsunaga S , Kishi T , Iwata N Memantine monotherapy for Alzheimer's disease: a systematic review and meta-analysis . PLoS ONE 2015 ;10 :e0123289 10.1371/journal.pone.0123289 25860130 29 Seitz DP , Brisbin S , Herrmann N Efficacy and feasibility of nonpharmacological interventions for neuropsychiatric symptoms of dementia in long term care: a systematic review . J Am Med Dir Assoc 2012 ;13 :503 –06. e2 . 10.1016/j.jamda.2011.12.059 22342481 30 Zetteler J Effectiveness of simulated presence therapy for individuals with dementia: a systematic review and meta-analysis . Aging Ment Health 2008 ;12 :779 –85 . 10.1080/13607860802380631 19023729 31 O'Neil ME , Freeman M , Portland V A systematic evidence review of non-pharmacological interventions for behavioral symptoms of dementia . Washington (DC ): Department of Veterans Affairs , 2011 . 32 Abraha I , Trotta F , Rimland JM Efficacy of non-pharmacological interventions to prevent and treat Delirium in older patients: a systematic overview. The SENATOR project ONTOP Series . PLoS ONE 2015 ;10 :e0123090 10.1371/journal.pone.0123090 26062023 33 Vélez-Díaz-Pallarés M , Lozano-Montoya I , Correa-Pérez A Non-pharmacological interventions to prevent or treat pressure ulcers in older patients: Clinical practice recommendations . The SENATOR-ONTOP series Eur Geriatr Med 2016 ;7 :142 –8 . 34 Lozano-Montoya I , Velez-Diaz-Pallares M , Abraha I Nonpharmacologic interventions to prevent pressure ulcers in older patients: an overview of systematic reviews (the software ENgine for the assessment and optimization of drug and non-drug therapy in older peRsons [SENATOR] definition of optimal evidence-based non-drug therapies in older people [ONTOP] series) . J Am Med Dir Assoc 2016 ;17 :370 e1 –70 e10 . 10.1016/j.jamda.2015.12.091 35 Velez-Diaz-Pallares M , Lozano-Montoya I , Abraha I Nonpharmacologic interventions to heal pressure ulcers in older patients: an overview of systematic reviews (the SENATOR-ONTOP series) . J Am Med Dir Assoc 2015 ;16 :448 –69 . 10.1016/j.jamda.2015.01.083 25737261 36 Rimland JM , Dell'Aquila G , O'Mahony D Meta-analysis of multifactorial interventions to prevent falls of older adults in care facilities . J Am Geriatr Soc 2015 ;63 :1972 –3 . 10.1111/jgs.13622 26390004 37 Rimland JM , Abraha I , Dell'Aquila G Effectiveness of non-pharmacological interventions to prevent falls in older people: a systematic overview. The SENATOR Project ONTOP Series . PLoS ONE 2016 ;11 :e0161579 10.1371/journal.pone.0161579 27559744 38 Abraha I , Cruz-Jentoft A , Soiza RL Evidence of and recommendations for non-pharmacological interventions for common geriatric conditions: the SENATOR-ONTOP systematic review protocol . BMJ Open 2015 ;5 :e007488 10.1136/bmjopen-2014-007488 39 Shea BJ , Grimshaw JM , Wells GA Development of AMSTAR: a measurement tool to assess the methodological quality of systematic reviews . BMC Med Res Methodol 2007 ;7 :10 10.1186/1471-2288-7-10 17302989 40 Savović J , Jones HE , Altman DG Influence of reported study design characteristics on intervention effect estimates from randomized, controlled trials . Ann Intern Med 2012 ;157 :429 –38 . 10.7326/0003-4819-157-6-201209180-00537 22945832 41 Abraha I , Cherubini A , Cozzolino F Deviation from intention to treat analysis in randomised trials and treatment effect estimates: meta-epidemiological study . BMJ 2015 ;350 :h2445 10.1136/bmj.h2445 26016488 42 Chan AW , Hróbjartsson A , Haahr MT Empirical evidence for selective reporting of outcomes in randomized trials: comparison of protocols to published articles . JAMA 2004 ;291 :2457 –65 . 10.1001/jama.291.20.2457 15161896 43 Guyatt GH , Oxman AD , Vist G GRADE guidelines: 4. Rating the quality of evidence--study limitations (risk of bias) . J Clin Epidemiol 2011 ;64 :407 –15 . 10.1016/j.jclinepi.2010.07.017 21247734 44 Robinson N , Lorenc A , Liao X The evidence for Shiatsu: a systematic review of Shiatsu and acupressure . BMC Complement Altern Med 2011 ;11 :88 10.1186/1472-6882-11-88 21982157 45 Lin LC , Yang MH , Kao CC Using acupressure and Montessori-based activities to decrease agitation for residents with dementia: a cross-over trial . J Am Geriatr Soc 2009 ;57 :1022 –9 . 10.1111/j.1532-5415.2009.02271.x 19507295 46 Nguyen QA , Paton C The use of aromatherapy to treat behavioural problems in dementia . Int J Geriatr Psychiatry 2008 ;23 :337 –46 . 10.1002/gps.1886 17918182 47 Fung JKKM , Tsang HWH , Chung RCK A systematic review of the use of aromatherapy in treatment of behavioral problems in dementia . Geriatr Gerontol Int 2012 ;12 :372 –82 . 10.1111/j.1447-0594.2012.00849.x 22433025 48 Forrester LT , Maayan N , Orrell M Aromatherapy for dementia . Cochrane Database Syst Rev 2014 ;(2) :CD003150 10.1002/14651858.CD003150.pub2 24569873 49 Fu CY , Moyle W , Cooke M A randomised controlled trial of the use of aromatherapy and hand massage to reduce disruptive behaviour in people with dementia . BMC Complement Altern Med 2013 ;13 :165 10.1186/1472-6882-13-165 23837414 50 Lin PW , Chan WC , Ng BF Efficacy of aromatherapy (Lavandula angustifolia) as an intervention for agitated behaviours in Chinese older persons with dementia: a cross-over randomized trial . Int J Geriatr Psychiatry 2007 ;22 :405 –10 . 10.1002/gps.1688 17342790 51 O'Connor DW , Eppingstall B , Taffe J A randomized, controlled cross-over trial of dermally-applied lavender (Lavandula angustifolia) oil as a treatment of agitated behaviour in dementia . BMC Complement Altern Med 2013 ;13 :315 10.1186/1472-6882-13-315 24219098 52 Smallwood J , Brown R , Coulter F Aromatherapy and behaviour disturbances in dementia: a randomized controlled trial . Int J Geriatr Psychiatry 2001 ;16 :1010 –13 . 10.1002/gps.473 11607948 53 Ballard CG , O'Brien JT , Reichelt K Aromatherapy as a safe and effective treatment for the management of agitation in severe dementia: the results of a double-blind, placebo-controlled trial with Melissa . J Clin Psychiatry 2002 ;63 :553 –8 . 10.4088/JCP.v63n0703 12143909 54 Burns A , Perry E , Holmes C A double-blind placebo-controlled randomized trial of Melissa officinalis oil and donepezil for the treatment of agitation in Alzheimer's disease . Dement Geriatr Cogn Disord 2011 ;31 :158 –64 . 10.1159/000324438 21335973 55 Akhondzadeh S , Noroozian M , Mohammadi M Melissa officinalis extract in the treatment of patients with mild to moderate Alzheimer's disease: a double blind, randomised, placebo controlled trial . J Neurol Neurosurg Psychiatr 2003 ;74 :863 –6 . 10.1136/jnnp.74.7.863 56 Burleigh S , Armstrong C On the scent of a useful therapy . J Dement Care 1997 ;5 :21 – 3 . 57 Gray SG , Clair AA Influence of aromatherapy on medication administration to residential-care residents with dementia and behavioral challenges . Am J Alzheimers Dis Other Demen 2002 ;17 :169 –74 . 10.1177/153331750201700305 12083347 58 Holmes C , Hopkins V , Hensford C Lavender oil as a treatment for agitated behaviour in severe dementia: a placebo controlled study . Int J Geriatr Psychiatry 2002 ;17 :305 –8 . 10.1002/gps.593 11994882 59 Jimbo D , Kimura Y , Taniguchi M Effect of aromatherapy on patients with Alzheimer's disease . Psychogeriatrics 2009 ;9 :173 –9 . 10.1111/j.1479-8301.2009.00299.x 20377818 60 Kilstoff K , Chenoweth L New approaches to health and well-being for dementia day-care clients, family carers and day-care staff . Int J Nurs Pract 1998 ;4 :70 –83 . 10.1046/j.1440-172X.1998.00059.x 9748936 61 Cameron H , du Toit S , Richard G Using lemon balm oil to reduce aggression and agitation in dementia: results of a pilot study . J Dement Care 2011 ;19 :36 –8 . 62 Cohen-Mansfield J Nonpharmacologic interventions for inappropriate behaviors in dementia: a review, summary, and critique . Am J Geriatr Psychiatry 2001 ;9 :361 –81 . 10.1097/00019442-200111000-00005 11739063 63 Viggo Hansen N , Jørgensen T , Ørtenblad L Massage and touch for dementia . Cochrane Database Syst Rev 2006 ;(4) :CD004989 10.1002/14651858.CD004989.pub2 17054228 64 Remington R Calming music and hand massage with agitated elderly . Nurs Res 2002 ;51 :317 –23 . 10.1097/00006199-200209000-00008 12352780 65 Moyle W , Murfield JE , O'Dwyer S The effect of massage on agitated behaviours in older people with dementia: a literature review . J Clin Nurs 2013 ;22 :601 –10 . 10.1111/j.1365-2702.2012.04234.x 23164052 66 Holliday-Welsh DM , Gessert CE , Renier CM Massage in the management of agitation in nursing home residents with cognitive impairment . Geriatr Nurs 2009 ;30 :108 –17 . 10.1016/j.gerinurse.2008.06.016 19345850 67 Forbes D , Blake Catherine M , Thiessen Emily J Light therapy for improving cognition, activities of daily living, sleep, challenging behaviour, and psychiatric disturbances in dementia . Cochrane Database Syst Rev 2014 ;(2) :CD003946 10.1002/14651858.CD003946.pub4http://onlinelibrary.wiley.com/doi/10.1002/14651858.CD003946.pub4/abstract 24574061 68 Ancoli-Israel S , Martin JL , Gehrman P Effect of light on agitation in institutionalized patients with severe Alzheimer disease . Am J Geriatr Psychiatry 2003 ;11 :194 –203 . 10.1097/00019442-200303000-00010 12611749 69 Burns A , Allen H , Tomenson B Bright light therapy for agitation in dementia: a randomized controlled trial . Int Psychogeriatr 2009 ;21 :711 –21 . 10.1017/S1041610209008886 19323872 70 Dowling GA , Mastick J , Hubbard EM Effect of timed bright light treatment for rest-activity disruption in institutionalized patients with Alzheimer's disease . Int J Geriatr Psychiatry 2005 ;20 : 738 –43 . 10.1002/gps.1352 16035127 71 Riemersma-van der Lek RF , Swaab DF , Twisk J Effect of bright light and melatonin on cognitive and noncognitive function in elderly residents of group care facilities: a randomized controlled trial . JAMA 2008 ;299 :2642 –55 . 10.1001/jama.299.22.2642 18544724 72 Chaudhury H , Hung L , Badger M The role of physical environment in supporting person-centered dining in long-term care: a review of the literature . Am J Alzheimers Dis Other Demen 2013 ;28 :491 –500 . 10.1177/1533317513488923 23687182 73 Barrick AL , Sloane PD , Williams CS Impact of ambient bright light on agitation in dementia . Int J Geriatr Psychiatry 2010 ;25 :1013 –21 . 10.1002/gps.2453 20104513 74 Padilla R Effectiveness of environment-based interventions for people with Alzheimer's disease and related dementias . Am J Occup Ther 2011 ;65 :514 –22 . 10.5014/ajot.2011.002600 22026319 75 Hickman SE , Barrick AL , Williams CS The effect of ambient bright light therapy on depressive symptoms in persons with Dementia . J Am Geriatr Soc 2007 ;55 :1817 –24 . 10.1111/j.1532-5415.2007.01428.x 17944896 76 Whear R , Coon JT , Bethel A What is the impact of using outdoor spaces such as gardens on the physical and mental well-being of those with dementia? A systematic review of quantitative and qualitative evidence . J Am Med Dir Assoc 2014 ;15 :697 –705 . 10.1016/j.jamda.2014.05.013 25037168 77 Luk KY , Lai KYC , Li CC The effect of horticultural activities on agitation in nursing home residents with dementia . Int J Geriatr Psychiatry 2011 ;26 :435 –6 . 10.1002/gps.2493 21412848 78 Gonzalez MT , Kirkevold M Benefits of sensory garden and horticultural activities in dementia care: a modified scoping review . J Clin Nurs 2014 ;23 :2698 –715 . 10.1111/jocn.12388 24128125 79 Connell BR , Sanford JA , Lewis D Therapeutic effects of an outdoor activity program on nursing home residents with dementia . J Hous Elderly 2007 ;21 :194 –209 . 10.1300/J081v21n03_10 80 Chatterton WB , Baker F Morgan K The singer or the singing: who sings individually to persons with dementia and what are the effects? Am J Alzheimers Dis Other Demen 2010 ;25 :641 –9 . 10.1177/1533317510385807 21131670 81 McDermott OC , Crellin N , Ridder HM Music therapy in dementia: a narrative synthesis systematic review . Int J Geriatr Psychiatry 2013 ;28 :781 –94 . 10.1002/gps.3895 23080214 82 Ueda TS , Suzukamo Y , Sato M Effects of music therapy on behavioral and psychological symptoms of dementia: a systematic review and meta-analysis . Ageing Res Rev 2013 ;12 :628 –41 . 10.1016/j.arr.2013.02.003 23511664 83 Vasionytė IM , Madison G Musical intervention for patients with dementia: a meta-analysis . J Clin Nurs 2013 ;22 :1203 –16 . 10.1111/jocn.12166 23574287 84 Wall M , Duffy A The effects of music therapy for older people with dementia . Br J Nurs 2010 ;19 :108 –13 . 10.12968/bjon.2010.19.2.46295 20220649 85 Whear R , Abbott R , Thompson-Coon J Effectiveness of mealtime interventions on behavior symptoms of people with dementia living in care homes: a systematic review . J Am Med Dir Assoc 2014 ;15 :185 –93 . 10.1016/j.jamda.2013.10.016 24405641 86 Groene RW Effectiveness of music therapy 1:1 intervention with individuals having senile dementia of the Alzheimer's type . J Music Ther 1993 ;30 :138 –57 . 10.1093/jmt/30.3.138 87 Guétin S , Portet F , Picot MC Effect of music therapy on anxiety and depression in patients with Alzheimer's type dementia: randomised, controlled study . Dement Geriatr Cogn Disord 2009 ;28 :36 –46 . 10.1159/000229024 19628939 88 Ledger AJ , Baker FA An investigation of long-term effects of group music therapy on agitation levels of people with Alzheimer's disease . Aging Ment Health 2007 ;11 :330 –8 . 10.1080/13607860600963406 17558584 89 Miura H , Kanayama Y , Mogi N Effect and significance of music therapy on elderly persons with mild dementia . 2005 ;5 :48 –57 . 90 Sung HC , Chang AM , Abbey J The effects of preferred music on agitation of older people with dementia in Taiwan . Int J Geriatr Psychiatry 2006 ;21 :999 –1000 . 10.1002/gps.1585 16998779 91 Sung HC , Chang AM , Lee W-L A preferred music listening intervention to reduce anxiety in older adults with dementia in nursing homes . J Clin Nurs 2010 ;19 :1056 –64 . 10.1111/j.1365-2702.2009.03016.x 20492050 92 Suzuki M , Kanamori M , Nagasawa S Music therapy-induced changes in behavioral evaluations, and saliva chromogranin a and immunoglobulin a concentrations in elderly patients with senile dementia . Geriatr Gerontol Int 2007 ;7 :61 –71 . 10.1111/j.1447-0594.2007.00374.x 93 Tuet RWK , Lam LCW A preliminary study of the effects of music therapy on agitation in Chinese patients with dementia . Hong Kong J Psychiatry 2006 ;16 :87 –91 . 94 Ikeda M , Suzuki M , Sawai D Effects of nursing intervention using rhythmic exercise for patients with severe senile dementia . Nurs Res 2006 ;39 :301 –13 . 95 Mihara B , Hosoya M , Mihara Y The effect of music therapy for elderly with dementia: a comparative study between large group and small group sessions . Japanese J Music Ther 2004 ;4 :208 –16 . 96 Raglio A , Bellelli G , Traficante D Efficacy of music therapy in the treatment of behavioral and psychiatric symptoms of dementia . Alzheimer Dis Assoc Disord 2008 ;22 :158 –62 . 10.1097/WAD.0b013e3181630b6f 18525288 97 Sung HC , Chang SM , Lee WL The effects of group music with movement intervention on agitated behaviours of institutionalized elders with dementia in Taiwan . Complement Ther Med 2006 ;14 :113 –19 . 10.1016/j.ctim.2006.03.002 16765849 98 Choi AN , Lee MS , Cheong KJ Effects of group music intervention on behavioral and psychological symptoms in patients with dementia: a pilot-controlled trial . Int J Neurosci 2009 ;119 :471 –81 . 10.1080/00207450802328136 19229716 99 Goka F The effects of combined music and reminiscence therapy for small groups of the elderly with senile dementia of Alzheimer type, and the efficacy of the evaluation method . Japanese J Music Ther 2005 ;5 :25 –38 . 100 Raglio A , Bellelli G , Traficante D Efficacy of music therapy treatment based on cycles of sessions: a randomised controlled trial . Aging Ment Health 2010 ;14 :900 –4 . 10.1080/13607861003713158 21069596 101 Smith GH A comparison of the effects of three treatment interventions on cognitive functioning of Alzheimer patients . Music Ther 1986 ;6 :41 –56 . 10.1093/mt/6.1.41 102 Suzuki M , Kanamori M , Watanabe M Behavioral and endocrinological evaluation of music therapy for elderly patients with dementia . Nurs Health Sci 2004 ;6 :11 –18 . 10.1111/j.1442-2018.2003.00168.x 14764189 103 Van de Winckel A , Feys H , De Weerdt W Cognitive and behavioural effects of music-based exercises in patients with dementia . Clin Rehabil 2004 ;18 :253 –60 . 10.1191/0269215504cr750oa 15137556 104 Svansdottir HB , Snaedal J Music therapy in moderate and severe dementia of Alzheimer's type: a case–control study . Int Psychogeriatr 2006 ;18 :613 –21 . 10.1017/S1041610206003206 16618375 105 Clark ME , Lipe AW , Bilbrey M Use of music to decrease aggressive behaviors in people with dementia . J Gerontol Nurs 1998 ;24 :10 –17 . 10.3928/0098-9134-19980701-05 9801526 106 Cooke M , Moyle W , Shum D A randomized controlled trial exploring the effect of music on quality of life and depression in older people with Dementia . J Health Psychology 2010 ;15 :765 –76 . 10.1177/1359105310368188 107 Nair B , Heim C , Krishnan C The effect of Baroque music on behavioural disturbances in patients with dementia . Australas J Ageing 2011 ;30 :11 –15 . 10.1111/j.1741-6612.2010.00439.x 21395934 108 Raglio A , Oasi O , Gianotti M Effects of music therapy on psychological symptoms and heart rate variability in patients with dementia. A pilot study . Current Aging Sci 2010 :242 –6 . 10.2174/1874609811003030242 109 Irish M , Cunningham CJ , Walsh JB Investigating the enhancing effect of music on autobiographical memory in mild Alzheimer's disease . Dement Geriatr Cogn Disord 2006 ;22 :108 –20 . 10.1159/000093487 16717466 110 Denney A Quiet music. An intervention for mealtime agitation? J Gerontol Nurs 1997 ;23 :16 –23 . 10.3928/0098-9134-19970701-09 111 Hicks-Moore SL Relaxing music at mealtime in nursing homes: effects on agitated patients with dementia . J Gerontol Nurs 2005 ;31 :26 –32 . 10.3928/0098-9134-20051201-07 112 Ragneskog H , Kihlgren M , Karlsson I Dinner music for demented patients: analysis of video-recorded observations . Clin Nurs Res 1996 ;5 :262 –77 . Discussion 78–82 10.1177/105477389600500302 8850771 113 Thomas DW , Smith M The effect of music on caloric consumption among nursing home residents with Dementia of the Alzheimer's type . Act Adapt Aging 2009 ;33 :1 –16 . 10.1080/01924780902718566 114 Kiepe MS , Barbara S , Thomas K Effects of dance therapy and ballroom dances on physical and mental illnesses: a systematic review . Arts Psychotherapy 2012 ;39 :404 –11 . 10.1016/j.aip.2012.06.001 115 Guzman-Garcia A , Hughes JC , James IA Dancing as a psychosocial intervention in care homes: a systematic review of the literature . Int J Geriatr Psychiatry 2013 ;28 :914 –24 . 10.1002/gps.3913 23225749 116 Hokkanen L , Rantala L , Remes AM Dance and movement therapeutic methods in management of dementia: a randomized, controlled study . J Am Geriatr Soc 2008 ;56 :771 –2 . 10.1111/j.1532-5415.2008.01611.x 18380687 117 Chung JC , Lai CK , Chung PM Snoezelen for dementia . Cochrane Database Syst Rev 2002 ;(4) :CD003152 .12519587 118 Baker R , Bell S , Baker E A randomized controlled trial of the effects of multi-sensory stimulation (MSS) for people with dementia . Br J Clin Psychology 2001 ;40 :81 –96 . 10.1348/014466501163508 119 Baker R. & Dowling Z. (1995) INTERACT. A New Measure of Response to Multi-Sensory Environments. Research Publication. Research and Development Support Unit, Poole Hospital, Dorset. 120 Baker R , Holloway J , Holtkamp CCM Effects of multi-sensory stimulation for people with dementia . J Adv Nurs 2003 ;43 :465 –77 . 10.1046/j.1365-2648.2003.02744.x 12919265 121 van Weert JC , van Dulmen AM , Spreeuwenberg PM Effects of snoezelen, integrated in 24 h dementia care, on nurse-patient communication during morning care . Patient Educ Couns 2005 ;58 :312 –26 . 10.1016/j.pec.2004.07.013 16054329 122 van Weert JC , van Dulmen AM , Spreeuwenberg PM Behavioral and mood effects of Snoezelen integrated into 24-hour dementia care . J Am Geriatr Soc 2005 ;53 :24 –33 . 10.1111/j.1532-5415.2005.53006.x 15667372 123 Bi X , Lv H , Chen BL Effects of transcutaneous electrical nerve stimulation on pain in patients with spinal cord injury: a randomized controlled trial . J Phys Ther Sci 2015 ;27 :23 –5 . 10.1589/jpts.27.23 25642029 124 Salami O , Lyketsos C , Rao V Treatment of sleep disturbance in Alzheimer's dementia . Int J Geriatr Psychiatry 2011 ;26 :771 –82 . 10.1002/gps.2609 20872779 125 Cameron M , Lonergan E , Lee H Transcutaneous electrical nerve stimulation (TENS) for dementia . Cochrane Database Syst Rev 2003 (3) :CD004032 .12917999 126 Scherder E , Knol D , van Someren E Effects of low-frequency cranial electrostimulation on the rest-activity rhythm and salivary cortisol in Alzheimer's disease . Neurorehabil Neural Repair 2003 ;17 :101 –8 . 10.1177/0888439003017002004 12814055 127 Scherder EJ , Bouma A , Steen L Influence of transcutaneous electrical nerve stimulation on memory in patients with dementia of the Alzheimer type . J Clin Exp Neuropsychol 1992 ;14 :951 –60 . 10.1080/01688639208402546 1360474 128 Scherder EJ , Deijen JB , Vreeswijk SH Cranial electrostimulation (CES) in patients with probable Alzheimer's disease . Behav Brain Res 2002 ;128 :215 –17 . 10.1016/S0166-4328(01)00323-0 11796166 129 Scherder EJ , Van Someren EJ , Bouma A Effects of transcutaneous electrical nerve stimulation (TENS) on cognition and behaviour in aging . Behav Brain Res 2000 ;111 :223 –5 . 10.1016/S0166-4328(00)00170-4 10840147 130 Scherder EJ , Van Someren EJ , Swaab DF Transcutaneous electrical nerve stimulation (TENS) improves the rest-activity rhythm in midstage Alzheimer's disease . Behav Brain Res 1999 ;101 :105 –7 . 10.1016/S0166-4328(98)00150-8 10342404 131 Van Someren EJ , Scherder EJ , Swaab DF Transcutaneous electrical nerve stimulation (TENS) improves circadian rhythm disturbances in Alzheimer disease . Alzheimer Dis Assoc Disord 1998 ;12 :114 –18 . 10.1097/00002093-199806000-00010 9651141 132 Hozumi S , Hori H , Okawa M Favorable effect of transcranial electrostimulation on behavior disorders in elderly patients with dementia: a double-blind study . Int J Neurosci 1996 ;88 :1 –10 . 10.3109/00207459608999809 9003961 133 Woods B , Aguirre E , Spector AE Cognitive stimulation to improve cognitive functioning in people with dementia . Cochrane Database Syst Rev 2012 ;(2 ) :CD005562 10.1002/14651858.CD005562.pub2 22336813 134 Aguirre E , Woods RT , Spector A Cognitive stimulation for dementia: a systematic review of the evidence of effectiveness from randomised controlled trials . Ageing Res Rev 2013 ;12 :253 –62 . 10.1016/j.arr.2012.07.001 22889599 135 Alves J , Magalhaes R , Thomas RE Is there evidence for cognitive intervention in Alzheimer disease? A systematic review of efficacy, feasibility, and cost-effectiveness . Alzheimer Dis Assoc Disord 2013 ;27 :195 –203 . 10.1097/WAD.0b013e31827bda55 23314062 136 Carrion C , Aymerich M , Baillés E Cognitive psychosocial intervention in dementia: a systematic review . Dement Geriatr Cogn Disord 2013 ;36 :363 –75 . 10.1159/000354365 24022505 137 Yu F , Rose KM , Burgener SC Cognitive training for early-stage Alzheimer's disease and dementia . J Gerontol Nurs 2009 ;35 :23 –9 . 10.3928/00989134-20090301-10 19326826 138 Olazarán J , Reisberg B , Clare L Nonpharmacological therapies in Alzheimer's disease: a systematic review of efficacy . Dement Geriatr Cogn Disord 2010 ;30 :161 –78 . 10.1159/000316119 20838046 139 Zientz J , Rackley A , Chapman SB Evidence-based practice recommendations for dementia: educating caregivers on Alzheimer's disease and training communication strategies . J Med Speech Lang Pathol 2007 ;15 :liii –lxiv . 140 Baines S , Saxby P , Ehlert K Reality orientation and reminiscence therapy. A controlled cross-over study of elderly confused people . Br J Psychiatry 1987 ;151 :222 –31 . 10.1192/bjp.151.2.222 3318991 141 Ferrario E , Cappa G , Molaschi M Reality orientation therapy in institutionalized elderly patients: preliminary results . Arch Gerontol Geriatr 1991 ;12 :139 –42 . 142 Onder G , Zanetti O , Giacobini E Reality orientation therapy combined with cholinesterase inhibitors in Alzheimer's disease: randomised controlled trial . Br J Psychiatry 2005 ;187 :450 –5 . 10.1192/bjp.187.5.450 16260821 143 Robichaud L , Hebert R , Desrosiers J Efficacy of a sensory integration program on behaviors of inpatients with dementia . Am J Occup Ther 1994 ;48 :355 –60 . 10.5014/ajot.48.4.355 8059869 144 Kim SY , Yoo EY , Jung MY A systematic review of the effects of occupational therapy for persons with dementia: a meta-analysis of randomized controlled trials . NeuroRehabilitation 2012 ;31 :107 –15 . 10.3233/NRE-2012-0779 22951704 145 Kurz AF , Leucht S , Lautenschlager NT The clinical significance of cognition-focused interventions for cognitively impaired older adults: a systematic review of randomized controlled trials . Int Psychogeriatr 2011 ;23 :1364 –75 . 10.1017/S1041610211001001 21740614 146 Blake M Group reminiscence therapy for adults with dementia: a review . Br J Community Nurs 2013 ;18 :228 –33 . 10.12968/bjcn.2013.18.5.228 23752322 147 Subramaniam P , Woods B The impact of individual reminiscence therapy for people with dementia: systematic review . Expert Rev Neurother 2012 ;12 :545 –55 . 10.1586/ern.12.35 22550983 148 Deponte A , Missan R Effectiveness of validation therapy (VT) in group: preliminary results . Arch Gerontol Geriatr 2007 ;44 :113 –17 . 10.1016/j.archger.2006.04.001 149 Wang JJ , Yen M , OuYang WC Group reminiscence intervention in Taiwanese elders with dementia . Arch Gerontol Geriatr 2009 ;49 :227 –32 . 10.1016/j.archger.2008.08.007 150 Haight BK , Gibson F , Michel Y The Northern Ireland life review/life storybook project for people with dementia . Alzheimers Dement 2006 ;2 :56 –8 . 10.1016/j.jalz.2005.12.003 19595856 151 Morgan S , Woods R Life review with people with dementia in care homes: a preliminary randomized controlled trial . 2010 ;1 :43 –60 . 152 Lai CK , Chi I , Kayser-Jones J A randomized controlled trial of a specific reminiscence approach to promote the well-being of nursing home residents with dementia . Int Psychogeriatr 2004 ;16 :33 –49 . 10.1017/S1041610204000055 15190995 153 Haslam C , Haslam SA , Jetten J The social treatment: the benefits of group interventions in residential care settings . Psychol Aging 2010 ;25 :157 –67 . 10.1037/a0018256 20230136 154 Politis AM , Vozzella S , Mayer LS A randomized, controlled, clinical trial of activity therapy for apathy in patients with dementia residing in long-term care . Int J Geriatr Psychiatry 2004 ;19 :1087 –94 . 10.1002/gps.1215 15481065 155 Neal M , Barton Wright P Validation therapy for dementia . Cochrane Database Syst Rev 2003 (3) :CD001394 10.1002/14651858.CD001394 12917907 156 Peoples . Validation therapy, versus reality orientation as treatment for disorientated institutionalised elderly [Masters dissertation] Akron : College of Nursing, University of Akron , 1982 . 157 Robb SS , Stegman CE , Wolanin MO No research versus research with compromised results: a study of validation therapy . Nurs Res 1986 ;35 :113 –18 .3081879 158 Toseland RW , Diehl M , Freeman K The impact of validation group therapy on nursing home residents with dementia . J Appl Gerontol 1997 ;16 :31 –50 . 10.1177/073346489701600102 159 Vasse E , Vernooij-Dassen M , Spijker A A systematic review of communication strategies for people with dementia in residential and nursing homes . Int Psychogeriatr 2010 ;22 :189 –200 . 10.1017/S1041610209990615 19638257 160 Cohen-Mansfield J Agitated behaviors in the elderly. II. Preliminary results in the cognitively deteriorated . J Am Geriatr Soc 1986 ;34 :722 –7 . 10.1111/j.1532-5415.1986.tb04303.x 3760436 161 Abraha I , Rimland JM , Lozano-Montoya I simulated presence therapy for dementia: a systematic review protocol . BMJ Open 2016 ;6 :e011007 10.1136/bmjopen-2015-011007 162 Kverno KS , Black BS , Nolan MT Research on treating neuropsychiatric symptoms of advanced dementia with non-pharmacological strategies, 1998-2008: a systematic literature review . Int Psychogeriatr 2009 ;21 :825 –43 . 10.1017/S1041610209990196 19586562 163 Teri L , Logsdon RG , McCurry SM Nonpharmacologic treatment of behavioral disturbance in dementia . Med Clin North Am 2002 ;86 :641 –56 . viii. 10.1016/S0025-7125(02)00006-8 12168563 164 Teri L , Gibbons LE , McCurry SM Exercise plus behavioral management in patients with Alzheimer disease: a randomized controlled trial . JAMA 2003 ;290 :2015 –22 . 10.1001/jama.290.15.2015 14559955 165 Proctor R , Burns A , Powell HS Behavioural management in nursing and residential homes: a randomised controlled trial . Lancet 1999 ;354 :26 –9 . 10.1016/S0140-6736(98)08237-3 10406361 166 Teri L , Huda P , Gibbons L STAR: a dementia-specific training program for staff in assisted living residences . Gerontologist 2005 ;45 :686 –93 . 10.1093/geront/45.5.686 16199404 167 McCurry SM , Gibbons LE , Logsdon RG Nighttime insomnia treatment and education for Alzheimer's disease: a randomized, controlled trial . J Am Geriatr Soc 2005 ;53 :793 –802 . 10.1111/j.1532-5415.2005.53252.x 15877554 168 Gitlin LN , Winter L , Dennis MP A biobehavioral home-based intervention and the well-being of patients with dementia and their caregivers: the COPE randomized trial . JAMA 2010 ;304 :983 –91 . 10.1001/jama.2010.1253 20810376 169 Beck CK , Vogelpohl TS , Rasin JH Effects of behavioral interventions on disruptive behavior and affect in demented nursing home residents . Nurs Res 2002 ;51 :219 –28 . 10.1097/00006199-200207000-00002 12131234 170 Suhr J Progressive muscle relaxation in the management of behavioural disturbance in Alzheimer's disease . Neuropsychol Rehabil 1999 ;9 :31 –44 . 10.1080/713755590 171 McCallion P , Toseland RW , Freeman K An evaluation of a family visit education program . J Am Geriatr Soc 1999 ;47 :203 –14 . 10.1111/j.1532-5415.1999.tb04579.x 9988292 172 Livingston G , Kelly L , Lewis-Holmes E A systematic review of the clinical effectiveness and cost-effectiveness of sensory, psychological and behavioural interventions for managing agitation in older adults with dementia . Health Technol Assess 2014 ;18 :1 –226 . v–vi 10.3310/hta18610 173 Gormley N , Lyons D , Howard R Behavioural management of aggression in dementia: a randomized controlled trial . Age Ageing 2001 ;30 :141 –5 . 10.1093/ageing/30.2.141 11395344 174 Teri L , Logsdon RG , Peskind E Treatment of agitation in AD: a randomized, placebo-controlled clinical trial . Neurology 2000 ;55 :1271 –8 . 10.1212/WNL.55.9.1271 11087767 175 Bourgeois MS , Burgio LD , Schulz R Modifying repetitive verbalizations of community-dwelling patients with AD . Gerontologist 1997 ;37 :30 –9 . 10.1093/geront/37.1.30 9046703 176 Moniz Cook ED , Swift K , James I Functional analysis-based interventions for challenging behaviour in dementia . Cochrane Database Syst Rev 2012 ;(2) :CD006929 10.1002/14651858.CD006929.pub2 22336826 177 Mador JE , Giles L , Whitehead C A randomized controlled trial of a behavior advisory service for hospitalized older patients with confusion . Int J Geriatr Psychiatry 2004 ;19 :858 –63 . 10.1002/gps.1165 15352143 178 Burgio L , Stevens A , Guy D Impact of two psychosocial interventions on white and African American family caregivers of individuals with dementia . Gerontologist 2003 ;43 :568 –79 . 10.1093/geront/43.4.568 12937335 179 Chenoweth L , King MT , Jeon Y-H Caring for Aged Dementia Care Resident Study (CADRES) of person-centred care, dementia-care mapping, and usual care in dementia: a cluster-randomised trial . Lancet Neurol 2009 ;8 :317 –25 . 10.1016/S1474-4422(09)70045-6 19282246 180 Farran CJ , Gilley DW , McCann JJ Psychosocial interventions to reduce depressive symptoms of dementia caregivers: a randomized clinical trial comparing two approaches . J Mental Health Aging 2004 ;10 :337 –50 . 181 Fossey J , Ballard C , Juszczak E Effect of enhanced psychosocial care on antipsychotic use in nursing home residents with severe dementia: cluster randomised trial . BMJ 2006 ;332 :756 –61 . 10.1136/bmj.38782.575868.7C 16543297 182 Gitlin LN , Winter L , Corcoran M Effects of the home environmental skill-building program on the caregiver-care recipient dyad: 6-month outcomes from the Philadelphia REACH Initiative . Gerontologist 2003 ;43 :532 –46 . 10.1093/geront/43.4.532 12937332 183 Gonyea JG , O'Connor MK , Boyle PA Project CARE: a randomized controlled trial of a behavioral intervention group for Alzheimer's disease caregivers . Gerontologist 2006 ;46 :827 –32 . 10.1093/geront/46.6.827 17169938 184 Huang HL , Shyu YI , Chen MC A pilot study on a home-based caregiver training program for improving caregiver self-efficacy and decreasing the behavioral problems of elders with dementia in Taiwan . Int J Geriatr Psychiatry 2003 ;18 :337 –45 . 10.1002/gps.835 12673611 185 Losada Baltar A , Izal Fernandez de Troconiz M , Montorio Cerrato I [Differential efficacy of two psychoeducational interventions for dementia family caregivers] . Rev Neurol 2004 ;38 :701 –8 .15122537 186 Moniz-Cook E , Elston C , Gardiner E Can training community mental health nurses to support family carers reduce behavioural problems in dementia? An exploratory pragmatic randomised controlled trial . Int J Geriatr Psychiatry 2008 ;23 :185 –91 . 10.1002/gps.1860 17621379 187 Zarit SH , Anthony CR , Boutselis M Interventions with care givers of dementia patients: comparison of two approaches . Psychol Aging 1987 ;2 :225 –32 . 10.1037/0882-7974.2.3.225 3268213 188 Gitlin LN , Winter L , Dennis MP Targeting and managing behavioral symptoms in individuals with dementia: a randomized trial of a nonpharmacological intervention . J Am Geriatr Soc 2010 ;58 :1465 –74 . 10.1111/j.1532-5415.2010.02971.x 20662955 189 Finnema E , Dröes RM , Ettema T The effect of integrated emotion-oriented care versus usual care on elderly persons with dementia in the nursing home and on nursing assistants: a randomized clinical trial . Int J Geriatr Psychiatry 2005 ;20 : 330 –43 . 10.1002/gps.1286 15799079 190 Schrijnemaekers V , van Rossum E , Candel M Effects of emotion-oriented care on elderly people with cognitive impairment and behavioral problems . Int J Geriatr Psychiatry 2002 ;17 :926 –37 . 10.1002/gps.681 12325052 191 Brodaty H , Arasaratnam C Meta-analysis of nonpharmacological interventions for neuropsychiatric symptoms of dementia . Am J Psychiatry 2012 ;169 :946 –53 . 10.1176/appi.ajp.2012.11101529 22952073 192 Eggenberger E , Heimerl K , Bennett MI Communication skills training in dementia care: a systematic review of effectiveness, training content, and didactic methods in different care settings . Int Psychogeriatr 2013 ;25 :345 –58 . 10.1017/S1041610212001664 23116547 193 Teri L , McCurry SM , Logsdon R Training community consultants to help family members improve dementia care: a randomized controlled trial . Gerontologist 2005 ;45 :802 –11 . 10.1093/geront/45.6.802 16326662 194 Burgio LD , Stevens A , Burgio KL Teaching and maintaining behavior management skills in the nursing home . Gerontologist 2002 ;42 :487 –96 . 10.1093/geront/42.4.487 12145376 195 Burgio LD , Allen-Burge R , Roth DL Come talk with me: improving communication between nursing assistants and nursing home residents during care routines . Gerontologist 2001 ;41 :449 –60 . 10.1093/geront/41.4.449 11490043 196 Done DJ , Thomas JA Training in communication skills for informal carers of people suffering from dementia: a cluster randomized clinical trial comparing a therapist led workshop and a booklet . Int J Geriatr Psychiatry 2001 ;16 :816 –21 . 10.1002/gps.436 11536349 197 Magai C , Cohen CI , Gomberg D Impact of training dementia caregivers in sensitivity to nonverbal emotion signals . Int Psychogeriatr 2002 ;14 :25 –38 . 10.1017/S1041610202008256 12094904 198 Staal JA , Sacks A , Matheis R The effects of Snoezelen (multi-sensory behavior therapy) and psychiatric care on agitation, apathy, and activities of daily living in dementia patients on a short term geriatric psychiatric inpatient unit . Int J Psychiatry Med 2007 ;37 :357 –70 . 10.2190/PM.37.4.a 18441625 199 Clare L , Linden DE , Woods RT Goal-oriented cognitive rehabilitation for people with early-stage Alzheimer disease: a single-blind randomized controlled trial of clinical efficacy . Am J Geriatr Psychiatry 2010 ;18 :928 –39 . 10.1097/JGP.0b013e3181d5792a 20808145 200 Gitlin LN , Winter L , Burke J Tailored activities to manage neuropsychiatric behaviors in persons with dementia and reduce caregiver burden: a randomized pilot study . Am J Geriatr Psychiatry 2008 ;16 :229 –39 . 10.1097/JGP.0b013e318160da72 18310553 201 Gitlin LN , Corcoran M , Winter L A randomized, controlled trial of a home environmental intervention: effect on efficacy and upset in caregivers and on daily function of persons with dementia . Gerontologist 2001 ;41 :4 –14 . 10.1093/geront/41.1.4 11220813 202 Gaugler JE , Kane RL Families and assisted living . Gerontologist 2007 ;47 (suppl 1 ):83 –99 . 10.1093/geront/47.Supplement_1.83 203 Collet J , de Vugt ME , Verhey FR Efficacy of integrated interventions combining psychiatric care and nursing home care for nursing home residents: a review of the literature . Int J Geriatr Psychiatry 2010 ;25 :3 –13 . 10.1002/gps.2307 19513988 204 Zimmerman S , Anderson WL , Brode S Systematic review: effective characteristics of nursing homes and other residential long-term care settings for people with dementia . J Am Geriatr Soc 2013 ;61 :1399 –409 . 10.1111/jgs.12372 23869936 205 Whall AL , Black ME , Groh CJ The effect of natural environments upon agitation and aggression in late stage dementia patients . Am J Alzheimers Dis Other Demen 1997 ;12 :216 –20 . 10.1177/153331759701200506 206 Hicks-Moore SL , Robinson BA Favorite music and hand massage: two interventions to decrease agitation in residents with dementia . Dementia 2008 ;7 :95 –108 . 10.1177/1471301207085369 207 Potter R , Ellard D , Rees K A systematic review of the effects of physical activity on physical functioning, quality of life and depression in older people with dementia . Int J Geriatr Psychiatry 2011 ;26 :1000 –11 . 10.1002/gps.2641 21905096 208 Thuné-Boyle IC , Iliffe S The effect of exercise on behavioral and psychological symptoms of dementia: towards a research agenda . Int Psychogeriatr 2012 ;24 :1046 –57 . 10.1017/S1041610211002365 22172121 209 Burgener SC , Yang Y , Gilbert R The effects of a multimodal intervention on outcomes of persons with early-stage dementia . Am J Alzheimers Dis Other Demen 2008 ;23 :382 –94 . 10.1177/1533317508317527 18453642 210 Rolland Y , Pillard F , Klapouszczak A Exercise program for nursing home residents with Alzheimer's disease: a 1-year randomized, controlled trial . J Am Geriatr Soc 2007 ;55 :158 –65 . 10.1111/j.1532-5415.2007.01035.x 17302650 211 Steinberg M , Leoutsakos JM , Podewils LJ Evaluation of a home-based exercise program in the treatment of Alzheimer's disease: the Maximizing Independence in Dementia (MIND) study . Int J Geriatr Psychiatry 2009 ;24 :680 –5 . 10.1002/gps.2175 19089875 212 Bernabei V , De Ronchi D , La Ferla T Animal-assisted interventions for elderly patients affected by dementia or psychiatric disorders: a review . J Psychiatr Res 2013 ;47 :762 –73 . 10.1016/j.jpsychires.2012.12.014 23369337 213 Churchill M , Safaoui J , McCabe BW Using a therapy dog to alleviate the agitation and desocialization of people with Alzheimer's disease . J Psychosoc Nurs Ment Health Serv 1999 ;37 :16 –22 . 214 McCabe BW , Baun MM , Speich D Resident dog in the Alzheimer's special care unit . West J Nurs Res 2002 ;24 :684 –96 . 10.1177/019394502320555421 12365768 215 Richeson NE Effects of animal-assisted therapy on agitated behaviors and social interactions of older adults with dementia . Am J Alzheimers Dis Other Demen 2003 ;18 :353 –8 . 10.1177/153331750301800610 14682084 216 Libin A , Cohen-Mansfield J Therapeutic robocat for nursing home residents with dementia: preliminary inquiry . Am J Alzheimers Dis Other Demen 2004 ;19 :111 –16 . 10.1177/153331750401900209 15106392 217 Motomura N , Yagi T , Ohyama H Animal assisted therapy for people with dementia . Psychogeriatrics 2004 ;4 :40 –2 . 10.1111/j.1479-8301.2004.00062.x 218 Sellers DM The evaluation of an animal assisted therapy intervention for elders with dementia in long-term care . Act Adapt Aging 2006 ;30 :61 –77 . 10.1300/J016v30n01_04 219 Brush JA , Meehan RA , Calkins MP Using the environment to improve intake for people with dementia . Alzheimer's Care Today 2002 ;3 :330 –8 . 220 Koss E , Gilmore G Environmental interventions and functional ability of AD patients . In: Vellas, B., Fitten, J. and Frisoni, G. eds. Research and practice in Alzheimer’s disease. Paris: Springer, 1998:185–191. 221 Frisoni GB , Gozzetti A , Bignamini V Special care units for dementia in nursing homes: a controlled study of effectiveness . Arch Gerontol Geriatr 1998 ;26 :215 –24 . 10.1016/S0167-4943(98)80031-9 18653138 222 Fontana Gasio P , Kräuchi K , Cajochen C Dawn-dusk simulation light therapy of disturbed circadian rest-activity cycles in demented elderly . Exp Gerontol 2003 ;38 :207 –16 . 10.1016/S0531-5565(02)00164-X 12543279

PMC005xxxxxx/PMC5372077.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01341710.1136/bmjopen-2016-013417Medical Education and TrainingResearch1506170917371709Perspectives on procedure-based assessments: a thematic analysis of semistructured interviews with 10 UK surgical trainees Shalhoub Joseph 1http://orcid.org/0000-0002-3498-2511Marshall Dominic C 2Ippolito Kate 3 1 Faculty of Medicine, Department of Surgery & Cancer, Imperial College London, London, UK 2 School of Medicine, Imperial College London, London, UK 3 Educational Development Unit, School of Professional Development, Imperial College London, London, UKCorrespondence to Joseph Shalhoub; j.shalhoub@imperial.ac.uk2017 24 3 2017 7 3 e01341712 7 2016 3 2 2017 6 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives The introduction of competency-based training has necessitated development and implementation of accompanying mechanisms for assessment. Procedure-based assessments (PBAs) are an example of workplace-based assessments that are used to examine focal competencies in the workplace. The primary objective was to understand surgical trainees' perspective on the value of PBA. Design Semistructured interviews with 10 surgical trainees individually interviewed to explore their views. Interviews were audio-recorded and transcribed; following this, they were open and axial coded. Thematic analysis was then performed. Results Semistructured interviews yielded several topical and recurring themes. In trainees' experience, the use of PBAs as a summative tool limits their educational value. Trainees reported a lack of support from seniors and variation in the usefulness of the tool based on stage of training. Concerns related to the validity of PBAs for evaluating trainees' performance with reports of ‘gaming’ the system and trainees completing their own assessments. Trainees did identify the significant value of PBAs when used correctly. Benefits included the identification of additional learning opportunities, standardisation of assessment and their role in providing a measure of progress. Conclusions The UK surgical trainees interviewed identified both limitations and benefits to PBAs; however, we would argue based on their responses and our experience that their use as a summative tool limits their formative use as an educational opportunity. PBAs should either be used exclusively to support learning or solely as a summative tool; if so, further work is needed to audit, validate and standardise them for this purpose. MEDICAL EDUCATION & TRAININGEDUCATION & TRAINING (see Medical Education & Training)SURGERY ==== Body Strengths and limitations of this study This study employed semistructured interviews; this has a number of advantages over a survey approach, such as avoiding superficial responses and offering the flexibility to follow-up on and seek detail and clarification on key points. There was a single interviewer (JS) for all of the semistructured interviews who has an understanding of the subject and the study. The study is limited by a single data collection method, namely transcribed and coded semistructured interviews. Selection bias may exist as a result of the four procedure-based assessments that the trainee brought with them to the semistructured interview. A further potential source of bias may result from sample selection; however, this study makes no claims on the basis of representativeness or generalisability of the data across trainees, but rather aims to offer in-depth insight into trainees' views. The authors' hold the view that with interpretive qualitative research ‘(r)esearchers should not strive to be objective and look for ways to reduce bias. Rather, they need to face head on the subjective nature of their role’. Introduction Surgical training has evolved over many years as the result of changing circumstances and expectations in healthcare, and in order to meet a number of political, economic, sociological, technological, legal and environmental challenges.1–4 The introduction of competency-based training has necessitated the development and implementation of accompanying mechanisms for the assessment and ‘sign off’ of attainment (or not) of the prescribed competencies. Among the mechanisms introduced were those that examined focal competencies in the workplace; workplace-based assessments (WBAs). WBAs are used in several healthcare systems including the UK, the USA, Canada, New Zealand and Australia.5 The tools currently in use for this purpose include case-based discussion, procedure-based assessment (PBA), mini-clinical evaluation exercise (mini-CEX), direct observation of procedural skill and the mini-peer assessment tool. PBAs involve a direct observation of an index procedure or operation with comments given on important steps, tasks or skills at the preoperative, intraoperative and postoperative stages of the procedure, considered to be essential for its safe and successful completion. PBAs were selected as the WBA to investigate as, owing to their nature and the procedures they assess, they hold a degree of inherent face validity for trainees and their assessors. It was hoped that surgical trainees would, therefore, be able to relate back to the operations which formed the basis of their assessments and to describe and discuss with clarity, illustrating their opinions with vivid examples. It has been argued that WBAs have been implemented in surgical training for summative purposes with limited evidence to support their use.6 There is a large body of theoretical and practice-based evidence to support the educational value of the key components of WBAs, namely authentic assessment,7 continuous feedback8–10 and self-reflection on practice,11 12 but little to convince surgeons of the actual impact of WBAs in their context. The widespread implementation of these tools prior to the existence of satisfactory evidence of their validity is unhelpful, particularly considering their growing influence on career progression. The uptake of WBAs in the UK, as measured by the number of validated WBAs per trainee, has increased steadily since their introduction in 2007. This accompanied the establishment of competency-based training as part of the ‘Modernising Medical Careers’ postgraduate medical training initiative.13 This growth in use may be due to increasingly positive attitudes regarding the educational value of WBAs. However, it has been postulated that it is largely as a result of WBAs becoming compulsory as an assessment of learning, to make a summative judgement that informs progression decisions, as opposed to a developmental, formative assessment for learning. Furthermore, targets have been set with regard to numbers of WBAs required per year of surgical training by the UK Joint Committee on Surgical Training (JCST). There are stipulations with regard to numbers of PBAs to be completed and also the performance levels to be achieved (as well as the number of PBAs to evidence this performance level). This specific use of PBAs is based on research by Marriot et al;14 however, the effectiveness of PBAs when used for a summative purpose extends beyond the scope of this evidence base. An example pro forma PBA has been included as online supplementary figure S1. 10.1136/bmjopen-2016-013417.supp1supplementary figure In the light of this growing usage, this study aims to contribute to the evidence base for the value of WBAs by adding insight into UK trainees' perspectives to previous work that has largely focused on quantitative outcomes and the views of trainers. By exploring the opinions of trainees on their PBA experiences, we hope to clearly highlight the existing educationally effective practice and the challenges currently faced so that this evidence can inform the future development of PBAs. Methods Interview questionnaire development An iterative process was undertaken to develop the schedule of questions for the semistructured interviews, with the aim of compiling a series of relevant, open questions relating to the key topics: questions brainstormed; grouped under headings according to the points that they addressed; groups on which to focus selected; de-duplication; questions refined; time check; refined further; trial-run pilot semistructured interview with a colleague; refined further and the final set of questions selected. Participant selection and recruitment Inclusion criteria were: (1) UK surgical trainees; (2) who gave written informed consent to participate; and (3) registered with and have been using the Intercollegiate Surgical Curriculum Programme (ISCP) for >1 year, hence have developed formal experience of WBAs in postgraduate surgical training. Surgical trainees who have assessed the interviewer (JS) and surgical trainees who have been assessed by the interviewer were excluded from the study. Recruitment for this study was by email sent to members of the UK Association of Surgeons in Training. A purposeful sample of 10 trainees with experience of using PBAs was selected, on the basis that they would provide rich information and useful insight into this aspect of training.15 The aim was not to achieve theoretical saturation but to gain in-depth insight into 10 trainees' unique experiences to better understand how trainees perceive and make sense of the PBAs they are required to complete.16 Data collection A qualitative research methodology was employed to address the research question. Semistructured interviews were completed by JS with the aim of establishing surgical trainees' views in relation to PBAs and their use. The semistructured interviews were guided by the interview schedule (figure 1) and limited to a 1-hour duration during which follow-up questions were permitted. Interviews were digitally audio-recorded and transcribed. Trainees were requested to bring two PBAs, which they felt had helped them to develop their practice, and two PBAs, which they felt did not. These acted as ‘stimulus materials’ for the interview. Figure 1 PBA semistructured interview schedule. PBA, procedure-based assessment. Data analysis The data were analysed using open and axial coding, a well-recognised qualitative approach.17 18 Codes were assigned to units of meaning (ie, portions of text of varying size such as words, sentences and phrases) in the transcripts. Open coding (fracturing of the data and grouping/categorising) was performed, followed by axial coding (rearranging the data in new ways).17 18 This coding was completed manually by JS and reviewed by DCM and KI. Multiple themes were identified in the transcripts and these were grouped under broader themes for discussion.17 We report our findings following the recommendations outlined in the Consolidated Criteria for Reporting Qualitative Research (COREQ);19 a completed checklist is supplied in online supplementary table S1. 10.1136/bmjopen-2016-013417.supp2supplementary table COREQ checklist Results and discussion Ten surgical trainees took part in the semistructured interview process between February and April 2014 totalling more than 374 min; characteristics of the surgical trainees are summarised in table 1. Table 1 Summary of the characteristics of the surgical trainees who participated in this study Number of participants 10 Age range (years) 29–36 Gender (M:F) 6:4 Stages of surgical training Core training 2 to specialty training 7 Surgical specialties Ear, nose and throat surgery Orthopaedic surgery General surgery (colorectal) General surgery (vascular) General surgery Years of qualification 2001–2009 Training regions Oxford Eastern Wales North West Thames West of Scotland East Midlands South Wales Defence/Wessex Duration of semistructured interview (min:s) 13:45–59:29 Systematic analysis of the thematically coded interview transcripts enabled identification of 68 themes, which were further categorised under 5 major themes (see online supplementary table S2): Use of PBAs to derive educational value, Use of PBAs for assessment, Trainer dependent, Strengths and benefits of PBAs, Limitations and areas for improvement. 10.1136/bmjopen-2016-013417.supp3supplementary table Summary of the themes which emerged from the semi-structured interviews Use of PBAs to derive educational value One of the proposed educational values of PBAs as WBA tools is their ability to identify the trainee in difficulty;20 however, this was not necessarily the experience of trainees in this study:I am sure that certainly in my last placement there were times where I was a failing trainee in the sense that I didn't feel that I was getting the hang of a particular operation or I was frustrated in my training but PBA didn't rescue me. ST3 general surgery This interviewee felt that PBAs are not being used in the manner in which they were designed and intended. In order for PBAs to be effective: (1) the trainer would have to observe this trainee undertaking the procedure; (2) the trainer would have to give feedback on the procedure; (3) this feedback would be specific and detailed; (4) the feedback would be formative, addressing this trainee's shortcomings in relation to the ‘particular operation’ and offer targeted direction to overcome any technical or non-technical deficiencies the trainee may feel that they have. Another trainee described the challenges related to imposing a quota necessary for PBAs to be completed and that leading to a box ticking, educationally invalid experience; this view was shared by all subsequent interviewees. These points are echoed by the results of a survey of 500 medical trainees conducted by the Joint Royal Colleges of Physicians Training Board (JRCPTB) where 67% of trainees stated that WBAs were too time-consuming, with a large proportion concerned that their supervisors did not have time to complete the assessments, often leaving trainees to complete them—including writing feedback.21 Furthermore, a survey of supervisors by the JRCPTB showed that two-thirds did not find WBAs effective in identifying underperforming trainees21 and that many felt WBAs to be an unproductive ‘tick box’ exercise. Bindal et al22 identified time pressure as a major source of dissatisfaction, and claimed that WBAs were widely viewed by trainees as a summative tick box exercise with little educational value.23 Interviewee 10 provided the perspective of a trainee who is nearing the end of their training, who requires global feedback and focused training to meet the requirements of independent practice. This is educationally at odds with the ‘point-by-point’ feedback mandated by PBAs:I am wanting to be assessed more globally. A sort of overall impression of “am I almost ready for consultant practice or not?” rather than a specific nitty gritty “did you do this? Yes/No” and “did you do it” or on a particular point “satisfactory” or “ready for CCT”. I don't know. I think that they task down into so many tiny little constituent parts that sometimes the global overview is lost. ST7 general (colorectal) surgery Indeed, from October 2013, the JRCPTB reverted to a multiple consultant report to offer structured feedback on trainees' overall clinical abilities from a range of supervisors.21 The use of assessments performed on multiple occasions, by multiple assessors, in this way allows for ‘triangulation’ of assessment, which is an important determinant of assessment reliability.20 23 24 Interviewee 10's desire to be assessed more globally resonates with concerns25 26 voiced internationally that the introduction of competency-based postgraduate medical training created a gap between the assessment of specific competencies and actual clinical practice, emphasising competency over expertise. In order to bridge this gap ten Cate and Scheele argue for the use of entrustable professional activities (EPAs), a process that involves assessing global performance in all the competencies required to carry out critical professional activities such as running an operating list or perioperative care.25 Essentially, EPAs assess an individual's readiness to be entrusted with professional responsibility. Use of PBAs for assessment Previous work has demonstrated the effectiveness of PBAs for summative assessment of specific procedures and performance correlates well with surgical experience.14 Based on this, the JCST has advised a minimum number of PBAs to demonstrate minimum competence in certain specified procedures. However, the summative role of PBAs appears to have expanded beyond this remit. Further quantifying numbers of PBAs may have the unintended consequence of trainees viewing all PBAs as summative assessments. A number of points and areas for concern were raised by participants in relation to the use of PBAs for assessment. First, some trainees felt that PBAs were not specific to a level of training. In particular, one senior trainee felt that they were not fit for purpose as an assessment tool for more senior trainees as, owing to their ‘point-by-point’ nature, a global overview is lost. Second, it was the opinion of some interviewees that the use of PBAs for summative assessment conflicted with and prevented their use as an effective formative learning tool. Third, in the experience of our cohort, many of their trainers had not been trained to complete PBAs, or told trainees to complete their own PBAs. This may reduce the validity of the PBA outcomes for formative or summative assessment. Interviewee 5 in particular raised many interesting points about the use of WBA as tools for formative feedback rather than summative assessment and the validity of the PBA as a tool for summative assessment:Whereas I suspect that if that aspect of their use was completely put to the side and you were just told this is a purely formative tool for your interest etc., I suspect trainees would apply them a bit more like they were designed …So I still fundamentally think that a report from a trainer over 6 months or a year or your progress and general competency in multiple fields is by far the most powerful tool although it's difficult to quantify numerically or whatever like a workplace based assessment does. So my personal view is workplace based assessments should firmly be formative tools that aid in the teaching process or the training process and perhaps their number can be used as a surrogate of engagement with the process but I do not think that this should be used as a summative tool. ST6 general (vascular) surgery The view of interviewee 5 was supported by interviewee 3 with regard to the importance of PBAs for formative purposes:I think the most important thing about all of these things is the feedback. It's not the mark you get at the end of the day. ST3 general (colorectal) surgery There appears to be a misalignment of conception between those who developed WBAs (for formative use) and the JCST which has implemented WBAs (for a combination of formative and summative purposes).27 This quote is one of a number of examples given that relate to concerns over implementation of PBAs, as opposed to, primarily, an issue with the content and format of PBAs. In a recent survey of orthopaedic surgeons, less than half of trainees and trainers had received training in the use of PBAs.28 Interestingly, 53% (67% of consultants and 44% of registrars) were unsure whether PBAs had a formative or summative assessment purpose, and 13% (10% of consultants and 15% of registrars) believed they were summative.28 The article concluded with the important reminders that all users should familiarise themselves with PBAs, that PBAs were designed to be formative and that successful completion of a PBA at any level does not constitute a ‘licence to operate’.28 Some argue that the key to the success of the PBA is careful implementation.29 Provided trainees receive feedback and have confidence in the assessment system, improvement in competence will take place.29 30 However, this link between feedback and actual improved practice is not given. Indeed, it has been reported that only 8% of feedback or areas for development identified through mini-CEX were converted into an action plan.11 31 Currently, trainers also fulfil the role of assessors for PBAs. Many trainers have very little or no training in assessment and, although training might help trainers to make better judgements in assessment, there remains the issue of the trainer being both a teacher and assessor, with the potential for conflict of interest. Being both teacher and assessor are components of an integrated educational programme, but it is preferable that teaching and summative assessments are not carried out by the same assessor.32 However, if PBAs were deployed in a purely formative way, the trainer who is responsible for a trainee's development could use PBAs and the feedback and action points from these to guide future training (ie, as assessment for learning). It appeared that there was an element of ‘gaming’ of WBAs by some trainees who have their eye on securing high summative scores. An example of ‘gaming’ of PBAs as described by some trainees include: selecting ‘doves’ (less critical trainers/assessors) to complete their PBAs; and selecting cases where they have performed particularly well for assessment (and not completing a PBA for cases where their performance was, in their view, less good). In 2009, Munsch,33 the Chairman of the UK JCST, stated that WBAs are “feedback tools to guide and direct training, rather than summative competency-based assessments.”27 To clarify the formative role of WBAs as assessments for learning, it has even been suggested that WBAs be renamed to workplace-based assessment for learning.34 Trainer dependent Eight of the 10 participants commented on obstruction from seniors, or a lack of support or engagement from trainers in the PBA process. For example, when questioned on the negatives of PBAs, one interviewee responded:The main obstacle or issue I find is from higher levels and not actually from me getting them achieved…finding consultant input to sit down and go through them with to sign them off was a difficulty…it's so difficult to get them… CT2 orthopaedic surgery The critical role of faculty in WBAs has been highlighted in the literature, as has the need for strategies to enhance their participation and training.11 Furthermore, since the quality of training is dependent on the effectiveness of the surgeon–teachers in their roles as educators and assessors, as well as proficient professional (surgeon) service providers, the evaluation of these qualities in the surgeon–teacher forms an important component of the WBA.35 There has been a move away from the presumption that all surgeons are good trainers and that having a trainee as a surgical consultant is a right. One of the implications of this study is to highlight that additional senior support is necessary for trainees to facilitate the use of WBAs. Strengths and benefits of PBAs Undoubtedly, PBAs have strengths and benefits; the interviews highlighted the huge educational potential of PBAs. A positive aspect, commented on by more than half the trainees, was the value of PBAs in documenting progression:I have had consultants who have said we are doing that operation on Thursday. I want you to go and prepare for it. It's an operation that you haven't seen much and I've gone away and done my homework and got some of the best training I could because both of us had shown up that day. You've done your homework about the operation, what it involves, all the steps and they are there ready to train you because they have set that aside beforehand. ST4 ear, nose and throat (ENT) surgery This approach could certainly be captured and encouraged if the formative role of PBAs were emphasised. Also, if the intent to complete a PBA was declared well in advance, for example, when the cases for a particular operating list are determined, this would improve the alignment of trainee development need with training opportunities. This, however, should not preclude the use of PBAs opportunistically, at short notice, if it is felt that a particular case, for example, an emergency procedure, presented an opportunity for feedback, reflection and generation of an action plan for learning. Another benefit of PBAs identified by this study is the effect they can have on standardising the assessors, as they are required to complete a pro forma assessing a range of aspects of the trainee's performance:So it can almost bring different assessors to the same level. So if you have got an assessor who doesn't look at the whole picture, it can make them look at the whole picture because it looks at everything. CT2 orthopaedic surgery The majority of trainees also identified that PBAs themselves can generate both training and reflection. A trainee may use a PBA to prompt a senior to instigate a training session. The majority of surgical training is ‘on-the-job’ and spontaneous. PBAs provide a method for formalising this learning and providing a record that it occurred. Further, when completed thoroughly, encouraging development over time with repeat assessments, PBAs may provide a very valuable tool. Pelgrim et al36 suggest a stepwise process to support this and obtain maximum effectiveness. Limitations and areas for improvement It was highlighted that, in the experience of our interviewees, PBAs are not being used as they were intended or designed. Interviewee 6 described having had virtually no experience of WBA being used as intended, and that the forms were predominantly filled in retrospectively by the trainee:It's a really difficult question. The only reason it is difficult…is because I have had little experience of them being used properly. ST4 general (vascular) surgery Such comments are only a limited reflection on the PBAs themselves and somewhat reflect the system that they are being used in and those individuals who use them (both trainees and trainers). Indeed, a lack of awareness of the educational rationale for PBAs and, therefore, poor implementation seems to be at the heart of the challenges described. Proponents of WBAs state that: “absence of evidence of benefit does not mean that WBAs, when undertaken correctly, do not work.”37 The trainees describe efforts to improve the way in which PBAs are used and to reverse a number of ‘bad habits’ that crept into the system when it was initially launched:They are having to undo a lot of damage that was done when workplace based assessments generally were introduced without really any particular training or fanfare. They were just dropped on trainees and trainers and I think unfortunately that has fostered a lot of resentment on both sides that is still being felt many years down the line. ST5 general (colorectal) surgery In addition to having a potentially negative impact on individual trainees, it is pertinent to consider the impact of the introduction, use and perception of these tools on the training system as a whole:My general impression is that perhaps not everybody or every trainer is using them in the manner to which they were intended and so I think they may be actually giving some sort of a slightly false impression of standardisation in training. ST5 general (colorectal) surgery As this interviewee suggested, rather than raising standards in postgraduate surgical training, PBAs could actually be blinding problems. Suggestions for improvement: the trainee perspective Suggestions for improving PBAs related to how they were used, as opposed to the design of the PBAs themselves. Recurring themes included improving the quality of the feedback and changing the ethos of the process, so that it is more educational and less stressful as detailed below. The use of PBAs Interestingly, some of the interviewees' proposed solutions related to PBAs to be used in the way they were intended and included providing more evidence that a formative trainee–trainer interaction had occurred, eliminating the tick box, blind sign-off:If the structuralise thing helps then you can use that as your form but to upload a recorded conversation between you and your consultant about what you learned because then there is clear evidence of that discussion, that training having taken place from both your angle and the consultants. You can't do it retrospectively particularly easily. You can't just be signed off. It would need to be something that you sat down and did. ST4 ENT surgery Here, the trainee spoke about a genuine dialogue between the trainee and trainer. Nicol12 argues that feedback should be conceptualised as a dialogue between the giver and receiver. This feedback dialogue carried out before and after the PBA observation would significantly maximise the benefit of PBAs and other WBAs and represents one of the few examples of the combination of written and immediate verbal feedback in postgraduate medical education. This is supported by a recent systematic review highlighting that feedback given in verbal and written form appears to be more beneficial than either of these forms alone.38 Half of the participants commented that trainees often fill in their own PBA forms, and all of the participants highlighted the limited or complete lack of review of the trainees' and trainers' free-text boxes by subsequent assessors, for example, at the Annual Review of Competence Progression. However, if the PBA process is to have educational value, it is important for trainees to read their trainers' comments, use this to inform reflections on their performance and document this. The majority of participants described PBAs as being stressful. PBAs were designed to support trainees' development and certainly not to cause them undue stress. Our findings indicate that it is the way that PBAs have been used and the pressure to accrue large numbers of them that is a cause of stress. This is compounded by the fact that they have been used for summative assessment, which is often inherently stressful. Improving the quality of feedback: a self-assessment Trainees suggested that using self-reflection to make the feedback more trainee-specific and using the PBA as a tool for prompting solicited and targeted trainer feedback could improve the relevancy and acceptability of the feedback the trainee receives, as the following quote describes.I think you should rate yourself and send that to the consultant and along the boxes he should tick either “agree” or “disagree” and then put comments. And I think that would be better feedback for your own strengths and weaknesses because an area that you think you have done really well, the consultant may actually think that you need to develop, you weren't slick or you weren't safe with your instruments or something. ST3 general (colorectal) surgery Strengths and limitations This study has a number of important strengths. A semistructured interview has many advantages over a survey approach, such as avoiding superficial responses and offering the flexibility to follow-up on and seek detail and clarification on key points and interesting comments raised by the trainees. There was a single interviewer (JS) for all of the semistructured interviews who has an understanding of the subject and the study. He was, therefore, able to pursue lines of questioning and explore areas of interest where the trainee showed a strength of feeling or opinion. There are a number of limitations to this study. First, the study involved a single data collection method, namely transcribed and coded semistructured interviews. Second, there was selection bias with regard to the selection of the four PBAs that the trainee brought with them to the semistructured interview. This bias is somewhat deliberate and was designed to highlight extremes, both in terms of positive and negative assessment and feedback experiences, and to prompt reflection. This reflection was evident in the trainees' responses. There exists a selection bias in relation to the themes which were drawn from the transcripts, as well as those themes identified for further discussion; however, the systematic data analysis approach aimed to identify those themes that appeared most important to interviewees. A further potential source of bias is JS, a surgical trainee himself, as interviewer. Third, with regard to sample selection, the study makes no claims on the basis of representativeness or generalisability across trainees, but rather aimed to offer in-depth insight into trainees' views. This is an evolving area of surgical training and further work is required to investigate the changing perspective of trainees over time. Future studies may also gain more insight into the value of PBAs at different stages of training and any associations between trainee understanding of purpose PBAs and their value. Conclusion A change of the current culture within medical education has certainly been observed and there is a recognition that WBAs are here to stay. This study contributes to evidence that the use of PBAs as a summative tool compromises their formative use as an educational opportunity. It also further highlights issues with senior support for the process and the way in which PBAs are being conducted. However, there are examples where PBAs have been of great value to the trainee. Further, training and understanding in the use of PBAs may allow them to be of greater benefit to the trainee. We suggest, therefore, that a fork in the road may have been reached whereby PBAs are either used exclusively as a formative educational tool, with resources allocated to ensure that this is done properly, or efforts must be made to gather evidence for their use to summatively assess competency. The authors would like to thank the surgical trainees who volunteered their time to participate in this study. Contributors: JS and KI were responsible for project planning and study design. JS conducted interviews and coded data. JS, DCM and KI were responsible for data analysis and interpretation of results. All authors contributed to the first draft of the manuscript and have reviewed the final version before submission. Funding: This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Competing interests: None declared. Ethics approval: Imperial College London Education Ethics Review Process (EERP1314-004). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Shalhoub J , Giddings CE , Ferguson HJ Developing future surgical workforce structures: a review of post-training non-Consultant grade specialist roles and the results of a national trainee survey from the Association of Surgeons in Training . Int J Surg 2013 ;11 :578 –83 . 10.1016/j.ijsu.2013.09.010 24075967 2 Fitzgerald JE , Giddings CE , Khera G Improving the future of surgical training and education: consensus recommendations from the Association of Surgeons in Training . Int J Surg 2012 ;10 :389 –92 . 10.1016/j.ijsu.2012.03.012 22449833 3 Vesey A The shape of your own training—time to get engaged 2014. http://www.jcst.org/blog/the-shape-of-your-own-training-time-to-get-engaged (accessed 12/10/2014 ). 4 Chartered Institute of Personnel and Development. PESTLE analysis 2013 http://www.cipd.co.uk/hr-resources/factsheets/pestle-analysis.aspx 5 Miller A , Archer J Impact of workplace based assessment on doctors’ education and performance: a systematic review . BMJ 2010 ;341 :c5064 10.1136/bmj.c5064 20870696 6 Bodgener S , Tavabie A Is there a value to case-based discussion? Educ Prim Care 2011 ;22 :223 –8 . 10.1080/14739879.2011.11494004 21781388 7 Miller GE The assessment of clinical skills/competence/performance . Acad Med 1990 ;65 (9 Suppl):S63 –7 . 10.1097/00001888-199009000-00045 2400509 8 Cantillon P , Sargeant J Giving feedback in clinical settings . BMJ 2008 ;337 :a1961 10.1136/bmj.a1961 19001006 9 Sargeant J , Mann K , Ferrier S Exploring family physicians’ reactions to multisource feedback: perceptions of credibility and usefulness . Med Educ 2005 ;39 :497 –504 . 10.1111/j.1365-2929.2005.02124.x 15842684 10 Sargeant J , Mann K , van der Vleuten C “Directed” self-assessment: practice and feedback within a social context . J Contin Educ Health Prof 2008 ;28 :47 –54 . 10.1002/chp.155 18366127 11 Norcini J , Burch V Workplace-based assessment as an educational tool: AMEE Guide No. 31 . Med Teach 2007 ;29 :855 –71 . 10.1080/01421590701775453 18158655 12 Nicol D From monologue to dialogue: improving written feedback in mass higher education . Assess Eval Higher Educ 2010 ;35 :501 –17 . 10.1080/02602931003786559 13 Shalhoub J , Santos C , Bussey M A descriptive analysis of the use of workplace-based assessments in UK surgical training . J Surg Educ 2015 ;72 :786 –94 . 10.1016/j.jsurg.2015.03.019 25980830 14 Marriott J , Purdie H , Crossley J Evaluation of procedure-based assessment for assessing trainees’ skills in the operating theatre . Br J Surg 2011 ;98 :450 –7 . 10.1002/bjs.7342 21254025 15 Savin-Baden M , Howell Major C Qualitative research: the essential guide to theory and practice . Routledge , 2012 . 16 O'Reilly M , Parker N ‘Unsatisfactory saturation’: a critical exploration of the notion of saturated sample sizes in qualitative research . Qual Res 2013 ;13 :190 –7 . 10.1177/1468794112446106 17 Corbin J , Strauss A Grounded theory research—procedures, canons and evaluative criteria . Z Soziol 1990 ;19 :418 –27 . 18 Strauss AL , Corbin JM Basics of qualitative research: grounded theory procedures and techniques . Newbury Park, CA : Sage Publications , 1990 . 19 Tong A , Sainsbury P , Craig J Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups . Int J Qual Health Care 2007 ;19 :349 –57 . 10.1093/intqhc/mzm042 17872937 20 Bookless LR , Jones AE , Phillips AW What evidence is there for the use of workplace-based assessment in surgical training? J Surg Educ 2015 ;72 :367 –8 . 10.1016/j.jsurg.2014.11.011 25595858 21 Board JRCoPT . Recommendations for specialty trainee assessment and review—incorporating lessons learnt from the workplace-based assessment pilot , 2014 https://www.jrcptb.org.uk/sites/default/files/April%202014%20Recommendations%20for%20specialty%20trainee%20assessment%20and%20review.pdf 22 Bindal T , Wall D , Goodyear HM Trainee doctors’ views on workplace-based assessments: are they just a tick box exercise? Med Teach 2011 ;33 :919 –27 . 10.3109/0142159X.2011.558140 22022902 23 Ali JM Getting lost in translation? Workplace based assessments in surgical training . Surgeon 2013 ;11 :286 –9 .23597666 24 Norcini J The validity of long cases . Med Educ 2001 ;35 :720 –1 . 10.1046/j.1365-2923.2001.01006.x 11489096 25 ten Cate O , Scheele F Competency-based postgraduate training: can we bridge the gap between theory and clinical practice? Acad Med 2007 ;82 :542 –7 . 10.1097/ACM.0b013e31805559c7 17525536 26 van Loon KA , Driessen EW , Teunissen PW Experiences with EPAs, potential benefits and pitfalls . Med Teach 2014 ;36 :698 –702 . 10.3109/0142159X.2014.909588 24804911 27 Nicholas R , Whitaker J Work-based assessments: do not forget the formative! J Surg Educ 2014 ;71 :281 . doi: 10.1016/j.jsurg.2014.01.016.24797840 28 Roushdi I , Tennent D Current usage patterns of procedure-based assessments in the orthopaedic community . Bull R Coll Surg Engl 2015 ;97 :e1 –3 . 10.1308/147363515X14134529300148 29 Mendes da Costa T Procedure-based assessments: an appropriate assessment tool? Bull R Coll Surg Engl 2014 ;96 :236 –8 . 10.1308/147363514X13990346756409 30 Smither JW , London M , Reilly RR Does performance improve following multisource feedback? A theoretical model, metaanalysis and review of empirical findings . Pers Psychol 2005 ;58 :33 –66 . 31 Holmboe ES , Yepes M , Williams F Feedback and the mini clinical evaluation exercise . J Gen Intern Med 2004 ;19 (Pt 2 ):558 –61 . 10.1111/j.1525-1497.2004.30134.x 15109324 32 Subcommittee PWBA . Workplace based assessment , 2005 . 33 Munsch C British surgeons’ experiences of mandatory online workplace-based assessment—reply to authors . J R Soc Med 2009 ;102 :357 –8 . 10.1258/jrsm.2009.090278 34 Beard J Workplace-based assessment: the need for continued evaluation and refinement . Surgeon 2011 ;9 (Suppl 1 ):S12 –13 . 10.1016/j.surge.2010.11.014 21549983 35 Norcini J , Talati J Assessment, surgeon, and society . Int J Surg 2009 ;7 :313 –17 . 10.1016/j.ijsu.2009.06.011 19573630 36 Pelgrim EA , Kramer AW , Mokkink HG The process of feedback in workplace-based assessment: organisation, delivery, continuity . Med Educ 2012 ;46 :604 –12 . 10.1111/j.1365-2923.2012.04266.x 22626052 37 Beard J , Marriott J , Purdie H Workplace based assessments as opportunities for learning. Response to: impact of workplace based assessment on doctors’ education and performance: a systematic review . BMJ 2010 ;341 10.1136/bmj.c5064 38 Ivers N , Jamtvedt G , Flottorp S Audit and feedback: effects on professional practice and healthcare outcomes . Cochrane Database Syst Rev 2012 ;6 :CD000259 10.1002/14651858.CD000259.pub3

PMC005xxxxxx/PMC5372078.txt

==== Front BMJ Open Diabetes Res CareBMJ Open Diabetes Res CarebmjdrcbmjdrcBMJ Open Diabetes Research & Care2052-4897BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjdrc-2016-00030610.1136/bmjdrc-2016-000306Cardiovascular and Metabolic Risk15061870Association of physiological and psychological health outcomes with physical activity and sedentary behavior in adults with type 2 diabetes Garcia Jeanette M 1Cox Daniel 2Rice David J 3 1 Department of Education and Human Services, University of Central Florida, Orlando, Florida, USA 2 Departments of Psychiatry, Internal Medicine, and Opthalmology, University of Virginia, Charlottesville, Virginia, USA 3 School of Nursing and Health Sciences, Florida Southern College, Lakeland, Florida, USA,Correspondence to Dr Jeanette M Garcia; Jeanette.garcia@ucf.edu2017 29 3 2017 5 1 e00030628 7 2016 24 1 2017 3 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Purpose To examine the association between change in moderate-to-vigorous physical activity (MVPA) and sedentary behavior (SB) over a 6-month period with physiological and psychological factors in adults with type 2 diabetes (T2D). Methods Participants included 26 middle-aged (mean age=56.1±10.8 years; 42% women), overweight/obese (mean body mass index (BMI) =37.22±8.78 kg/m2) adults who had been diagnosed with T2D within the past 5 years (mean HbA1c=7.81%). Participants underwent a physical examination, blood tests, and psychological questionnaires, including a self-report questionnaire that assessed the consumption of high glycemic and low glycemic load foods. Participants wore an Actigraph accelerometer for 7 days to assess MVPA and SB. All measures were collected at baseline and at the 6-month follow-up. Spearman rank correlations and regression models were conducted to examine the relationship between activity variables, and the association of activity measures with health outcomes at the 6-month follow-up. Results Decreases in duration of SB bouts and increases in MVPA were associated with decreased levels of HbA1c (p<0.05). Over 50% of the variance in HbA1c levels could be attributed to changes in MVPA and SB. Conclusions MVPA and SB were independently associated with diabetes-related health outcomes. Results suggest that emphasis should be placed on increasing MVPA while decreasing SB, particularly duration of SB bouts. This suggests that even small changes in daily behavior may contribute to improvement in diabetes-related health outcomes. Physical Activity and HealthPsychosocial FactorsSedentary BehaviorsA1C ==== Body Key messages Physical activity and sedentary behavior are distinct behaviors associated with different diabetes-related outcomes. Decreases in uninterrupted bouts of sedentary behavior are associated with reduction in HbA1c levels. Increased moderate-to-vigorous physical activity is associated with a reduction in HbA1c levels and improvements in psychosocial outcomes such as a reduction in depressive symptoms. Small changes in sedentary behavior, such as breaking up uninterrupted bouts of sedentary behavior with a few minutes of light activity, may be significant enough to improve diabetes-related outcomes. Introduction There has been a rise in the prevalence of type 2 diabetes (T2D) in the USA, which may be correlated with the increasing levels of obesity, poor dietary habits, and physical inactivity.1 In fact, an estimated 29.1 million Americans are currently living with T2D, with 1.4 million new cases being diagnosed each year.2 Over 422 million cases of T2D have been diagnosed worldwide.1 Diabetes is reported to be the 7th leading cause of death in the USA, and if left untreated, can lead to numerous health complications such as blindness, neuropathy, and infection. Through a combination of medication and healthy lifestyle behaviors (diet, physical activity), however, T2D can be effectively managed and individuals can live a full and healthy life.2 In response to the increase in prevalence of T2D, public health recommendations have promoted increases in moderate-to-vigorous physical activity (MVPA).3 However, there is emerging evidence that attention should also be focused on sedentary behavior (SB), which is directly related to chronic disease, independent of physical activity.4 5 In fact, SB has been associated with obesity, abnormal glucose metabolism, and the metabolic syndrome.6 In addition, a recent meta-analysis by Biswas et al7 found a 34% higher risk for all-cause mortality for prolonged sedentary time, even after adjusting for MVPA. Another meta-analysis by Wilmot et al8 found similar results, reporting that greater levels of SB were associated with a twofold increase in relative risk of T2D (RR=2.12 (1.61–2.78)) as compared to the group with the lowest levels of SB. In addition to the effects of total sedentary time, the manner in which it is accumulated may also be important. Single bouts of prolonged, uninterrupted SB has been shown to decrease insulin sensitivity in adults, while additional evidence shows that breaking up bouts of SB may lead to improved metabolic profiles, however, few studies have examined these associations in individuals with T2D.3 9 In addition to physiological outcomes, evidence suggests that psychological outcomes may be connected to activity behavior, such that increases in MVPA may be associated with improved mood and psychological well-being in adults.10 Although the literature on SB and psychological well-being is less known, there have been a few studies suggesting that increases in SB may be associated with adverse behavioral outcomes such as depression.11 Poorer psychological health may be associated with poorer prognosis in individuals with T2D.12 Therefore, the purpose of this study was to examine the association between changes/improvements in MVPA and SB over a period of 6 months with physiological and psychological health outcomes for individuals recently diagnosed with T2D. Methods Participants were 26 adults, ranging in age from 33 to 77 years old, with T2D (mean age: 56.1±10.8 years, 42% women, mean of 2.1±1.7 years since diagnosis) who were part of a larger study13 designed to examine the effects of a lifestyle intervention on diabetes outcomes. Out of the 26 adults, 14 participants were part of the intervention arm, which consisted of an intensive lifestyle modification group that emphasized better nutritional choices and increased exercise, and the remaining 12 were part of the control group, which focused on diabetes support and education. These 26 participants were included in this smaller analysis because they had complete objective measures of physical activity and SB at baseline and postintervention. Additional inclusion criteria of the parent study were: have been diagnosed with T2D within the past 5 years, between 24 and 80 years of age, have an Hba1c level of >7.0%, and the approval of their treating physician to participate in the study and signed an informed consent approved by the University of Virginia Institutional Review Board. Participants completed a physical examination, blood tests, a 6 min walk test to assess physical fitness, psychological questionnaires, and wore devices to assess their MVPA and SB for 1 week at baseline and at 6-month follow-up. Details of these measures are presented below. Physiological measures All physiological measures were assessed during a clinic visit at baseline and then at the 6-month follow-up visit. Height was measured to the nearest millimeter using a wall-mounted stadiometer. Weight was assessed to the nearest 0.1 kg using a spring scale that was calibrated daily. Metabolic markers were assessed using venous blood samples, following an overnight fast. Blood tests included: HbA1c TOSOH G7 (HPLC), CRP-HS Abbott Architect (latex immunoassay), and Insulin Immulite 2000 (sandwich immunoassay). Physical activity/sedentary behavior The ActiGraph GT3X+ device was used to assess physical activity (ActiGraph, LLC, Pensacola, Florida, USA). The ActiGraph has the ability to detect normal human motion while filtering out high-frequency vibrations that would artificially increase movement data, and has been validated for use in adults in laboratory and field studies.14 Data were collected in 1 min epochs and postprocessed, using ActiLife software, into counts per minute (CPM). Participants were instructed to wear the devices for 7 days, and to only remove them when they were sleeping or engaging in water-based activities (eg, swimming). To be included in the analysis, participants were required to have at least 3 days of at least 600 min of data per day, with at least two weekdays and one weekend. Times of ≥60 min of accelerometer CPM values =0 were considered times when the device was not worn and were excluded from the analysis.15 Values ≥2296 CPM were classified as moderate-to-vigorous physical activity (MVPA) and values <100 CPM were classified as SB.14 16 Bouts of SB were defined as average consecutive minutes spent in sedentary time, and breaks in sedentary time were defined as interruptions (>100 CPM) in sedentary bouts. Average steps per day were also recorded in participants. Aerobic fitness Participants were instructed to complete a 6 min walk test to assess exercise tolerance.17 This test is self-paced, and measures the distance an individual is able to walk over a total of 6 min on a hard, flat surface. The primary outcome is the total distance covered in meters that participants traveled over the 6 min time period. A lower score, which would reflect less distance covered in 6 min, would indicate lower aerobic fitness. This instrument has high test–retest reliability, with an intraclass correlation (ICC) of 0.80 reported in a group of older adults who had chronic heart failure and associated comorbidities including diabetes and hypertension.18 Moderate to high relationships have been reported (r=0.56–0.88) between the 6 min walk test and peak oxygen uptake, assessed by maximal exercise testing, in older populations.19 20 Psychosocial questionnaires Quality of life Quality of life was assessed with the 26 item WHO Quality of Life (WHOQOL-BREF) Questionnaire.21 This questionnaire is a modified version of the original 100 item WHO Quality of Life (WHOQOL-100) Questionnaire, and consists of four of the following domains: physical health (seven items), psychological health (six items), social relationships (three items), and environmental health (eight items). The response items on the questionnaire ranged from 1 to 5. Similar to WHOQOL-100, domain scores are calculated by multiplying the mean of all items in the domain by 4. Previous studies have shown high correlations between the WHOQOL-BREF and the original WHOQOL-100, ranging from 0.89 (domain 3) to 0.95 (domain 1). Internal consistency was good, with Cronbach's alphas ranging from 0.66 (domain 3) to 0.84 (domain 1) in the current investigation. Empowerment Diabetes empowerment was assessed using the 8-item short form of the original 28-item Diabetes Empowerment Scale (DES-SF).22–23 The original questionnaire consisted of eight conceptual dimensions (assessing the need for change, developing a plan, overcoming barriers, asking for support, supporting oneself, coping with emotion, motivating oneself, and making diabetes care choices appropriate for one's priorities and circumstances). The 8-item version was created by choosing from the 28 items that had the highest item to subscale correlation from each of the eight conceptual dimensions. Response options ranged from 1 (strongly disagree) to 5 (strongly agree). The questionnaire starts with the phrase ‘In general, I believe that I…’ and is followed by eight statements such as ‘…know what part(s) of taking care of my diabetes that I am dissatisfied with’ and ‘…am able to turn m diabetes goals into a workable plan’. The DES-SF demonstrated good reliability with α=0.84. Depressive symptoms Depressive symptoms were assessed using a shorter version of the Patient Health Questionnaire (PHQ).24 The PHQ-9 is the 9-item depression module from the original PHQ. This questionnaire can be used as a screening diagnostic tool for depressive symptoms, and to assess depression severity. The questionnaire opens with, ‘Over the last 2 weeks, how often have you been bothered by any of the following problems?’ and then lists nine statements that assess depressive symptoms such as ‘little interest or pleasure in doing things’ and ‘feeling down, depressed, or hopeless’. The PHQ-9 score can range from 0 to 27, since each of the 9 items can be scored as 0 (not at all) to 3 (nearly every day). Previous studies have reported excellent internal reliability, citing Cronbach's alphas of 0.86 and 0.89.25 Diabetes-related stress Diabetes-related stress was assessed using the 5-item Problem Areas in Diabetes (PAID) Scale.26 This scale was created as a short-form version of the original 20-item PAID Questionnaire. The PAID-5 has a 5-point response option ranging from 0 (not a problem) to 4 (serious problem). Total scores on the PAID-5 can range from 0 to 20, with higher scores suggesting greater diabetes-related emotional distress. McGuire et al27 found that the PAID-5 had satisfactory sensitivity (94%) and specificity (89%) for recognition of diabetes-related emotional distress. Additionally, this measure reported good internal reliability, with Cronbach's alphas ranging from 0.83 to 0.86.27 Statistical analysis Descriptive statistics were calculated to determine how many of the participants met the PA recommendations of 30 min of daily MVPA, recommended by the American Heart Association, at baseline and 6 months. Physiological and psychological health factors were assessed by repeated measures t-tests to examine the change in values over the 6-month period. Additionally, repeated measures t-tests were also used to examine the differences between activity measures at baseline and at the 6-month follow-up. Spearman Rank correlations were conducted to examine the association among activity variables. Linear regression models, controlling for age, gender, body mass index (BMI), were conducted to examine whether activity variables (MVPA, SB, and fitness) independently predicted diabetes-related outcomes. All analyses were conducted in SAS V.9.3 with a significance level set α <0.05. Results On average, participants were obese, engaged in low levels of MVPA, and were largely sedentary during the day. Approximately 50% of participants were on diabetes-related medications. At baseline, only two out of the 26 (8%) met the recommended daily MVPA levels of 30 min per day, and three (12%) met the recommended MVPA levels at 6 months. Over the 6-month period, participants increased their MVPA, on average, by 6 min per day, while decreasing their total time in SB by around 20 min. In addition, participants decreased their uninterrupted bouts of SB by almost 10 min. While trends existed for increases in average steps per day (5601.73 vs 6603.95, p=0.07) and greater MVPA levels at 6 months (20.8 vs 27.04 MVPA min/day, p=0.06), no significant differences existed for aerobic fitness or any of the SBs. Participant physiological and psychological health factors at baseline and 6 months are displayed in table 1. HbA1c levels significantly decreased at 6 months compared with baseline levels (7.52 vs 8.01, p=0.01). Additionally, feelings of empowerment over diabetes significantly improved at 6 months (33.15 vs 26.5, p<0.0001). No significant changes in any of the other physiological or psychological variables occurred over the 6 months. Table 1 Baseline and 6 month physiological and psychological health variables Variables Baseline 6-month follow-up BMI 37.61 (8.84) 36.31 (9.4) Insulin 17.52 (11.73) 17.84 (17.54) HbA1c 8.1 (0.83) 7.52 (1.26)** CRP 7.36 (7.4) 6.08 (6.3) Systolic BP 128.63 (17.97) 127.67 (15.12) Quality of life 50.67 (6.86) 52.67 (6.56) Depression 3.33 (3.05) 2.89 (3.14) Diabetes concern 7.21 (4.83) 8.26 (6.86) Diabetes empowerment 26.5 (9.44) 33.15 (5.46)** **p≤0.01. BMI, body mass index; BP, blood pressure. Relationship among activity and sedentary variables Correlations among changes in activity and sedentary variables are displayed in table 2. Spearman correlations indicated that there were significant correlations in activity variables with change in average steps per day associated with a change in MVPA (r=0.85, p<0.0001) and a trend toward an association with change in aerobic fitness (r=0.34, p<0.1). Increases in SB per day was associated with increases in uninterrupted bouts of SB (r=0.44, p<0.05). Inverse correlations existed between activity and sedentary variables with decreases in MVPA associated with increases in SB (r=−0.48, p<0.05) and increases in bouts of SB (r=−0.78, p<0.0001). Finally, a decrease in average steps per day was associated with increases in SB (r=−0.42, p<0.05) and bouts of SB (r=−0.65, p<0.01). Table 2 Relationship among change in activity variables Variables Δ MVPA Δ Steps Δ Fitness Δ SB Δ SB bouts Δ SB breaks Δ MVPA – 0.85*** 0.16 −0.48* −0.78*** −0.1 Δ Steps 0.85*** – 0.34 −0.42* −0.65*** −0.12 Δ Fitness 0.16 0.34† – 0.12 −0.07 0.03 Δ SB −0.48* −0.42* 0.12 – 0.44* −0.22 Δ SB bouts −0.78*** −0.65*** −0.07 0.44* – 0.2 Δ SB breaks −0.1 −0.12 0.03 −0.22 0.2 – *p<0.05; ***p<0.001; †p<0.01. MVPA, moderate-to-vigorous physical activity; SB, sedentary behavior. Regression results for activity variables Regression models indicated that increases in MVPA (β=−0.04, SE=.01, p<0.01) and average steps walked per day (β=−0.003, SE=0.00, p<0.01) were associated with lower HbA1c levels at the 6-month follow-up. Table 3 displays the regression results for the activity behaviors. For the psychological variables, an increase in MVPA was associated with decreases in depression (β=−0.07, SE=0.03, p<0.05) and self-reported concerns about diabetes (β=−0.16, SE=0.07, p<0.05). Finally, an increase in aerobic fitness was associated with an increase in overall quality of life (β=0.01, SE=.01, p<0.05). Table 3 Regression results for activity behaviors Variables Δ MVPA Δ Total steps Δ Fitness Physiological  BMI 0.01 (0.11) −0.00 (0.00) −0.01 (0.01)  Insulin −0.04 (0.21) −0.00 (0.00) 0.01 (0.01)  Hba1c −0.04 (0.01)** −0.003 (0.00)** −0.00 (0.02)  CRP −0.02 (0.07) −0.00 (0.00) −0.01 (0.00)  Systolic BP 0.04 (0.18) 0.00 (0.00) −0.01 (0.01) Psychological  Quality of life 0.05 (0.08) 0.0005 (0.00) 0.01 (0.005)*  Depression −0.07 (0.03)* −0.00 (0.00) −0.00 (0.00)  Diabetes concern −0.16 (0.07)* −0.01 (0.00)* −0.00 (0.01)  Diabetes empowerment −0.02 (0.06) 0.00 (0.00) −0.00 (0.00) *p<0.05; **p<0.01. BP, blood pressure; BMI, body mass index; MVPA, moderate-to-vigorous physical activity. Regression results for SBs Regression models indicated that reductions in longer bouts of uninterrupted SB was associated with reduced HbA1c levels at the 6-month follow-up (β=0.17, SE=.08, p<0.05). Results for SBs are displayed in table 4. Increased breaks in SB was associated with increased insulin levels (β=0.44, SE=.16, p<0.05). Finally, although not significant, there was a trend toward longer bouts of uninterrupted SB being associated with greater concern of diabetes (β=0.68, SE=0.47, p<0.1). Table 4 Regression results for sedentary behaviors Variables Δ Total sedentary time Δ Duration of bouts Δ Number of breaks Physiological  BMI −0.04 (0.03) −0.63 (0.63) 0.06 (0.08)  Insulin −0.05 (0.1) 0.94 (1.59) 0.44 (0.16)*  HbA1c 0.004 (0.00) 0.017 (0.08)* 0.02 (0.01)  CRP −0.02 (0.02) −0.1 (0.42) 0.04 (0.05)  Systolic BP −0.02 (0.06) −0.32 (1.23) 0.07 (0.15) Psychological  Quality of life −0.01 (0.03) 0.12 (0.5) −0.04 (0.06)  Depression 0.01 (0.01) 0.17 (0.28) 0.03 (0.03)  Diabetes concern 0.07 (0.06) 0.68 (0.47)† 0.07 (0.06)  Diabetes empowerment −0.01 (0.02) 0.00 (0.00) −0.05 (0.05)  LGL −0.06 (0.08) −1.5 (1.39) −0.05 (0.18)  HGL 0.05 (0.06) 0.63 (1.29) −0.03 (0.16) *p<0.05. †p<0.1. BP, blood pressure; BMI, body mass index. Since bouts of SB and the activity variables were associated with HbA1c levels, bouts and MVPA were entered simultaneously into a multiple regression model, along with BMI (no other demographic variable was correlated with HbA1c). Increased MVPA (β=−0.04, SE=0.01, p=0.01) and decreased bouts of SB (β=0.05, SE=.02, p=0.05) were both significantly associated with decreased HbA1c levels, while BMI was not quite significant, although a trend remained (β=0.05, SE=0.03, p=0.06). The overall model was significant (p=0.03), and the addition of these variables accounted for nearly 50% of the variance in HbA1c levels. Discussion Similar to previous studies, MVPA and SB were only moderately correlated with each other, indicating that MVPA and SB are distinctly different behaviors.4 7 28 In addition, participation in MVPA does not fully protect individuals from adverse health effects from high levels of SB as evidenced by the independent effect of SB and duration of sedentary bouts on HbA1c in the regression model. Despite the small sample size, the increased duration of uninterrupted bouts of SB and decreased activity levels were associated with increased HbA1c levels after 6 months, accounting for 50% of the variance in the model. This finding is informative to future research given that HbA1c levels are used to diagnose T2D and evaluate the success of its treatment. These results that suggest HbA1c levels are influenced by MVPA have been supported by previous findings.29–31 Church et al29 reported the effects of a 9 month aerobic and resistance training trial on 262 sedentary individuals with T2D, finding that the exercise training group experienced a significant reduction in HbA1c levels compared with the control group. Fewer studies have examined objective SB and HbA1c levels; however, studies examining insulin resistance and SB have found mixed results.30–31 Helmerhorst et al30 found that time spent in SB was associated with fasting insulin; however, another study by Ekelund et al31 found that MVPA, but not sedentary time, was associated with increased insulin resistance. With the significant decrease in HbA1c levels over the 6-month period in the current study, nine of the 26 (35%) participants achieved HbA1c levels <7.0. This suggests that developing interventions that focus on increasing MVPA and breaking up long bouts of uninterrupted SB may help to control T2D and improve their metabolic health. Interestingly, SB was not associated with any of the psychological health outcomes, while MVPA and average steps per day were associated with decreased levels of depression and concern over diabetes. This finding is supported by previous research that found increases in activity were associated with improved mood and decreased depression.10 32 Previous literature has suggested that intensity is also associated with decreased depression which may explain why actively engaging in activity, but not SB, was associated with improved mood.32–34 Additionally, it could be speculated that participants felt better about the prognosis of their T2D, since they were actively engaging in behaviors that are known to improve their health such as increased MVPA. Since SB tends to be more of a passive behavior, participants may have been more aware of behaviors that increased MVPA, rather than behaviors that decreased sedentary time. Additional studies to compare the effects of PA and SB on psychological health are warranted. It would be of interest for future studies to actively track SB, and provide participants feedback for decreasing prolonged uninterrupted sedentary time and/or overall SB throughout the day. Similar to evidence suggesting the use of activity trackers to promote PA behavior through feedback, feedback on SB may allow participants to make even larger changes to decrease their SB. Results of this study indicate that MVPA and SB are two distinct behaviors that are associated with different health outcomes status. These findings suggest that special emphasis should be placed on decreasing SB and breaking up bouts of SB during the day to improve diabetes-related physiological and psychological outcomes. This particular sample was obese, engaged in a low amount of MVPA, and accumulated nearly 10 hours of SB with the average duration of a bout of SB lasting 90 min. However, small changes in activity behavior such as decreasing the duration of uninterrupted SB, was associated with significant improvements in HbA1c levels. For individuals who have not been active or suffer from chronic health problems, undertaking an exercise program may not be feasible, however, by just decreasing the amount of time sitting on a couch, individuals with T2D may experience positive health benefits such as a decrease in HbA1c levels. Implications Results from this study carry several implications regarding future lifestyle interventions for obese individuals with T2D. This study found that SB and physical activity largely affect HbA1c levels, suggesting that to improve physiological profiles in individuals with T2D, interventions must focus on increasing MVPA and decreasing SB and prolonged bouts of SB behavior. Although much emphasis has been placed on increasing MVPA in these populations, alterations in SB patterns may lead to physiological improvements in several diabetes-related health outcomes as well. Future studies should further examine whether educating individuals with T2D on the importance of reducing SB to help improve their health could potentially lead to improvements in psychological outcomes, similar to MVPA. The significant relationships between lower HbA1c levels and physical activity, despite only modest increases in MVPA and decreases in sedentary bouts, suggest that even just small increases in activity and decreases in SB could bring about significant positive health improvements. These changes could include taking a 5 min walk during an 8 hour desk office job, or even breaking up long bouts of SB by standing and stretching every 30 min. Additionally, the use of commercial pedometers that alert wearers when there has been no movement in the past 60 min may be beneficial in reminding individuals to break up SB bouts. Such small changes may be particularly appealing to an obese, sedentary population as an initial behavioral strategy. 10.1136/annrheumdis-2016-210131.supp1supplementary data The study authors would like to thank the research staff who assisted with the study, and the participants who made this study possible. The authors also would like to acknowledge John R. Sirard for the use of his accelerometers which were used to assess activity behavior. Contributors: DC was the principle investigator on the project, helped to conceptualize the design of the project, supervised the data collection, and helped to prepare and revise the final manuscript. JMG helped to conceptualize the design of the project, performed the data analysis, was involved in the interpretation of data, helped to prepare and revise the final document, and, as the corresponding author, oversaw the contributions of all of the authors. DJR assisted with the data analysis and interpretation of the data, helped to prepare and revise the document, and was responsible for overseeing the revisions of the final document. All of the authors agree that they are accountable for the accuracy and integrity of all parts of the project. Funding: JMG was supported by the Cardiovascular Epidemiology Training Grant T32HL098048 in Behavior, the Environment, and Global Health from the National Heart, Lung, and Blood Institute, National Institutes of Health. The other authors received no specific grant from any funding agency in public, commercial, or not-for-profit sectors. Competing interests: None declared. Ethics approval: The study received ethics approval by University of Virginia Institutional Review Board. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Ownership of the data from the project will be shared between the three authors. Storage and security of the data will be the responsibility of the corresponding author, JMG, and the data will be housed on a dedicated hard drive using password-protected program. Any requests for access to the data will be reviewed by all three authors and the release of any portion of the data must be agreed on by all three of the authors. ==== Refs References 1 Global Report on Diabetes . Geneva : World Health Organization , 2016 . 2 American Diabetes Association website. http://www.diabetes.org/diabetes-basics/statistics/?loc=superfooter (accessed Jul 2016 ). 3 Pate RR , Pratt M , Blair SN Physical activity and public health: a recommendation from the procedure Centers for Disease Control and Prevention and the American College of Sports Medicine . JAMA 1995 ;273 :402 –7 .7823386 4 Chastin SFM , Palarea-Albaladejo J , Dontje ML Combined effects of time spent in physical activity, sedentary behaviors, and sleep on obesity and cardio-metabolic health markers: a novel compositional data analysis approach . PLoS ONE 2015 ;10 :e0139984 10.1371/journal.pone.0139984 26461112 5 Larsen BA , Martin L , Strong DR Sedentary behavior and prevalent diabetes in Non-Latino Whites, Non-Latino Blacks, and Latinos: findings from The National Health Interview Survey . J Public Health (Oxf) 2015 ;37 :634 –40 . 10.1093/pubmed/fdu103 25538143 6 Owen N , Healy GN , Matthews CE Too much sitting: the population-health science of sedentary behavior . Exerc Sport Sci Rev 2010 ;38 :105 –13 . 10.1097/JES.0b013e3181e373a2 20577058 7 Biswas A , Oh PI , Faulkner GE Sedentary time and its association with risk for disease incidence, mortality, and hospitalization in adults: a systematic review and meta-analysis . Ann Intern Med 2015 ;162 :123 –34 . 10.7326/M14-1651 25599350 8 Wilmot EG , Edwardson CL , Achana FA Sedentary time in adults and the association with diabetes, cardiovascular disease and death: systematic review and meta-analysis . Diabetologia 2012 ;55 :2895 –905 . 10.1007/s00125-012-2677-z 22890825 9 Healy GN , Dunstan DW , Salmon J Breaks in sedentary time . Diabetes Care. 2008 ;31 :661 –6 . 10.2337/dc07-2046 18252901 10 Penedo FJ , Dahn JR Exercise and well-being: a review of mental and physical health benefits associated with physical activity . Curr Opin Psychiatry 2005 ;18 :189 –93 .16639173 11 Teychenne M , Ball K , Salmon J Sedentary behavior and depression among adults: a review . Int J Behav Med 2010 ;17 :246 –54 . 10.1007/s12529-010-9075-z 20174982 12 Grisby AB , Anderson RJ , Freedland KE Prevalence of anxiety in adults with Diabetes: a systematic review . J Psychosom Res 2002 ;53 :1053 –60 .12479986 13 Cox DJ , Taylor AG , Singh H Glycemic load, exercise, and monitoring blood glucose (GEM): a paradigm shift in the treatment of type 2 diabetes mellitus . Diabetes Res Clin Pract 2016 ;111 :28 –35 . 10.1016/j.diabres.2015.10.021 26556234 14 Matthews CE , Chen KY , Freedson PS Amount of time spent in sedentary behaviors in the United States, 2003–2004 . Am J Epidemiol 2008 ;167 :875 –81 . 10.1093/aje/kwm390 18303006 15 Trost SG , McIver KL , Pate RR Conducting accelerometer-based activity assessments in field-based research . Med Sci Sports Exerc 2005 ;37 :S531 –43 .16294116 16 Matthew CE Calibration of accelerometer output for adults . Med Sci Sports Exerc 2005 ;37 :S512 –22 .16294114 17 Balke B A simple field test for the assessment of physical fitness . Rep Cuv Aeromed Res Inst US 1963 ;(53) :1 –8 . 18 Ingle L A review of the six-minute walk test: Its implication as a self-administered assessment tool . Eur J Cardiovasc Nurs 2009 ;8 :232 –4 . 10.1016/j.ejcnurse.2009.01.002 19223237 19 Faggiano P , D'Aloia A , Gualeni A Assessment of oxygen uptake during the 6-minute walking test in patients with heart failure: preliminary experience with a portable device . Am Heart J 1997 ;134 :203 –6 .9313598 20 Guyatt GH , Thompson PJ , Berman LB How should we measure function in patients with chronic heart and lung disease? . J Chronic Dis 1985 ;38 :517 –24 .4008592 21 The World Health Organization Quality of Life (WHOQOL)-BREF . World Health Organization , 2004 . 22 The WHOQOL Group . Development of the World Health Organization WHOQOL-BREF Quality of Life Assessment . Psychol Med 1998 ;28 :551 –8 .9626712 23 Anderson RM Diabetes Empowerment Scale (DES) . Ann Arbor, MI : University of Michigan, Diabetes Research & Training Center, G1116—Towsley Center—0201 , 48109–0201. 24 Anderson RM , Funnell MM , Fitzgerald JT The Diabetes Empowerment Scale: a measure of psychosocial self-efficacy . Diabetes Care. 2000 ;23 :739 –43 .10840988 25 Patient Health Questionnaire (PHQ-9), PHQ9 Copyright © Pfizer Inc. A2662B 10–04–2005 . 26 Kroenke K , Spitzer RL , Williams JB The PHQ-9: validity of a brief depression severity measure . J Gen Intern Med 2001 ;16 :606 –13 .11556941 27 McGuir BE , Morrison TG , Hermanns N Short-form measures of diabetes-related emotional distress. The Problem Areas in Diabetes Scale (PAID)-5 and PAID-1 . Diabetologia 2010 ;53 :66 –9 . 10.1007/s00125-009-1559-5 19841892 28 Shuval K , Finley CE , Barlow CE Independent and joint effects of sedentary time and cardiorespiratory fitness on all-cause mortality: the Cooper Center Longitudinal Study . BMJ 2015 ;5 :e008956 10.1136/bmjopen-2015-008956 29 Church TS , Blair SN , Cocrehan S Effects of aerobic and resistance training on hemoglobin A1c levels in patients with type 2 diabetes: a randomized controlled trial . JAMA 2010 ;304 :2253 –62 . 10.1001/jama.2010.1710 21098771 30 Helmerhorst HJF , Wijndaele K , Brage S Objectively measured sedentary time may predict insulin resistance independent of moderate-and vigorous-intensity physical activity . Diabetes 2009 ;58 :1776 –9 . 10.2337/db08-1773 19470610 31 Ekelund U , Brage S , Griffin SJ Objectively measured moderate-and vigorous-intensity physical activity but not sedentary time predicts insulin resistance in high-risk individuals . Diabetes Care 2009 ;32 :1081 –6 . 10.2337/dc08-1895 19252168 32 Carek PJ , Labstein SE , Carek SM Exercise for the treatment of depression and anxiety . 33 Kennedy MM Effect of exercise intensity on mood in step aerobics . J Sports Med Phys Fitness 1997 ;27 :200 –4 . 34 Hamer M , Endrighi R , Poole L Physical activity, stress reduction, and mood: insight into immunological mechanisms . Methods Miol Biol 2012 ;934 :89 –102 .

PMC005xxxxxx/PMC5372080.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01468110.1136/bmjopen-2016-014681Patient-Centred MedicineResearch150617221704Evaluating CollaboRATE in a clinical setting: analysis of mode effects on scores, response rates and costs of data collection Barr Paul J 1Forcino Rachel C 1Thompson Rachel 1Ozanne Elissa M 1Arend Roger 2Castaldo Molly Ganger 13O'Malley A James 1Elwyn Glyn 1 1 The Dartmouth Institute for Health Policy & Clinical Practice, Lebanon, New Hampshire, USA 2 Dartmouth-Hitchcock Patient and Family Advisory Council, Lebanon, New Hampshire, USA 3 Dartmouth Master of Health Care Delivery Science Program, Hanover, New Hampshire, USA▸ Additional material is available. To view please visit the journal online (http://dx.doi.org/10.1136/bmjopen-2016-014681). Correspondence to Dr Glyn Elwyn; glynelwyn@gmail.com2017 24 3 2017 7 3 e01468110 10 2016 6 1 2017 13 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Background Shared decision-making (SDM) has become a policy priority, yet its implementation is not routinely assessed. To address this gap we tested the delivery of CollaboRATE, a 3-item patient reported experience measure of SDM, via multiple survey modes. Objective To assess CollaboRATE response rates and respondent characteristics across different modes of administration, impact of mode and patient characteristics on SDM performance and cost of administration per response in a real-world primary care practice. Design Observational study design, with repeated assessment of SDM performance using CollaboRATE in a primary care clinic over 15 months of data collection. Different modes of administration were introduced sequentially including paper, patient portal, interactive voice response (IVR) call, text message and tablet computer. Participants Consecutive patients ≥18 years, or parents/guardians of patients <18 years, visiting participating primary care clinicians. Main measures CollaboRATE assesses three core SDM tasks: (1) explanation about health issues, (2) elicitation of patient preferences and (3) integration of patient preferences into decisions. Responses to each item range from 0 (no effort was made) to 9 (every effort was made). CollaboRATE scores are calculated as the proportion of participants who report a score of nine on each of the three CollaboRATE questions. Key results Scores were sensitive to mode effects: the paper mode had the highest average score (81%) and IVR had the lowest (61%). However, relative clinician performance rankings were stable across the different data collection modes used. Tablet computers administered by research staff had the highest response rate (41%), although this approach was costly. Clinic staff giving paper surveys to patients as they left the clinic had the lowest response rate (12%). Conclusions CollaboRATE can be introduced using multiple modes of survey delivery while producing consistent clinician rankings. This may allow routine assessment and benchmarking of clinician and clinic SDM performance. PRIMARY CAREshared decision-makingpatient-reported experience measurepatient-reported measurementmode effectsGordon and Betty Moore Foundationhttp://dx.doi.org/10.13039/1000009363929 ==== Body Strengths and limitations of this study This study deployed a range of survey administration modes within 24 hours of clinic visits. By consistently measuring performance of the same clinicians, we were able to assess effects of survey administration mode on clinician rank order and CollaboRATE score. We accounted for patient characteristics in our analysis by including patient demographic and clinical data. Data collection for 15 months led to respondent and organisational fatigue, impacting response rates. Cost analysis relied on retrospective cost estimates. Introduction Assessing whether patients have experienced shared decision-making (SDM) in clinical settings has become a priority for multiple stakeholders.1 In the US context, the Consumer Assessment of Healthcare Providers and Systems Clinician and Group (CG-CAHPS) survey is a widely used patient reported experience measure (PREM).2 However, the measurement of SDM is tangential to the overall survey goals.3 In addition, recall bias may undermine the validity of results given that patients are asked to assess experiences over the preceding 6 months.4 Furthermore, response rates are significantly below the 40% target set by the Agency for Healthcare Research and Quality. Alternative PREMs of SDM are designed primarily for research and not clinical use.5–7 To address this gap we developed CollaboRATE, a 3-item PREM of SDM.8 9 CollaboRATE assesses three core SDM tasks during a clinic visit: (1) information and explanation about health issues, (2) elicitation of patient preferences and (3) integration of patient preferences into decisions. CollaboRATE has demonstrated psychometric soundness,9 and has also been used in a demonstration project to assess differences in SDM across 34 clinical teams in England.10 In addition, Blue Shield of California is using CollaboRATE to assess SDM as part of a preauthorisation process for joint replacement.11 A research study at a Veterans Administration site found positive correlations between CollaboRATE scores and patient assessments of satisfaction and communication.12 CollaboRATE's brevity minimises respondent burden and facilitates varied data collection modes, including email, interactive voice response (IVR) automated telephone calls, short message service (SMS) text messages, electronic kiosk, tablet computer and paper. Tourangeau identified five administration-mode factors related to survey response: (1) respondent contact method (eg, in person or email); (2) survey medium (eg, paper or electronic); (3) administration method (interviewer or self-completed); (4) sensory input (eg, visual or oral) and (5) response mode (eg, handwritten, keyboard, voice).13 Data analysis needs to account for mode effects and differences in response rates, especially if results are intended as measures of performance. However, to date, there has been limited research on differences in validity, reliability and administration costs by mode in patient-reported measurement.14 Mixed modes of CG-CAHPS survey administration, including options to respond over the phone, have yielded response rates near 40%.15 These mixed-mode response rates are higher than those achieved by mail-back surveys alone15 while collecting comparable data.16 In a systematic review of survey response rates, email delivery was associated with response rates of 15%,17 similar to the 14% response rate for CG-CAHPS delivered by email.16 SMS administration has associated response rates ranging from 49% to 98%,18–21 and data quality comparable to that collected by person-led phone interviews or IVR.21 The cost of PREM administration has significant implications, yet receives little attention.16 22 23 Achieving a reasonable response rate, at a realistic cost, is a key requirement if PREMs are to become sustainable sources of clinical performance data. This project was designed to assess feasibility and identify challenges to the routine implementation and real-time delivery of CollaboRATE. In this paper, we aim to evaluate CollaboRATE as a measure of clinician performance with regard to response rates, CollaboRATE SDM scores at the level of individual clinicians, respondent characteristics across modes, impact of mode and patient characteristics on SDM scores, and cost of data collection per patient response. To do so, we conducted a single-site demonstration project where we sequentially introduced different modes of patient-reported data collection including paper, patient portal, IVR calls, SMS and electronic tablet in the clinic. We hypothesise variation in CollaboRATE scores by clinician as well as lower CollaboRATE scores in data collection modes completed outside the clinic setting. Methods Participants Setting The study was conducted with two clinical teams, consisting of 15 clinicians, in a primary care clinic of a rural academic medical centre in New Hampshire, with 16 000 registered patients. Inclusion criteria Consecutive patients 18 years or older visiting the participating clinical teams Monday through Friday beginning in April 2014 were eligible to complete CollaboRATE. Parents or guardians of patients under age 18 were eligible to complete CollaboRATE on behalf of their children. Measures CollaboRATE CollaboRATE consists of three items (box 1).8 9 Box 1 CollaboRATE items Thinking about the appointment you have just had: How much effort was made to help you understand your health issues? How much effort was made to listen to the things that matter most to you about your health issues? How much effort was made to include what matters most to you in choosing what to do next? Responses to each item can range from 0 (No effort was made) to 9 (Every effort was made) for a maximum total of 27. Responses to CollaboRATE were scored in a binary fashion; participants who responded to all questions with a ‘9’ were considered to have experienced SDM and all others were not. Participants with one or more missing responses on CollaboRATE were excluded from analyses. We calculated the proportion of participants who reported a score of 9 on each of the three CollaboRATE questions for each clinician.9 24 This scoring technique is a strategy for aiding interpretation while avoiding potential ceiling effects common among clinician performance assessments completed by patients.9 Incomplete CollaboRATE surveys were not included in analysis. Data collection Modes of CollaboRATE survey administration We adopted five modes of CollaboRATE administration. In each mode, eligible patients were given opportunities to complete the survey within 24 hours of their clinic visits. Modes were implemented sequentially beginning in April 2014, each for a 3-month period (table 1). Clinicians were made aware of data collection at the start of the study. At the end of the study, we provided participating clinicians with a summary of the results, their individual scores and a brief presentation on how to achieve SDM in routine clinical settings. Table 1 Modes of CollaboRATE administration in chronological order Mode Description Paper survey A paper-based version of CollaboRATE was given to patients by administrative staff as they left the clinic following their visits. Administrative staff added patient identifiers to the surveys to enable linkage to medical records data. Patients were asked to place completed surveys in a locked receptacle in the clinic. Patient portal CollaboRATE was delivered using an online patient portal (MyChart), part of the clinic's electronic medical record. The questionnaire was programmed by the medical centre's information systems department. As clinical encounters were completed, emails containing a web link to the CollaboRATE questionnaire were sent to those patients who had portal accounts. Interactive voice response (IVR) CollaboRATE was delivered to patients using an interactive voice response telephone system programmed by the medical centre's information systems department. An automated telephone call was made to each patient's cell phone at 19:00 on the day of their clinic visit. Before initiating the survey, the respondent was asked to confirm that they were the individual who had visited the clinic that day. On confirmation, numerical keypad responses to CollaboRATE questions were requested. Short message service (SMS text messages) Text messages, programmed by the medical centre's information systems department, were sent to patient cell phones at 19:00 on the day of their clinical visits. The first message introduced the survey and offered opt-out opportunities. Remaining messages each contained a single CollaboRATE question and response instructions. Subsequent CollaboRATE questions were triggered by each reply, sending a total of four text messages. Tablet and mail Using tablet computers, research assistants offered patients an opportunity to complete an online version of CollaboRATE hosted in Qualtrics (Provo, Utah, USA) as they left the clinic. Patients were asked for their name, age, gender and to indicate the clinician visited. Patients who declined the tablet opportunity were asked to respond by completing a paper-based survey to be returned in a postage-paid envelope. Participant characteristics For the first four modes, CollaboRATE responses were linked to patient demographic and clinical data, including patient age, gender, number of health conditions, visit type (annual wellness or other), marital status and socioeconomic status (SES). SES was estimated by the percentage of people living in the patient's zip code with income beneath the federal poverty level using data from the American Community Survey.25 26 These data were extracted from patients' medical records, anonymised by medical centre staff and shared with the research team. Data linkage was not undertaken for the tablet mode; clinician, patient age and gender were self-reported in that mode. Data analysis Respondent characteristics and response rates Response rates were based on the number of patients attending the clinic during the data collection periods. Not all patients were registered with a patient portal or had a cellular telephone number in their record. We therefore calculated adjusted response rates for patient portal, IVR and SMS modes, including only patients who could have received CollaboRATE. We compared the demographic characteristics of respondents to non-respondents for each mode. Pearson's χ2 tests and Student's t-tests were used for categorical and continuous variables, respectively. Logistic regression analysis was used to confirm the descriptive findings comparing respondents to non-respondents and to examine whether an interaction between age and number of comorbidities predicted response, where response was the binary outcome variable and independent variables included age, number of comorbidities, age multiplied by number of comorbidities, gender, age, and whether the visit was for a wellness check-up. We also calculated the median time to achieve 25 completed CollaboRATE surveys per clinician per mode, which has been used as a minimum for obtaining stable score estimates.24 Studies which have calculated the number of patient responses needed to obtain reliable information have found that around 25 completed questionnaires are needed.27 Clinician SDM performance by mode To assess clinician performance across modes, we estimated a logistic regression model including clinician indicators and all available patient characteristics as covariates with the CollaboRATE top score indicator as the dependent variable. We used the model to estimate the CollaboRATE top score (accounting for patient characteristics) for each clinician in each mode and used these estimates to rank clinicians' performance and assessed concordance of the estimated rankings across modes. Clinicians who did not reach 25 patient responses in any of the four modes were excluded from the results.24 27 Effects of administration mode and patient demographics on CollaboRATE scores To evaluate the effects of survey administration mode and patient characteristics on CollaboRATE scores, we conducted mixed effects logistic regression analysis with CollaboRATE score as the dependent variable. Fixed effects included administration mode and patient characteristics. Clinicians were included as a random effect to account for clustering of patients by clinician, allowing inferences to be generalisable beyond the clinicians in our study. The unit of analysis was the CollaboRATE score per individual visit. To address potential clustering of CollaboRATE responses by patients who had more than one visit, we also conducted the mixed effects logistic regression analysis described above including only the first CollaboRATE response from each patient. Cost analysis by survey administration mode The costs associated with each administration mode were estimated to calculate a cost per fully completed survey. Data included development costs (programming, testing and planning) and field costs (piloting, supplies and labels, technical and vendor support, management and staff time). We estimated the cost per completed survey by dividing the total cost for each mode by the number of completed surveys. Estimated costs per mode do not include costs associated with an institutional electronic medical record, patient portal or survey collection platform but do include variable costs such as technical support and charges associated with sending SMS messages and IVR calls. To estimate the likely long-term cost per completed survey, we repeated the analysis excluding research and set-up costs. Analysis was conducted in Stata V.13 and SPSS V.21. An α level of ≤0.05 was considered statistically significant. Results There were 4421 patients who completed the CollaboRATE survey over the 15-month study period (April 2014–October 2015), resulting in an overall CollaboRATE top score of 68% for the clinic. Only 19 incomplete CollaboRATE responses were recorded. Among eligible patients, the average response rate across modes was 25%, with the highest response rate in the combined tablet and mail mode (41%) followed by the patient portal mode (34%). The administration mode had significant effects on CollaboRATE scores, with the highest score recorded during the paper survey administration (81%), followed by the patient portal (71%), with the lowest overall CollaboRATE score recorded in the IVR survey administration mode (61%) (table 2). Table 2 Response rate and CollaboRATE scores across modes Mode Response rate (n) CollaboRATE score* (%) Clinician score range (%) Paper  All patients 12% (541/4692) 81 72–93 Patient portal  All patients 21% (1019/4939) 71 59–83  Eligible patients† 34% (1019/3015) IVR  All patients 19% (893/4814) 61 42–75  Eligible patients† 25% (893/3589) SMS  All patients 17% (757/4520) 65 46–82  Eligible patients† 23% (757/3329) Tablet and mail  All patients 41% (1211/2943) 66 53–83 *CollaboRATE score represents the proportion of respondents marking 9 on all three items (totalling 27/27). †These calculations exclude patients who did not have patient portal accounts or phone numbers on file at which to receive the CollaboRATE survey. IVR, interactive voice response; SMS, short message service. The fastest median time to achieve 25 completions per clinician was accomplished in the tablet/mail-in mode (17 days), followed by paper in-clinic (25 days), patient portal (25 days), IVR (31 days) and SMS (40 days). Clinician rank Figure 1 demonstrates that clinician rank order by proportion of CollaboRATE top scores was identical across the four administration modes (adjusted by patient case mix; see online supplementary appendices 1 and 2), though the average top score differed across modes. Figure 1 Clinician scores by mode, adjusted for patient characteristics^*. ^While 15 clinicians participated in all four data collection modes, only eight reached 25 patient responses in all modes; therefore, eight of 15 clinicians are shown here. *During the electronic tablet/postal mail phase, responses were not linked to the electronic medical record; as a result, patient demographic data were unavailable. 10.1136/bmjopen-2016-014681.supp1supplementary appendices Respondent characteristics by administration mode Respondents versus non-respondents by mode Demographic characteristics of all patients attending the clinic were similar across modes (table 3). Respondents were slightly older than non-respondents across all modes and represented the overall clinic population with regard to gender. Respondents were slightly more likely than non-respondents to be seen for an annual wellness visit in the patient portal and SMS modes. Respondents also had more comorbid conditions than non-respondents in the paper mode. Logistic regression analysis confirmed these descriptive findings and showed no significant interaction between age and number of comorbid health conditions in predicting response to the CollaboRATE survey (OR 1.00; 95% CI 0.99 to 1.00; full results available on request). Table 3 Respondent and non-respondent characteristics by mode Paper Patient portal IVR SMS Tablet and mail†‡ Respondents (n=541) Non-respondents (n=4151) Respondents (n=1019) Eligible non-respondents (n=1996) Respondents (n=893) Eligible non-respondents (n=2696) Respondents (n=757) Eligible non-respondents (n=2572) Respondents (n=1211) Female 65.4% 65.6% 68.8% 67.6% 63.2% 64.5% 64.2% 64.9% 57.1% Mean age (SD) 54.09 (15.73) 47.12* (21.54) 54.12 (15.93) 46.27* (10.98) 44.61 (18.77) 41.97* (20.02) 46.56 (18.58) 43.76* (19.78) 49.57 (18.38) Number of morbidities *  0 17.7% 31.4% 59.6% 62.7% 35.9% 37.2% 38.4% 37.7% –  1 35.9% 42.2% 32.5% 31.1% 43.7% 44.7% 40.8% 42.1% –  2 25.1% 16.7% 7.5% 6.0% 15.0% 13.1% 16.0% 14.7% –  3 13.7% 6.3% 0.5% 0.2% 4.1% 3.6% 3.4% 4.4% –  4 or more 7.6% 3.4% 0% 0.1% 1.2% 1.4% 1.3% 1.1% –  Annual wellness visit 20.3% 17.4% 25.3% 19.3%* 19.4% 16.6% 22.6% 17.5% – *Significantly different from respondent group, p<0.05. †During the electronic tablet/postal mail phase, responses were not linked to the electronic medical record; as a result, non-respondent patient characteristics are unavailable. ‡An intervention was introduced midway through data collection. IVR, interactive voice response; SMS, short message service. Respondent characteristics by mode Among respondents, gender (χ2=11.32, p=0.02), age (χ2=49.88, p<0.001), reason for visit (χ2=11.04, p=0.01) and number of comorbidities (χ2=288.39, p<0.001) showed statistically significant variation across the modes. Respondents to the SMS and IVR modes were on average 8–10 years younger than respondents to the paper and patient portal modes. There were differences in respondent health status between modes; in the paper mode ∼46% of respondents had two or more comorbidities compared with 21% in SMS and IVR modes and 7% in the patient portal mode. Effect of administration mode and patient demographics on scores The results of a mixed effects logistic regression demonstrate significantly lower CollaboRATE scores in the patient portal (OR 0.60, 95% CI 0.45 to 0.80), IVR (OR 0.45, 95% CI 0.34 to 0.59) and SMS (OR 0.51, 95% CI 0.38 to 0.67) modes when compared with the paper mode. CollaboRATE scores increased slightly with patient age (OR 1.01 per year of age, 95% CI 1.01 to 1.02) but no other patient characteristics were associated with CollaboRATE scores (full results in online supplementary appendix 3). The estimated SD for the clinician random effect was 0.34, implying there was substantial unexplained heterogeneity in the clinicians' ratings even after adjusting for observed differences between the patient cases. Our sensitivity analysis including only the first CollaboRATE response from each unique patient finds very similar results (see online supplementary appendix 4). While the annual visit predictor becomes significant (OR 1.65, 95% CI 1.15 to 2.37) in this analysis limited to the first CollaboRATE response from each patient during the 15-month study period, all other fixed effects retain similar magnitude and significance and the estimated SD for the clinician random effect remains consistent at 0.33. Cost of data collection Owing to high personnel involvement in administering the paper and tablet modes, these modes were the most expensive (see online supplementary appendix 5). The cost per unique response was US$20.56 and US$16.71 for the paper and tablet modes respectively. The SMS, IVR and patient portal modes were significantly less expensive at US$10.28, US$8.87 and US$6.39 per unique response, respectively. When estimating the variable cost per completed response (excluding set-up costs and research related costs), these decrease to US$16.38 and US$13.46 for the personnel-intensive modes (tablet and paper modes) and US$3.36, US$2.83 and US$2.20 for the SMS, IVR and patient portal modes. Discussion Principal findings Clinician SDM performance rankings were stable across data collection modes, though they were sensitive to mode effects, as were response rates. We were able to achieve the highest response rate (41%) by using tablet computers administered by research staff in-clinic, although this approach was among the costliest. The lowest response rate involved clinic staff giving paper surveys to patients as they left the clinic (12%). Mode of administration and respondent characteristics were associated with CollaboRATE scores, with the lowest score in the IVR mode and the highest in the paper mode. We speculate that this finding may be explained, at least in part, by patients' perceived distance from the clinic experience and the perceived anonymity of the response. Across all modes, respondents were slightly older than non-respondents. Younger patients appeared to prefer, or were more comfortable than older patients, completing CollaboRATE using SMS and IVR, where respondents were ∼10 years younger than respondents from other modes. Across all modes, the clinician identity and mode of survey administration had the greatest impact on CollaboRATE scores, while increased patient age was associated with a small but statistically significant increase in CollaboRATE score. If fixed institutional costs are excluded, the use of existing technological systems such as patient portals or SMS capabilities can reduce the cost of obtaining completed survey responses. The variable costs associated with collecting 25 responses per clinician are US$55 for patient portal, US$71 for IVR, US$84 for SMS, US$337 for paper in-clinic and US$410 for tablet, indicating that the modes requiring the least personnel involvement that maintained reasonable response rates (SMS, IVR, patient portal) cost the least. While the tablet mode had high response rates and was the most time-efficient data collection method, the personnel time required made it the most expensive mode. Strengths and limitations We consider the work to have several strengths. First, we deployed a range of modes to deliver a PREM of SDM, daily and within 24 hours of the clinic visit. By consistently measuring the performance of the same clinicians, we were also able to assess mode effects on clinician rank order and CollaboRATE score, and demonstrate the feasibility of this approach to performance measurement. We collected CollaboRATE routinely in a clinic setting and experienced the related practical challenges: competing priorities for administrative, information systems, and clinical staff made survey implementation challenging. We faced significant challenges embedding these survey administration methods in institutional information systems, where programming SMS, patient portal and IVR methods were by necessity secondary to other institutional deadlines and demands. However, we were able to access and link patient demographic and clinical data with CollaboRATE responses, which allowed us to account for patient characteristics. The study also had limitations. Data collection for 15 months, using sequential administration modes, led to respondent and organisational fatigue, impacting response rates. Clinic staff were burdened by paper survey tasks. We received reports that patients who had previously completed the CollaboRATE survey perceived little value in repeating their evaluation. Our study design also did not account for potential repeated measures; since CollaboRATE's reference period is the patient's most recent clinical encounter, we assumed each CollaboRATE response to be independent and did not account for clustering by patient. Our sensitivity analysis excluding all but a patient's initial CollaboRATE response found results consistent with our original conclusions, suggesting only minimal potential influence of patient-level clustering on our results. Order of mode implementation was not randomised, and we were unable to account for potential order effects in our analysis. In addition, significant changes in clinic management impeded paper administration of CollaboRATE. SMS and IVR administration modes were hampered because the clinic record system did not consistently distinguish between cellular and other telephone numbers. Selection bias, enabled by low response rates in some data collection modes and the potential for unmeasured demographic differences between respondents and non-respondents, may contribute to variation in scores between modes. However, we believe the risk of selection bias is low given the similar measured demographic characteristics of respondents and non-respondents. Our cost analysis methods also relied on retrospective cost estimates. Context in existing literature CollaboRATE provided a consistent estimate of the level of SDM practiced by clinicians, with no difference in rank order of clinicians across modes despite a shift in CollaboRATE top score. These findings are similar to those from the review by Hood et al14 where mode was not related to differences in the precision of 36-Item Short Form Health Survey (SF-36) scores, but survey administration (self-report or interviewer-led) and sensory stimuli (auditory, visual or both) were the factors that impacted the absolute mean scores. Using the Tourangeau13 framework, it appears that collecting data in the clinic setting results in the highest scores, compared with the lower scores obtained using the IVR or SMS modes. Lower scores when using IVR are also found by Rodriguez et al23 who reported primary care patients had significantly lower scores on the Ambulatory Care Experience Survey by IVR, after adjusting for patient characteristics, compared with mail and internet administration. In addition, our findings match others who report lower evaluations when patients are asked to respond outside clinic settings.28–30 Only the in-clinic electronic tablet/mail-in mode achieved the 40% response rate recommended for CAHPS.31 The patient portal response rate of eligible email recipients was 34%, double the average response rate found in a recent review,17 which supports the view that technological methods yield improved response rates, provided respondent burden is avoided.14 The use of SMS was not as successful; unreliable cellular coverage was a factor, and it remains unclear whether this mode can provide representative data given demographic differences in the use of cellphones and text messages. Our cost analyses correspond with previous reports that the highest cost per completed survey is associated with in-clinic data collection.16 22 23 However, when set-up and research costs were excluded, all modes of administration were below US$20 per completed response and were <US$5 per completed response for three modes, namely patient portal, SMS and IVR. Implications Patient experience measurement is a relatively recent addition to healthcare settings and is widely viewed as a key method of assessing value.32 Nevertheless, there is a significant risk that the measurement of patient experience becomes viewed as a burden, where surveys are lengthy and administered out of context. Findings from our study suggest that CollaboRATE can overcome these challenges as it consists of only three items that can be delivered via multiple modes at a cost comparable to existing survey administration. In addition, clinician ranking was consistent, regardless of data collection mode. The decision of which mode is best for data collection will depend on the extent to which data collectors can harness technological tools and access patient information such as email addresses or telephone numbers to facilitate contact while maximising patient confidentiality. If multiple modes are used, it is important for administrators to account for mode effects when estimating CollaboRATE scores. We now face the challenge of studying CollaboRATE in more settings to test the generalisability of our findings and establish benchmarks for clinic-level and clinician-level performance. There is also a need to investigate automated collection, analysis and visualisation of the data so that the results can facilitate the core purpose of providing timely feedback to improve clinical practice. Clinicians and clinics who score low on CollaboRATE will want to know what they can do to achieve SDM and in turn provide to patients higher quality healthcare. We acknowledge the following individuals for their invaluable contributions to the project: Erik Olson and Shelley Sanyal for programming data collection and reporting systems for IVR/SMS and online patient portal data collection modes, respectively; Margaret Menkov for data management and medical record data extraction; Jef Hale, Scott Farr and Charles Goff for IT oversight; Chelsea Worthen, Ethan Berke and Michelle L'Heureux for clinic leadership. Twitter: Follow Paul Barr @BarrPaulJ Contributors: PJB contributed to design of the work, drafting the article and final approval of the version to be published. RCF contributed to design of the work, data analysis and interpretation, drafting the article and final approval of the version to be published. RT contributed to design of the work, critical revision of the article and final approval of the version to be published. EMO contributed to design of the work, data collection, data analysis and interpretation, critical revision of the article and final approval of the version to be published. RA contributed to the design of the work and final approval of the version to be published. MGC contributed to design of the work, data collection, critical revision of the article and final approval of the version to be published. AJO contributed to design of the work, data interpretation, critical revision of the article and final approval of the version to be published. GE contributed to conception and design of the work, drafting the article and final approval of the version to be published. Funding: This work was funded by the Gordon and Betty Moore Foundation, grant number 3929. Competing interests: GE reports personal fees from Emmi Solutions LLC, personal fees from National Quality Forum, personal fees from Washington State Health Department, personal fees from Shared Decision Making 3rd edition, personal fees from Groups (Radcliffe Press), outside the submitted work; and GE has initiated and led the Option Grid patient decision aids Collaborative, which produces and publishes patient knowledge tools in the form of comparison tables (http://optiongrid.org/). GE has been a member of teams that have developed measures of shared decision-making and care integration. These tools and measures are published and are available for use. For further information see http://www.glynelwyn.com/. Ethics approval: The study received ethics approval by Dartmouth Committee for the Protection of Human Subjects (#24529). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Wagner E , Coleman K , Reid R Guiding transformation: how medical practices can become patient-centered medical homes . Commonwealth Fund , 2012 http://www.commonwealthfund.org/~/media/Files/Publications/Fund%20Report/2012/Feb/1582_Wagner_guiding_transformation_patientcentered_med_home_v2.pdf (accessed 7 Dec 2016 ). 2 Dyer N , Sorra JS , Smith SA Psychometric properties of the Consumer Assessment of Healthcare Providers and Systems (CAHPS®) Clinician and Group Adult Visit Survey . Med Care 2012 ;50 :S28 –34 . 10.1097/MLR.0b013e31826cbc0d 23064274 3 U.S. Department of Health & Human Services . The CAHPS Program. http://www.ahrq.gov/cahps/about-cahps/cahps-program/index.html (accessed 24 Aug 2016 ). 4 U.S. Department of Health & Human Services . Clinician & Group Survey and Instructions. http://www.ahrq.gov/cahps/surveys-guidance/cg/instructions/index.html (accessed 24 Aug 2016 ). 5 Kasper J , Heesen C , Köpke S Patients’ and observers’ perceptions of involvement differ. Validation study on inter-relating measures for shared decision making . PLoS ONE 2011 ;6 :e26255 10.1371/journal.pone.0026255 22043310 6 Scholl I , Koelewijn-van Loon M , Sepucha KR Measurement of shared decision making—a review of instruments . Z Evid Fortbild Qual Gesundhwes 2011 ;105 :313 –24 . 10.1016/j.zefq.2011.04.012 21620327 7 Barr PJ , Elwyn G Measurement challenges in shared decision making: putting the ‘patient’ in patient-reported measures . Health Expect 2016 ;19 :993 –1001 . 10.1111/hex.12380 26111552 8 Elwyn G , Barr PJ , Grande SW Developing CollaboRATE: a fast and frugal patient-reported measure of shared decision making in clinical encounters . Patient Educ Couns 2013 ;93 : 102 –7 . 10.1016/j.pec.2013.05.009 23768763 9 Barr PJ , Thompson R , Walsh T The psychometric properties of CollaboRATE: a fast and frugal patient-reported measure of the shared decision-making process . J Med Internet Res 2014 ; 16 :e2 10.2196/jmir.3085 24389354 10 Thompson R , Elwyn G Patient-reported measurement of shared decision making: The development, validation, and pilot implementation of CollaboRATE. In: International Conference on Communication in Healthcare . Amsterdam , 2014 . 11 Rickert J An innovative patient-centered total joint replacement program . Heal Aff Blog 2016 http://healthaffairs.org/blog/2016/03/28/an-innovative-patient-centered-total-joint-replacement-program/ (accessed 8 Jul 2016 ). 12 Meterko M , Radwin L , Bokhour B Measuring patient-centered care: a brief measure for use in point-of-care assessments . White Paper. US Department of Veterans Affairs, Office of Patient-centered Care and Cultural Transformation 2015 . 13 Tourangeau R , Rips L , Rasinski K The psychology of survey response . Cambridge University Press , 2000 . 14 Hood K , Robling M , Ingledew D Mode of data elicitation, acquisition and response to surveys: a systematic review . Health Technol Assess 2012 ;16 :1 –162 . 10.3310/hta16270 15 Elliott MN , Zaslavsky AM , Goldstein E Effects of survey mode, patient mix, and nonresponse on CAHPS hospital survey scores . Health Serv Res 2009 ;44 :501 –18 . 10.1111/j.1475-6773.2008.00914.x 19317857 16 Bergeson SC , Gray J , Ehrmantraut LA Comparing Web-based with Mail Survey Administration of the Consumer Assessment of Healthcare Providers and Systems (CAHPS(®)) Clinician and Group Survey . Prim Health Care 2013 ;3 :1000132 10.4172/2167-1079.1000132 24078901 17 Gwaltney CJ , Shields AL , Shiffman S Equivalence of electronic and paper-and-pencil administration of patient-reported outcome measures: a meta-analytic review . Value Health 2008 ;11 :322 –33 . 10.1111/j.1524-4733.2007.00231.x 18380645 18 Kongsted A , Leboeuf-Yde C The Nordic back pain subpopulation program—individual patterns of low back pain established by means of text messaging: a longitudinal pilot study . Chiropr Osteopat 2009 ;17 :11 10.1186/1746-1340-17-11 19919715 19 Axén I , Bodin L , Bergström G The use of weekly text messaging over 6 months was a feasible method for monitoring the clinical course of low back pain in patients seeking chiropractic care . J Clin Epidemiol 2012 ;65 :454 –61 . 10.1016/j.jclinepi.2011.07.012 22169083 20 Christie A , Dagfinrud H , Dale Ø Collection of patient-reported outcomes;—text messages on mobile phones provide valid scores and high response rates . BMC Med Res Methodol 2014 ;14 :52 10.1186/1471-2288-14-52 24735061 21 Schober MF , Conrad FG , Antoun C Precision and disclosure in text and voice interviews on smartphones . PLoS ONE 2015 ;10 :e0128337 10.1371/journal.pone.0128337 26060991 22 California HealthCare Foundation . Measuring and improving patient experience in the safety net . 2011 http://www.chcf.org/~/media/MEDIA%20LIBRARY%20Files/PDF/PDF%20M/PDF%20MeasuringImprovingPatientExperienceSummary.pdf 23 Rodriguez HP , von Glahn T , Rogers WH Evaluating patients’ experiences with individual physicians: a randomized trial of mail, internet, and interactive voice response telephone administration of surveys . Med Care 2006 ;44 :167 –74 . 10.1097/01.mlr.0000196961.00933.8e 16434916 24 Makoul G , Krupat E , Chang CH Measuring patient views of physician communication skills: development and testing of the Communication Assessment Tool . Patient Educ Couns 2007 ;67 :333 –42 . 10.1016/j.pec.2007.05.005 17574367 25 U.S. Census Bureau . American FactFinder. http://factfinder.census.gov/faces/nav/jsf/pages/index.xhtml (accessed 26 Aug 2016 ). 26 Colla CH , Lewis VA , Gottlieb DJ Cancer spending and accountable care organizations: Evidence from the Physician Group Practice Demonstration . Healthc (Amst) 2013 ;1 :100 –7 . 10.1016/j.hjdsi.2013.05.005 25072017 27 Chisholm A , Askham J What do you think of your doctor? A review of questionnaires for gathering patients’ feedback on their doctor . Oxford, UK : Picker Institute , 2006 . 28 Bendall-Lyon D , Powers T Time does not heal all wounds . Mark Heal 2001 ;21 :10 –14 . 29 Jensen HI , Ammentorp J , Kofoed PE User satisfaction is influenced by the interval between a health care service and the assessment of the service . Soc Sci Med 2010 ;70 :1882 –7 . 10.1016/j.socscimed.2010.02.035 20382459 30 Bjertnaes OA The association between survey timing and patient-reported experiences with hospitals: results of a national postal survey . BMC Med Res Methodol 2012 ;12 :13 10.1186/1471-2288-12-13 22335801 31 U.S. Department of Health & Human Services . Frequently Asked Questions About CAHPS|Agency for Healthcare Research & Quality. http://www.ahrq.gov/cahps/faq/index.html (accessed 18 Jan 2016 ). 32 Nelson EC , Mohr JJ , Batalden PB Improving health care, Part 1: the clinical value compass . Jt Comm J Qual Improv 1996 ;22 :243 –58 .8743061

PMC005xxxxxx/PMC5372084.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01307610.1136/bmjopen-2016-013076Diabetes and EndocrinologyResearch150618431704Effectiveness of chronic care models for the management of type 2 diabetes mellitus in Europe: a systematic review and meta-analysis Bongaerts Brenda W C 12Müssig Karsten 234Wens Johan 5Lang Caroline 6Schwarz Peter 6Roden Michael 234Rathmann Wolfgang 12 1 Institute for Biometrics and Epidemiology, German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University Düsseldorf, Düsseldorf, Germany 2 German Center for Diabetes Research (DZD e.V.), Partner Düsseldorf, Düsseldorf, Germany 3 Department of Endocrinology and Diabetology, Medical Faculty, Heinrich Heine University Düsseldorf, Düsseldorf, Germany 4 Institute for Clinical Diabetology, German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University Düsseldorf, Düsseldorf, Germany 5 Department of Medicine and Health Sciences, Primary and Interdisciplinary Care Antwerp, University of Antwerp, Antwerp, Belgium 6 Department of Medicine III, Division of Prevention and Care of Diabetes, University of Dresden, Dresden, GermanyCorrespondence to Dr Brenda Bongaerts; brenda.bongaerts@ddz.uni-duesseldorf.de2017 20 3 2017 7 3 e01307618 6 2016 20 12 2016 23 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives We evaluated the effectiveness of European chronic care programmes for type 2 diabetes mellitus (characterised by integrative care and a multicomponent framework for enhancing healthcare delivery), compared with usual diabetes care. Design Systematic review and meta-analysis. Data sources MEDLINE, Embase, CENTRAL and CINAHL from January 2000 to July 2015. Eligibility criteria Randomised controlled trials focussing on (1) adults with type 2 diabetes, (2) multifaceted diabetes care interventions specifically designed for type 2 diabetes and delivered in primary or secondary care, targeting patient, physician and healthcare organisation and (3) usual diabetes care as the control intervention. Data extraction Study characteristics, characteristics of the intervention, data on baseline demographics and changes in patient outcomes. Data analysis Weighted mean differences in change in HbA1c and total cholesterol levels between intervention and control patients (95% CI) were estimated using a random-effects model. Results Eight cluster randomised controlled trials were identified for inclusion (9529 patients). One year of multifaceted care improved HbA1c levels in patients with screen-detected and newly diagnosed diabetes, but not in patients with prevalent diabetes, compared to usual diabetes care. Across all seven included trials, the weighted mean difference in HbA1c change was −0.07% (95% CI −0.10 to −0.04) (−0.8 mmol/mol (95% CI −1.1 to −0.4)); I2=21%. The findings for total cholesterol, LDL-cholesterol and blood pressure were similar to HbA1c, albeit statistical heterogeneity between studies was considerably larger. Compared to usual care, multifaceted care did not significantly change quality of life of the diabetes patient. Finally, measured for screen-detected diabetes only, the risk of macrovascular and mircovascular complications at follow-up was not significantly different between intervention and control patients. Conclusions Effects of European multifaceted diabetes care patient outcomes are only small. Improvements are somewhat larger for screen-detected and newly diagnosed diabetes patients than for patients with prevalent diabetes. Type 2 diabetes mellitusManaged careSystematic reviewMeta-analysisEuropeEuropean Commissionhttp://dx.doi.org/10.13039/5011000007802012 12 03 ==== Body Strengths and limitations of this study This is the first systematic review providing a comprehensive overview of studies that have evaluated the effectiveness of multifaceted diabetes care programmes addressing all their components together, rather than separately. The focus in this systematic review was on European multifaceted diabetes care programmes only, to meet the need for efficient and established programmes to providing optimal chronic care due to the burden of increasing diabetes prevalence in Europe. There is an important lack of studies which evaluate the effectiveness of implementing all chronic care model-components simultaneously. Overall, the studies included in this systematic review provided insufficient details to fully understand the intensity of the intervention, and there was only little overlap in the wide range of outcome measures evaluated. Introduction Chronic disease management relies on the assumption that providing optimal chronic care requires changes of patients and professionals with regard to behaviour, culture, and communication.1 2 Indeed, with ageing of the population and the growing prevalence of chronic diseases, initiatives to improving quality of chronic care require more than evidence about effective diagnostic procedures and treatments in comparison to acute disorders.3 Aimed at describing essential elements for improving outcomes in care of chronic diseases, the chronic care model (CCM) was developed in the mid-1990s and was further refined in 1997.2 4 5 This primary care-based model is based on the assumption that improvements in care require an approach that incorporates patients, healthcare providers and system level interventions.4 6 The CCM comprises six interrelated components deemed essential for providing high-quality care to patients with chronic disease: (1) healthcare organisation (ie, providing leadership for securing resources and removing barriers to care), (2) self-management support (ie, facilitating skills-based learning and patient empowerment), (3) decision support (ie, providing guidance for implementing evidence-based care), (4) delivery system design (ie, coordinating care processes), (5) clinical information systems (ie, tracking progress through reporting outcomes to patients and providers and (6) community resources and policies (ie, sustaining care by using community-based resources and public health policy).7 The current literature indicates a widespread application of the CCM to multiple illnesses, and various studies have provided a rigorous evaluation of its individual components.5 8–14 In general, these studies have reported positive effects on patient outcomes and processes of care. The reported effect sizes, however, are relatively small, and many outcomes are flawed by a considerable level of statistical heterogeneity.10 13–25 An aspect that complicates the assessment of effectiveness of chronic care programmes is their inherent multicomponent nature.14 20 25 While some authors found that the total number of CCM elements incorporated in the interventions did not influence patient outcomes,9 10 others concluded that interventions containing more than one CCM component were more successful at improving the quality of care than single-component interventions.11 24 26 27 To date, no summative reviews have evaluated to which extent the complete CCM—thus all six components combined in interventions—improves diabetes care. As such, the aim of the current review was to systematically identify studies of diabetes care assessing the effect of interventions addressing all six components of the CCM. We subsequently aimed to pool the effect of these models on biochemical outcomes (HbA1c, cholesterol levels, blood pressure, body mass index (BMI), fasting glucose, triglyceride and creatinine levels), patient-reported outcomes (health-related quality of life) and diabetes complications (macrovascular and microvascular complications, hypoglycaemia, cardiovascular risk, medication use and processes of care) in adult patients with type 2 diabetes compared to usual diabetes care by means of a meta-analysis. Methods Our systematic review was based on a protocol with input from experts in diabetes care, statistical methods and primary care. The protocol was composed according to the PRISMA-P guidelines (see online supplementary file S1).28 10.1136/bmjopen-2016-013076.supp1supplementary file Data sources and searches We identified studies by searching MEDLINE, Embase, CINAHL and CENTRAL from January 2000 until July 2015. Search syntaxes were developed in consultation with the Cochrane Metabolic and Endocrine Disorders Group by adapting and combining published search strategies from previous systematic reviews on chronic (diabetes) care management.10 12 Given that the CCM—and its terminology—had been introduced in the late 1990s, we restricted the search to publications from January 2000 onwards. In addition, reference lists of eligible studies and systematic reviews on multifaceted diabetes care were searched by hand to identify additional studies. The full MEDLINE search strategy is available in the online supplementary file S2. 10.1136/bmjopen-2016-013076.supp2supplementary file Study selection One reviewer (BWCB) identified potentially relevant studies for inclusion by screening title and abstract of all citations that resulted from our literature search. Two reviewers (BWCB and WR) then screened the full text of these articles. Only randomised controlled trials were considered eligible for inclusion. Non-randomised studies were excluded, as were studies written in a language other than English. Since this systematic review was part of a large European project on managed diabetes care that aimed at developing chronic care management standards and guidance for Europe,29 we further excluded all non-European CCM trials. Trials eligible for inclusion had to comply with the following inclusion criteria. Type of participants Individuals, regardless of gender and ethnicity, diagnosed with type 2 diabetes and with or without comorbidities. Type of intervention Previous systematic reviews on multifaceted chronic care have reported that randomised-controlled-trial-interventions are generally described poorly and incomprehensively, which complicates mapping the individual elements of the intervention to the six CCM components. To avoid mapping difficulties, we have reformulated the following inclusion criteria for the interventions: The intervention had to be described as a multifaceted CCM or programme that (1) was designed specifically for individuals with type 2 diabetes, (2) was based on guidelines, (3) provided multidisciplinary care, (4) addressed patient empowerment, (5) provided quality management (eg, patient registry systems, recording of process measurements and adherence to guidelines, achievement of treatment goals), (6) was delivered in primary or secondary care and (7) had a minimum duration of 6 months. The control intervention had to be defined as usual diabetes care as recommended in that particular country (eg, regular follow-up with the required health professional and a full diabetes annual review). Type of outcome measures We considered three categories of outcome measures: (1) biochemical outcomes, including HbA1c, cholesterol levels, blood pressure, BMI, fasting glucose, triglyceride and creatinine levels, (2) patient-reported outcomes, including health-related quality of life, and (3) diabetes complications, including macrovascular and microvascular complications, hypoglycaemia, cardiovascular risk, medication use and processes of care. Any disagreements between the two reviewers regarding the inclusion or exclusion of studies were resolved by consensus. Data extraction and quality assessment Using a standard structured data abstraction form, one reviewer (BWCB) performed the data extraction which was confirmed by a second reviewer (WR). The extracted data included study design, length of intervention/follow-up, sample size, inclusion and exclusion criteria, mean or median age of the included sample, percentage males, study setting (ie, primary or secondary care), intervention details and mean differences in change for various outcomes. When important information or outcome data were missing, trial authors of the included studies were contacted. When unavailable, the particular data were not included in the analyses. The standard Cochrane EPOC Risk of Bias Tool was used to assess risk of bias for each of the selected studies.30 Since all included studies were cluster-randomised controlled trials, additional attention was given to potential sources of bias specific to cluster-randomised trials: (1) recruitment bias: did recruitment of diabetes patients take place before or after randomisation of the clusters?, (2) did the intervention and control group differ in baseline characteristics?, (3) did any of the clusters drop out during follow-up?, (4) was clustering accounted for in the statistical analyses? If a certain domain could not be classified as ‘high’ or ‘low’ risk of bias due to inadequate reporting, it was deemed ‘unclear’ risk of bias. Data synthesis and analysis Owing to heterogeneity of the study populations and duration of the interventions, and owing to the small overlap in outcomes of the individual trials, an extensive meta-analysis and meta-regression of all reported outcome variables was not possible. The available data only allowed to statistically pool the results for HbA1c concentrations and total cholesterol levels. Review Manager (RevMan 5.2.0; the Cochrane Collaboration) was used to compute the weighted mean difference in change in HbA1c and total cholesterol between intervention and control groups, employing the generic inverse variance method. To incorporate between-study and within-study variance, we used a random effects model for estimating the weighted mean differences in change between intervention and control group across the included trials.31 Mean differences were pooled separately for the different types of diabetes patients (prevalent, screen-detected and newly diagnosed), and subsequently for the entire patient population. The consistency of the findings across the studies was assessed using forest plots. We evaluated statistical heterogeneity by calculating the I2 statistic, a measure independent of the number of studies and effect size metric.32 All outcome variables other than HbA1c and total cholesterol, we analysed descriptively. Results Figure 1 summarises the identification of relevant studies and the numbers of excluded and included studies. The search of the electronic databases identified 9464 abstracts of studies published between January 2000 and July 2015. After excluding duplicate citations (n=1227) and studies unrelated to the current review′’s topic (n=7801), we considered 436 articles for full-text review. Of these, 424 studies failed to meet our explicit inclusion criteria. In total, 12 articles met our inclusion criteria and were included in the current review.33–44 No relevant studies were retrieved by hand-search. Figure 1 Flow chart summarising the identification of studies for inclusion in the review. Study characteristics The 12 included articles33–44 reported on eight unique cluster randomised controlled trials,33 35 39–41 43–45 carried out between 1989 and 2011. Two of these trials, Addition-Denmark40 and Addition-Cambridge,35 had not individually reported any follow-up results in sequel to their study protocols. Their 5-year data however were pooled in the Addition-Europe study46 together with the 5-year data of the Addition-Netherlands39 and Addition-Leicester43 trials. For the remainder of the ‘Methods’ section, we will describe the design features and assess risk of bias for the Addition-Denmark and Addition-Cambridge trials based on their published protocol, yet for the ‘Results’ section we will have to resort to the pooled five-year data from the Addition-Europe study. This means that although we identified eight unique trials,33 35 39–41 43–45 there are just seven publications to extract data from.33 39 41 43–46 All trials had recruited either general practitioners or physician practices which represented the cluster level (level of randomisation). In one study,45 however, first-level clusters were formed by district (characterised as urban, rural and mixed) and second-level clusters by the physicians. The total number of patients with type 2 diabetes enrolled by the physicians amounted to 9529, of whom 8921 (94%) had been included in the analyses. The objective of each trial was the structured multifaceted management of diabetes, and the interventions were aimed at improving the patients’ cardiovascular risk profile44 45 and metabolic control,33 35 39 40 43 44 and assessing the effect of multifaceted care on the occurrence of cardiovascular events,35 39 40 43 overall mortality41 and risk factors for clinical complications.41 Interventions focused on all aspects of the CCM including more regular and frequent consultations, annual screening for diabetes complications, patient education/advice, guideline-based clinical treatment and physician education, regular/annual feedback reports to physicians, referrals, record keeping, formation of multidisciplinary (primary care provider) teams, delegation of routine diabetes tasks to a trained practice nurse, patient and physician reminders and patient–physician communication and decision-making. The interventions were largely delivered by general practitioners and physicians, yet specialised nurses or practice nurses were also involved in the intervention-programme as part of the practice team and to (partly) replace the physician in providing diabetes care.33 35 39 40 43 44 Two main aspects differed among the trials: the type of diabetes patient enrolled and the duration of the intervention. Three trials33 44 45 had included patients with prevalent diabetes and intervened for 1-year. The average diabetes duration in these studies ranged from 5.8 to 9.5 years. One trial41 had enrolled patients with newly diagnosed type 2 diabetes and assessed outcome measures after 6 years of intervention. Finally, there were four trials35 39 40 43 that first had initiated a diabetes screening programme and subsequently had recruited those with screen-detected diabetes to participate in the intervention study. Follow-up measurements were assessed at 1-year and at 5 years. Table 1 presents an overview of interventions and findings of the included publications. Tables 2 and 3 present the baseline patient characteristics for the trials that recruited patients with prevalent diabetes33 44 45 and for the trials that recruited patients with screen-detected39 43 46 and newly diagnosed diabetes,41 respectively. Table 1 Characteristics of the included cluster randomised controlled trials Study Comparison Effect on end points* Notes Cleveringa et al. (2008)33 Intervention: Patient consultation by a practice nurse+use of a computerised decision support system+guideline-based care+physician support by practice nurse+interdisciplinary care by a specialist team+individualised treatment advice+patient education+physician feedback+recall system+regular patient consultations by practice nurse+physician feedback versus Usual diabetes care (not further specified) Biochemical outcomes HbA1c (0) Total cholesterol (+, i) HDL-cholesterol (0) LDL-cholesterol (+, i) Systolic blood pressure (+, i) Diastolic blood pressure (+, i) 10-year CHD risk (+, i) Diabetes complications and processes of care HbA1c below target value† (+, i) Total cholesterol below target value† (+, i) LDL-cholesterol below target value† (+, i) Systolic blood pressure below target value† (+, i) All treatment targets reached† (+, i) At baseline, patients in the intervention group had higher HDL-cholesterol levels, were more often smoker and more often had a history of CHD. Statistical analyses were conducted by intention-to-treat and for missing follow-up data the last observation was carried forward. Comparisons between the intervention and control group were adjusted for cluster structure. Sönnichsen et al. (2008)45 Intervention: Physician education+guideline-based care+patient education+use of a clinical information system tool+interdisciplinary care by a specialist team+patient reminders+physician reminders+goal setting+shared decision-making patient and physician+regular consultations versus Usual diabetes care (not further specified) Biochemical outcomes HbA1c (0) Total cholesterol (+, i) HDL-cholesterol (0) LDL-cholesterol (0) Systolic blood pressure (0) Diastolic blood pressure (0) Body mass index (+, i) Triglycerides (0) Creatinine (0) Diabetes complications and processes of care To the guidelines adherent: – number of eye examinations† (+, i) – number of foot examinations† (+, i) – provision of patient education† (+, i) – regular HbA1c checks† (+, i) At baseline, patients in the intervention group had a higher BMI and higher cholesterol levels than patients in the control group. Statistical analyses were conducted by intention-to-treat and for missing follow-up data the last observation was carried forward. Comparisons between the intervention and control groups were adjusted for cluster structure and baseline characteristics. Frei et al. (2010)44 Intervention: Specialist team involving a practice nurse+practice nurse education+physician education+physician support by practice nurse+regular independent patient consultations by practice nurse+use of a clinical information system tool+guideline-based care+physician feedback+patient information leaflets+self-management support for patient+patient treatment groups versus Usual diabetes care (not further specified) Biochemical outcomes HbA1c (0) Total cholesterol (0) HDL-cholesterol (0) LDL-cholesterol (+, i) Systolic blood pressure (+, i) Diastolic blood pressure (+, i) Body mass index (0) Fasting blood glucose (0) Patient-reported outcomes Diabetes complications and processes of care Number GP visits† (0) Change in antidiabetic therapy (0) Change in antihypertensive therapy (0) Change in lipid-lowering therapy (0) There were no baseline differences in patient characteristics between intervention and control group. Statistical analyses were conducted by intention-to-treat and for missing follow-up data the last observation was carried forward. Comparisons between intervention and control group were adjusted for cluster structure and baseline characteristics. Webb et al. (2010)43 Intervention: Structured patient education+lifestyle advice and self-management with ongoing (bimonthly) professional support+individualised management+guideline-based care+shared decision-making patient and healthcare professional+annual screening for diabetic complications+care delivered by a specialist team (specialty doctor, diabetes nurse educator, and a dietician)+patient reminders+physician reminders versus Usual diabetes care (not further specified) Biochemical outcomes HbA1c (+, i) Total cholesterol (+, i) LDL-cholesterol (+, i) HDL-cholesterol (0) Systolic blood pressure (+, i) Diastolic blood pressure (+, i) Body mass index (+, i) Weight (+, i) Waist circumference (0) Triglycerides (0) 5-year CHD risk (+, i) 5-year CVD risk (+, i) Patient-reported outcomes Health-related quality of life (0) Diabetes complications and processes of care Hypoglycaemia† (+, i) Use of antihypertensive drugs† (+, i) Use of lipid-lowering drugs† (+, i) Use of antiplatelet therapy† (+, i) Use of metformin† (0) Use of sulfonylurea† (0) At baseline, more patients in the intervention group were taking antihypertensive medication when entering the study and had higher total and LDL-cholesterol levels. Statistical analyses were conducted by intention-to-treat. It was not reported whether or not data were missing and how missing data were handled. Comparisons between intervention and control group were adjusted for cluster structure and baseline characteristics (except quality of life which had not been measured at baseline). Janssen et al. (2009)39 Intervention: Physician education+diabetes nurse education+lifestyle advice+guideline based care+physician support by diabetes nurse+evaluation and feed-back sessions diabetes nurse+frequent patient consultations with diabetes nurse+shared decision-making patient, physician and diabetes nurse+physician reminders+patient reminders versus Usual diabetes care (not further specified) Biochemical outcomesHbA1c (+, i) Total cholesterol (+, i) LDL-cholesterol (+, i) HDL-cholesterol (0) Systolic blood pressure (+, i) Diastolic blood pressure (+, i) Body mass index (+, i) Fasting blood glucose (+, i) Triglycerides (0) Patient-reported outcomes Health-related quality of life (0) Diabetes complications and processes of care Hypoglycaemia† (0) There were no baseline differences in patient characteristics between the intervention and control group. Statistical analyses were conducted by intention-to-treat and for missing follow-up data the last observation was carried forward. Comparisons between the intervention and control group were adjusted for baseline characteristics, and clustering at practice level. Griffin et al. (2011)46 This study combined the data after five years of a multifaceted care intervention from the (1) Addition-Denmark study (Lauritzen et al40), (2) the Addition-Netherlands study (Janssen et al39), (3) the Addition-Cambridge study (Echouffo et al35) and (4) the Addition-Leicester study (Webb et al43) in a meta-analysis. Biochemical outcomes HbA1c (+, i) Total cholesterol (+, i) LDL-cholesterol (+, i) HDL-cholesterol (0) Systolic blood pressure (+, i) Diastolic blood pressure (+, i) Body mass index (0) Weight (0) Waist circumference (0) Triglycerides (0) Creatinine (+, c) Patient-reported outcomes Health-related quality of life (0) Diabetes complications and processes of care All-cause mortality (0) CVD mortality (0) Myocardial infarction (0) Stroke (0) Revascularisation procedures (0) Hypoglycaemia† (0) Meeting target values for:  HbA1c (+, i)  Blood pressure (+, i)  Total cholesterol (+, i) Use of any glucose-lowering drugs (+, i) Change in any antihypertensive drugs (+, i) Change in any cholesterol-lowering drugs (+, i) Baseline characteristics were well matched between intervention and control group. In Denmark however, more patients were identified in practices assigned to the intervention arm then in those assigned to control arm. And in the intervention group, more patients had a history of ischaemic heart disease. Statistical analyses were conducted by intention-to-treat and patients with missing outcome values at baseline were excluded from the analyses. Those with missing outcome baseline values were included according to the missing indicator method. Comparisons between intervention and control group were adjusted for cluster structure and baseline characteristics. Olivarius et al. (2001)41 Intervention: Patient follow-up every three months+annual screening for diabetes complications+shared decision-making patient and physician+physician feedback+goal setting+clinical guidelines+physician education+patient leaflets and folders+lifestyle advise+protocol based care+physician recall system versus Usual diabetes care (not further specified) Biochemical outcomes HbA1c (+, i) Total cholesterol (+, i) Systolic blood pressure (+, i) Diastolic blood pressure (0) Weight (0) Fasting blood glucose (+, i) Triglycerides (0) Creatinine (0) Diabetes complications and processes of care Overall mortality† (0) Severe hypoglycaemia† (0) Diabetic retinopathy† (0) Non-fatal myocardial infarction† (0) Non-fatal stroke† (0) Peripheral neuropathy† (0) Microalbuminuria† (0) Angina pectoris† (0) Intermittent claudication† (0) Number of consultations† (+, i) Number of referrals to diabetes clinic† (−, i) Number of hospital admissions† (0) Use of metformin† (+, i) Use of other glucose-lowering drugs† (0) Use of antihypertensive drugs† (0) Use of lipid-lowering drugs† (0) At baseline, more patients in the intervention group were excluded because of severe somatic disease than in the control group. Furthermore, occupation and smoking habits differed between the two groups. Statistical analyses were conducted by intention-to-treat. It was not reported whether or not data were missing or how missing data were handled. Comparisons between intervention and control group were adjusted for cluster structure and baseline characteristics. *+=positive effect; 0=no effect; −=negative effect; i=favouring intervention group; u=favouring control (usual care) group. The effects of the intervention are represented by the difference in change from baseline to follow-up between the intervention and control group. † The effect of the intervention is represented by a difference in proportions of patients at follow-up between the intervention and control group. CHD, coronary heart disease; CVD, cardiovascular (heart) disease; GP, general practitioner; T2DM, type 2 diabetes mellitus. Table 2 Baseline patient characteristics of the included cluster randomised controlled trials studying patients with prevalent diabetes   Cleveringa et al33 * Sönnichsen et al45 † Frei et al44 ‡ Intervention Control Intervention Control Intervention Control N 1699 1692 649 840 162 164 Follow-up duration (years) 1 1 1 1 1 1 Type of diabetes patients Prevalent diabetes Prevalent diabetes Prevalent diabetes Country Netherlands Austria Switzerland Baseline patient characteristics  Age (years) 65.2±11.3 65.0±11.0 65.4±10.4 65.5±10.4 65.7±10.4 68.3±10.6  Sex (% men) 48.2 49.8 51.0 53.1 54 60  Ethnicity (% Caucasian) 97.7 97.6 − − − −  Diabetes duration (years) 5.8±5.7 5.4±5.8 7.0±6.5 9.5±7.4 10.3±7.8  Current smoking (% yes) 22.6 16.6 13.4 14 9  Body mass index (kg/m2) 30.0±5.3 30.2±5.3 30.4±5.1 29.7±4.9 30.5±5.3 30.7±5.9  Systolic blood pressure (mm Hg) 149±22 149±21 141±19 139±17 140±18 138±17  Diastolic blood pressure (mm Hg) 83±11 82±11 83±11 82±10 83±10 79±10  UKDPS CHD risk (%) 22.5±16.5§ 21.7±15.8§ − − − −  HbA1c (%) 7.1±1.3 7.0±1.1 7.46±1.53 7.34±1.31 7.8±1.5 7.6±1.1  Total cholesterol (mmol/L) 5.0±1.0 4.9±1.1 5.15±1.14 5.02±1.09 5.0±1.2 4.7±1.1  HDL-cholesterol (mmol/L) 1.36±0.36 1.32±0.35 1.35±0.39 1.32±0.36 1.2±0.3 1.3±0.4  LDL-cholesterol (mmol/L) 2.8±0.92 2.8±0.95 2.87±0.96 2·87±0·91 2·8±1·1 2·5±1·1  Fasting glucose (mmol/L) 8·0±2·4 7·8±2·2 − − 8·4±2·5 7·7±2·2  Creatinine (μmol/L) 87.5±27.7 85.9±22.5 84.9±30.9 84.9±34.5 − −  Triglycerides (mmol/L) 1.8±1.1 1.8±1.3 2.14±1.82 2.00±1.73 − −  Urinary albumin (mg/L) − − − − − −  Quality of life: PCS¶ 43.9±10.9  Quality of life: MCS¶ 50.1±11.3  History of myocardial infarction (%) 47.1 63.3 8.4 − −  History of stroke (%) 7.0 − −  Diabetic retinopathy (%) 2.9 3.3 - - 9.3 8.1  Peripheral neuropathy (%) − − − − 18.6 13.4 Values are mean±sd, or percentages. Bold font indicates that the particular baseline characteristic differed statistically significantly between the intervention and control group. *The information on BMI, fasting glucose, creatinine, triglycerides and retinopathy was obtained through contacting the authors. †The information on diabetes duration, smoking, history of myocardial infarction and history of stroke was obtained from the publication describing baseline characteristics of the total study population and stratified by sex (Flamm et al60). ‡The quality of life summary scores for the physical and mental component were obtained from the publication describing baseline characteristics of the total study population (Frei et al61). Peripheral neuropathy is represented by ‘pathological foot status’ and diabetic retinopathy is represented by ‘annual eye exam: pathological’. §Values concern the 10-year UKDPS CHD risk. ¶Quality of life was assessed with the 36-item Short Form Health Survey (SF-36). CHD, coronary heart disease; MCS, Mental Component Summary Score; PCS, Physical Component Summary Score; UKPDS, UK Prospective Diabetes Study. Table 3 Baseline patient characteristics of the included cluster randomised controlled trials studying patients with screen-detected and newly diagnosed diabetes   Webb et al43 Janssen et al39 Griffin et al46 Olivarius et al41 Intervention Control Intervention Control Intervention Control Intervention Control N 146 199 255 243 1678 1379 649 614 Follow-up duration (years) 1 1 1 1 5 5 6 6 Type of diabetes patients Screen-detected diabetes Screen-detected diabetes Screen-detected diabetes Newly diagnosed diabetes Country UK Netherlands UK, Netherlands, Denmark Denmark Baseline patient characteristics  Age (years) 59.4±10.0 60.0±10.0 60.1±5.4 59.9±5.1 60.3±6.9 60.2±6.8 65.5 (55.3–74.0) 65.3 (56.3–73.5)  Sex (% men) 56.9 58.3 51.8 56.0 58.5 57.3 52.4 53.1  Ethnicity (% Caucasian) 52.7 62.3 98.0 98.7 95.8 93.4 − −  Diabetes duration (years) 0 0 0 0 0 0 0 0  Current smoking (% yes) 15.2 10.2 26.3 21.4 26.9 27.8 35.5 34.5  Body mass index (kg/m2) 31.0±5.9 31.5±5.7 31.2±5.1 30.4±4.6 31.6±5.6 31.6±5.6 29.4 (26.2–33.0) 28.8 (26.0–32.3)  Systolic blood pressure (mm Hg) 145.7±18.5 148.4±20.5 166±23 163±23 148.5±22.1 149.8±21.3 150 (130–164) 148 (130–160)  Diastolic blood pressure (mm Hg) 87.8±10.4 89.5±10.7 90±11 89±10 86.1±11.1 86.5±11.3 85 (80–90) 85 (80–90)  UKPDS CHD risk (%) 8.5±5.8† 9.3±7.1* − − − − − −  HbA1c (%) 7.2±1.5 7.3±1.8 7.3±1.6 7.4±1.7 7.0±1.6 7.0±1.5 10.2 (8.6–11.6) 10.2 (8.7–11.9)  Total cholesterol (mmol/L) 5.3±1.2 5.6±1.3 5.6±1.1 5.6±1.1 5.5±1.1 5.6±1.2 6.2 (5.4–7.1) 6.2 (5.5–7.2)  HDL-cholesterol (mmol/L) 1.2±0.4 1.2±0.3 1.1±0.4 1.1±0.3 1.2 (1.0–1.5) 1.2 (1.0–1.5) − −  LDL-cholesterol (mmol/L) 3.2±1.0 3.5±1.0 3.7±1.0 3.7±1.0 3.4±1.0 3.5±1.0 − −  Fasting glucose (mmol/L) − − 7.8±2.3 8.1±2.8 − − 13.8 (10.7–17.0) 13.7 (10.7–17.0)  Creatinine (μmol/L) − − − − 83.4±17.1 84.9±18.6 90 (81–101) 88 (79–100)  Triglycerides (mmol/L) 2.1±1.9 2.1±1.4 1.9±1.0 2.0±1.6 1.6 (1.2–2.3) 1.7 (1.2–2.4) 2.03 (1.44–2.91) 1.98 (1.39–2.95)  Urinary albumin (mg/L) − − − − − − 11.7 (6.0–32.5) 11.8 (5.7–27.5)  Quality of life: PCS† 39.0 (37.4–40.5) 38.5 (37.1-40.0) No summary scores reported − − − −  Quality of life: MCS† 38.2 (35.2–41.2) 39.2 (36.5–41.9) No summary scores reported − − − −  History of myocardial infarction (%) 15.8* 10.6‡ − − 6.8 6.1 6.6 7.7  History of stroke (%) − − 2.9 1.9 3.5 4.2  Diabetic retinopathy (%) − − − − − − 5.0 4.5  Peripheral neuropathy (%) − − − − − − 18.8 19.7 Values are mean±sd, or median (IQR) or percentages. Bold font indicates that the comparison between the intervention and control group was statistically significant. *Values concern the 5-year UKDPS CHD risk. †Quality of life was assessed with the 12-item Short Form Health Survey (SF-12) in the study by Webb et al, and with the 36-item Short Form Health Survey (SF-36) in the study by Janssen et al. ‡Defined as ‘pre-existing CVD’, including myocardial infarction, stroke and angina. CHD, coronary heart disease; MCS, Mental Component Summary Score; PCS, Physical Component Summary Score; UKPDS, UK Prospective Diabetes Study. Data quality assessment Figure 2 summarises the risk of bias for the trials included in this review. Although the Addition-Denmark40 and the Addition-Cambridge35 trials had not published 1-year data, they did provide 5-year data for the Addition-Europe meta-analysis46 and were thus included in the risk of bias assessment. However, since not having published actual trial data, we could not assess the domains of incomplete outcome data, selective reporting and other bias, which resulted in the occurrence of blanks in figure 2. Figure 2 Risk of bias graph. Review authors' judgements about each risk of bias item presented as percentages across all included studies. Studies included are Cleveringa et al (2008);33 Sönnichsen et al (2008),45 Frei et al (2010),44 Olivarius et al (2001),41 Janssen et al (2009),39 Webb et al (2010),43 Lauritzen et al (2000)40 and Echouffo et al (2009).35 The studies from Lauritzen and Echouffo were included in the risk of bias assessment since their 5-year follow-up data had been included in the Addition-Europe meta-analysis by Griffin et al.46 As the Addition-Europe publication only reported pooled data, no comprehensive overview of results was available for the studies by Lauritzen and Echouffo, which resulted in the blanks in the risk of bias graph. Seven trials had at least one domain judged as unclear risk of bias. Five trials had at least one domain judged as high risk of bias. Only one study44 had explicitly described that their physicians were unaware of being allocated to the intervention or control group when recruiting eligible patients. For the remaining studies, prior knowledge of treatment allocation cannot be ruled out (recruitment bias). Furthermore, the Addition studies35 39 40 43 were the only trials in which patients remained unaware of group assignment throughout the study. In four studies,35 39 40 43 outcome assessment was performed completely blinded for patient allocation. In one study,45 only laboratory outcomes were assessed blinded, whereas clinical outcomes were obtained by contacting the general practitioner, introducing possible bias. No substantial baseline differences between the intervention and control groups existed with regard to the outcomes of interest. Biochemical outcomes All studies had assessed biochemical outcomes at follow-up, including HbA1c level, blood lipid levels, blood pressure and BMI. HbA1c levels All studies assessed HbA1c values at follow-up. For six33 39 43–46 of the seven study populations glycaemic control at baseline was moderate to good, as expressed by the mean HbA1c concentrations ranging from 7.0 to 7.8% (53 to 62 mmol/mol). The three trials with prevalent type 2 diabetes patients33 44 45 observed no statistically significant difference in change in HbA1c levels between the intervention and control group after 1-year of intervention (figure 3). There was no statistical heterogeneity between these three trials (I2=0%) and the weighted mean difference in change between intervention and control groups was −0.06% (95% CI −0.13 to 0.01) (−0.7 mmol/mol (95% CI −1.4 to 0.1)), in favour of the intervention group. Using a similarly short intervention period, yet studying patients with screen-detected type 2 diabetes, the Addition-Leicester trial43 observed a significant difference in change in HbA1c between the two trial arms of −0.20% (95% CI −0.32 to −0.08) (−2.2 mmol/mol (95% CI −3.4 to −0.9)). Whereas the Addition-Netherlands authors39 did not report the actual difference in HbA1c change between the two groups, they stated in their paper that the improvement in HbA1c was significantly better in the intervention group, compared to the control group. The pooled 5-year data from all four Addition-trials46 showed a somewhat smaller, yet significantly greater improvement in HbA1c concentration in intervention patients, compared to control patients (−0.08% (95% CI −0.14 to −0.02)) (−0.9 mmol/mol (95% CI −1.5 to −0·2)) (figure 3). Finally, the effect of multifaceted care in Danish patients with newly diagnosed diabetes41 after 6 years of intervention was comparable to that in screen-detected patients after 5 years of intervention46 (−0.06% (95% CI −0.08 to −0.03)) (−0.7 mmol/mol (95% CI −0.9 to −0.3)). Figure 3 Mean difference in change (95% CI) in HbA1c levels (%) after multifaceted care between intervention and control groups. Results are stratified by type of diabetes patient. IV, generic inverse variance method. The three studies including patients with prevalent diabetes has an intervention duration of 1-year. The methodology for calculating the difference in change between intervention and control group that Cleveringa et al.33 have used (subtracting the HbA1c change over time for the control group from the change over time for the intervention group) was the opposite of that used by the other trials (subtracting the HbA1c change over time for the intervention group from the change over time for the control group). Since this would result in a misleading visual presentation of the findings from Cleveringa et al.,33 we have recalculated their HbA1c results according to the methodology used by the other studies. The study of Webb et al.43 had an intervention duration of one year and the study of Griffin et al.46 combined the 5-year intervention data from all four Addition studies, including the five-year data from Webb et al.43 The study including patients with newly detected diabetes had an intervention duration of six years. Pooling all seven trials, multifaceted care improved HbA1c concentration with −0.07% (95% CI −0.10 to −0.04) (−0.8 mmol/mol (95% CI −1.1 to −0.4)) (figure 3). Statistical heterogeneity across the seven trials was small to moderate (I2=21%). Cholesterol levels Figure 4 presents the mean differences in change in total cholesterol levels for all seven trials. Of the three trials that studied prevalent diabetes patients, only the Dutch trial33 observed multifaceted care to significantly improve total cholesterol concentrations. In the remaining two studies,44 45 cholesterol levels were similar between intervention and control arm. Statistical heterogeneity across the three studies was low (I2=12%) and their weighted mean difference in change between intervention and control groups amounted to −0.14 mmol/L (95% CI −0.22 to −0.07). Similar to HbA1c, the effect of multifaceted care on cholesterol seemed larger in screen-detected patients than in patients with prevalent diabetes. After 1-year of intervention, Addition-Leicester43 found a mean difference in change between the intervention and control group of −0.56 mmol/L (95% CI −0.87 to −0.25). The pooled 5-year data from all four Addition trials also showed a significantly greater improvement in total cholesterol levels in intervention patients, compared to control patients (−0.27 mmol/L (95% CI −0.34 to −0.20)). Finally, in Danish patients with newly diagnosed diabetes,41 6 years of multifaceted care had caused cholesterol levels to improve (−0.15 mmol/L (95% CI −0.29 to −0.01)). Figure 4 Mean difference in change (95% CI) in total cholesterol levels (mmol/L) after multifaceted care between intervention and control groups. Results are stratified by type of diabetes patient. IV, generic inverse variance method. The three studies including patients with prevalent diabetes has an intervention duration of 1-year. The methodology for calculating the difference in change between intervention and control group that Cleveringa et al.33 have used (subtracting the HbA1c change over time for the control group from the change over time for the intervention group) was the opposite of that used by the other trials (subtracting the HbA1c change over time for the intervention group from the change over time for the control group). Since this would result in a misleading visual presentation of the findings from Cleveringa et al.,33 we have recalculated their HbA1c results according to the methodology used by the other studies. The study of Webb et al.43 had an intervention duration of one year and the study of Griffin et al.46 combined the 5-year intervention data from all four Addition studies, including the five-year data from Webb et al.43 The study including patients with newly detected diabetes had an intervention duration of six years. Pooling all trials, the effect of multifaceted care on improvement of total cholesterol resulted in a weighted difference in change between intervention and control patients of −0.20 mmol/L (95% CI −0.28 to −0.11); I2=64%. In addition to improvements in total cholesterol levels, HDL-cholesterol levels appeared to be unaffected by multifaceted care in patients with prevalent diabetes.33 44 45 LDL-cholesterol levels on the other hand did improve (see online supplementary figure S1 and S2). The Dutch33 and the Swiss44 study found significantly better improvements in LDL-cholesterol for the intervention group, when compared to the control group. The Addition-Netherlands39 and Addition-Leicester43 studies observed that multifaceted care significantly improved LDL-cholesterol levels after 1-year, while HDL-cholesterol remained largely unchanged. Similar results were reported for 5 years of intervention by the Addition-Europe study.46 The Danish study41 with newly diagnosed diabetes patients had not measured HDL and LDL-cholesterol levels. 10.1136/bmjopen-2016-013076.supp4supplementary figures Blood pressure Two33 44 out of the three trials with patients with prevalent diabetes reported a difference in change in diastolic and systolic blood pressure, both being in favour of the intervention group (see online supplementary figures S3 and S4). Better improvements in blood pressure were also seen in intervention patients with screen-detected diabetes, compared to control patients.39 43 46 Improvements after 1-year of intervention43 were larger than those after 5 years of intervention.46 In patients with newly diagnosed diabetes41 6 years of multifaceted care significantly improved systolic, but not diastolic, blood pressure when compared to usual diabetes care. Similar to HbA1c and total cholesterol, the results for blood pressure were stronger for patients with screen-detected and newly diagnosed diabetes than for those with prevalent, long-standing diabetes. Body mass index With regard to the studies on prevalent diabetes, only the Austrian study45 found a significant difference in change in BMI between the intervention group and control group after 1-year of intervention (see online supplementary figure S5). In screen-detected diabetes patients,39 43 multifaceted care resulted in a significantly higher reduction in BMI, compared to usual diabetes care. Furthermore, Addition-Leicester43 reported a higher reduction in BMI and body weight (kg) for the intervention group compared to the control group, but observed no difference in reduction of waist circumference. After an intervention duration of 5 years, the pooled reduction in weight and waist circumference, but not in BMI, in screen-detected diabetes was significantly higher in the intervention group compared to the control group.46 The Danish trial41 with newly diagnosed diabetes patients observed no difference in weight change after 6 years of intervention, yet BMI had not been measured. For further biochemical outcomes, see online supplementary file S3 and figures S6–S8. 10.1136/bmjopen-2016-013076.supp3supplementary file Patient-reported outcomes The effect of a multifaceted care intervention on the patients’ quality of life accounted for the only patient-reported outcome assessed by the included trials. Health-related quality of life Quality of life was reported by five33 39 43 44 46 of the seven trials, most of which had used the 36-item Short Form Health Survey (SF-36) to assess the different domains of health-related quality of life. In patients with prevalent diabetes,33 44 significant changes over time were absent for all scores of the SF-36 subscales for the intervention and control arms. A superior effect of multifaceted care was observed only on the SF-36 subscale ‘health change’ in the Dutch trial with prevalent diabetes patients.33 For the two Addition studies reporting results after 1-year of intervention,39 43 as for the pooled 5-year data by Addition-Europe,46 no significant changes in the physical and mental summary scores of the SF-36, or the abbreviated SF-12 version that was used in the Addition-Leicester trial,43 could be demonstrated. Diabetes complications Only few trials had reported diabetes complications, including cardiovascular disease and mortality. Closely related to the prevention and occurrence of complications, some studies evaluated the effect of their intervention on processes of care, such as reaching target values for HbA1c and receiving regular eye and foot examinations. Macrovascular and microvascular complications Macrovascular and microvascular diabetes complications during follow-up were reported by the two studies41 46 with the longer intervention periods. The Addition-Europe study46 had assessed myocardial infarction, stroke, coronary and peripheral revascularisation procedures, cardiovascular death and total mortality and non-traumatic amputation in screen-diagnosed diabetes patients. Although the estimated HRs for these events all favoured the intervention group, none of the estimates reached statistical significance. In newly diagnosed diabetes patients,41 multifaceted care had not resulted in differences between intervention and control group regarding the risk of diabetic retinopathy, peripheral neuropathy, microalbuminuria, non-fatal myocardial infarction and stroke, angina pectoris or intermittent claudication at 6 years. Processes of care Only three studies assessed processes of care or process quality measures.33 45 46 The Dutch study33 with prevalent diabetes patients observed that multifaceted care resulted in significantly more patients reaching treatment targets (18.9%) than usual diabetes care (13.4%) (treatment targets were defined as HbA1c ≤7% (53 mmol/mol), systolic blood pressure ≤140 mm Hg, total cholesterol ≤4.5 mmol/L and LDL-cholesterol ≤2.5 mmol/L). Process quality measures at 1-year, defined as the percentage of patients receiving guideline-adherent foot-examinations, eye-examinations and HbA1c-examinations, were reported by the Austrian study with prevalent diabetes patients45 to be significantly higher in the intervention group. The pooled 5-year results from the four Addition studies46 showed that in both trial arms more patients had values below target thresholds for HbA1c (<7% (53 mmol/mol)), blood pressure (≤135/85 mm Hg) and cholesterol level (<4.5 mmol/L), yet proportions were higher in the intervention group than in the control group. For further diabetes complications and related outcomes, see online supplementary file S3. Discussion This review assessed the effectiveness of chronic disease management models for type 2 diabetes on the improvement of patient outcomes, in Europe. In general, the effects of multifaceted care on patient outcomes were rather small and their magnitude seemed to differ according to the type of diabetes patient being studied. Our analysis suggested that in comparison to usual diabetes care, multifaceted care improves HbA1c levels for patients with screen-detected diabetes and patients with newly diagnosed diabetes, but not for patients with prevalent type 2 diabetes. Similar findings were observed for total cholesterol, LDL-cholesterol, BMI and body weight. The resulting improvements in blood pressure seemed less strongly related to the type of diabetes patient studied. Other outcomes, such as fasting glucose levels, triglycerides, quality of life and diabetes complications, had been reported inconsequently, and results varied widely across the included trials. The few cluster randomised controlled trials that we identified from the literature were relatively heterogeneous with regard to the individual components of the implemented intervention, duration of the intervention, type of diabetes patient and reported outcomes. For each trial, methodological quality was acceptable and there were very low rates of dropout among the enrolled patients. Still, details on the randomisation procedure were frequently missing as well as information concerning concealment of allocation from general practitioners and physicians in advance to recruitment of eligible patients. Since the currently performed meta-analysis included only a small number of trials, caution is warranted not to overinterpret its results. The χ2 statistic for example, indicating homogeneity of the effect of the intervention on HbA1c and total cholesterol, has low power when based on only few, and small-sized, studies.47 When interpreting the data, we thus prefer to look at the direction of the individual effect estimates and CIs, rather than let the calculated statistics guide our conclusions. As such, given the current literature, it is not possible to draw an unequivocal conclusion about the effectiveness of chronic multifaceted care on diabetes patient outcomes. Overall, previous systematic reviews have reported that an integrated approach to diabetes care versus usual diabetes care may improve clinical and biochemical outcomes,9 10 19 20 23 24 48 including HbA1c levels, blood pressure and blood lipid concentrations. Those reviews that included a meta-analysis reported mean differences in HbA1c reduction between intervention and control groups ranging from −0.14% (95% CI −0.25 to −0.05) to −0.5% (95% CI −0.6 to −0.3). Mean differences in total cholesterol have only been estimated by one meta-analysis, which reported a reduction of −0.24 mmol/L (95% CI −0.41 to −0.06) in favour of the intervention group.10 This study also reported a mean difference in diastolic blood pressure reduction of −1.3 mm Hg (95% CI −0.21 to −0.6) and a mean difference in systolic blood pressure reduction of −2.2 mm Hg (95% CI −3.5 to −0.9), comparable with the summary estimate for systolic blood pressure from Elissen et al (−2.8 mm Hg (95% CI −4.7 to −0.9)).20 All other outcomes of multifaceted care interventions were described narratively. Improvements have been observed for frequency of retinopathy screening,20 48 49 screening for peripheral polyneuropathy and foot lesions,20 48 49 proteinuria measurements49 and the monitoring frequency of lipid and HbA1c levels.49 In addition, there seems to be an economic benefit of integrated diabetes care.50 Yet, other systematic reviews have found no impact on patients outcomes and processes of care18 25 49 or have disputed the clinical relevance of statistically significant findings.19 A comparison of the reported effect estimates with our summary estimates for HbA1c and total cholesterol warrants caution, given the varying number of CCM elements the estimates were based on, the heterogeneity among the included diabetes patients, the different restrictions to geographical region and the number of included studies in each review. The novelty of the current systematic review is that it provides a comprehensive overview of diabetes care trials that have evaluated the effectiveness of the all the six components of the CCM combined, instead of one or more components. Overall, we found there is an important lack of studies which evaluate the implementation of all six CCM-components simultaneously. In the current literature, findings on the issue of whether multifaceted chronic care is to be preferred over single-faceted care are conflicting.9–12 24–26 51 However, improving the management of a complex disease like diabetes is a challenging goal which, we believe, may not be achieved by targeting single care aspects only. Another novel aspect of the current review is the focus on state-of-the-art diabetes management in Europe only. The narrow view relates to the enormous burden that type 2 diabetes represents in Europe, in individual and in societal terms.52 The prevalence of diabetes in Europe is expected to increase from 59.8 million adults in 2015 to 71.1 million in 2040.53 As reflected by recent guidelines for the management of patients with type 2 diabetes,54 healthcare providers have increasingly focused at improving and controlling cardiovascular risk factors to improve patient outcomes, including hyperglycaemia, overweight or obesity, elevated blood pressure and dyslipidemia. Results from the Steno-2 trial support the view that even in high-risk patients with type 2 diabetes multifaceted care has the potential to reduce the risk of complications and mortality.55 Randomising 160 patients with type 2 diabetes and persistent microalbuminuria to an intensive multifactorial treatment and conventional therapy, the authors found that the multifactorial treatment was associated with a lower risk of cardiovascular events after 13.3 years of follow-up, as well as with a lower risk of death from cardiovascular disease, compared to conventional treatment. And while the CCM has been proposed as a tool to improve the quality of diabetes care and, subsequently, patient outcomes, the current review indicates that at least the existing programmes have not been as successful in this respect as intended. The challenge thus remains to translate results from landmark studies like Steno-2, into primary care, where the majority of type 2 diabetes patients are being treated. When aiming to improve chronic healthcare, it has been proposed that only assessing the effects of a multifaceted care intervention on patient outcomes is not sufficient. In order to gain insights into why and when certain interventions are effective, it is also important to focus on barriers and facilitators to the implementation process of the intervention and their effect on the interplay between intervention and outcomes.56 This latter aspect is usually not evaluated or reported on by randomised controlled trials implementing a multifaceted care intervention.57 As such, it has not yet been possible to analyse the relationships between context, mechanisms and outcomes of multifaceted diabetes care interventions and to subsequently provide meaningful insights into how these have influenced the outcomes achieved.57 There are some limitations of our work that need to be considered. First, many studies provided insufficient detail in their methods section to fully understand the intensity of (specific components of) the intervention. This complicated our appraisal of whether all components of the CCM were fully covered. Also, the different interventions that the trials have used to represent a given component of the CCM have possibly resulted in some heterogeneity across the trials. In addition to the insufficiently described interventions, standards for usual diabetes care were not elaborated on in any of the trials. Online versions of diabetes care guidelines were found to be published in the country’s native language and represented current versions only. However, most European countries define their standards according to the recommendations made by the joint task force convened by the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD).54 58 Indeed, identified guidelines from the Netherlands, Austria, and the UK did comply with the ADA/EASD recommendations. We do therefore not expect that practices of usual diabetes care in the individual trials have differed to the extent of causing a significant increase in heterogeneity. Second, whereas the aim of the current review was to investigate the effectiveness of CCMs in Europe, the trials available for this review only represented the Western part of Europe. Countries with the highest prevalence of diabetes lie in Eastern Europe, that is, Turkey, Montenegro, Macedonia and Serbia.52 The top-three countries in Western Europe with the highest diabetes prevalence are Germany, Spain and Italy,52 none of which were represented in this review. And third, the procedure of selecting relevant studies for the current review was largely performed by only one person. However, two reviewers subsequently screened the full text of all potentially relevant papers such that the final decision on inclusion was based on two opinions. In conclusion, the available scientific evidence regarding the effectiveness of multifaceted chronic care programmes for type 2 diabetes in older patients in Europe is low. In general, the current findings support the concept of the CCM, yet the improvements in patient outcomes and processes of care are only small. While key aspects of type 2 diabetes can be improved by a multifactorial intervention, it is not yet clear if these improvements will subsequently lower diabetes-related complications, such as cardiovascular disease and overall mortality. Furthermore, the effect of the interventions seemed, at least partly, to depend on the type of diabetes patient, which could suggest effect modification by disease duration and/or disease severity. Another aspect that could add to the differences in effectiveness between the individual interventions is the degree in which they facilitate changes in social behaviour. This implies that more attention in trials should be spent to factors like adherence to treatment strategies, level of self-management skills and patients’ knowledge on their disease. These traits need to be positively affected before an improvement in clinical measures can even occur,1 yet studies generally reveal little on person-centred factors. And finally, there is a lack of knowledge (on information) on effective methods to address important pragmatic questions about improvement of care, for example, which specific mechanism or procedure of a CCM works, for which patients and under which circumstances?59 Future research would need to incorporate the measurement of context, mechanisms and outcomes of multifaceted care into study designs in order to deliver the full extent of insights needed to improve chronic diabetes care and, ultimately, patient outcomes. The authors thank Trials Search Co-ordinator Maria-Inti Metzendorf and Professor Bernd Richter (MD) from the Cochrane Metabolic and Endocrine Disorders group (University Hospital Düsseldorf, Germany) for their valuable assistance, guidance and advice offered while developing the literature search strategy. The authors thank the trial authors of the Dutch Diabetes Care Implementation Study, the Swiss Chronic CARE for diAbeTes Study (CARAT) and the Danish Diabetes Care in General Practice study for kindly providing us additional trial results. Furthermore, we are grateful to Professor Oliver Kuß (PhD) from the Institute for Biometrics and Epidemiology of the German Diabetes Center (Düsseldorf, Germany) for his useful contributions to developing the review protocol. Contributors: BWCB designed the review by writing the review protocol, identified studies for inclusion, extracted and interpreted the data and drafted and revised the article. KM contributed to the review protocol and to the discussion. He further revised the draft paper for intellectual content. JW involved in conception of the review and he contributed to the review protocol, to interpretation of the data and to the discussion. Furthermore, JW revised the draft paper for intellectual content. CL contributed to the review protocol and to the discussion, and she revised the draft paper for intellectual content. PS conceived and initiated the review, contributed to the review protocol and contributed to the interpretation of the data, to the discussion and to revision of the draft paper. MR involved in conception of the review and revised the draft paper for intellectual content. WR contributed to the review protocol, identified studies for inclusion, extracted and interpreted the data and revised the draft paper for intellectual content. All authors approved the final completed article. Funding: The MANAGE-CARE project—of which this systematic review was part—was supported by grants from the European Commission (Grant Agreement 2012 12 03). The funding body had no influence on the design and conduct of the study, interpretation of the data and contents and publication of this manuscript. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Lemmens KM , Nieboer AP , van Schayck CP A model to evaluate quality and effectiveness of disease management . Qual Saf Health Care 2008 ;17 :447 –53 . 10.1136/qshc.2006.021865 19064661 2 Wagner EH , Austin BT , Von Korff M Organizing care for patients with chronic illness . Milbank Q 1996 ;74 :511 –44 . 10.2307/3350391 8941260 3 Clark CM , Fradkin JE , Hiss RG Promoting early diagnosis and treatment of type 2 diabetes: The National Diabetes Education Program . JAMA 2000 ;284 :363 –5 . 10.1001/jama.284.3.363 10891969 4 Wagner EH , Davis C , Schaefer J A survey of leading chronic disease management programs: are they consistent with the literature? Manag Care Q 1999 ;7 :56 –66 .10620960 5 Wagner EH , Austin BT , Davis C Improving chronic illness care: translating evidence into action . Health Aff (Millwood) 2001 ;20 :64 –78 . 10.1377/hlthaff.20.6.64 11816692 6 Glasgow RE , Orleans CT , Wagner EH Does the chronic care model serve also as a template for improving prevention? Milbank Q 2001 ;79 :579 –612 , iv–v 10.1111/1468-0009.00222 11789118 7 http://www.improvingchroniccare.org, 09-07-2015 . 8 Bodenheimer T , Wagner EH , Grumbach K Improving primary care for patients with chronic illness . JAMA 2002 ;288 :1775 –9 . 10.1001/jama.288.14.1775 12365965 9 Tsai AC , Morton SC , Mangione CM A meta-analysis of interventions to improve care for chronic illnesses . Am J Manag Care 2005 ;11 :478 –88 .16095434 10 Si D , Bailie R , Weeramanthri T Effectiveness of chronic care models-oriented interventions to improve quality of diabetes care: a systematic review . Prim Health Care Res Dev 2008 ;9 :25 –40 . 11 Stellefson M , Dipnarine K , Stopka C The chronic care model and diabetes management in US primary care settings: a systematic review . Prev Chronic Dis 2013 ;10 :E26 10.5888/pcd10.120180 23428085 12 Zwar N , Harris M , Griffiths R A systematic review of chronic disease management . Research Centre for Primary Health Care and Equity, School of Public Health and Community Medicine, University of New South Wales , 2006 . 13 Tricco AC , Ivers NM , Grimshaw JM Effectiveness of quality improvement strategies on the management of diabetes: a systematic review and meta-analysis . Lancet 2012 ;379 :2252 –61 . 10.1016/S0140-6736(12)60480-2 22683130 14 Shojania KG , Ranji SR , McDonald KM Effects of quality improvement strategies for type 2 diabetes on glycemic control: a meta-regression analysis . JAMA 2006 ;296 :427 –40 . 10.1001/jama.296.4.427 16868301 15 Glazier RH , Bajcar J , Kennie NR A systematic review of interventions to improve diabetes care in socially disadvantaged populations . Diabetes Care 2006 ;29 :1675 –88 . 10.2337/dc05-1942 16801602 16 Loveman E , Royle P , Waugh N Specialist nurses in diabetes mellitus . Cochrane Database Syst Rev 2003 ;(2) :CD003286 .12804458 17 Norris SL , Chowdhury FM , Van Le K Effectiveness of community health workers in the care of persons with diabetes . Diabet Med 2006 ;23 :544 –56 . 10.1111/j.1464-5491.2006.01845.x 16681564 18 Renders CM , Valk GD , Griffin S Interventions to improve the management of diabetes mellitus in primary care, outpatient and community settings . Cochrane Database Syst Rev 2001 ;(1) :CD001481 .11279717 19 Egginton JS , Ridgeway JL , Shah ND Care management for type 2 diabetes in the United States: a systematic review and meta-analysis . BMC Health Serv Res 2012 ;12 :72 10.1186/1472-6963-12-72 22439920 20 Elissen AM , Steuten LM , Lemmens LC Meta-analysis of the effectiveness of chronic care management for diabetes: investigating heterogeneity in outcomes . J Eval Clin Pract 2013 ;19 :753 –62 .22372830 21 Housden L , Wong ST , Dawes M Effectiveness of group medical visits for improving diabetes care: a systematic review and meta-analysis . CMAJ 2013 ;185 :E635 –644 . 10.1503/cmaj.130053 23939218 22 Ivers NM , Tricco AC , Taljaard M Quality improvement needed in quality improvement randomised trials: systematic review of interventions to improve care in diabetes . BMJ Open 2013 ;3 :pii: e002727 10.1136/bmjopen-2013-002727 23 Pimouguet C , Le Goff M , Thiebaut R Effectiveness of disease-management programs for improving diabetes care: a meta-analysis . CMAJ 2011 ;183 :E115 –127 . 10.1503/cmaj.091786 21149524 24 Shojania KG , Ranji SR , Shaw LK Closing the quality gap: a critical analysis of quality improvement strategies (Vol 2: Diabetes Care) . Rockville, MD: Agency for Healthcare Research and Quality; 2004. September. Technical Review no. 9. AHRQ publication no. 04-0051-2 . 25 Baptista DR , Wiens A , Pontarolo R The chronic care model for type 2 diabetes: a systematic review . Diabetol Metab Syndr 2016 ;8 :7 10.1186/s13098-015-0119-z 26807158 26 Boaz A , Baeza J , Fraser A Effective implementation of research into practice: an overview of systematic reviews of the health literature . BMC Res Notes 2011 ;4 :212 10.1186/1756-0500-4-212 21696585 27 Brusamento S , Legido-Quigley H , Panteli D Assessing the effectiveness of strategies to implement clinical guidelines for the management of chronic diseases at primary care level in EU Member States: a systematic review . Health Policy 2012 ;107 :168 –83 . 10.1016/j.healthpol.2012.08.005 22940062 28 Moher D , Shamseer L , Clarke M Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement . Syst Rev 2015 ;4 :1 10.1186/2046-4053-4-1 25554246 29 http://www.managecare-project.eu/ . 30 Higgins JP , Altman DG , Sterne JAC Chapter 8: Assessing risk of bias in included studies . In: Higgins JP , Green S , eds. Cochrane Handbook for Systematic Reviews of Interventions , Version 510 [updated March 2011] The Cochrane Collaboration , 2011 http://www.cochrane-handbook.org 31 DerSimonian R , Laird N Meta-analysis in clinical trials . Control Clin Trials 1986 ;7 :177 –88 . 10.1016/0197-2456(86)90046-2 3802833 32 Higgins JP , Thompson SG Quantifying heterogeneity in a meta-analysis . Stat Med 2002 ;21 :1539 –58 . 10.1002/sim.1186 12111919 33 Cleveringa FGW , Gorter KJ , van den Donk M Combined task delegation, computerized decision support, and feedback improve cardiovascular risk for type 2 diabetic patients: a cluster randomized trial in primary care . Diabetes care 2008 ;31 :2273 –5 . 10.2337/dc08-0312 18796619 34 Cleveringa FGW , Minkman MH , Gorter KJ Diabetes Care Protocol: effects on patient-important outcomes. A cluster randomized, non-inferiority trial in primary care . Diabet Med 2010 ;27 :442 –50 . 10.1111/j.1464-5491.2010.02968.x 20536517 35 Echouffo-Tcheugui JB , Simmons RK , Williams KM The ADDITION-Cambridge trial protocol: a cluster—randomised controlled trial of screening for type 2 diabetes and intensive treatment for screen-detected patients . BMC Public Health 2009 ;9 :136 10.1186/1471-2458-9-136 19435491 36 Flamm M , Panisch S , Winkler H Effectiveness of the Austrian disease management programme ‘Therapie Aktiv’ for type 2 diabetes regarding the improvement of metabolic control, risk profile and guideline adherence: 2 years of follow up . Wien Klin Wochenschr 2012 ;124 :639 –46 . 10.1007/s00508-012-0226-x 22872229 37 Flamm M , Panisch S , Winkler H Impact of a randomized control group on perceived effectiveness of a disease management programme for diabetes type 2 . Eur J Public Health 2012 ;22 :625 –9 . 10.1093/eurpub/ckr147 21990343 38 Frei A , Senn O , Chmiel C Implementation of the chronic care model in small medical practices improves cardiovascular risk but not glycemic control . Diabetes Care 2014 ;37 :1039 –47 . 10.2337/dc13-1429 24513589 39 Janssen PG , Gorter KJ , Stolk RP Randomised controlled trial of intensive multifactorial treatment for cardiovascular risk in patients with screen-detected type 2 diabetes: 1-year data from the ADDITION Netherlands study . Br J Gen Pract 2009 ;59 :43 –8 . 10.3399/bjgp09X394851 19105915 40 Lauritzen T , Griffin S , Borch-Johnsen K The ADDITION study: proposed trial of the cost-effectiveness of an intensive multifactorial intervention on morbidity and mortality among people with type 2 diabetes detected by screening . Int J Obes Relat Metab Disord 2000 ;24 (Suppl 3 ):S6 –11 . 10.1038/sj.ijo.0801420 11063279 41 Olivarius NF , Beck-Nielsen H , Andreasen AH Randomised controlled trial of structured personal care of type 2 diabetes mellitus . BMJ 2001 ;323 :970 –5 . 10.1136/bmj.323.7319.970 11679387 42 Sönnichsen AC , Winkler H , Flamm M The effectiveness of the Austrian disease management programme for type 2 diabetes: a cluster-randomised controlled trial . BMC Fam Pract 2010 ;11 :86 10.1186/1471-2296-11-86 21054881 43 Webb DR , Khunti K , Srinivasan B Rationale and design of the ADDITION-Leicester study, a systematic screening programme and randomised controlled trial of multi-factorial cardiovascular risk intervention in people with type 2 diabetes mellitus detected by screening . Trials 2010 ;11 :16 10.1186/1745-6215-11-16 20170482 44 Frei A , Chmiel C , Schläpfer H The Chronic CARe for diAbeTes study (CARAT): a cluster randomized controlled trial . Cardiovasc Diabetol 2010 ;9 :23 10.1186/1475-2840-9-23 20550650 45 Sönnichsen AC , Rinnerberger A , Url MG Effectiveness of the Austrian disease-management-programme for type 2 diabetes: study protocol of a cluster-randomized controlled trial . Trials 2008 ;9 :38 10.1186/1745-6215-9-38 18565213 46 Griffin SJ , Borch-Johnsen K , Davies MJ Effect of early intensive multifactorial therapy on 5-year cardiovascular outcomes in individuals with type 2 diabetes detected by screening (ADDITION-Europe): a cluster-randomised trial . Lancet 2011 ;378 :156 –67 . 10.1016/S0140-6736(11)60698-3 21705063 47 Deeks JJ , Higgins JP , Altman DG Chapter 9: Analysing data and undertaking meta-analyses . In: Higgins JP , Green S , eds. Cochrane Handbook for Systematic Reviews of Interventions Version 510 [updated March 2011] The Cochrane Collaboration , 2011 http://www.cochrane-handbook.org . 48 Knight K , Badamgarav E , Henning JM A systematic review of diabetes disease management programs . Am J Manag Care 2005 ;11 :242 –50 .15839184 49 Norris SL , Nichols PJ , Caspersen CJ The effectiveness of disease and case management for people with diabetes. A systematic review . Am J Prev Med 2002 ;22 :15 –38 . 10.1016/S0749-3797(02)00423-3 50 de Bruin SR , Heijink R , Lemmens LC Impact of disease management programs on healthcare expenditures for patients with diabetes, depression, heart failure or chronic obstructive pulmonary disease: a systematic review of the literature . Health Policy 2011 ;101 :105 –21 . 10.1016/j.healthpol.2011.03.006 21592607 51 Bodenheimer T , Wagner EH , Grumbach K Improving primary care for patients with chronic illness: the chronic care model, Part 2 . JAMA 2002 ;288 :1909 –14 . 10.1001/jama.288.15.1909 12377092 52 International Diabetes Federation . IDF Diabetes Atlas . 6th edn Brussels, Belgium : International Diabetes Federation , 2013 http://www.idf.org/diabetesatlas . 53 International Diabetes Federation . IDF Diabetes Atlas . 7th edn Brussels, Belgium : International Diabetes Federation , 2015 http://www.idf.org/diabetesatlas . 54 Inzucchi SE , Bergenstal RM , Buse JB Management of hyperglycemia in type 2 diabetes, 2015: a patient-centered approach: update to a position statement of the American Diabetes Association and the European Association for the Study of Diabetes . Diabetes Care 2015 ;38 :140 –9 . 10.2337/dc14-2441 25538310 55 Gaede P , Lund-Andersen H , Parving HH Effect of a multifactorial intervention on mortality in type 2 diabetes . N Engl J Med 2008 ;358 :580 –91 . 10.1056/NEJMoa0706245 18256393 56 Berwick DM The science of improvement . JAMA 2008 ;299 :1182 –4 . 10.1001/jama.299.10.1182 18334694 57 Busetto L , Luijkx KG , Elissen AM Context, mechanisms and outcomes of integrated care for diabetes mellitus type 2: a systematic review . BMC Health Serv Res 2016 ;16 :18 10.1186/s12913-015-1231-3 26772769 58 Inzucchi SE , Bergenstal RM , Buse JB Management of hyperglycemia in type 2 diabetes: a patient-centered approach: position statement of the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD) . Diabetes Care 2012 ;35 :1364 –79 . 10.2337/dc12-0413 22517736 59 Davidoff F , Batalden P , Stevens D Publication guidelines for quality improvement studies in health care: evolution of the SQUIRE project . BMJ 2009 ;338 :a3152 .19153129 60 Flamm M , Winkler H , Panisch S Quality of diabetes care in Austrian diabetic patients willing to participate in a DMP - at baseline . Wien Klin Wochenschr 2011 ;123 :436 –43 .21691756 61 Frei A , Herzog S , Woitzek K Characteristics of poorly controlled Type 2 diabetes patients in Swiss primary care . Cardiovasc Diabetol 2012 ;11 :70 .22704274

PMC005xxxxxx/PMC5372095.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01415310.1136/bmjopen-2016-014153Mental HealthResearch1506171217101712Cross-sectional survey on defensive practices and defensive behaviours among Israeli psychiatrists Reuveni I 1Pelov I 1Reuveni H 2Bonne O 1Canetti L 13 1 Department of Psychiatry, Hadassah Hebrew University Medical Center, Jerusalem, Israel 2 Department of Health Systems Management, Faculty of Health Sciences, Ben-Gurion University of the Negev, Beer-Sheva, Israel 3 Department of Psychology, Hebrew University, Jerusalem, IsraelCorrespondence to Dr I Reuveni; reuveni@hadassah.org.ilIR and IP contributed equally to this article. 2017 20 3 2017 7 3 e01415311 9 2016 17 1 2017 8 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective Psychiatry is a low-risk specialisation; however, there is a steady increase in malpractice claims against psychiatrists. Defensive psychiatry (DP) refers to any action undertaken by a psychiatrist to avoid malpractice liability that is not for the sole benefit of the patient's mental health and well-being. The objectives of this study were to assess the scope of DP practised by psychiatrists and to understand whether awareness of DP correlated with defensive behaviours. Methods A questionnaire was administered to 213 Israeli psychiatry residents and certified psychiatrists during May and June 2015 regarding demographic data and experience with malpractice claims, medicolegal literature and litigation. Four clinical scenarios represented defensive behaviours and reactions (feelings and actions) to malpractice claims. Results Forty-four (20.6%) certified psychiatrists and four (1.9%) residents were directly involved in malpractice claims, while 132 (62.1%) participants admitted to practising DP. Residents acknowledged the practice of DP more than did senior psychiatrists (p=0.038). Awareness of DP correlated with unnecessary hospitalisation of suicidal patients, increased unnecessary follow-up visits and prescribing smaller drug dosages than required for pregnant women and elderly patients. Conclusions This study provides evidence that DP is well established in the routine clinical daily practice of psychiatrists. Further studies are needed to reveal whether DP effectively protects psychiatrists from malpractice suits or, rather, if it impedes providing quality psychiatric care and represents an economic burden that leads to more harm for the patient. Defensive MedicineDefensive Psychiatry (DP)Medical Malpractice ==== Body Strengths and limitations of this study This is the first survey among Israeli psychiatrists to assess the scope of defensive practices and behaviours carried out in the public and private sectors. Defensives practices were assessed in four clinical domains (suicidal, pregnant and elderly patients, and medication initiation/change) to target major issues in the field of psychiatry liable to medicolegal actions. The study identified specific actions taken by psychiatrists to avoid malpractice liability. Psychiatrists' perceptions of defensive medicine, contrary to objective data, may be biased and lead physicians to overstate the frequency of performing defensive medicine. Introduction Defensive medicine comprises medical actions that deviate from sound medical practice, performed primarily to reduce exposure to malpractice liability or to provide legal protection in the case of a malpractice lawsuit.1–3 Defensive psychiatry (DP) refers to any action undertaken by a psychiatrist mainly to avoid malpractice liability, rather than for the sole benefit of the patient's mental health and well-being.4 There are two main forms of defensive medical behaviours described in the literature: (1) Assurance behaviour (‘positive defensive medicine’), which involves ordering diagnostic tests and/or treatments, referrals to other physicians and additional services of marginal medical value merely to prevent or limit liability. An example in psychiatry would be a patient with suicidal ideation who could and should be treated as an outpatient, but is hospitalised merely for defensive reasons. (2) Avoidance behaviour (‘negative defensive medicine’) refers to the physician’s reluctance to be involved in the treatment of high-risk patients or procedures.2 4 An example of this is the reluctance to prescribe medication to pregnant women suffering from affective or anxiety disorders, although there are clear indications to begin pharmacological treatment. The practice of defensive medicine places a great economic burden on society; in addition, it is not supported by evidence-based studies and can be harmful due to complications from unnecessary tests and procedures.1 3 5–7 Various studies have tried to evaluate the cost of defensive medicine over the years. Kessler and McClellan5 showed that defensive medicine is responsible for up to 9% of total health spending. Others showed lower percentages (around 1.5% of the total health expenditure). Mello and colleagues6 estimated the total annual cost of medical liability in 2008 to be more than 55 billion dollars in the USA. Nonetheless, it is a widespread phenomenon rooted in various fields of medicine. Studdert et al2 showed that up to 93% of physicians in a high-risk environment practise defensive medicine. Asher et al 3 demonstrated, in a nationwide survey in Israel, that defensive behaviours are common (up to 60% prevalence) in eight medical disciplines, four of which are not considered to be at high risk for litigation. Another study showed that 97% of obstetricians and gynaecologists felt that their daily work practice was affected by concerns about being sued for medical negligence.8 Psychiatry is considered a low-risk specialisation.9 10 However, data from recent years demonstrated that there is a steady increase in reports of medical negligence, claims of malpractice and reports at the state board level against psychiatrists.4 11 Some of the allegations made in litigation cases in the field of psychiatry include incorrect diagnosis, incorrect or ineffective treatment, medication errors, improper detention while hospitalised, doctor–patient boundary violations and inadequate assessment and management of suicidal patients.4 10 12 Research regarding DP is scarce and mostly limited to suicide assessments. Aims of study The primary aim of this study was to assess the scope of defensive medicine practised by Israeli psychiatrists in the public and private sectors. The secondary aim was to understand how one’s awareness of defensive practices correlates with applying defensive behaviours and the psychological impact of past malpractice claims. Methods Two hundred and thirteen Israeli certified psychiatrists and residents in psychiatry volunteered to complete a cross-sectional survey on defensive practices and attitudes. This study was approved by the Israel Psychiatric Association and was administered during the Triannual Congress of the Israel Psychiatric Association in May 2015. The survey was completely anonymous and included an introduction with an explanation regarding the nature of the survey and definition for defensive medicine. There is only one previous questionnaire of defensive medicine in the psychiatric literature, reported by Passmore and Leung;13 therefore, we started with a replication of the previous questionnaire and added more details to examine the scope of DP as well as its application. Our research team consulted with experienced psychiatrists holding academic positions in university medical centres on important topics in the field in order to develop the questionnaire. It was pretested on 17 psychiatry residents (mean age =33.8, SD=3.6) working in public hospitals, who volunteered to participate in the survey. Following the pretest, minor changes were made according to the research team's suggestions. The survey took ∼10 min to complete. The final questionnaire asked about demographic data (age, gender, professional position within department, work experience and work in public and private practices) as well as personal experience with malpractice claims and exposure to medicolegal literature and litigation. To survey the extent of defensive medicine, we asked a direct question: ‘Do you practice defensive medicine?’ Admission of practising defensive medicine with at least half of the physician’s patients was considered as acknowledgment of defensive medicine by the participant. In addition, we asked participants about defensive behaviours in various clinical scenarios. Specifically, we inquired about 13 behaviours in four domains (see online supplementary material). These domains were chosen by our research team to target major issues in the field of psychiatry and day-to-day practice, as well as high-risk cases for medicolegal actions. The four scenarios of possible defensive practice were: (1) treatment of suicidal patients, (2) treatment of pregnant women, (3) initiating or changing drug treatment and (4) treatment of elderly patients. Participants scored the practice of specific defensive behaviours on a 5-point Likert scale as follows: 5 ‘with every patient’, 4 ‘with most patients’, 3 ‘with half of the patients’, 2 ‘with a few patients’ and 1 ‘with no patient’. To assess the internal reliability, we calculated Cronbach's α of the 13 items that measured defensive behaviours, resulting in good internal consistency: α=0.67. Participants were also asked about their feelings (anxious, restless, angry, loss of energy or tired, guilty and mistrustful) and functioning (sleep problems and interference with work, family or social activities) in the period they were involved in malpractice claims. 10.1136/bmjopen-2016-014153.supp1supplementary methods Statistical analysis Continuous data were analysed using the t-test for independent samples or Pearson correlation coefficients. Reported p values are two-sided. All analyses were performed using the IBM SPSS V.21.0 (IBM Corp, 2012) statistical software. Results The demographic characteristics of the psychiatrists surveyed are presented in table 1. In our sample, both sexes were almost equally represented; about three-quarters were certified psychiatrists and slightly less than half were in a management position. Most participants (77.9%) worked in a public hospital and more than half (53.5%) had a private practice. Of the 213 psychiatrists, only 48 (22%) were directly involved in malpractice claims. Among them, 44 (91.7%) were certified psychiatrists and four (8.3%) were residents. Table 1 Demographic characteristics of the sample of Israeli psychiatrists Age M=48.00 (SD=11.82) Gender  Male 117 (54.9%)  Female 96 (45.1%) Experience  Resident 44 (20.7%)  Certified 169 (79.3%) Department position  Resident 44 (20.7%)  Consultant 67 (31.4%)  Department head 102 (47.9%) Place of work  Public hospital 166 (77.9%)  HMO 41 (19.2%)  Private practice 114 (53.5%) History of malpractice claims  Yes 48 (22.5%)  No 165 (77.5%) Reading medicolegal literature and litigation  Yes 104 (48.8%)  No 109 (51.2%) HMO, health maintenance organisation. There was a small but significant negative correlation between the age of the participants and acknowledgement of defensive practice (r=−0.14, p=0.049), meaning that there is a tendency in younger participants to admit that they practise defensive medicine. In treating suicidal patients, female psychiatrists were more prone to consult with a senior psychiatrist than male practitioners (female psychiatrists: M=3.41, SD=1.27; male psychiatrists: M=2.85, M=1.20; t(171)=−2.96; p=0.004) and more prone to refer to another mental health professional (female psychiatrists: M=3.28, SD=0.87; male psychiatrists: M=2.90, SD=1.13; t(180)=−2.59; p=0.010). There were no significant differences in the acknowledgement or practice of defensive medicine when examining the department position, place of work or reading medicolegal literature. Participants reporting a history of malpractice claims were more prone to acknowledge defensive practice (M=3.02, SD=0.96) compared with those who did not (M=2.60, SD=0.85) and this difference was significant (t(192)=2.82; p=0.005). There were significant differences between residents in psychiatry and certified psychiatrists for the practice of defensive medicine (table 2). Residents acknowledged the practice of defensive medicine more than did experienced psychiatrists (p=0.038). For suicidal patients, residents were more prone to advise hospitalisation (p=0.017) and to consult with senior psychiatrists (p<0.001) than certified psychiatrists. Residents avoided the prescription of drugs to pregnant patients more than did experienced psychiatrists (p=0.025). For elderly patients treated with antipsychotics, certified psychiatrists explained the risks of cerebrovascular diseases more than did residents (p=0.009). Table 2 Defensive medicine: comparison between residents and certified psychiatrists   Certified Resident M (SD) M (SD) t df p Value Acknowledgement of defensive practice 2.62 (0.93) 2.90 (0.73) 2.11 196 0.038 Defensive behaviours Suicidal patients  Advises unwarranted hospitalisation 2.82 (1.09) 3.34 (1.28) 2.46 199 0.017  Increases follow-up 3.51 (1.10) 3.75 (1.04) 1.30 199 0.195  Initiates contact with family 4.05 (0.83) 3.96 (0.87) −0.51 178 0.611  Consults senior psychiatrist 2.84 (1.14) 4.62 (0.75) 10.16 171 <0.001  Refers to another professional 3.03 (1.03) 3.35 (1.06) 1.47 180 0.144  Prescribes medication without indication 1.71 (0.72) 2.00 (0.89) 1.83 180 0.069 Changing or initiating new medication  Informs about severe yet rare side effects 3.45 (1.29) 3.30 (1.27) −0.69 208 0.490  Records that explained about side effects 3.49 (1.25) 3.27 (1.28) −1.02 207 0.308  Informs of increased risk of suicidality 2.74 (1.44) 2.52 (1.37) −0.88 205 0.380 Pregnant patients  Avoids medication altogether 2.56 (1.04) 3.09 (1.08) 2.25 186 0.025  Prescribes a smaller dosage 2.44 (1.21) 2.46 (1.25) 0.062 191 0.950 Elderly patients  Informs of cerebrovascular diseases risk 3.00 (1.30) 2.37 (1.38) −2.65 191 0.009  Prescribes a smaller dosage 3.95 (0.93) 3.95 (0.66) −0.02 195 0.986 In our sample, 62.1% of participants admitted practising defensive medicine with at least half of their patients (table 3), and this was very common in all four surveyed domains. To understand the relationship between acknowledging defensive practice and actually practising defensive medicine, we computed the correlation between the answer to ‘Do you practice defensive medicine?’ and self-reports of defensive practice behaviours in the four clinical scenarios mentioned above. As shown in table 3, participants felt they were employing defensive procedures when treating suicidal patients when they advised hospitalisation, even if not necessary, or increased the frequency of follow-up visits when not warranted. They also felt that they were practising defensive medicine when prescribing smaller drug dosages than required in the treatment of pregnant women and elderly patients. Other behaviours, even if very frequent, were apparently not considered defensive medicine by our subjects; thus, they did not correlate with the acknowledgement of defensive practice. Table 3 Defensive medicine: frequencies and correlations with acknowledgement of defensive practice and with anxiety Per cent of defensive practice or behaviours Correlation with acknowledgement of defensive practice Correlation with anxiety Acknowledgement of defensive practice 62.1% 1.00 0.30* Defensive behaviours Suicidal patients  Advises unwarranted hospitalisation 54.2% 0.34† 0.47†  Increases follow-up 75.6% 0.23† 0.25‡  Initiates contact with family 93.3% −0.11 0.10  Consults senior psychiatrist 52.6% 0.14 0.20  Refers to another professional 65.9% 0.01 0.17  Prescribes medication without indication 10.4% 0.09 −0.03 Changing or initiating new medication  Informs about severe yet rare side effects 72.9% −0.10 0.18  Records that explained about side effects 74.2% 0.06 0.08  Informs of increased risk of suicidality 47.3% −0.01 0.23‡ Pregnant patients  Avoids medication altogether 46.8% 0.05 0.39†  Prescribes a smaller dosage 45.6% 0.27† 0.05 Elderly patients  Informs of cerebrovascular diseases risk 55.4% −0.06 0.18  Prescribes a smaller dosage 91.9% 0.23† 0.01 *p<0.05. †p<0.001. ‡p<0.10. Of the 58 participants who reported how they were affected by malpractice claims, 36 felt anxious, 33 angry, 26 restless, 16 distrustful, 14 guilty, while 14 reported loss of energy or fatigue, 16 had sleeping problems and 11 reported impaired functioning in work, family relations or social activities. Since anxiety and anger were the most reported psychological symptoms, we further calculated Pearson correlation coefficients between levels of anxiety or anger and measures of defensive behaviour and defensive practice. We found positive associations between the level of anxiety and acknowledging defensive practice, advising hospitalisation in suicidal patients, and avoiding drug prescription in suicidal patients. We also found an almost significant trend between levels of anxiety and increased follow-up in suicidal patients and telling the patient about increased suicidal symptoms before starting selective serotonin reuptake inhibitors (table 3). The only significant correlation with anger was advising hospitalisation in suicidal patients (r=0.37 p=0.006). Discussion This is the first study to describe defensive medicine among psychiatrists that reports on the extent of the phenomenon as well as identifies specific actions taken by psychiatrists to avoid malpractice liability. This study demonstrates that defensive medicine is a well-rooted common practice among psychiatrists as the prevalence of defensive medicine was 62.1%. These results are similar to a prior study conducted among psychiatrists in the UK, focusing on four specified actions: admitting patients to the hospital when their condition could be managed as an outpatient, placing patients on a higher level of observation than warranted, writing in patients' records specific remarks such as ‘not suicidal’ and dictating letters more than necessary for managing the patient's illness. Overall, three-quarters of psychiatrists have performed at least one of the four actions within the past month.13 Together with our study, these results are surprising, especially in light of previous reports asserting that psychiatry is a low-risk specialty.9 A recent study portrayed psychiatry as responsible for only 1% of compensations paid during the period of the study (2005–2014), with the lowest risk of recurrence of lawsuits.14 Nevertheless, Jena and colleagues9 reported an annual probability of 2.6% for psychiatrists being sued in the USA. Data also show that the proportion of physicians facing malpractice claims in low-risk specialties is about 36% by age 45, and rises up to 75% at age 65. Another interesting finding was the tendency of younger psychiatrists to report defensive behaviours and practise defensive medicine (table 2), despite malpractice typically characterising older psychiatrists. Martin-Fumadó et al12 reported that, among Spanish doctors, the mean time from specialisation to motivating a claim was 12 years. Similarly, Reich et al10 reported that the risk of disciplinary action increased with the years of practice. This recognition of defensive medical practice among young doctors may be unique to psychiatry, as studies of defensive medicine in other specialties failed to show correlation with age.8 In the USA, a study regarding defensive medicine among high-risk specialist physicians showed that being in practice for over 30 years is a risk factor for practising defensive medicine.2 Regarding psychiatry specifically, Passmore and Leung13 demonstrated that there is a higher propensity of junior trainees to admit patients to the hospital and to place patients on higher levels of observation than is necessary. A US study found that walk-in psychiatric patients are more likely to be admitted if treated by a less experienced psychiatrist (first or second year residents) compared with more experienced staff (third year residents and attending physicians).15 We did not find any gender differences in the acknowledgement of defensive practice; however, in treating suicidal patients, female doctors consulted more often with a senior psychiatrist than male doctors, and were also significantly more prone to refer to another mental health professional. This finding is noteworthy, especially in the light of a recent review that found that male doctors have nearly 2.5 times the odds of medicolegal actions compared with female doctors.16 Studdert et al14 reported that 82% of paid malpractice claims involved male physicians. It is noteworthy that the only gender differences in defensive practices were items related to seeking support from seniors or colleagues. This might reflect differences in gender attitudes in the practice of the profession more than defensive medicine per se, but there is a lack of literature in the field; thus, more research is required. Beyond describing the scope of the phenomena of DP, we aimed to understand how one’s awareness of practising defensive medicine was related to actually applying defensive behaviours (table 3). We considered the psychiatrist's answer to the question ‘Do you practice defensive medicine?’ as evidence of conscious practice of defensive medicine, as previous studies have done.17 However, reporting specific behaviours in different scenarios is not necessarily conscious or unconscious behaviour. We aimed to ascertain conscious versus unconscious behaviour by calculating the correlation between the answer to the previous question and self-reports of defensive practice behaviours in the four clinical scenarios. We assumed that a positive correlation between the self-acknowledgement of defensive practices and any specific behaviours (eg, increasing follow-up with suicidal patients) is good evidence that participants are conscious of practising defensive medicine (the more they admit defensive medicine, the more they increase follow-up). Conversely, no correlation implicates that they do not consider that behaviour as practising defensive medicine, for example, prescribing medication without indication to suicidal patients is apparently an unconscious defensive practice because it was not associated with self-acknowledgement of defensive medicine. As expected, psychiatrists who were directly involved in malpractice claims were more prone to acknowledge defensive practice, as seen in other studies of defensive medicine in psychiatry13 and other medical specialties.3 When treating a suicidal patient, most physicians reported practising defensive behaviours on the questionnaire, except for prescribing medication without indication. Interestingly, only two items, ‘advising hospitalisation even if unwarranted’ and ‘increased frequency of follow-up even if not necessary’, were significantly correlated with acknowledgement of practising defensive medicine and higher levels of anxiety concerning malpractice claims. Psychiatrists cannot always predict or prevent patient suicide, even if they provide the best medical care possible.18 Nonetheless, the consequences of not preventing such an act hold a tremendous risk for malpractice liability.19 This inherently affects the considerations and treatment decisions made by the treating psychiatrist, who mostly uses ‘positive’ defensive medicine and assurance behaviours. Paradoxically, a referral to unnecessary hospitalisation can lead to reports about improper detention, one of the main reasons for lawsuits in this field.10 This again raises the question whether defensive medicine is effective in preventing the physician’s malpractice liability. Moreover, only these two defensive behaviours were recognised as defensive medicine and not, for example, the referral of the suicidal patient to another professional. This finding led us to question to what extent psychiatrists are aware of their defensive practices. When examining the relationship between acknowledgement of defensive practice and treatment of pregnant women, we found that psychiatrists prescribe smaller dosages of medication than is customary. They may not recognise that they are practising defensive medicine when avoiding prescribing proper dosages to these patients. However, psychiatrists who experience more anxiety about malpractice claims avoid giving pregnant women medication altogether, implying that this behaviour is indeed related to defensive medicine. This practice raises special concerns as it is known that untreated depression or anxiety during pregnancy may negatively impact the mother and fetus, and residual depression due to inefficient dosage of psychotropic treatment may result in dual exposures for the fetus to medication and untreated depression.20 Prescribing smaller drug dosages than customary was also evident in the treatment of elderly patients. Certified psychiatrists, more often than residents, attributed this to the propensity of antipsychotic drugs to cause cerebrovascular diseases. This practice correlated with acknowledgement of defensive medicine, but not with higher levels of anxiety. These reported prescribing habits are in accordance with the FDA black box warning and guidelines from 2005 and may not represent malpractice concerns, but rather a more clinical standpoint in treating this population. One alarming finding is that 10.4% of participants in the survey stated that they prescribed medication without indication. Since it has no significant correlation with acknowledgement of defensive practice or anxiety, we concluded that this may be an expression of unconscious defensive behaviour, though there may be other, patient-specific as well as doctor-specific, factors that influence whether or not to prescribe medication. Bradley21 demonstrated in his study that 44.3% of doctors who reported that they were prescribing medication in order to preserve the doctor–patient relationship experienced discomfort. This was described in terms of avoiding litigation or reports, and also as avoiding damage to the doctor–patient relationship, avoiding conflict and ‘keeping the peace’. Finally, the study showed that among participants of the study, malpractice claims raised distressful feelings suggesting that they are the ‘second victims’, meaning that they, the caregivers, may be extremely distressed by the mistakes they have committed.22 Furthermore, the caregiver subjected to legal proceedings may experience the ‘clinical judicial syndrome’ (CJS) that comprises a series of physical, psychological and behavioural symptoms.23 The associations between levels of anxiety or anger after malpractice reports among study participants, and measures of defensive behaviours and defensive practices are in line with Pellino and Pellino’s24 assertion that defensive medicine, the concept of ‘second victim’ and CJS are indeed an intertwined phenomena. Limitations The survey was voluntary; therefore, it is possible that physicians who agreed to participate were more prone to acknowledge practising defensive medicine, thus biasing the results. Furthermore, physicians were recruited among attendants to a psychiatric congress, which may limit the generalisation of the findings. This study described psychiatrists' perceptions of defensive medicine and not objective data that may describe this phenomenon. Self-reports of defensive medicine may be biased, and may lead physicians to overstate the frequency of performing defensive medicine. Objective methods for measuring defensive medicine are extremely difficult to employ.1 3 It is often difficult to identify the difference between liability-related motivation and other factors that influence clinical decision-making. For example, in this study, practices that could be viewed as defensive medicine, such as ‘initiate contact with the family of a suicidal patient’ and ‘inform of severe yet rare side effects of new drug treatment’, characterised a high percentage of doctors, but were not correlated with defensive medicine or anxiety levels. These clinical decisions might reflect the policy of the institution where the physician works, the personal attitude of the physician, a less patriarchal standpoint that sees the patient's right to decide about their treatment and more. There are no available data on the respondent's annual incomes and malpractice premiums. In Israel, a premium is paid by the employer as part of the national work contracts with the Israeli Medical Association. Thus, since this is not ‘out-of-pocket money’, both these factors probably do not influence the daily practice of defensive medicine in our cohort. Conclusions Defensive medicine is well established in routine clinical psychiatry, despite this specialty having a low risk for malpractice lawsuits. Coping with the defensive medicine is a challenging task. Although there are several suggested strategies, evidence is lacking and there is not one efficient solution to resolve this issue.25 Any solution should include changes in the physician’s, as well as the patient’s, perspective and behaviour. This may include giving more information to the public regarding the recommended care and relevant diagnostic and treatment options in clinical situations prone to defensive medicine. Also, developing and applying clinical practice guidelines targeting risky clinical situations may aid doctors when faced with difficult situations. Finally, making reforms in the liability and compensation systems available today may be a way to preserve the beneficial effects of defensive medicine while diminishing its hazardous effects. Further studies are necessary to determine if the putative effect of defensive medicine impedes high quality clinical psychiatric care of patients, to examine the economic burden of DP on healthcare providers and explore different strategies to cope with this troubling phenomenon. Contributors: All authors contributed extensively to the work presented in this paper. IR and IP designed the study, collected data and wrote the manuscript. HR conceived the study and edited the prepared manuscript. OB supervised the data collection and analysis and edited the manuscript. LC performed the statistical analysis, interpreted the data and wrote the manuscript. Funding: This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Supporting data are available on request. Please contact the corresponding author reuveni@hadassah.org.il. ==== Refs References 1 Anderson RE Billions for defense: the pervasive nature of defensive medicine . Arch Intern Med 1999 ;159 :2399 –402 . 10.1001/archinte.159.20.2399 10665887 2 Studdert DM , Mello MM , Sage WM Defensive medicine among high-risk specialist physicians in a volatile malpractice environment . JAMA 2005 ;293 :2609 –17 . 10.1001/jama.293.21.2609 15928282 3 Asher E , Greenberg-Dotan S , Halevy J Defensive medicine in Israel—a nationwide survey . PLoS ONE 2012 ;7 :e42613 10.1371/journal.pone.0042613 22916140 4 Simon RI Defensive psychiatry and the disruption of treatment boundaries . Isr J Psychiatry Relat Sci 2000 ;37 :124 –31 .10994296 5 Kessler DP , Mcclellan M Do doctors practice defensive medicine? Q J Econ 1996 ;111 :353 –90 . 10.2307/2946682 6 Mello MM , Chandra A , Gawande AA National costs of the medical liability system . Health Aff 2010 ;29 :1569 –77 . 10.1377/hlthaff.2009.0807 7 Hermer LD , Brody H Defensive medicine, cost containment, and reform . J Gen Intern Med 2010 ;25 :470 –3 . 10.1007/s11606-010-1259-3 20143176 8 Asher E , Dvir S , Seidman DS Defensive medicine among obstetricians and gynecologists in tertiary hospitals . PLoS ONE 2013 ;8 :e57108 10.1371/journal.pone.0057108 23483896 9 Jena AB , Seabury S , Lakdawalla D Malpractice risk according to physician specialty . N Engl J Med 2011 ;365 :629 –36 . 10.1056/NEJMsa1012370 21848463 10 Reich J , Schatzberg A An empirical data comparison of regulatory agency and malpractice legal problems for psychiatrists . Ann Clin Psychiatry 2014 ;26 :91 –6 .24812648 11 Thompson JW , Boudreaux JM , Artecona J Medical malpractice: Psychiatry . In: Payne-James J , Byard RW , eds. Encyclopedia of Forensic and legal medicine . 2nd edn Cambridge, MA : Academic Press , 2016 :484 –8 . 12 Martin-Fumadó C , Gómez-Durán EL , Rodríguez-Pazos M Medical professional liability in psychiatry . Actas Esp Psiquiatr 2015 ;43 :205 –12 .26631303 13 Passmore K , Leung WC Defensive practice among psychiatrists: a questionnaire survey . Postgrad Med J 2002 ;78 :671 –3 . 10.1136/pmj.78.925.671 12496324 14 Studdert DM , Bismark MM , Mello MM Prevalence and characteristics of physicians prone to malpractice claims . N Engl J Med 2016 ;374 :354 –62 . 10.1056/NEJMsa1506137 26816012 15 Meyerson AT , Moss JZ , Belville R Influence of experience on major clinical decision. Training implications . Arch Gen Psychiatry 1979 ;36 :423 –7 . 10.1001/archpsyc.1979.01780040065007 426609 16 Unwin E , Woolf K , Wadlow C Sex differences in medico-legal action against doctors: a systematic review and meta-analysis . BMC Med 2015 ;13 :172 10.1186/s12916-015-0413-5 26268807 17 Silberstein E , Shir-Az O , Reuveni H Defensive medicine among plastic and aesthetic surgeons in Israel . Aesthet Surg J 2016 ;36 : NP299 304 . 10.1093/asj/sjw094 27277272 18 Sher L Preventing suicide . QJM 2004 ;97 :677 –80 . 10.1093/qjmed/hch106 15367739 19 Bleich A , Baruch Y , Hirschmann S Management of the suicidal patient in the era of defensive medicine: focus on suicide risk assessment and boundaries of responsibility . Isr Med Assoc J 2011 ;13 :653 –6 .22279696 20 Byatt N , Deligiannidis KM , Freeman MP Antidepressant use in pregnancy: a critical review focused on risks and controversies . Acta Psychiatr Scand 2013 ;127 :94 –114 . 10.1111/acps.12042 23240634 21 Bradley CP Factors which influence the decision whether or not to prescribe: the dilemma facing general practitioners . Br J Gen Pract 1992 ;42 :454 –8 .1472390 22 Wu AW Medical error: the second victim. The doctor who makes the mistake needs help too . BMJ 2000 ;320 :726 –7 .10720336 23 Hurtado Hoyo E , Gutiérrez Zaldívar H , Iraola N Síndrome Clínico Judicial . Rev Asoc Med Argent 2006 ;119 :8 –16 . 24 Pellino IM , Pellino G Consequences of defensive medicine, second victims, and clinical-judicial syndrome on surgeons’ medical practice and on health service . Updates Surg 2015 ;67 :331 –7 . 10.1007/s13304-015-0338-8 26650202 25 Sclar D , Housman M Medical malpractice and physician liability: examining alternatives to defensive medicine . Harv Health Pol Rev 2003 ;4 :75 –84 .

PMC005xxxxxx/PMC5372097.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01288610.1136/bmjopen-2016-012886Mental HealthResearch15061712172416931711Spin in RCTs of anxiety medication with a positive primary outcome: a comparison of concerns expressed by the US FDA and in the published literature Beijers Lian 1Jeronimus Bertus F 12Turner Erick H 34de Jonge Peter 12Roest Annelieke M 1 1 University of Groningen, University Medical Center Groningen, Department of Psychiatry, Interdisciplinary Center Psychopathology and Emotion regulation (ICPE), Groningen, The Netherlands. 2 University of Groningen, Department of Developmental Psychology, Groningen, The Netherlands 3 Behavioral Health and Neurosciences Division, VA Portland Health Care System, Portland, Oregon, USA 4 Departments of Psychiatry and Pharmacology, Oregon Health & Science University, Portland, Oregon, USACorrespondence to Dr Bertus F Jeronimus; b.f.jeronimus@rug.nl2017 29 3 2017 7 3 e0128861 6 2016 28 11 2016 6 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives This study aimed to determine the presence of spin in papers on positive randomised clinical trials (RCTs) of antidepressant medication for anxiety disorders by comparing concerns expressed in the Food and Drug Administration (FDA) reviews with those expressed in the published paper. Methods For every positive anxiety medication trial with a matching publication (n=41), two independent reviewers identified the concerns raised in the US FDA reviews and those in the published literature. Spin was identified when concerns or limitations were expressed by the FDA (about the efficacy of the study drug) but not in the corresponding published paper. Concerns mentioned in the papers but not by the FDA were scored as ‘non-FDA’ concerns. Findings Only six out of 35 (17%) of the FDA concerns pertaining to drug efficacy were reported in the papers. Two papers mentioned a concern that fit the FDA categories, but was not mentioned in the corresponding FDA review. Eighty-seven non-FDA concerns were counted, which often reflected general concerns or concerns related to the study design. Conclusions Results indicate the presence of substantial spin in the clinical trial literature on drugs for anxiety disorders. In papers describing RCTs on anxiety medication, the concerns raised by the authors differed from those raised by the FDA. Published papers mentioned a large number of generic concerns about RCTs, such as a lack of long-term research and limited generalisability, while they mentioned few concerns about drug efficacy. These results warrant the promotion of independent statistical review, reporting of patient-level data, more study of spin, and an increased expectation that authors report FDA concerns. RCTsSpinBiasFood Drug AdministrationPSYCHIATRY ==== Body Strengths and limitations of this study Novel study of spin regarding efficacy-related concerns in drug randomised clinical trials with positive primary outcomes. Food and Drug Administration (FDA) reviews are used as an objective source to compare with the published literature. The estimate may be conservative because FDA concerns required agreement between FDA reviewers, which was not always the case. No other types of spin were examined. Introduction It is becoming increasingly apparent that the results of scientific research are not always accurately reported. Results are frequently presented more positively than they actually are.1 2 For example, trials are more likely to be published if they are positive, and authors may selectively report outcome measures in such a way that negative trials appear positive.3–7 Reporting strategies, intentional or unintentional, which mislead readers in their evaluation of the beneficial effects of experimental interventions or their safety, are called ‘spin’.8 9 Examples include the overinterpretation of results10 and exaggerated claims in press releases and media coverage of studies.11 Although awareness of spin is increasing, most studies of spin in the scientific literature have focused on randomised clinical trials (RCTs) whose primary outcomes are negative.7 8 12–16 However, the fact that a primary outcome is positive does not preclude spin.11 Researchers might, for example, emphasise the most favourable-looking variable or manipulate figures.3 5 16 In papers about RCTs with positive primary outcomes, the most prominent types of spin are arguably the exaggeration of drug efficacy and concerns regarding trial outcome reliability. Examples include when the study drug had no effect on secondary end points or when drug efficacy was shown only for men or only for women. Treatment decisions are influenced by the way such concerns are presented. One way to detect spin in publications on positive trials is to compare the concerns raised about a medication trial in the reporting paper with those reported by an objective, independent third party. The US Food and Drug Administration (FDA) can play such a role. Before a drug can be marketed in the USA, it must be approved by the FDA. Before starting any US trial, pharmaceutical companies must notify the FDA of its details, such as the primary end point and statistical analytic method. When the trial is completed, FDA medical and statistical reviewers inspect the sponsor's results, trying to replicate them using the patient-level data submitted by the sponsor. The reviewers then write a summary report on each of the trials, together with a recommendation as to whether the drug meets the FDA's criteria for marketing approval. These reports are examined by higher level reviewers (team leader, division director, sometimes office director), culminating in the decision whether to approve the drug.17 These reports are combined into an FDA drug approval package (hereafter called FDA review), which is made publicly available.18 The existence of multiple reviews increases the odds that any concerns regarding drug efficacy results will be expressed. Spin can be measured by comparing concerns mentioned in the FDA drug approval packages with those mentioned in corresponding trial publications. The objective of this study is to examine spin in the reporting of positive RCTs of antidepressant medication for anxiety disorders, by comparing concerns expressed in the FDA reviews with those expressed in the published paper. Method FDA reviews In a previous paper, Roest et al7 identified a cohort of phase 2 and 3 double-blind placebo-controlled clinical trials that had been registered with the FDA in pursuit of marketing approval; the drugs were second-generation antidepressants for the treatment of five anxiety disorders (generalised anxiety disorder, panic disorder, social anxiety disorder, post-traumatic stress disorder and obsessive-compulsive disorder). Nine drugs were approved for these indications by the FDA: seven selective serotonin reuptake inhibitors (SSRIs): paroxetine, paroxetine controlled release, sertraline, fluoxetine, fluvoxamine, fluvoxamine controlled release and escitalopram, and two serotonin norepinephrine reuptake inhibitors (SNRIs): venlafaxine extended release and duloxetine. FDA reviews were downloaded from the FDA's website (http://www.accessdata.fda.gov/scripts/cder/drugsatfda/index.cfm), or, if these reviews were not available for download, requested from the FDA's Freedom of Information Office (http://www.accessdata.fda.gov/scripts/foi/FOIRequest/requestinfo.cfm). The reader may access these reviews at http://doi.org/10.6083.M4H9949P. In that previous paper, the FDA's regulatory decisions were previously classified as (1) positive (clearly supporting efficacy) or (2) not positive.7 The current study included only positive trials (κ=41), that is, trials with statistically significant results on the primary end point(s). Since the FDA review for fluoxetine in the treatment of panic disorder became available after the publication of Roest et al, this study was able to include one additional positive trial. These 42 trials were covered in 21 FDA reviews (one for each drug indication combination, each including 1–4 trials) (see figure 1). Further details on the number of trials included in each FDA review can be found in supplementary Table 1 of Roest et al.7 Figure 1 Flow chart of trial and paper selection. Published papers Matching journal publications for the 41 trials were identified, as reported in Roest et al,7 by systematically searching PubMed, EMBASE and the Cochrane Central Register of Controlled Trials without language restrictions, with a search cut-off date of 19 December 2012. The title field was searched for the name of the drug and the type of anxiety disorder and any field was searched for the word placebo. The matching paper for the additional (fluoxetine) trial included in this study was found by hand search. One trial was not published. The other 41 trials were described in 38 papers because six trials were published in pooled analyses (see figure 1). Concerns For each trial with a matching publication (κ=41), we scored concerns in that publication and in the FDA review. A concern was defined as an expressed limitation regarding a specific trial. Two independent researchers (AMR and LB) first scored concerns expressed in the FDA reviews. Subsequently, the concerns in the papers were scored independently by researchers LB and BFJ. Discrepancies were resolved by consensus (LB, BFJ, AMR). The inter-rater agreement was expressed in terms of Cohen's κ, which, following Landis and Koch,19 is regarded as moderate from 0.41 to 0.60, substantial from 0.61 to 0.80 and almost perfect above 0.81. Concerns in the FDA reviews Concerns in the FDA reviews were counted when they were expressed by at least two FDA parties (eg, medical reviewer, statistical reviewer, team leader, division director). Only in exceptional cases were concerns counted when they were mentioned only by the medical reviewer: (1) when the statistical review was missing and the concern was also reported in other FDA reviews, or (2) when no detailed evaluation on a specific trial was present for the statistical reviewer and team leader. The concerns found in the FDA reviews were sorted into 11 categories (figure 2). Figure 2 FDA concerns in FDA review packages and related trial papers. *In specific cases, the FDA allowed for multiple primary end points. +Such that the FDA excluded results from analysis. FDA, Food and Drug Administration. Concerns in the papers The FDA concerns were counted as present or absent in the abstract or discussion section of each paper. Additionally, since FDA concerns might not be the only relevant ones, concerns mentioned in the papers but not by the FDA were scored as ‘non-FDA’ concerns and grouped into categories (see figure 3). Concerns in the papers were counted when they were either mentioned in the limitations section or clearly formulated as a concern. This included any formulation that contained signal words like ‘unfortunately’ or ‘cause for concern’, or was framed as a future research question. Figure 3 Non-FDA concerns in trial papers. *Concerns that occurred only once were classified as ‘Other. FDA, Food and Drug Administration. Spin Spin was identified when concerns about drug efficacy expressed by the FDA were absent from the published reports. Results FDA concerns A total of 38 concerns were identified in the 41 trials (89% directly, the rest after debate). Inter-rater reliability was substantial (κ=0.73). Eleven of these concerns pertained to the lack of evidence for a dose–response relationship. The second and third largest categories of concerns pertained to the lack of statistical significance on the primary end point(s) according to the observed cases (OC) and last observation carried forward (LOCF) analyses, respectively. The other categories occurred only one to three times (see figure 2), but these were more serious concerns, including lack of efficacy in one of the sexes, lack of significant findings for all dose groups and a lower efficacy of the study drug compared with previously approved drugs. Paper concerns In our review of the 38 papers, we identified a total of 86 concerns (62%) directly; the rest were debated (inter-rater reliability substantial, κ=0.64). Almost all (99%) concerns were mentioned in the discussion paragraph of the paper. Four of these concerns were also mentioned in the abstract. One concern occurred in the abstract but not in the discussion section of the paper. Spin in the papers In three instances, one FDA concern was present for both trials published in pooled analyses. To keep the comparison between FDA and paper concerns conservative, such a pair was counted as a single FDA concern, leaving 35 unique concerns. Only six of these 35 (17%) FDA concerns were found in the trial papers (covering three categories), as outlined in figure 2. The concern category most frequently present in the papers was the lack of a dose–response relationship (four cases in papers vs 11 in FDA reviews). Two papers mentioned a concern that fit the FDA categories, but was not mentioned in the corresponding FDA review (not shown in figure 2). These concerns pertained to the lack of a dose–response relationship and the lack of significance in OC analyses. Non-FDA concerns in the papers Six of the concerns mentioned in the papers were also scored by the FDA (7%), and two concerns from the papers fell into the FDA categories. This left 78 additional concerns not mentioned by the FDA, yielding an average of two concerns per reviewed paper. As can be seen in figure 3, the most abundant of the concerns not mentioned by the FDA pertained to the lack of long-term research (68% of papers). The second most common concern focused on the lack of generalisability, usually due to study exclusion criteria. Fourteen concerns occurred only once and were scored as ‘other’ concerns. Discussion This paper demonstrates, within a cohort of positive RCTs for anxiety disorders, the existence of spin. The FDA commonly expressed concerns about drug efficacy, but only a fraction of these (17%) were conveyed in the corresponding journal publications. The published papers mentioned 12 times more non-FDA concerns (n=78) than FDA concerns (n=6), but the former consisted largely of generic concerns applicable to virtually any RCT, such as a lack of long-term research and limited trial generalisability. Other non-FDA concerns often reflected design choices, such as lack of an active comparator. Publication of such generic concerns, when coupled with a lack of stated concerns about drug efficacy, could itself be interpreted as a type of spin. Strengths and limitations The results of this study should be interpreted in the light of the following strengths and limitations. Our definition of spin excluded other types of spin, for example, linguistic spin5 and graphical misrepresentation of data,3 so this study arguably presents an overly conservative picture. Nonetheless, we feel that misrepresentation of concerns about drug efficacy, in the context of RCTs with positive results, may be a more important type of spin because of its more direct relevance to clinical practice. Failure to transparently convey efficacy concerns might distort the drug's apparent risk–benefit ratio, leading to undue clinician enthusiasm and potentially inappropriate treatment. The strength of the study lies in our having undertaken a straightforward comparison using an objective third party. However, we cannot be sure that the FDA reviews included mention of all relevant concerns about drug efficacy, hence our search for concerns in published papers. Also, while the FDA reviewers are objective third-party evaluators of drug efficacy, they did not always report the same concerns. For example, high dropout rates, an issue sometimes mentioned by statistical reviewers, was never raised by medical reviewers or team leaders. Our requirement that specific FDA concerns be expressed in two or more reviews might have led to an underestimation of the total number of FDA concerns. Finally, a minority of the publications preceded (rather than followed) the FDA approval date, leaving open the possibility that the FDA had not yet made the sponsor aware of its efficacy concerns, which could explain the sponsor’s not mentioning them in the paper. However, our aim was not to ascertain whether authors conveyed concerns expressed to them by the FDA, but rather whether they had thoughtfully reflected on their trial's limitations and conscientiously reported them. Implications We observed that most FDA concerns about drug efficacy were not mentioned in trial papers, whereas most non-FDA concerns were vague and could apply to virtually any clinical trial. While we agree that the lack of long-term research is a problem, a plethora of non-efficacy-related concerns might serve to obfuscate or distract from issues of greater relevance to clinicians. This raises ethical questions, considering that problems associated with prescription drug use (eg, poisoning) are third on the list of causes of death in Europeans and Americans above age 65, after heart disease and cancer.1 20 21 Anxiety disorders have an estimated year prevalence of 12% and antidepressants are the primary pharmacological treatment for anxiety.21 About 10% of Europeans currently use SSRIs or SNRIs on a daily basis.22–24 Previous studies have shown that bias and spin occur in many scientific fields.3 4 7 8 12–15 17 To further determine the extent to which concerns about drug efficacy are misrepresented in published papers, the method employed here can be applied to other fields. If misrepresentation of concerns proves to be widespread, a number of measures can be taken. One possible approach is to have patient-level data analysed and reviewed by an independent party, as has been carried out for at least one controversial trial.25 26 Another is for journal reviewers and editors to require authors of publications to mention FDA concerns, though this could be carried out only for studies published after FDA approval. Conclusion This study indicates that, in the context of anxiety medication trials with positive primary outcomes, published papers substantially misrepresent concerns about drug efficacy. The majority of concerns raised by the FDA, an independent third party, were not reported in matching publications. The concerns raised in the papers by sponsors, who have a financial interest in the outcome of the trial, were often vague and applicable to clinical drug trials in general. Such concerns may be overrepresented at the expense of issues of greater importance to clinicians, such as (problems with) drug efficacy. Patient-level data reporting, independent review and increased expectation that authors convey FDA concerns could help reduce spin. Before such measures are implemented, however, this phenomenon should be investigated in related fields. Contributors: The study was conceived by AMR, EHT and PdJ. The FDA reports were reviewed by LB and AMR. The papers were reviewed by LB and BFJ. The data were analysed by LB. The initial draft of the manuscript was written by LB and BFJ. EHT, PdJ and AMR edited the manuscript. All authors contributed significantly to the study and approved the final version of the manuscript. Funding: This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: FDA reviews used in this study can be downloaded from the following website http://doi.org/10.6083.M4H9949P. ==== Refs References 1 Gøtzsche PC , Smith R , Rennie D Deadly medicines and organised crime: how big pharma has corrupted healthcare . London : Radcliffe , 2013 . 2 Ioannidis JPA Why most published research findings are false . PLoS Med 2005 ;2 :e124 10.1371/journal.pmed.0020124 16060722 3 Al-Marzouki S , Roberts I , Marshall T The effect of scientific misconduct on the results of clinical trials: a Delphi survey . Contemp Clin Trials 2005 ;26 :331 –7 . 10.1016/j.cct.2005.01.011 15911467 4 Boutron I , Altman DG , Hopewell S Impact of spin in the abstracts of articles reporting results of randomized controlled trials in the field of cancer: the SPIIN randomized controlled trial . J Clin Oncol 2014 ;32 :4120 –6 . 10.1200/JCO.2014.56.7503 25403215 5 Marco CA , Larkin GL Research ethics: ethical issues of data reporting and the quest for authenticity . Acad Emerg Med 2000 ;7 :691 –4 . 10.1111/j.1553-2712.2000.tb02049.x 10905651 6 Turner EH , Rosenthal R Efficacy of antidepressants . BMJ 2008 ;336 :516 –17 . 10.1136/bmj.39510.531597.80 18319297 7 Roest AM , de Jonge P , Williams CD Reporting bias in clinical trials investigating the efficacy of second-generation antidepressants in the treatment of anxiety disorders: a report of 2 meta-analyses . JAMA Psychiatry 2015 ;72 :500 –10 . 10.1001/jamapsychiatry.2015.15 25806940 8 Boutron I , Dutton S , Ravaud P Reporting and interpretation of randomized controlled trials with statistically nonsignificant results for primary outcomes . JAMA 2010 ;303 :2058 –64 . 10.1001/jama.2010.651 20501928 9 Lazarus C , Haneef R , Ravaud P Classification and prevalence of spin in abstracts of non-randomized studies evaluating an intervention . BMC Med Res Methodol 2015 ;15 :85 10.1186/s12874-015-0079-x 26462565 10 Ochodo EA , de Haan MC , Reitsma JB Overinterpretation and misreporting of diagnostic accuracy studies: evidence of ‘spin ’. Radiology 2013 ;267 :581 –8 . 10.1148/radiol.12120527 23360738 11 Yavchitz A , Boutron I , Bafeta A Misrepresentation of randomized controlled trials in press releases and news coverage: a cohort study . PLoS Med 2012 ;9 :e1001308 .22984354 12 Gewandter JS , McKeown A , McDermott MP Data interpretation in analgesic clinical trials with statistically nonsignificant primary analyses: an ACTTION systematic review . J Pain 2015 ;16 :3 –10 . 10.1016/j.jpain.2014.10.003 25451621 13 Linden A Identifying spin in health management evaluations . J Eval Clin Pract 2011 ;17 :1223 –30 . 10.1111/j.1365-2753.2010.01611.x 21210897 14 Lockyer S , Hodgson R , Dumville JC ‘Spin’ in wound care research: the reporting and interpretation of randomized controlled trials with statistically non-significant primary outcome results or unspecified primary outcomes . Trials 2013 ;14 :371 10.1186/1745-6215-14-371 24195770 15 Patel S V , Chadi SA , Choi J The use of ‘spin’ in laparoscopic lower GI surgical trials with nonsignificant results: an assessment of reporting and interpretation of the primary outcomes . Dis Colon Rectum 2013 ;56 :1388 –94 . 10.1097/01.dcr.0000436466.50341.c5 24201393 16 Fletcher RH , Black B ‘Spin’ in scientific writing: scientific mischief and legal jeopardy . Med Law 2007 ;26 :511 –25 .17970249 17 Turner EH , Matthews AM , Linardatos E Selective publication of antidepressant trials and its influence on apparent efficacy . N Engl J Med 2008 ;358 :252 –60 .18199864 18 Turner EH How to access and process FDA drug approval packages for use in research . BMJ 2013 ;347 :f5992 10.1136/bmj.f5992 24126858 19 Landis JR , Koch GG The measurement of observer agreement for categorical data . Biometrics 1977 ;33 :159 –74 . 10.2307/2529310 843571 20 Hall AJ , Logan JE , Toblin RL Patterns of abuse among unintentional pharmaceutical overdose fatalities . JAMA 2008 ;300 :2613 –20 . 10.1001/jama.2008.802 19066381 21 Kessler RC , Aguilar-Gaxiola S , Alonso J The global burden of mental disorders: an update from the WHO World Mental Health (WMH) surveys . Epidemiol Psichiatr Soc 2009 ;18 :23 –33 . 10.1017/S1121189X00001421 19378696 22 Mojtabai R , Olfson M National patterns in antidepressant treatment by psychiatrists and general medical providers: results from the national comorbidity survey replication . J Clin Psychiatry 2008 ;69 :1064 –74 . 10.4088/JCP.v69n0704 18399725 23 Organisation for Economic Co-operation and Development (OECD), Health Statistics 2016, pharmaceutical consumption. http://www.oecd.org (accessed 7 Oct 2016 ). 24 Bauer M , Monz BU , Montejo AL Presribing patterns of antidepressants in Europe: results from the Factors Influencing Depression Endpoints Research (FINDER) study . Eur Psychiatry 2008 ;23 :66 –73 . 10.1016/j.eurpsy.2007.11.001 18164600 25 Keller MB , Ryan ND , Strober M Efficacy of paroxetine in the treatment of adolescent major depression: a randomized, controlled trial . J Am Acad Child Adolesc Psychiatry 2001 ;40 :762 –72 . 10.1097/00004583-200107000-00010 11437014 26 Le Noury J , Nardo JM , Healy D Restoring Study 329: efficacy and harms of paroxetine and imipramine in treatment of major depression in adolescence . BMJ 2015 ;351 :h4320 10.1136/bmj.h4320 26376805

PMC005xxxxxx/PMC5372098.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01401210.1136/bmjopen-2016-014012Respiratory MedicineProtocol1506173117041701Evaluation of an alternative care provider clinic for severe sleep-disordered breathing: a study protocol for a randomised controlled trial Ip-Buting Ada 1Kelly Jenny 1Santana Maria J 12Penz Erika D 3Flemons W Ward 145Tsai Willis H 245Fraser Kristin L 45Hanly Patrick J 45Pendharkar Sachin R 1245 1 W21C Research and Innovation Centre, Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada 2 Department of Community Health Sciences, Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada 3 College of Medicine, University of Saskatchewan, Saskatoon, Saskatchewan, Canada 4 Department of Medicine, Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada 5 Foothills Medical Centre Sleep Centre, University of Calgary, Calgary, Alberta, CanadaCorrespondence to Dr Sachin R Pendharkar; Sachin.pendharkar@ucalgary.ca2017 29 3 2017 7 3 e01401224 8 2016 21 2 2017 6 3 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction Despite the high prevalence of sleep-disordered breathing (SDB) and the significant health consequences associated with untreated disease, access to diagnosis and treatment remains a challenge. Even patients with severe SDB (severe obstructive sleep apnoea or hypoventilation), who are at particularly high risk of adverse health effects, are subject to long delays. Previous research has demonstrated that, within a sleep clinic, management by alternative care providers (ACPs) is effective for patients with milder forms of SDB. The purpose of this study is to compare an ACP-led clinic (ACP Clinic) for patients with severe SDB to physician-led care, from the perspective of clinical outcomes, health system efficiency and cost. Methods and analysis The study is a randomised, controlled, non-inferiority study in which patients who are referred with severe SDB are randomised to management by a sleep physician or by an ACP. ACPs will be supervised by sleep physicians for safety. The primary outcome is positive airway pressure (PAP) adherence after 3 months of therapy. Secondary outcomes include: long-term PAP adherence; clinical response to therapy; health-related quality of life; patient satisfaction; healthcare usage; wait times from referral to treatment initiation and cost-effectiveness. The economic analysis will be performed using the perspective of a publicly funded healthcare system. Ethics and dissemination Ethics approval was obtained from the Conjoint Health Research Ethics Board (ID: REB13-1280) at the University of Calgary. Results from this study will be disseminated through presentations at scientific conferences and publication in peer-reviewed journals. Trial registration number NCT02191085; Pre-results. SLEEP MEDICINE ==== Body Strengths and limitations of this study This is a randomised controlled non-inferiority trial comparing alternative care provider care to traditional physician-led care for patients with severe sleep-disordered breathing. The primary outcome of treatment adherence is objective and important for this patient population, who are at greater risk of medical complications. The comprehensive evaluation strategy includes clinical and health system outcomes, including its effect on timely access to care and economic impacts. This is a single centre study that uses non-physician healthcare providers who may not be available in other sleep clinics or jurisdictions. Introduction Background and rationale Sleep-disordered breathing (SDB) is common and has significant medical consequences. The most common type of SDB, obstructive sleep apnoea (OSA) affects up to 24% of men and 9% of women.1 Untreated severe OSA has been associated with an increased risk of cardiovascular disease, including hypertension, stroke, and fatal and non-fatal cardiovascular events.2–5 Additionally, patients with OSA are at increased risk for motor vehicle crashes,6 use more healthcare resources,7 and may experience reduced survival compared with those without OSA.8 The economic impact of OSA in the USA has been estimated at US$3.5 billion/year, with the mean annual cost of healthcare usage and treatment of medical consequences of OSA exceeding age-matched and sex-matched controls twofold to threefold.9 Treatment of OSA with continuous positive airway pressure (CPAP) reduces cardiovascular risk, motor vehicle crashes and healthcare usage, and is cost-effective.2 9–12 Hypoventilation, defined as a daytime partial pressure of arterial carbon dioxide ≥45 mm Hg or an increase in nocturnal transcutaneous carbon dioxide ≥10 mm Hg occurs in up to 20% of patients with OSA,13 and can also occur in obese patients and in association with other common respiratory diseases such as chronic obstructive pulmonary disease. Patients with hypoventilation use more healthcare resources than the general population14 and are at an increased risk of acute respiratory failure, hospital admission and death.15 Treatment of hypoventilation with positive airway pressure (PAP) therapy, with or without supplemental oxygen, reduces the number of days in hospital and may reduce physician visits.14 16 17 However, ideal treatment for those with hypoventilation requires overnight monitoring of PAP initiation to determine if CPAP alone is safe and effective, or if additional therapy (non-invasive ventilation (NIV), supplemental O2) is required. In contrast, those patients without hypoventilation can be safely treated with home auto-CPAP titration. The difficulty in providing timely access to sleep specialists is widespread. Delayed access for sleep care have been reported in Canada,18 the USA, Europe and Australia.19 These delays are particularly important for patients with severe SDB (severe OSA and/or hypoventilation) due to the increased risk of adverse clinical outcomes. Current strategies to improve timely access include the use of home sleep apnoea testing (HSAT) or telemedicine consultation with a sleep physician;20–23 these strategies may be particularly beneficial for patients residing in rural areas or those with mobility concerns. However, their success relies on the supply of sleep physicians for consultation, which is inadequate in many jurisdictions.24 The use of trained non-physician alternative care providers (ACPs) or primary care physicians to manage patients with SDB has been proposed to improve access to care and to reduce wait times. ACPs, such as nurses and respiratory therapists, have been shown to be effective substitutes for sleep physicians in the management of uncomplicated patients.25 26 Management by primary care physicians has also been demonstrated to be non-inferior to sleep specialist care.27 28 The above studies have identified potential roles for ACPs and primary care physicians in the management of uncomplicated OSA, but no study has examined whether anyone but a sleep physician can manage patients with more severe forms of SDB such as hypoventilation. It is likely, given the complexity of managing patients who may need more advanced PAP therapies and possibly supplemental oxygen, that these patients are best cared for within a sleep clinic and not in the primary care setting. It is unknown whether non-physician providers within a sleep clinic can safely and effectively diagnose and treat complex patients with severe SDB. Furthermore, the impact on wait times or the quality of patient care have not been evaluated. Given the high risk of adverse health outcomes related to untreated severe SDB, and prompted by wait times that far exceed the current Canadian guideline of 4 weeks from referral to assessment, we designed an ACP Clinic for patients with severe SDB. This paper provides a description of the model and the protocol for its implementation and evaluation. The rationale for publishing this protocol is to highlight the importance of improving access to care for this patient population, to consider an expanded scope of practice and service delivery for ACPs within a sleep clinic, and to describe our comprehensive evaluation of clinical and health system outcomes. Objectives The study is a randomised, controlled, non-inferiority study evaluating the 3-month and 1-year outcomes of an ACP Clinic for patients referred with suspected severe SDB. Patients will be randomised to one of two treatment arms—standard management by sleep physicians or management by ACPs in the ACP Clinic. The study hypotheses are that, compared with a traditional physician-led approach, the ACP Clinic will: result in similar PAP treatment adherence, subjective and objective response to therapy, health-related quality of life (HRQL) and patient satisfaction 3 months and 1 year after treatment initiation; reduce the time from referral to treatment initiation; result in similar number of sleep physician visits or diagnostic tests during the first year of treatment initiation; be cost-effective during the first year of treatment initiation. Method and analysis Study setting The Foothills Medical Centre (FMC) Sleep Centre is a publicly funded, tertiary academic sleep centre in Calgary, Alberta, with a catchment area of ∼2 million people in Alberta and British Columbia. There are eight sleep specialist physicians at the FMC Sleep Centre (six respirologists, one psychiatrist and one neurologist). The FMC Sleep Centre receives ∼2500 referrals annually, of which 60% are for SDB and 30% are for severe SDB. Twenty-five polysomnograms (PSG) and 75 HSAT are performed each week and all tests are interpreted by a sleep physician. All newly referred patients with suspected SDB undergo HSAT prior to the initial assessment by a sleep physician. Consistent with current Canadian guidelines,29 patients are prioritised based on severity of SDB on HSAT, medical comorbidity, daytime sleepiness and whether they work in a safety-critical occupation. Patients are assigned to sleep physicians based on the suspected diagnosis. Any physician at the FMC Sleep Centre may assess patients with uncomplicated OSA, but prior to this study, patients with suspected severe SDB were only scheduled to see respirologists. Patients may undergo PSG at the physician's discretion if the diagnosis is uncertain or if there is a concern about ambulatory PAP auto-titration. Consistent with current clinical guidelines,30 31 patients with mild or moderate SDB may be offered PAP therapy or referral to a dentist for oral appliance therapy, depending on symptoms and patient preference. However, PAP is recommended as first-line therapy for all patients with severe SDB. Patients are also counselled on lifestyle modification (eg, weight loss, avoidance of excessive alcohol and sedative use) as appropriate. Conventional practice at the FMC Sleep Centre is that ACPs conduct follow-up assessments, either in the clinic or by telephone, guided by sleep physician-approved protocols. For patients with uncomplicated OSA, we have previously demonstrated that this is an effective follow-up model.26 The primary sleep physician could be reconsulted by the ACP as necessary; physician reassessment could occur by review of the case with the ACP or direct follow-up with the patient. Alternative care providers All ACPs are respiratory therapists (RT) who have completed a 2-year accredited respiratory therapy training programme in Canada. Completion of this programme includes attendance of 300 hours of classroom and laboratory-based learning on respiratory physiology and over 800 hours of education and supervised clinical experience with invasive and non-invasive mechanical ventilation in a variety of medical contexts.32 In addition, ACPs at the FMC Sleep Centre have at least 5 years of experience assessing and managing patients with SDB. All ACPs are registered with the provincial respiratory therapy professional college in Alberta, Canada, which regulates RTs based on established standards of practice and continuing education requirements. The scope of ACP practice at the FMC Sleep Centre is defined by a physician-approved policy that complies with provincial regulations for Registered RTs. ACP activities include: initial education about PAP therapy; follow-up assessment of patients on PAP; and ordering of arterial blood gas and HSAT. ACPs can also make adjustments to PAP equipment, including humidity, ramp and expiratory pressure release settings and small pressure changes to CPAP without a physician prescription. Additionally, regular review of ACP protocols for the management of patients on PAP and education on respiratory and non-respiratory sleep disorders occur bi-weekly at ACP-focused case conferences. Eligibility criteria Patients who are referred to the FMC Sleep Centre will be eligible to participate in the study if they meet at least one of the following inclusion criteria: Respiratory Disturbance Index (RDI) ≥30 events/hour on HSAT; mean nocturnal oxygen saturation ≤85% on HSAT; suspected sleep hypoventilation syndrome, defined by an RDI ≥15 events/hour on HSAT and partial pressure of carbon dioxide ≥45 mm Hg on arterial blood gas while awake. Patients already on supplemental oxygen, in whom HSAT is insensitive for the diagnosis of OSA, will be recruited if the airflow channel on the HSAT indicates severe SDB or if the investigators determine that the clinical suspicion of severe SDB is high. Patients will be excluded from the study if they have a suspected concomitant sleep disorder other than SDB such as insomnia or narcolepsy, have previously been treated with PAP therapy for SDB, have primary health insurance from outside Alberta due to difficulties in collecting healthcare usage data, or if they fail to provide consent to participate in the study. Randomisation and blinding Participants will be randomly assigned to either the standard management by a sleep physician or to the ACP Clinic with a 1:1 allocation using a computer generated randomisation schedule. The allocation sequence will be concealed until the patient is eligible and consented to the study. The research associates will allocate patients according to the randomisation schedule and the booking clerk will schedule the patient with the appropriate provider according to the FMC Sleep Centre scheduling policies. The allocation of the participants will be stored in a database, which is only accessible to the research associate. If the patient prefers to be assessed by a particular provider at the time of initial scheduling, then the allocation would be discontinued and the patient would be excluded from the study. The research associate and booking clerks will not be blinded to the study. The ACPs and sleep physicians, including investigators, will not be able to identify study patients as both groups will be conducting clinics comprised of a combination of study and non-study patients and there will be no other indications that the patient is a study participant. Owing to the nature of the study, the participants will not be blinded to the study. Intervention Figure 1 presents an overview of study flow. Patients randomised to standard management will receive usual care as described above (see the ‘Study setting’). Patients will undergo an initial assessment by a sleep physician, who will establish a management plan with the patient that may include PSG, initiation of therapy for SDB and/or clinical follow-up. Follow-up may be delegated to an ACP, who will manage patients within their scope of practice as defined in existing physician-approved protocols. As is routine practice at the FMC Sleep Centre, ACPs will be able to refer patients back to the primary sleep physician for persistent symptoms of SDB or management of clinical issues outside of their scope of practice. Figure 1 Patient flow diagram. Patients randomised to care in the ACP Clinic will have an initial assessment by an ACP, during which time the management plan will be established by the ACP and the patient. Since this patient population is medically complex, the management plan will be reviewed with a sleep physician immediately after the assessment. The sleep physician will meet briefly with the patient to review any additional medical or sleep-related concerns before confirming the proposed plan for further testing and/or treatment. The presence of a sleep physician in the clinic was also deemed important in case a patient was unstable at the time of assessment (eg, severe hypoxemia, decompensated respiratory failure). Follow-up will occur with the ACP to review test results, discuss and initiate treatment and to assess treatment response. The ACP will be able to refer patients back to the primary sleep physician as in the usual care group. In both groups, as is standard practice at the FMC Sleep Centre, HSAT requisitions will be completed by physicians or ACPs and interpreted by a sleep physician. PSG requisitions and interpretation, and prescriptions for PAP therapy will be completed by the primary sleep physician. Outcomes A summary of study outcomes and when they will be measured is presented in table 1. Table 1 Outcome measures collection points Baseline 3 months 1 year Adherence to therapy PAP adherence ✓ ✓ Daytime sleepiness Epworth Sleepiness Scale ✓ ✓ ✓ Health-related quality of life/utility score Health-Utilities Index Sleep Apnoea Quality of Life Index ✓ ✓ ✓ Patient satisfaction Visit-Specific Satisfaction Instrument 9 ✓ ✓ Demand for sleep provider visits and diagnostic testing Number of ACP visits Number of sleep physician visits Number of HSATs and PSGs Physician time spent per patient during ACP clinic (time-in-motion study)  −  − ✓−* Healthcare usage Number of outpatient physician visits Number of hospitalisations Number of emergency department/urgent care visits ✓ Healthcare costs Healthcare usage costs HSAT and PSG Treatment costs ✓ Wait times Time from referral to initiation of therapy − − −* *Outcomes will be collected throughout the study. ACP, alternative care provider; HSAT, home sleep apnoea testing; PAP, positive airway pressure; PSGs, polysomnograms. Primary outcome The primary outcome is PAP adherence after 3 months of therapy. Treatment adherence will be reported in terms of average nightly use and dichotomised based on whether patients used PAP therapy for at least 4 hours a night for at least 70% of nights.33 Adherence downloads from the preceding 4 weeks will be obtained from each patient's PAP machine. PAP adherence was chosen as a primary outcome for several reasons. First, many outcome measures such as daytime sleepiness, quality of life or functional status are related to PAP use.33–35 In addition, many of the outcome measures used in other studies are subjective; patients often underestimate their symptoms leading to relative insensitivity of symptom scores as measures of treatment effectiveness.36–38 Second, treatment of severe SDB with PAP is associated with reductions in the risk of cardiovascular disease, development of metabolic disorders such as diabetes, healthcare usage and mortality.2 10 12 15 39 40 These are important clinical outcomes for individual patients34 and for health system usage and cost. Given that the population under study is at particularly high risk for these adverse medical consequences, we determined that treatment adherence was of a higher priority than other outcomes for a comprehensive evaluation of this novel model of care for severe SDB. Finally, PAP adherence is commonly used as an outcome measure in studies examining service delivery models for patients with SDB,20 28 41–45 and has been identified as an indicator of high-quality care for SDB.46 Secondary outcomes A number of secondary outcomes, related to clinical effectiveness, healthcare usage, system efficiency and cost, will be analysed. Additional details on the secondary outcomes are available in the online supplementary material. The secondary outcomes include: 10.1136/bmjopen-2016-014012.supp1supplemental data Long-term PAP adherence PAP adherence will be measured after 1 year of therapy. Daytime sleepiness The Epworth Sleepiness Scale is a validated patient questionnaire assessing daytime sleepiness. Health-related quality of life (HRQL) HRQL will be measured using general and disease-specific instruments. The Health-Utilities Index (HUI) and the short-form Sleep Apnoea Quality of Life Index (SAQLI) will be used to measure general HRQL and disease-specific HRQL, respectively. Patient satisfaction The Visit-Specific Satisfaction Instrument (VSQ-9) is a validated measure of patient satisfaction with an outpatient visit.47 Demand for sleep provider visits and diagnostic sleep testing Data on demand will include patient visits to physicians and ACPs, and all HSAT and PSG testing. We will also capture written communication between physicians and ACPs regarding the management of study patients. Healthcare usage Data on healthcare use will include outpatient physician visits, hospitalisations and emergency department or urgent care visits from referral to 1 year following treatment initiation. Cost effectiveness Costs captured from referral to 1 year after treatment initiation will be summarised in the following categories: sleep physician and ACP visits; sleep investigations; treatment costs and healthcare usage costs. Cost-effectiveness will be measured from the perspective of a publicly funded healthcare system. Wait times Time to treatment initiation will be evaluated for patients in each arm. Although wait times for initial assessment may be shorter for the ACP Clinic based on the addition of ACP supply, time to treatment initiation incorporates delays related to PSG and additional physician or ACP visits before a treatment decision is made. Time to treatment initiation is also a clinically relevant outcome, particularly for patients with severe SDB. Statistical analysis This study will use a modified intention to treat analysis, in which results will be analysed for patients who are randomised and have treatment adherence data 3 months after initiating PAP therapy. Paired t-tests will be used to compare clinical outcomes from baseline to 3 months and 1 year and unpaired t-tests will be used to compare time to treatment initiation and measures of patient demand. Multiple logistic regression will be used to identify predictors of study outcomes using variables identified as predictive on univariate regression. Outcomes will be transformed into binary variables based on clinically relevant cut-offs. The economic analysis will be performed using the perspective of a publicly funded healthcare system. An estimate of mean utility scores based on the HUI questionnaire will be calculated and the average quality-adjusted life-year (QALY) will be calculated for each study arm. Using bootstrapping, we will compare the difference in mean costs and QALYs between the study arms to generate an incremental cost-effectiveness ratio for an ACP-led clinic compared with traditional physician-led care. Sensitivity analyses will be performed to evaluate the effect of changes in physician fees, RT salaries, diagnostic testing costs and treatment costs. All study outcomes will be analysed in different subgroups to clarify the impact of ACP-led care for patients with different patient flow pathways or clinical treatments. Prespecified subgroups include: patients who undergo PSG versus patients who do not undergo PSG; patients treated with bi-level PAP versus patients treated with CPAP; patients who are treated with oxygen versus patients treated without oxygen. Sample size The study is powered to assess the non-inferiority of ACP-led care compared with usual care by sleep physicians. A non-inferiority margin of −1 hour of PAP adherence was determined by consensus of the investigators and has been used in previous studies comparing different models of care for SDB.28 The study will require 138 patients (69 in each arm) to achieve 90% power with a type I error of 0.05, using this non-inferiority margin and a SD of 2 hours of nightly CPAP use.48 The SD used in the sample size calculation was based on the results of two previous studies.33 34 To account for withdrawals and loss to follow-up (∼15% at the FMC Sleep Centre), recruitment will continue until 3-month adherence data is available for 150 patients. Trial status Patient recruitment began in October 2014 and was completed in August 2016. We are in the process of collecting baseline measurements, 3-month and 1-year follow-up measurements and anticipate the completion of 3-month and 1-year data collection by December 2016 and December 2017, respectively. Ethics and dissemination The ethics approval process involved reviewing the study with respect to content and compliance with applicable research and safety regulations. In addition to the initial approval of the study, the CHREB will review the study on an annual basis. Any modification to the study protocol will require a formal amendment to the protocol and submitted to the CHREB for approval. In addition, the study is registered under Clinicaltrials.gov (ID: NCT02191085). Changes or updates to the study, including the study protocol, must be made at least every 12 months. Prior to the completion of the study, the record must be reviewed every 6 months. The principal investigator and the attending physician are responsible for assessing, reporting and managing solicited and spontaneously reported adverse events and other unintended effects of trial interventions or trial conduct. Data quality and management To promote data quality, the research associates will randomly select 10% of eligible patients and review the data entered and collected. Range check for data values will be performed for the entire data set. In accordance with research ethics board approval, all data will be stored on a secure network drive within University of Calgary firewalls and will be accessed by the principal investigator and research associate using password-protected study computers through a Virtual Private Network. Identifying information will be replaced with a unique identifier in any data that is reviewed by investigators, with the associated identifying data stored in a separate password-protected file and accessible only to the research associate and principal investigator for the purposes of reconciling data errors. The clinical members of the research team may have access to clinical information on study patients as part of clinical care, but will not have access to identifiable patient records within the research database. The final data set, without any identifiable information, will be accessible to the principal investigator. The research team may access the data set on request. Consent and withdrawal Patients will be recruited by a research associate who is not involved in the clinical management of SDB. Written informed consent will be obtained from patients prior to any participation in the study. The research associates will contact patients prior to the 3-month and 1-year follow-up appointments to promote participant retention and complete follow-up. Participants can withdraw from the study at any time without any consequences. At time of withdrawal, the participant will be asked if additional healthcare usage data can be collected. If the participant chooses not to participate in further data collection, data contributed up to the point of withdrawal will be retained but no further data will be collected. Dissemination Results will be disseminated through publications in peer-reviewed journals; one manuscript will report the 3-month clinical data, and a second manuscript will include longer term clinical and health system outcomes after 1 year. Publications from this research will add to the emerging literature on novel models of care for SDB, and in particular will address whether ACPs can be used to manage a more severe subset of patients. This study lays the foundation for additional research to explore the optimal way to deliver care in the context of such a highly prevalent disease. The results of this study will also be shared with operational leaders at the FMC Sleep Centre, to determine whether the ACP Clinic model is effective for patients. We will also disseminate the protocol and the study findings to other centres within and outside of Canada through the Canadian Sleep and Circadian Network, and provide insights on our experience to those groups looking to implement such a programme. Discussion In many jurisdictions, providing timely access to care for patients with SDB is a challenge due to a shortage of sleep physicians. Building on the results of previous studies of alternative models of care delivery for OSA, this study aims to determine whether ACPs can manage patients with severe SDB. If the study aims are met, the role of ACPs in the management of severe SDB will be more clearly understood in terms of their impact on clinical, economic and health system outcomes. The results of this study will help sleep clinicians and health system administrators to determine the optimal scope of practice of ACPs. As is the case in many healthcare systems with limited access, sleep clinics may adopt strategies that prioritise patients with severe disease above those with milder disease. This strategy allows patients who are at higher risk of medical complications to be assessed sooner and is particularly important when wait times are long. It could be argued that ACPs are best used to manage milder patients as has been demonstrated in previous studies, thus creating capacity for sleep physicians to assess severe patients. However, given the burden of SDB in the population, a significant proportion of whom have severe disease, and insufficient supply of sleep physicians in many jurisdictions, it is probable that delays for severe patients are also long. Furthermore, if ACP capacity could be increased, it is conceivable that wait times for less severe patients might paradoxically be shorter than for more severe patients. Thus, a strategy that aims to directly improve access for higher risk patients, such as using ACPs to manage patients with severe SDB, is preferable to one in which ACPs only assess patients who are of lower priority. Many healthcare systems have insufficient resources to address this imbalance of supply and demand. This study proposes to mitigate this problem through the use of non-physician healthcare providers, under the hypothesis that ACPs can improve access and are cost-effective. However, when proposing a novel model of care, it is essential to demonstrate clinical effectiveness in addition to evaluating the impacts on the healthcare system. This study will comprehensively evaluate an ACP-led pathway for the management severe SDB, including clinical outcomes as well as demand for healthcare resources and costs. Limitations The proposed study has several limitations. First, it is possible that patient preference for care by a sleep physician may influence study outcomes. Our clinical experience with ACP care at the FMC Sleep Centre has been that patients do not object to this management pathway. Furthermore, a previous study of nurse-led care for OSA demonstrated that patient satisfaction did not suffer with non-physician management and in fact was higher with respect to certain aspects of the healthcare encounter.25 The research associate will reassure patients that ACP care will be supervised by a sleep physician and that referral back to a sleep physician can be initiated at any point at the patient's request. Finally, we have selected objective outcomes such as PAP adherence, wait times and healthcare costs to minimise potential confounding by patient perceptions of the care they receive. A second potential limitation relates to the use of multiple instruments to collect patient reported outcomes. It is possible that patients will experience significant burden from these questionnaires, leading to incomplete or inaccurate data. However, previous work has demonstrated that ‘questionnaire fatigue’ is not significant in an outpatient setting.49 50 Additionally, we have performed a small time trial for our battery of instruments with patient engagement researchers and observed that the entire set of questionnaires takes ∼15 min to complete. Consequently, we do not anticipate that the completion of questionnaires will lead to any delays on the day of the visit. Our expectation of no significant adverse impact on clinic flow is consistent with previous literature.50 Third, this study uses RTs as alternative care providers. RTs are highly trained in respiratory care and ventilation but are not a recognised health profession in some parts of the world. Thus, the study results might not be generalisable to jurisdictions in which the RT role does not exist. We recognise that an ACP in any clinical setting should have expertise in managing the patient population; in this regard, only RTs with additional expertise in the management of SDB are employed at the FMC Sleep Centre. Similarly, nurses and other healthcare professions should be comfortable with respiratory care and ventilation, and may require specific additional training to achieve this. Finally, patients who are randomised to the ACP Clinic arm could experience an adverse outcome related to ACP care, such as the need for an emergency department visit or hospitalisation due to unrecognised illness at the time of assessment. This important patient safety risk will be mitigated by the assignment of each ACP Clinic patient to a primary sleep physician, who will review the management plan with the ACP and meet each patient at the time of the clinic visit. Furthermore, our experience with the delegation of follow-up care to ACPs at the FMC Sleep Centre suggests that trained RTs appropriately identify complex patients, and obtain guidance either through direct communication or by scheduling a follow-up clinic visit with the sleep physician. While we have mandated physician supervision during the ACP Clinic, we recognise that in other jurisdictions, advanced practice nurses or nurse practitioners may have the expertise to manage these patients independently. Contributors: MJS, EDP, WWF, WHT, KLF, PJH contributed to study conception/design. EDP, WHT, SRP provided statistical expertise, sample size calculation and will lead the analysis. AI-B and JK contributed to data acquisition and drafted this protocol manuscript. All authors revised the paper and approved the final version. Funding: This work was supported by The Lung Association—Alberta & NWT and by the Canadian Sleep and Circadian Network (funded by the Canadian Institutes of Health Research, grant number 339739-CDP). Disclaimer The funding source had no role or input in the design of this study and will not have any role or input during its execution, analyses and interpretation of the data, or in the dissemination of the findings. Competing interests: None declared. Ethics approval: Ethics approval was obtained for this study from the Conjoint Health Research Ethics Board (CHREB) at the University of Calgary (Ethics ID: REB13-1280). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Study data is not yet available. ==== Refs References 1 Young T , Palta M , Dempsey J The occurrence of sleep-disordered breathing among middle-aged adults . N Engl J Med 1993 ;328 :1230 –5 . 10.1056/NEJM199304293281704 8464434 2 Marin JM , Carrizo SJ , Vicente E Long-term cardiovascular outcomes in men with obstructive sleep apnoea-hypopnoea with or without treatment with continuous positive airway pressure: an observational study . Lancet 2005 ;365 :1046 –53 . 10.1016/S0140-6736(05)74229-X 15781100 3 Peppard PE , Young T , Palta M Prospective study of the association between sleep-disordered breathing and hypertension . N Engl J Med 2000 ;342 :1378 –84 . 10.1056/NEJM200005113421901 10805822 4 Redline S , Yenokyan G , Gottlieb DJ Obstructive sleep apnea–hypopnea and incident stroke: the sleep heart health study . Am J Respir Crit Care Med 2010 ;182 :269 –77 . 10.1164/rccm.200911-1746OC 20339144 5 Yaggi HK , Concato J , Kernan WN Obstructive sleep apnea as a risk factor for stroke and death . N Engl J Med 2005 ;353 :2034 –41 . 10.1056/NEJMoa043104 16282178 6 George C , Nickerson P , Hanly P Sleep apnoea patients have more automobile accidents . Lancet 1987 ;330 :447 10.1016/S0140-6736(87)90974-3 7 Ronald J , Delaive K , Roos L Health care utilisation in the 10 years prior to diagnosis in obstructive sleep apnea syndrome patients . Sleep 1999 ;22 :225 –9 .10201067 8 Punjabi NM , Caffo BS , Goodwin JL Sleep-disordered breathing and mortality: a prospective cohort study . PLoS Med 2009 ;6 :e1000132 10.1371/journal.pmed.1000132 19688045 9 Tarasiuk A , Reuveni H The economic impact of obstructive sleep apnea . Curr Opin Pulm Med 2013 ;19 :639 –44 . 10.1097/MCP.0b013e3283659e1e 24060978 10 Bahammam A , Delaive K , Ronald J Health care utilisation in males with obstructive sleep apnea syndrome two years after diagnosis and treatment . Sleep 1999 ;22 :740 –7 .10505819 11 George CF Reduction in motor vehicle collisions following treatment of sleep apnoea with nasal CPAP . Thorax 2001 ;56 :508 –12 . 10.1136/thorax.56.7.508 11413347 12 Martínez-García MÁ , Galiano-Blancart R , Román-Sánchez P Continuous positive airway pressure treatment in sleep apnea prevents new vascular events after ischemic stroke . CHEST 2005 ;128 :2123 –9 . 10.1378/chest.128.4.2123 16236864 13 Kaw R , Hernandez AV , Walker E Determinants of hypercapnia in obese patients with obstructive sleep apnea: a systematic review and metaanalysis of cohort studies . CHEST 2009 ;136 :787 –96 . 10.1378/chest.09-0615 19567489 14 Berg G , Delaive K , Manfreda J The use of health-care resources in obesity-hypoventilation syndrome . CHEST 2001 ;120 :377 –83 . 10.1378/chest.120.2.377 11502632 15 Priou P , Hamel J-F , Person C Long-term outcome of noninvasive positive pressure ventilation for obesity hypoventilation syndrome . CHEST 2010 ;138 :84 –90 . 10.1378/chest.09-2472 20348200 16 Povitz M , Bolo CE , Heitman SJ Effect of treatment of obstructive sleep apnea on depressive symptoms: systematic review and meta-analysis . PLoS Med 2014 ;11 :e1001762 10.1371/journal.pmed.1001762 25423175 17 Povitz M , Hanly PJ , Pendharkar SR Treatment of sleep disordered breathing liberates obese hypoxemic patients from oxygen . PLoS ONE 2015 ;10 :e0140135 10.1371/journal.pone.0140135 26451835 18 Rotenberg BW , George CF , Sullivan KM Wait times for sleep apnea care in Ontario: a multidisciplinary assessment . Can Respir J 2010 ;17 :170 10.1155/2010/420275 20808975 19 Flemons WW , Douglas NJ , Kuna ST Access to diagnosis and treatment of patients with suspected sleep apnea . Am J Respir Crit Care Med 2004 ;169 :668 –72 . 10.1164/rccm.200308-1124PP 15003950 20 Kuna ST , Gurubhagavatula I , Maislin G Noninferiority of functional outcome in ambulatory management of obstructive sleep apnea . Am J Respir Crit Care Med 2011 ;183 :1238 –44 . 10.1164/rccm.201011-1770OC 21471093 21 Hirshkowitz M , Sharafkhaneh A A telemedicine program for diagnosis and management of sleep-disordered breathing: the fast-track for sleep apnea tele-sleep program . Semin Respir Crit Care Med 2014 ;35 :560 –70 . 10.1055/s-0034-1390069 25353101 22 Witmans MB , Dick B , Good J Delivery of pediatric sleep services via telehealth: the Alberta experience and lessons learned . Behav Sleep Med 2008 ;6 :207 –19 . 10.1080/15402000802371312 18853305 23 Mulgrew AT , Fox N , Ayas NT Diagnosis and initial management of obstructive sleep apnea without polysomnography: a randomized validation study . Ann Intern Med 2007 ;146 :157 –66 . 10.7326/0003-4819-146-3-200702060-00004 17283346 24 Pendharkar SR , Bischak DP , Rogers P Using patient flow simulation to improve access at a multidisciplinary sleep centre . J Sleep Res 2015 ;24 :320 –7 . 10.1111/jsr.12257 25431022 25 Antic NA , Buchan C , Esterman A A randomized controlled trial of nurse-led care for symptomatic moderate–severe obstructive sleep apnea . Am J Respir Crit Care Med 2009 ;179 :501 –8 . 10.1164/rccm.200810-1558OC 19136368 26 Pendharkar SR , Dechant A , Bischak DP An observational study of the effectiveness of alternative care providers in the management of obstructive sleep apnoea . J Sleep Res 2016 ;25 :234 –40 . 10.1111/jsr.12358 26503454 27 Chai-Coetzer CL , Luo YM , Antic NA Predictors of long-term adherence to continuous positive airway pressure therapy in patients with obstructive sleep apnea and cardiovascular disease in the SAVE study . Sleep 2013 ;36 :1929 –37 . 10.5665/sleep.3232 24293768 28 Sánchez-de-la-Torre M , Nadal N , Cortijo A Role of primary care in the follow-up of patients with obstructive sleep apnoea undergoing CPAP treatment: a randomised controlled trial . Thorax 2015 ;70 :346 –52 . 10.1136/thoraxjnl-2014-206287 25766689 29 Fleetham J , Ayas N , Bradley D Canadian Thoracic Society 2011 guideline update: diagnosis and treatment of sleep disordered breathing . Can Respir J 2011 ;18 :25 10.1155/2011/506189 21369547 30 Fleetham J , Ayas N , Bradley D Canadian Thoracic Society guidelines: diagnosis and treatment of sleep disordered breathing in adults . Can Respir J 2006 ;13 :387 10.1155/2006/627096 17036094 31 Epstein LJ , Kristo D , Strollo PJ Jr , et al Clinical guideline for the evaluation, management and long-term care of obstructive sleep apnea in adults . J Clin Sleep Med 2009 ;5 :263 –76 .19960649 32 CSRT position statement: the respiratory therapist's role in mechanical ventilation: Canadian Society of Respiratory Therapists , 2013 . 33 Kribbs NB , Pack AI , Kline LR Objective measurement of patterns of nasal CPAP use by patients with obstructive sleep apnea . Am Rev Respir Dis 1993 ;147 :887 –95 . 10.1164/ajrccm/147.4.887 8466125 34 Weaver TE , Maislin G , Dinges DF Relationship between hours of CPAP use and achieving normal levels of sleepiness and daily functioning . Sleep 2007 ;30 :711 .17580592 35 Antic NA , Catcheside P , Buchan C The Effect of CPAP in normalizing daytime sleepiness, quality of life, and neurocognitive function in patients with moderate to severe OSA . Sleep 2011 ;34 :111 –19 .21203366 36 Ahmed I , Thorpy MJ Determinants and measures of daytime sleepiness . In: Pagel FJ , Pandi-Perumal RS , eds. Primary care sleep medicine: a practical guide . New York, NY : Springer New York , 2014 :99 –106 . 37 Leclerc G , Lacasse Y , Page D Do obstructive sleep apnea syndrome patients underestimate their daytime symptoms before continuous positive airway pressure treatment? Can Respir J 2014 ;21 :216 –20 . 10.1155/2014/681952 24712013 38 Guimarães C , Martins MV , Vaz Rodrigues L Epworth Sleepiness Scale in obstructive sleep apnea syndrome—an underestimated subjective scale . Rev Port Pneumol 2012 ;18 :267 –71 . 10.1016/j.rppnen.2012.06.005 22743061 39 Campos-Rodriguez F , Pena-Grinan N , Reyes-Nunez N Mortality in obstructive sleep apnea-hypopnea patients treated with positive airway pressure . CHEST 2005 ;128 :624 –33 . 10.1378/chest.128.2.624 16100147 40 Marin JM , Soriano JB , Carrizo SJ Outcomes in patients with chronic obstructive pulmonary disease and obstructive sleep apnea: the overlap syndrome . Am J Respir Crit Care Med 2010 ;182 :325 –31 . 10.1164/rccm.200912-1869OC 20378728 41 Bouloukaki I , Giannadaki K , Mermigkis C Intensive versus standard follow-up to improve continuous positive airway pressure compliance . Eur Respir J 2014 ;44 :1262 –74 . 10.1183/09031936.00021314 24993911 42 Pamidi S , Knutson KL , Ghods F The impact of sleep consultation prior to a diagnostic polysomnogram on continuous positive airway pressure adherence . CHEST 2012 ;141 :51 –7 . 10.1378/chest.11-0709 21700685 43 Rosen CL , Auckley D , Benca R A multisite randomized trial of portable sleep studies and positive airway pressure autotitration versus laboratory-based polysomnography for the diagnosis and treatment of obstructive sleep apnea: the HomePAP study . Sleep 2012 ;35 :757 –67 . 10.5665/sleep.1870 22654195 44 Hoy CJ , Vennelle M , Kingshott RN Can intensive support improve continuous positive airway pressure use in patients with the sleep apnea/hypopnea syndrome? Am J Respir Crit Care Med 1999 ;159 :1096 –100 . 10.1164/ajrccm.159.4.9808008 10194151 45 Parthasarathy S , Subramanian S , Quan SF A multicenter prospective comparative effectiveness study of the effect of physician certification and center accreditation on patient-centered outcomes in obstructive sleep apnea . J Clin Sleep Med 2014 ;10 :1 –7 . 46 Aurora RN , Collop NA , Jacobowitz O Quality measures for the care of adult patients with obstructive sleep apnea . J Clin Sleep Med 2015 ;11 :357 –83 . 10.5664/jcsm.4556 25700878 47 Ware JE , Hays RD Methods for measuring patient satisfaction with specific medical encounters . Med Care 1988 ;26 :393 –402 . 10.1097/00005650-198804000-00008 3352332 48 Julious SA Sample sizes for clinical trials with normal data . Stat Med 2004 ;23 :1921 –86 . 10.1002/sim.1783 15195324 49 Santana M-J , Feeny DH Using the health utilities index in routine clinical care: process, feasibility, and acceptability: a randomized controlled trial . Patient 2009 ;2 :159 –67 . 10.2165/11313620-000000000-00000 22273167 50 Santana MJ , Feeny D , Weinkauf J The use of patient-reported outcomes becomes standard practice in the routine clinical care of lung–heart transplant patients . Patient Relat Outcome Meas 2010 ;1 :93 –105 . 10.2147/PROM.S11943 22915956

PMC005xxxxxx/PMC5372099.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01412410.1136/bmjopen-2016-014124NursingResearch1506171516861724Listening to paediatric primary care nurses: a qualitative study of the potential for interprofessional oral health practice in six federally qualified health centres in Massachusetts and Maryland Bernstein Judith 12Gebel Christina 12Vargas Clemencia 13Geltman Paul 1Walter Ashley 12Garcia Raul 1Tinanoff Norman 13 1 Center for Research to Evaluate and Eliminate Dental Disparities, Boston University Henry M. Goldman School of Dental Medicine, Boston, Massachusetts, USA 2 Department of Community Health Sciences, Boston University School of Public Health, Boston University, Boston, Massachusetts, USA 3 Department of Orthodontics and Pediatric Dentistry, University of Maryland School of Dentistry, Baltimore, Maryland, USACorrespondence to Dr Judith Bernstein; jbernste@bu.edu2017 29 3 2017 7 3 e0141242 9 2016 17 1 2017 14 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives To explore the opportunities for interprofessional collaboration (IPC) to improve paediatric oral health in federally qualified health centres (FQHCs), to identify challenges to IPC-led integration of oral health prevention into the well-child visit and to suggest strategies to overcome barriers. Sample Nurse managers (NMs), nurse practitioners (NPs), paediatric clinical staff and administrators in six FQHCs in two states were interviewed using a semistructured format. Design Grounded theory research. Topics included feasibility of integration, perceived barriers and strategies for incorporating oral health into paediatric primary care. Measurements Qualitative data were coded and analysed using NVivo 10 to generate themes iteratively. Results Nurses in diverse roles recognised the importance of oral health prevention but were unaware of professional guidelines for incorporating oral health into paediatric encounters. They valued collaborative care, specifically internal communication, joint initiatives and training and partnering with dental schools or community dental practices. Barriers to IPC included inadequate training, few opportunities for cross-communication and absence of charting templates in electronic health records. Conclusions NMs, NPs and paediatric nursing staff all value IPC to improve patients' oral health, yet are constrained by lack of oral health training and supportive charting and referral systems. With supports, they are willing to take on responsibility for introducing oral health preventive measures into the well-child visit, but will require IPC approaches to training and systems changes. IPC teams in the health centre setting can work together, if policy and administrative supports are in place, to provide oral health assessments, education, fluoride varnish application and dental referrals, decrease the prevalence of early childhood caries and increase access to a dental home for low-income children. Pediatric oral healthInterprofessional relationsCollaborative practiceNational Institute of Dental and Craniofacial Researchhttp://dx.doi.org/10.13039/100000072U54 DE019275 ==== Body Strengths and limitations of this study In-depth, open-ended interviews with nurses working in different roles and levels of authority in federally qualified health centres in two states allowed frank discussion of benefits and challenges to inclusion of oral health prevention topics in well-child visits. Engagement of clinic administration and staff was facilitated by procedures for protection of confidentiality, which allowed sensitive material about attitudes and knowledge to emerge. Collaborating clinics were selected to represent urban, rural and suburban clinics in states with different systems of financing and allocation of health resources, allowing for saturation of themes. Qualitative approaches are effective for identification of key issues for consideration in developing strategic interventions, but are not designed for generalisability beyond the specific settings that were studied. Background Early childhood caries (ECC) remains a pervasive, burdensome problem and the most common chronic childhood illness.1 2 Evidence-based approaches to reduce ECC include identifying high-risk children at an early age, coordinating care and referrals to a dental home by age one, and integrating oral health into paediatric primary care through anticipatory guidance and behaviour modification counselling.3–5 It is suggested that paediatric oral health is the next frontier for prevention.6 The aim of this qualitative study was threefold: (1) to investigate what nurses working with high-risk paediatric patients in a structured clinical setting believe and know about their potential to engage fully in collaborative practice to promote children's oral health; (2) to identify the barriers and challenges to role expansion that nurses experience in low-resource primary paediatric care settings and (3) to elicit their recommendations for enhancing their role in paediatric oral health prevention. Robust approaches to prevention require interdisciplinary and interprofessional collaboration (IPC).7 8 A recent multiorganisational report calls for the ‘partnering of primary care and dental health professionals to reduce the burden of oral disease’,6 and existing literature supports the importance of integrating dental and medical care in the primary care system.4 7 9 10 Worldwide, WHO policy, established collectively with the FDI Science Commission and the International Association for Dental Research, calls for an interprofessional effort to scale up capacity to produce oral health personnel, including dental hygienists, nurses and auxiliaries, providing for equitable distribution of these auxiliaries to the primary care level and ensuring proper service backup by dentists through appropriate referral systems.11 Nurses are uniquely positioned to foster IPC. Cooperation between nurse practitioners (NPs) and dieticians, for example, can be used to expand the dental workforce. IPC teams work together to provide oral health assessments, education, fluoride varnish application and dental referrals, decrease the prevalence of ECC and increase access to a dental home for low-income children.10 Referrals are important because children aged 2–5 years who receive a recommendation for a dental home are more likely to have a dental visit.12 13 Fewer than half of children on Medicaid receive a preventative dental visit each year,14 and one in seven patients (paediatric and adult) at federally qualified health centres (FQHCs) reported their most recent dental visit was >5 years ago.15 In a study of paediatric primary care physicians, 21% reported not screening for oral health due to lack of resources for referring patients to dental care and 42% reported difficulty integrating dental procedures into their practices.16 In another study, 96% reported that the most common method for dental referral was to provide the name of a dentist without assistance in making the appointment.17 IPC models, which are mandated by federal legislation18 and supported by dental and medical professionals,19 have the potential to increase oral health screening, referral and service capacity.20 They also aid in meeting accreditation standards and encourage implementation of practice guidelines,21 increase quality and efficiency22 and improve access for uninsured and underserved populations.23 We conducted a multimethod case study in six FQHCs with a medical home designation to investigate challenges to implementation of dental, medical and nursing collaboration, with the primary goal to understand how all roles within a clinic setting, including nurses, NPs, medical assistants, physicians, dental clinicians (where applicable), nurse managers (NM) and other administrators can work together to improve oral health.24 This current report describes findings from the subset of interviews with nurses across settings and nursing roles, describes their perspectives and challenges and the supports they would need to enhance the feasibility of IPC. Methods This study was reviewed and classified under exempt status by the Institutional Review Boards at Boston University and University of Maryland, where coinvestigators were located, and at University of Baltimore, where interviewers were located. Participants completed an informed consent to be interviewed by telephone and to have the conversation recorded and transcribed. The research team of seven, four women and three men, included five senior researchers. Design We selected three FQHCs in state 1 and three in state 2 to represent a continuum of oral health integration into paediatric services and a range of geographic locations (rural, small community and urban), organisational structures, patient populations, workforce composition and financial resources across a continuum of oral health integration into paediatric services. Contextual, organisational and professional data were collected from August 2014 through March 2015. We then conducted key informant interviews over a 6-month period to analyse contextual, organisational and professional factors that may facilitate or hinder IPC in paediatric oral health. Research team and reflexivity Each clinic administrator identified potential interviewees who were knowledgeable about conditions in their paediatric well-child clinics, explained the purpose of the study and asked if they were willing to be interviewed by telephone. Two independent interviewers from the faculty of the University of Baltimore (WW, DM) were provided with a list of interested interviewees from each clinic. They used this list to send introductory emails describing their qualifications, and then called to introduce themselves, review the study purpose, ascertain willingness to participate and conduct an informed consent process. The research team included dental and medical clinical faculty with advanced degrees (PhD, MS, MMSc and MPH) in related fields. Theoretical framework Methodologic orientation: domains of inquiry were established based on the four content areas of the Consolidated Framework for Implementation Research: the nature of the intervention (in this case, introduction of oral health prevention into the well-child visit), characteristics of children and their families, characteristics, attitudes and experiences of nurse clinicians and NMs (the inner context) and policy concerns (the outer context). The theoretic framework thus used content analysis and grounded theory. Study quality was addressed through use of the Consolidated Criteria for Reporting Qualitative Research (CORE-Q). Sample Clinic decision makers, nursing, medical and dental clinicians and support personnel were identified for interview by each clinic director. Selection of potential participants as role experts was purposive, designed to capture the expertise of paediatric primary care nurses across job descriptions (NM/charge nurse, registered nurse (RN) and NP) and within each clinic named in each state. In all, 42 contacts were named and 39 interviewed. All 10 of the nurses who were named by the administrators in the six clinics agreed to participate and completed interviews, and therefore constitute the sample for this investigation. Among them, there were five NMs, three NPs and two RNs, with all clinics represented. This sample was large enough to allow a broad generation of themes but small enough to limit burden and increase the likelihood that the clinics would agree to participation. Setting Confidential 1-hour interviews were conducted with one of the two University of Baltimore interviewers over the telephone number of participants' choice, at their convenience. Measures Interviewers used a semistructured interview schedule with cues and prompts to elicit relevance of oral health for general health, current oral health practices, degree of integration and feasibility of oral health activities as part of paediatric primary care, acceptable methods for integration, perceived facilitators and barriers for integration and potential strategies to address barriers. Questions were generated building on previously piloted items.6 The interview schedule is available as an online supplementary file. 10.1136/bmjopen-2016-014124.supp1supplementary file Protection of confidentiality The goal of this study was to promote full disclosure. Clinic administrators agreed to participation based on expectation of protection of confidentiality. For this reason, the sites that participated are not identified, and the individuals who were interviewed are identified in quotes only by their roles in clinic operation. Analytic strategy The data analytic process was designed to explore how factors related to the professional training and expertise of nurses and the hierarchical clinic culture they work in might affect nurses' ability and willingness to engage in IPC. Audio recordings were transcribed. Two experienced qualitative data analysts and a senior faculty member of the research team at Boston University independently coded three key informant interviews and assigned initial codes using inductive coding methods. The three team members then met to discuss the proposed code list and resolve any differences in interpretation. A list of agreed-on first-level codes were entered into a master code list using NVivo 10 software with accompanying code definitions so coding was consistent across interviews. Codes were assigned to three interview transcripts by two qualitative data analysts. Inter-rater reliability was assessed and the three interviews recoded independently until inter-rater reliability was satisfactory (κ coefficient >0.70). The remainder of the interviews were coded, and new codes were discussed, defined and added to the master list until the list reached saturation. Once all data were coded, the three team members met to assess theme characterisation from final codes, referring to interview field notes as well as transcripts. Interviewees were not contacted for clarifications and did not receive feedback. Results Ten of the 39 interviews from the parent study were with nursing participants in three roles: NM, NPS and RN. Three overarching themes emerged from these interviews: (1) recognition by nurses of the importance of oral health to paediatric general health and development, as a sine qua non for engagement in IPC; (2) facilitators and barriers to IPC and (3) nursing recommendations for enhancing IPC. Nursing responses largely reflected the sentiments of other paediatric and administrative staff, but nurses, overall, were much more willing to see integration of oral health content as a nursing responsibility, but only if appropriate training was made available. The nursing interviews generated concrete, useful suggestions for addressing barriers. We present here major and minor themes, which emerged with clarity and consistency. Recognition by nurses of patients' needs and a role for IPC in paediatric oral health Nurses across different roles and levels of authority made a clear connection between oral health and children's physical, mental and social development:I think it's very important because…the mouth is opening to the rest of the body…We have kids that I know of that have a hard time sitting in school because their mouths hurt so much. [NM] Interviewees were concerned about the consequences of lack of prevention:There are a lot of kids who have so many caries that need to be fixed or extractions that need to be done that they end up having to be fully sedated for the procedures. [NP] Some of the interviewees thought that parental neglect was responsible for the poor oral hygiene and high rates of ECCs that are common in their patients:Neglect is on top of the list….Parents/teachers don't have a clue, you know, baby bottle tooth decay, just awful, rampant. They don't admit that they put the baby to bed with the bottle but you know that they do because you just see the powder in the mouth clearly. [NP] Nurses described a low level of oral health literacy among parents who use services at a FQHC:A lot of them just, unfortunately…don't realize that they need preventative dental services just like they need well child visits. [NP] The lack of oral health awareness was particularly noted among the diverse immigrant families who use FQHCs:I think the patients that we get who are immigrants don't understand [that] having multiple bad teeth affects their eating, their sleeping, how they do in school. [NP] Interviewees reflected on when their own children were young and the problems they had themselves with brushing and getting dental care:Baby teeth… “They are going to fall out anyway.” And I think for some parents, that may be the point of reference that they are coming from…Dentists try to preserve baby teeth and keep the alignment [for] their second teeth. And I'm not sure that that's a message that the parents…have gotten. [NM] There was also understanding of the competing family priorities for FQHC patients:Life is expensive and there's bills for everything and dental services seems like a luxury for some that can't afford it, and so that's the one thing that gets forgotten. [NP] And: A large majority of our parents want their kids to be well and to not have issues. I'm just not sure that they have the resources to know where they need to go, to go to the dentist for their kids. [NM] Oral health content was seen to be consistent with the concept of a medical home that is an organising principle for FQHCs:Bringing in oral health, dental. It definitely supports the model of the patient center medical home. Where we live we have terrible winters…And if you only have to make one stop, I think that would be something that people would like. [NM] There was broad acceptance of responsibility for prevention:When we see kids especially in the course of well child visits, we discuss last dental appointment, oral hygiene, brushing the teeth because, unfortunately, some kids and parents are unaware that that's necessary. Talking about sugary drinks and foods is not only a part of the physical examination but it's also a part of anticipatory guidance. [NP] And that responsibility extends to charting:I always document if they brush their teeth [and] that they've been to the dentist in the last six months, [and] diet in terms of limiting juice, no soda, limit candy and high-sweetened snacks. [NP] Several interviewees were in favour of a collaborative approach to healthcare delivery:I mean, you can't separate the mouth from the child and the dentist from the medical provider. It all needs to be integrated. [NP] And: I really think in terms of education that this would be opportunity for the nurses to step forward and do some more stuff, the stuff the pediatricians don't need to directly do. And I'd like to see that happen. [NM] Facilitators to integration of oral health into paediatric well-childcare Interviewees mentioned many benefits of incorporating oral health preventive services into paediatric practice:If we can see them young and talk about it, hopefully we're having them make behavioral change so they don't suffer the same problems as other adults in their families do. [NP] The paediatric well-child visit was seen as an optimal setting for prevention:It just seems to me that kids have a better relationship with their primary care provider than they do their dentist. The dentist's office is often a scary place to be for children. [NM] And: We do fluoride treatments for our kids already. And I think that a lot of parents respect our pediatricians, and then if it came from the pediatrician, it would hold more weight. [NM] Many nurses also thought about efficiencies that might result:I think it's beneficial if you do as much as you can when they're here in front of you. Our families often have difficulty making it to multiple appointments. So there's probably something we could incorporate during a visit to …emphasize the point that we also consider oral health when we're considering the children's overall health. [Pediatric Nurse Practitioner] Others, especially at the RN level, were sceptical about ability to deliver, and concerned about delegating any tasks to extenders:I don't see any of our staff having time or a person just devoted directly to dental…[and] I don't think that assistants should be scheduling appointments at all, but I do feel that they can certainly send us messages where we can contact the patient and follow-up. But I don't feel that they are necessarily medically trained. [RN] Barriers to IPC There was a general lack of awareness of professional guidelines for preventive education and services in paediatrics or paediatric dentistry:I'm assuming that they have them. I don't know what they are currently. [NP] Space is always an issue in FQHCs:Space will probably be the biggie. But again, if we were going to incorporate it in each appointment and we didn't need special stuff like a dental chair, we'd be okay about that. [NM] Time is limited:They only get primarily 15 min with each patient, so I know that it's very hard for them to fit [in] every single thing that they have to do with a patient…A lot of our patients have health disparities, mental health issues, and dental probably isn't top of the list, but I believe with more time they would definitely incorporate that. [RN] Electronic medical record systems posed a significant barrier to integration in each of the clinics that were studied:If the dentist was concerned about nutritional status…we don't often get that information. While we are all connected under the health center, we are sort of together but we are separate. [RN] Respondents expressed a strong desire to have cross-communication and saw this as a logical next step, since both systems report on a patient's health. However, nurses reported not knowing about the activities of their colleagues:Well, I haven't seen what happens on the dental side because they are not here at my site. [NP] Charting for paediatric and dental clinicians was in entirely different systems:If you as the provider…actually had medical records that could talk to each other, like EPIC does, then [the dentist] could see what was going on medically…and you could see what they're doing; that would be absolutely wonderful. [NM] The lack of templates for oral health was also mentioned:I have to ‘free text’ that they've been to the dentist, that they brush their teeth, …that all has to be added on by me. If there was a template where we could just click on that, it would be great. [NP] Respondents spoke of their lack of oral health knowledge and the need for training in this area:I really feel like I probably need some additional training; Nurse-Practitioners are deficient in dental training. [NP] And: I would really like to get some additional hands-on training for the dental examination. I mean, I can go in there and look around in the mouth…but I may miss what I'm looking for. [NP] Nurses and NPs were open to collaboration with extenders, as long as they received the proper training:I think training just has to match what you're asking people to do. So if you're asking them to do a new procedure, they need the training to do that. [NP] And: Then the medical assistant could do the fluoride application at the end of the visit, as well as give them a toothbrush, and possibly some kind of book for the child, or coloring page, or something. [NM] Nurses were concerned about the lack of systems in place for tracking referral outcomes:That is a terrible deficiency that I hope they make the appointment and then they go to the clinic and that's where I lose it, unfortunately, in the loop. [NP] Additional resources are needed:Visuals mean everything. Kids see pictures….You can hand out type-written stuff forever, but when a child looks up and sees pictures of bad teeth, they respond to it. And parents look at it and they, you know, just a simple picture with caption: This is baby bottle tooth decay. This is what happens when you don't brush. [NP] Colocation of paediatric and dental services would help:I definitely think co-location would help. It's very hard for our parents….The parents know this site already and they know how they're going to get there. [NM] But what was needed even more was administrative leadership and champions:I do think it would be helpful to designate a team leader, a person that's kind of the advocate, who helps with training or setting up training. [NP] Discussion The success of IPC depends on confluence of three domains: interpersonal processes, communication channels inside the organisation, in this case, a clinic and in the organisation's external environment.24 25 Respondents in this study reported efforts towards dental, medical and nursing collaboration, as well as structural, cultural and educational barriers that complicate, stall or prevent the positive influence of these efforts. Structural barriers Efforts towards interdisciplinary team-based collaboration can be complicated by professional boundaries, sometimes referred to as ‘professional silos’ or ‘the silo effect’,26 27 as well as challenges from professional cultures, defined as a profession's ‘values, beliefs, attitudes, customs and behaviours’.28 Differences in power to make decisions were noticeable; RNs were more likely to feel sceptical about instituting change, while NPs and NMs appeared to be more receptive to integration of oral health into the well-child visit, perhaps because they felt more able to communicate their needs upward in the administrative chain. Cultural barriers The surgeon general has called for a shift in the perception that oral health is solely the responsibility of dentists, encouraging all clinicians—dental, medical and nursing—to participate in improving oral health to improve overall health status and adoption of collaborative, interdisciplinary approaches to care.7 Nurses felt the need for systematic training to take on these new responsibilities. Educational barriers The reported lack of confidence in expertise is certainly due to the scarcity of training in oral health noted here and described by the DentaQuest Foundation.29 The surgeon general has called for changes to clinical curricula to include oral health and multidisciplinary training in continuing medical education offerings.30 For example, providers need further education in assessment and use of assessment tools in order to improve accuracy in visible exams in well-childcare.31 In one study by AlYousef et al,32 87.5% of medical students rated their training in oral health assessments as fair or poor, though 90% believed the role of providers counselling children in oral health and referring them to treatment was important. The knowledge deficit that we identified could be remedied by teaching students’ oral examination as part of standard head, ears, eyes, nose and throat (HEENT) examinations.33 The NPs in this study who were unaware of guidelines, did not use an assessment tool and seemed uncertain as to what such a tool might contain, would certainly benefit from additional training in order to make such efforts standard practice. IPC in education has been shown to improve NPs' core competencies in oral health34 and could be part of a solution to overcoming the barriers described in this study. Innovative models are already in use, as well as guidance for how to integrate IPC models into curricula.35 In the field of nursing, options for IPE education could include service learning projects as well as inclusion of intervention strategies for oral health in standard curricula.36 Toolkits and resources are available for nurse educators for competencies development.37 Projects and dissemination could be core components of overall educational infrastructures, which include competencies to ‘prioritise oral health disease prevention and health promotion, provided evidence-based oral healthcare in a variety of practice settings, and collaborate in interprofessional teams across the healthcare system’.38 The Oral Health Nursing Education and Practice (OHNEP) programme at New York University's College of Nursing calls for a national nursing action plan with short-range, mid-range and long-range nursing strategies to advance the nation's oral health agenda; a National Nursing Workgroup on Oral Health; and train-the-trainer workshops for faculty and students, among other efforts.37 38 Advantages and limitations of this study This study details through the lens of nursing practice the need for IPC in paediatric oral health, and identifies specific areas that require attention to enable nurses in the FQHC setting to collaborate across professional boundaries. As such, it adds to a growing body of literature on the need for IPC to improve patients' oral health status. Qualitative research adds depth and suggests important themes, but is necessarily limited in breadth by small samples derived from six clinics in three states. The 2015 White Paper on oral health into paediatric primary care6 sets a clear path for us in stating that ‘Oral health is the next frontier’. The IPC approaches needed to explore that frontier require changes in training and systems, and much work needs to be done to establish feasible, affordable strategies to reach the preventive goals set by major organisations in the USA and worldwide; this study, while limited by scope, suggests some useful directions. Although the study was broad in the sense that it included clinics of different geographical settings and size in two states, generalisability of results is limited by the small sample interviewed from each clinic, and the small number of clinics that served as the setting for this study. Conclusions NMs, NPs and paediatric nursing staff interviewed in this study all valued IPC to improve patients' oral health, yet reported feeling constrained by lack of oral health training and supportive charting and referral systems. With supports, they would take on responsibility for introducing oral health preventive measures into the well-child visit, but will require IPC approaches to training and systems changes if they are to take an active role in oral health prevention. IPC teams in the health centre setting can work together, if policy and administrative supports are in place, to provide oral health assessments, education, fluoride varnish application and dental referrals, decrease the prevalence of ECCs and increase access to a dental home for low-income children. Caution should be exercised in generalising from this study, but in the clinics that were included, nurses clearly had the potential and willingness to work towards improvements in children's oral health if supports were provided. Contributors: JB, CG, CV, PG, AW, RG and NT all participated in study design and implementation, interpretation of study findings and writing of this report. Funding: This study was supported in part by a cooperative agreement with the National Institute of Dental and Craniofacial Research, National Institutes of Health (U54-DE019275). Competing interests: None declared. Ethics approval: Boston University's Institutional Review Board was the lead university on this project. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Small cells and a need to protect confidentiality of participating nurses and their clinics preclude release of identifiable tapes or transcripts, and Nvivo codes include identification of clinic and participant. The authors are happy to respond to queries from researchers interested in this topic. ==== Refs References 1 Institute of Medicine . Advancing oral health in America 2011 http://books.nap.edu/openbook.php?record_id=13086 (accessed 30 Aug 2016 ). 2 Department of Health and Human Services, & U.S. Public Health Service . Oral health in America: a report of the surgeon general. National Institute of Dental and Craniofacial Research, National Institutes of Health 2000 http://www.nidcr.nih.gov/DataStatistics/SurgeonGeneral/Documents/hck1ocv.@www.surgeon.fullrpt.pdf (accessed 30 Aug 2016 ). 3 Kagihara LE , Niederhauser VP , Stark M Assessment, management, and prevention of early childhood caries . J Am Acad Nurse Pract 2009 ;21 :1 –10 . 10.1111/j.1745-7599.2008.00367.x 19125889 4 Crall JJ Development and integration of oral health services for preschool-age children . Pediatr Dent 2005 ;27 :323 –30 .16317973 5 Hale KJ , and the American Academy of Pediatrics Section on Pediatric Dentistry . Oral health risk assessment timing and establishment of the dental home . Pediatrics 2003 ;111 (Pt 1 ):1113 –16 . 10.1542/peds.111.5.1113 12728101 6 Hummel J , Philips K , Holt B Oral health: an essential component of primary care (White Paper.) . Seattle, WA : Qualis Health 2015 http://www.nasuad.org/sites/nasuad/files/White-Paper-Oral-Health-Primary-Care.pdf (accessed 30 Aug 2016 ). 7 US Department of Health and Human Services . Oral health in America: a report of the surgeon general . J Calif Dent Assoc 2000 ;28 :685 –95 .11324049 8 Wagner Y , Heinrich-Weltzien R Evaluation of a regional German interdisciplinary oral health programme for children from birth to 5 years of age . Clin Oral Invest 2017 ;21 :225 –35 . 10.1007/s00784-016-1781-8 9 Health Resources and Services Administration . Integration of oral health and primary care practices . Rockville, MD : US Department of Health and Human Services , 2014 http://www.hrsa.gov/publichealth/clinical/oralhealth/primarycare/integrationoforalhealth.pdf (accessed 30 Aug 2016 ). 10 Biordi DL , Heitzer M , Mundy E Improving access and provision of preventive oral health care for very young, poor, and low-income children through a new interdisciplinary partnership . Am J Pub Health 2015 ;105 (Suppl 2 ):e23 –9 . 10.2105/AJPH.2014.302486 11 Petersen PE World Health Organization global policy for improvement of oral health—World Health Assembly 2007 . Internat Dent J 2008 ;58 :115 –21 . 10.1111/j.1875-595X.2008.tb00185.x 12 Beil HA , Rozier RG Primary health care providers’ advice for a dental checkup and dental use in children . Pediatrics 2010 ;126 :e435 –441 . 10.1542/peds.2009-2311 20660547 13 Zilversmit L , Kane DJ , Rochat R Factors associated with receiving treatment for dental decay among Medicaid-enrolled children younger than 12 years of age in Iowa, 2010 . J Pub Health Dent 2015 ;75 :17 –23 . 10.1111/jphd.12066 25131658 14 Mann C Update on CMS oral health initiative and other oral health related items (CMCS Informational Bulletin) . Baltimore, Maryland : Center for Medicaid and CHIP Services , 2014 http://www.medicaid.gov/Federal-Policy-Guidance/downloads/CIB-07-10-2014.pdf (accessed 30 Aug 2016 ). 15 Jones E , Shi L , Hayashi AS Access to oral health care: the role of federally qualified health centers in addressing disparities and expanding access . Am J Pub Health 2013 ;103 :488 –93 . 10.2105/AJPH.2012.300846 23327254 16 Close K , Rozier RG , Zeldin LP Barriers to the adoption and implementation of preventive dental services in primary medical care . Pediatrics 2010 ;125 :509 –17 . 10.1542/peds.2009-1008 20123767 17 dela Cruz GG , Rozier RG , Slade G Dental screening and referral of young children by pediatric primary care providers . Pediatrics 2004 ;114 :e642 –52 . 10.1542/peds.2004-1269 15520094 18 Edelstein BL The roles of federal legislation and evolving health care systems in promoting medical-dental collaboration . J Calif Dent Assoc 2014 ;42 :19 –23 .25080685 19 Cohen LA Expanding the physician's role in addressing the oral health of adults . Am J Pub Health 2013 ;103 :408 –12 . 10.2105/AJPH.2012.300990 23327256 20 Taylor E , Marino D , Thacker S Expanding oral health preventative services for young children: a successful interprofessional model . J Allied Health 2014 ;43 :e5 –9 .24598903 21 Jackson JT , Quinonez RB , Kerns AK Implementing a prenatal oral health program through interprofessional collaboration . J Dent Educ 2015 ;79 :241 –8 .25729017 22 Formicola AJ , Ro M , Marshall S Strengthening the oral health safety net: delivery models that improve access to oral health care for uninsured and underserved populations . Am J Pub Health 2008 ;98 (9 Suppl):S86 –8 .18687629 23 Beetstra S , Derksen D , Ro M A “health commons” approach to oral health for low-income populations in a rural state . Am J Pub Health 2008 ;98 (9 Suppl):S89 –90 . 10.2105/AJPH.98.Supplement_1.S89 18687630 24 Bernstein HH , Dhepyasuwan N , Connors K , CORNET Investigators . Evaluation of a national bright futures oral health curriculum for pediatric residents . Acad Pediatr 2013 ;13 :133 –9 . 10.1016/j.acap.2012.10.010 23257667 25 San Martín-Rodríguez L , Beaulieu MD , D'Amour D The determinants of successful collaboration: a review of theoretical and empirical studies . J Interprof Care 2005 ;19 (Suppl 1 ):132 –47 . 10.1080/13561820500082677 16096151 26 Vanderbilt AA , Dail MD , Jaberi P Reducing health disparities in underserved communities via interprofessional collaboration across health care professions . J Multidiscip Healthc 2015 ;8 :205 –8 . 10.2147/JMDH.S74129 25960659 27 Weller J , Boyd M , Cumin D Teams, tribes and patient safety: overcoming barriers to effective teamwork in healthcare . Postgrad Med J 2014 ;90 :149 –54 . 10.1136/postgradmedj-2012-131168 24398594 28 Hall P Interprofessional teamwork: professional cultures as barriers . J Interprof Care 2005 ;19 (Suppl 1 ):188 –96 . 10.1080/13561820500081745 16096155 29 Mitchell-Royston L , Nowak A , Silverman J Integration of oral health into pediatric medical primary care in community health centers . AAPD, The Pediatric Oral Health Research and Policy Center, Dentaquest Foundation , 2014 :1 –19 . http://www.aapd.org/assets/1/7/Dentaquest_Year_1_Final_Report.pdf (accessed 30 Aug 2016 ). 30 Peterson-Sweeney K , Stevens J Optimizing the health of infants and children: their oral health counts! J Pediatr Nurs 2010 ;25 :244 –9 . 10.1016/j.pedn.2009.03.003 20620804 31 Dumas SA , Weaver KE , Park SY Accuracy of visible plaque identification by pediatric clinicians during well-child care . Clin Pediatr 2013 ;52 :645 –51 . 10.1177/0009922813483876 32 AlYousef Y , Damiano P , Weber-Gasparoni K Medical students’ child oral-health-related knowledge, practices and attitudes . Eur J Dent Educ 2013 ;17 :218 –24 . 10.1111/eje.12041 24127762 33 Haber J , Hartnett E , Allen K Putting the mouth back in the head: HEENT to HEENOT . Am J Pub Health 2015 ;105 :437 –41 . 10.2105/AJPH.2014.302495 25602900 34 Hallas D , Fernandez JB , Herman NG Identification of pediatric oral health core competencies through interprofessional education and practice . Nurs Res Practice 2015 ;2015 : 360523 10.1155/2015/360523 35 Haber J , Spielman AI , Wolff M Interprofessional education between dentistry and nursing: the NYU experience . J Calif Dent Assoc 2014 ;42 :44 –51 .25080689 36 Smith GA , Riedford K Epidemiology of early childhood caries: clinical application . J Pediatr Nurs 2013 ;28 :369 –73 . 10.1016/j.pedn.2012.04.005 22608944 37 Dolce MC , Aghazadeh-Sanai N , Mohammed S Integrating oral health into the interdisciplinary health sciences curriculum . Dent Clin North Am 2014 ;58 :829 –43 . 10.1016/j.cden.2014.07.002 25201545 38 Dolce MC , Haber J , Shelley D Oral health nursing education and practice program . Nurs Res Pract 2012 ;2012 : 149673 10.1155/2012/149673 22685642

PMC005xxxxxx/PMC5372100.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01468010.1136/bmjopen-2016-014680PaediatricsResearch1506171916961699170417241715Cross-sectional observational assessment of quality of newborn care immediately after birth in health facilities across six sub-Saharan African countries de Graft-Johnson Joseph 1Vesel Linda 2Rosen Heather E 3Rawlins Barbara 1Abwao Stella 1Mazia Goldy 1Bozsa Robert 4Mwebesa Winifrede 5Khadka Neena 1Kamunya Rosemary 6Getachew Ashebir 7Tibaijuka Gaudiosa 8Rakotovao Jean Pierre 9Tekleberhan Alemnesh 10 1 Maternal and Child Survival Program, Washington, DC, USA 2 Innovations for Maternal, Newborn and Child Health, Concern Worldwide, New York, New York, USA 3 Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, USA 4 ICF International, Rockville, Maryland, USA 5 Save the Children, Washington, DC, USA 6 JHPIEGO Corporation, Baltimore, Maryland, USA 7 Consultant Obstetrician Gynecologist, Addis Ababa, Ethiopia 8 Maternal and Child Survival Program, Dar es Salaam, Tanzania 9 Maternal and Child Health Integrated Program, Antananarivo, Madagascar 10 Maternal and Child Survival Program, Addis Ababa, EthiopiaCorrespondence to Dr Joseph de Graft-Johnson; jjohnson@savechildren.org2017 27 3 2017 7 3 e01468010 10 2016 17 2 2017 27 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective To present information on the quality of newborn care services and health facility readiness to provide newborn care in 6 African countries, and to advocate for the improvement of providers' essential newborn care knowledge and skills. Design Cross-sectional observational health facility assessment. Setting Ethiopia, Kenya, Madagascar, Mozambique, Rwanda and Tanzania. Participants Health workers in 643 facilities. 1016 health workers were interviewed, and 2377 babies were observed in the facilities surveyed. Main outcome measures Indicators of quality of newborn care included (1) provision of immediate essential newborn care: thermal care, hygienic cord care, and early and exclusive initiation of breast feeding; (2) actual and simulated resuscitation of asphyxiated newborn infants; and (3) knowledge of health workers on essential newborn care, including resuscitation. Results Sterile or clean cord cutting instruments, suction devices, and tables or firm surfaces for resuscitation were commonly available. 80% of newborns were immediately dried after birth and received clean cord care in most of the studied facilities. In all countries assessed, major deficiencies exist for essential newborn care supplies and equipment, as well as for health worker knowledge and performance of key routine newborn care practices, particularly for immediate skin-to-skin contact and breastfeeding initiation. Of newborns who did not cry at birth, 89% either recovered on their own or through active steps taken by the provider through resuscitation with initial stimulation and/or ventilation. 11% of newborns died. Assessment of simulated resuscitation using a NeoNatalie anatomic model showed that less than a third of providers were able to demonstrate ventilation skills correctly. Conclusions The findings shared in this paper call attention to the critical need to improve health facility readiness to provide quality newborn care services and to ensure that service providers have the necessary equipment, supplies, knowledge and skills that are critical to save newborn lives. neonatal resuscitationessential newborn carenewborn carebreastfeedingcross-sectional observational assessmentUnited States Agency for International Developmenthttp://dx.doi.org/10.13039/100000200GHS-00-08-00002-00 ==== Body Strengths and limitations of this study The study helps close the evidence gap on the quality of newborn care. The study team collected and analysed directly observed data during childbirth. The study was conducted in multiple countries, which allowed the researchers to examine newborn care practices across these countries. The country samples are neither representative at the national level, nor are they similar across countries. No data on user perspectives regarding quality of care, which is an essential component of quality, were collected because it was outside the scope of this assessment. Introduction Globally, 2.7 million newborns die each year, largely as a result of birth asphyxia, complications of preterm birth and infections.1 This burden of death is disproportionately concentrated in low-income and middle-income countries (LMICs). Sub-Saharan Africa has the highest neonatal mortality rates (NMRs),1 experiencing a mortality reduction of 28% since 1990.2 Despite the impressive efforts undertaken around the Millennium Development Goals (MDG) 4 to reduce child mortality, the decline of neonatal mortality has been much slower than for mortality among children aged 1–59 months; consequently, neonatal deaths now contribute to about 45% of total under-5 deaths in 2015.1 Recognising the importance of addressing this problem, in 2014 the World Health Assembly endorsed the Every Newborn Action Plan (ENAP)—a road map for ending preventable newborn deaths and stillbirths—with a target for all countries to attain 12 or fewer neonatal deaths per 1000 live births by 2030.3 This target is included also in the Sustainable Development Goals.4 Additionally, the ENAP's first two strategic objectives focus on strengthening and investing in quality of care (QoC) around the time of birth as mechanisms to achieving the mortality target.3 Evidence suggests that the majority of these deaths can be prevented, particularly those occurring at birth and immediately thereafter, with simple, evidence-based essential newborn care (ENC) interventions conducted by skilled providers and supported with available commodities.5 These immediate interventions include ensuring that the newborn is kept warm with immediate drying and skin-to-skin contact, giving appropriate stimulation for newborns unable to breathe, providing additional neonatal resuscitation measures as necessary, and ensuring infection prevention, early initiation of exclusive breast feeding and hygienic cord care.2 5 6 Timeliness of service provision and quality of interventions are essential to prevent newborn deaths and consequent disability.7 8 However, shortfalls in the quality of maternal and child health in LMICs9–12 have resulted from a serious lack of attention to the evaluation and quality improvement of facility newborn care.13–15 Health facilities play a crucial role in reducing neonatal mortality, and thus it is critical that they have the capacity and QoC to meet demand and move towards greater improvements in health and survival for newborn infants. However, evidence suggests that provider skills and competencies within health facilities are insufficient to increase neonatal survival.16 Intrapartum complications, including birth asphyxia, account for nearly a quarter of all neonatal deaths.17 Up to 10% of babies require support to initiate breathing in the first minutes of life.2 6 18 Stimulation, including drying and rubbing, is required to initiate breathing in ∼10 million babies at birth globally every year. This technology-free practice should be adequately performed by all skilled birth attendants. Approximately six million babies require resuscitation with bag and mask ventilation.18 Basic resuscitation, including bag and mask ventilation with room air, can have a big impact on survival and can be achieved with basic training and competencies.4 6 18 Given proper training and adequately reprocessed functional equipment, neonatal resuscitation can be performed in low-resource settings.19 Basic resuscitation is adequate for most babies to survive; neonatal resuscitation training can avert 30% of deaths of full-term babies and 5–10% of preterm babies, saving hundreds of thousands of newborn lives. Less than 1% of asphyxiated babies require more advanced resuscitation.18 An increased focus on quality improvement for patient care within the health system and in formulation and implementation of health policies is critically important.12 20 Identifying gaps in and barriers to quality newborn care in facilities is one of the critical steps in the improvement process. To guide QoC improvement activities for maternal and newborn care at health facilities in selected countries, the Maternal and Child Health Integrated Program (MCHIP), with funding from the US Agency for International Development, conducted a multicountry health facility assessment that included actual observations of clinical care for mothers and newborn infants at health facilities in six sub-Saharan African countries: Ethiopia, Kenya, Madagascar, Mozambique, Rwanda and Tanzania (including Zanzibar). This paper presents information on key aspects of health facility readiness to provide ENC services, including resuscitation and the QoC at the assessed facilities. Methods Setting and sample The facility surveys were conducted in the six countries with ongoing MCHIP-supported projects. Health facility surveys were conducted between 2009 and 2012 in Ethiopia, Kenya, Madagascar, Mozambique, Rwanda and Tanzania (including Zanzibar). MCHIP collaborated with the Ministry of Health and other relevant stakeholders in each country to determine appropriate sampling for participating facilities based on the purpose of the survey in that country. Facility sampling took into account the number of facilities offering antenatal care and delivery services, facility client caseloads, and the fieldwork logistics involved. Surveys were conducted as part of a cross-sectional assessment or an existing survey or process. In Kenya, the survey was integrated into the 2010 Demographic Health Survey (DHS) Service Provision Assessment (SPA). Higher level facilities, such as district and regional hospitals, were generally selected through a census approach, whereas lower level facilities, such as health centres, were selected through random sampling procedures.21 Study participants included women in labour, newborns and healthcare providers working in the labour and delivery ward on the day of the assessment. The sample size for the required number of deliveries was estimated for individual countries to be at least 250 based on 50% prevalence for study indicators of interest, with 80% power and 95% precision, assuming a design effect of 2 for clustering of observations by provider and health facility, with adequate power to measure at least 18% change in key variables over time. Since the prevalence of the indicators of interest in the study population was unknown, 50% was used to generate the largest sample size. Data collection Data collection included observations of normal and complicated births, an inventory of labour and delivery supplies, and interviews with and knowledge tests for health workers attending deliveries. Data were collected on the quality of services and facility readiness to deliver routine and emergency care for pregnant and delivering women and for newborns. However, this paper only reports on the ENC component of the study. A series of five data collection tools were used to gather information on the quality of immediate newborn care. A clinical checklist was used to observe deliveries within 1 hour after birth to evaluate the quality of immediate ENC and newborn resuscitation. Provider interviews included questions on appropriate equipment needed for immediate newborn care after birth, components of immediate newborn care, signs and symptoms of newborn sepsis, and actions to be taken to care for low birthweight (LBW; <2.5 kg) babies. Each question had multiple answers, so respondents were asked to provide as many answers as came to mind and were probed if necessary. Given the limited number of actual asphyxiated babies, NeoNatalie anatomic models were used to assess providers' decision-making and ventilation skills in newborn resuscitation in a subset of countries: Madagascar, Rwanda and Tanzania. It is used to create scenarios of a baby with various breathing conditions. The chest of the simulator rises when the appropriate bag and mask ventilation technique is applied. A facility inventory was also conducted, and observations were made of key infrastructural components, storage capabilities, supplies and equipment essential for newborn healthcare services. Box 1 further describes the assessment tools and their purpose. Box 1 Survey tools and their purpose 1. Facility inventory checklist: assessed conditions of infrastructure, and availability and condition of commodities, supplies, and equipment. 2–3. Clinical practice observation tools for antenatal care (2) and labour and delivery (3): examined provider implementation of the evidence-based practices for routine care and screening, prevention, and management of major obstetric and newborn complications at the time of birth, including postpartum haemorrhage, pre-eclampsia/eclampsia, newborn asphyxia and respectful provider/client interactions. 4. Health worker interview and knowledge test: assessed health worker knowledge of how to identify, manage and treat common maternal and newborn health (MNH) complications, including newborn resuscitation skills. 5. Key informant interview and document review: reviewed national policies related to MNH care included the country's essential drug lists or formularies, clinical practice policies and guidelines, and curricula/syllabi on relevant topics for both preservice and in-service education for health professions. The antenatal care, labour and delivery tools were adapted from existing tested obstetric tools.22 Additional items were added based on the contents of the WHO's Managing Complications in Pregnancy and Childbirth (MCPC) document.23 The newborn resuscitation observation items were based on the Helping Babies Breathe algorithm (http://www.helpingbabiesbreathe.org). The tools were pretested at health facilities, and the validity and reliability of the tools were assessed during each of the data collectors training conducted in the study countries. Observation scores by trained observers were compared with trainers' scores, and inter-rater reliability was examined. The inventory tool was adapted from the DHS SPA tool, a well-tested tool.24 All the study tools can be accessed at http://www.mchip.net/QoCSurveys. Data collectors were trained to observe the entire labour and birthing process, starting from the first stage of labour through the first hour postdelivery for normal birth or until completion of the management of birth asphyxia or any maternal complication beyond the first hour. If data collectors did not arrive until the second stage of labour or during a newborn resuscitation, they were required to record from the time of their arrival to the same end point as stated above. This was an observational study only and no intervention was expected according to the study protocol. However, observers, who were all clinicians, were trained to first call for the service providers' supervisor if and when they witnessed any life-threatening actions by a health worker. If there was no supervisor or any other senior health worker available, they had the personal choice to step out of their data collector role and assist the provider in their professional role as a licenced and practising clinician in their respective countries. The decision was the data collector's and not a required intervention mandated by the study protocol. It was discussed with and accepted by the Johns Hopkins University Institutional Review Board (IRB). A local researcher supervised the survey in each country. Staff from MCHIP headquarters collaborated with the local researchers, local MCHIP staff, the Ministry of Health and other relevant country-specific organisations for planning and implementation. Skilled birth attendants, including doctors, nurses and midwives, were trained as data collectors by MCHIP staff who were doctors, nurses, midwives, and monitoring and evaluation specialists. During the 1–2 weeks training period, data collectors were introduced to the aim and objectives of the study, research tools, standardisation of clinical observation skills, and use of HTC smartphones or Samsung Galaxy tablets, including hands-on practice using one of these tools in the field. Data collectors worked in teams of three to four people over 2–4 days to collect data in each facility. To minimise bias, the data collectors did not visit health facilities where they were employed or supervised. Data were transmitted daily to a cloud-based, password-protected server, with data monitored and cleaned on an ongoing basis. The data entry system incorporated forced data quality checks before transmission to the central server. Both on-site and off-site monitoring of data flow was conducted given the two-directional capability of the electronic data processing system. Indicators and data analysis Evidence-based, globally accepted guidelines for clinical practice for immediate newborn care, described in the WHO's MCPC, were used as the gold standard for QoC.23 Indicators of quality of newborn care included (1) observation of ENC: dry baby with towel/cloth immediately, discard wet towel/cloth and cover with dry towel/cloth, place newborn skin-to-skin, tie or clamp cord when pulsations stop or 2–3 min after birth, cut cord with clean blade or scissors, help mother initiate breast feeding within first hour of birth, and practice infection prevention measures (hand washing and use of sterile gloves); (2) observations and/or simulations of neonatal resuscitation: dry, position, ventilate and adjust mask if needed; and (3) health worker knowledge of newborn care: immediate newborn care, signs of sepsis, and care of LBW babies and associated equipment and supplies. Indicators of facility readiness to provide newborn care included the percentage of facilities with the following equipment necessary for immediate newborn care: disposable cord ties/clamps, clean dry towel or blanket to wrap the baby, sterile/clean scissors or blade, bag and mask, suction device, and resuscitation table or firm surface for conducting newborn resuscitation. Descriptive analyses were conducted, including frequencies, means, and simple and cross-tabulations of quality and process indicators. The facility's readiness, mainly availability of equipment and supplies needed for immediate newborn care and resuscitation, was assessed. Data were disaggregated by type of facility to assess potential differences, which were displayed if significant. All analyses were conducted using Stata V.12 (StataCorp, College Station, Texas, USA). Ethical approval and informed consent Verbal informed consent was obtained from all study participants. As this was a minimal risk observational study conducted in settings with low literacy and obtained from women in active labour, verbal rather than written consent was used. In cases where a woman was ill or unable to give consent due to an emergency obstetric complication, the next of kin provided informed consent on her behalf. Results Characteristics of sampled health facilities and healthcare providers The assessments were conducted in a total of 643 health facilities with varying proportions of hospitals (referral level) and health centres. Table 1 shows the proportion of hospitals versus health centres sampled in each of the countries. There was a slightly bigger proportion of hospitals than health centres in the overall sample, ranging from 28% in Tanzania to 100% in Ethiopia. More health centres were sampled in Tanzania because of prioritisation of MCHIP project areas and focus on support to health centres with high delivery caseloads. Table 1 Number of facilities sampled per country Sample Ethiopia* Kenya Madagascar Mozambique Rwanda Tanzania Total Facilities visited 19 409 36 46 72 61 643 Hospitals 100% 52% 75% 46% 58% 28% 53% Health centres 0% 48% 25% 54% 42% 72% 47% *Only hospitals were sampled in Ethiopia due to low numbers of deliveries in lower level health facilities. A total of 2689 births were observed across the six countries. Of these, the study team observed newborn care for 2377 newborns and actual resuscitation of 230 newborns through the first hour after birth. Newborn care was not observed for some births for various reasons, including births that occurred when the observer was with another mother or had left the labour ward or facility. A total of 514 newborn resuscitation simulations were conducted. A knowledge test was administered to a total of 1016 healthcare providers to assess information used for decision-making related to newborn care. Information on their clinical qualifications, training and experience providing newborn care and supervision was also gathered (table 2). Table 2 Number of observations, knowledge tests and simulations per country Sample Ethiopia* Kenya Madagascar Mozambique Rwanda Tanzania Total Observations of care  Deliveries 192 626 347 525 293 706 2689  Newborn and postnatal care for mother and baby 115 571 336 508 225 622 2377  Newborn resuscitation 18 44 48 22 43 49 224 Knowledge and skills test administered  Health workers interviewed 79 210 139 186 145 257 1016  Newborn resuscitation simulations NA† NA† 132 NA† 137 245 514 *Only hospitals were sampled in Ethiopia due to low numbers of deliveries in lower level health facilities. †Newborn resuscitation simulations were not done in Kenya, Ethiopia and Mozambique. NA, not available. Table 3 illustrates the characteristics of 1016 service providers who were interviewed, by country. They comprised 93% skilled birth attendants, the majority of whom were nurses or midwives, with predominantly female interviewees in all countries. Table 3 Characteristics of interviewed health workers Interviewed health worker characteristics Ethiopia Kenya Madagascar Mozambique Rwanda Tanzania Pooled total Cadre*  Physician† 7.6% 4.3% 25.2% 0.5% 13.1% 3.5% 7.7%  Nurse/midwife‡ 78.5% 94.8% 69.8% 84.4% 86.2% 84.4% 85.1%  Student/unskilled§ 7.6% 1.0% 0.7% 9.7% 0.7% 6.2% 3.9%  Other/missing 6.3% 0.0% 4.3% 5.4% 0.0% 5.8% 3.2% Gender  Male 32.9% 33.8% 13.7% 0.5% 26.2% 5.4% 17.7%  Female 67.1% 66.2% 86.3% 98.9% 73.8% 91.8% 81.6%  Missing 0.0% 0.0% 0.0% 0.5% 0.0% 2.7% 0.7% Total interviews 79 210 139 186 145 257 1016 *Health cadres' definitions were provided by each country. †Doctor: general practitioners, obstetricians, gynaecologists and other specialists such as paediatricians, residents or assistant medical officers. ‡Nurse/midwife: Bachelor of Science and Diploma in Nursing/Midwifery, registered and enrolled nurses/midwives, nursing officers, maternal and child health aides, paramedics, or health officers. §Student/unskilled: medical attendants, health assistants or traditional birth attendants. Immediate ENC practices Results shown in figure 1 reveal areas of strength and weakness in performance of ENC practices by providers caring for the 2377 babies observed in the facilities surveyed. Coverage differed between and within countries. The most frequently observed practice, performed for 94% of newborns observed, was cutting the cord with a clean blade or scissors, followed by immediate drying with a towel/cloth (79%). However, only 43% of all mothers (ranging from 19% to 79%) were provided assistance to initiate breast feeding within the first hour of life, and only 45% of newborns (ranging from 10% to 59%) were placed skin-to-skin immediately after delivery. Figure 1 Observations of immediate essential newborn care. Use of high-level disinfected or sterile gloves for vaginal examinations during labour was high (93% of all observations) but hand washing prior to delivery was observed for only 37% of cases; providers were more likely to wash their hands after delivery (78%). The majority of facilities had soap and water (89%) and clean or sterile gloves (73%) available in the delivery area. Neonatal resuscitation Figure 2 illustrates providers' observed actions for 209 asphyxiated babies; 21 babies (9% of total cases) were excluded because they were classified as stillbirths or had missing data. Of the 209 asphyxiated babies with data, 106 received stimulation (ie, drying and/or back rubbing) and positioning of their heads to ensure the patency of their airways (left side of figure 2). The remaining 103 babies were not dried immediately, did not receive back rubs, or did not have their heads positioned to ensure their airways were opened. Forty-five out of the 106 (42%) babies who received stimulation and proper positioning of their necks recovered, while 61 were still not breathing normally. Fifty of these 61 received ventilation with bag and mask; of these, 45 recovered and 5 were declared dead. The remaining 11 (of 61) did not receive bag and mask ventilation; 8 started breathing spontaneously and 3 were declared dead. Figure 2 Neonatal resuscitation management tree for all countries. Thirty of the 103 babies who were not dried and/or received no back rubs or whose necks were not extended to open their airways (right side of figure 2) initiated breathing spontaneously, 1 was declared dead, and 72 continued to be asphyxiated. Forty-four of the 72 asphyxiated babies received bag and mask ventilation, and 36 recovered while 8 died. Of the remaining 26 who received no bag and mask ventilation, 21 recovered and 5 died. Overall, out of the 209 observed asphyxiated babies, 185 (89%) newborns recovered, 22 died and 2 had missing outcome data. Availability of ENC equipment and supplies Table 4 presents the availability of equipment needed for ENC and management of birth asphyxia. The presence of disposable cord ties or clamps was below 70% when pooled across all countries, with a range between 36.4% and 99.5%. The largest gaps were found for towels and blankets necessary to dry and wrap newborns (or cover them if placed skin-to-skin with their mothers) and promote thermoregulation, which were present in only 40.5% of all facilities, with a great deal of variability across countries (8.0–53.2%), apart from Rwanda, which proved to have a greater supply (80.6%). The majority of facilities were equipped with sterile scissors or blades crucial to maintaining hygienic cord care and preventing infections. Overall, the availability of all three ENC supplies was poor, at only 30.8% of all facilities (ranging from 8.2% to 52.7%). Results were similar at hospital and health centre levels. Table 4 Availability of selected equipment/supplies for immediate ENC and newborn resuscitation ENC and newborn resuscitation equipment/supplies Ethiopia Kenya Madagascar Mozambique Rwanda Tanzania Total Hospital Health centre Hospital Health centre Hospital Health centre Hospital Health centre Hospital Health centre Hospital Health centre Hospital Health centre Pooled Disposable cord ties or clamps 70% NA 99% 100% 39% 25% 73% 77% 59% 51% 46% 79% 68% 74% 70% Towel or blanket to wrap baby 46% NA 59% 15% 59% 28% 31% 30% 82% 71% 9% 6% 47% 22% 41% Sterile scissors or blade 94% NA 98% 100% 80% 100% 94% 89% 100% 100% 88% 85% 93% 90% 92% Bag and mask 90% NA NA NA 50% 13% 82% 80% 89% 68% 100% 55% 83% 61% 57% Suction device 88% NA 95% 71% 93% 53% 100% 100% 94% 83% 91% 76% 94% 82% 91% Resuscitation table for newborn 100% NA 65% 37% 43% 39% 94% 77% 74% 74% 88% 73% 75% 66% 72% NA, not available. Facilities exhibited high availability of suction devices (pooled mean=90.7%). Availability of resuscitation tables averaged 72.4% but varied across countries. Meanwhile, newborn-specific bag and mask availability was high in most facilities, particularly hospitals (81.9–100%)—apart from Madagascar, with a low of 13% for health centres—which brought the pooled average to 57.1%. It should be noted that differences were observed by the type of facility. Availability of resuscitation equipment was higher in hospitals in Tanzania and Madagascar than in health centres. All countries apart from Ethiopia also reported having tube and masks, although at a slightly lower availability than bag and masks (43.5%). The study assessed the availability of other newborn-related equipment and supplies, including the availability of weighing scales and antibiotics for newborn infection treatment. Almost all the facilities had weighing scales (pooled mean=95.4%), but only about 50% and 60% of facilities across all countries had adequate supplies of injectable ampicillin and gentamicin, and oral amoxicillin, respectively. Health worker knowledge Table 5 illustrates health workers' knowledge on four key areas related to newborn care. Gaps in knowledge on newborn care were observed in all countries and varied considerably for each indicator and country. Mean percentage scores for knowledge on basic equipment and supplies needed to provide immediate care after birth ranged from 25% to 50%, with great variations by country and type of equipment. Only 11% of respondents in Kenya mentioned all of the essential equipment, the highest percentage of all countries. A range of 36–62% in mean scores was found for health worker knowledge related to immediate newborn care practices. For knowledge on identification of newborn sepsis and care of LBW babies, the mean scores ranged from 27% to 65% and 32% to 54%, respectively. Most providers knew that LBW babies needed thermal protection but were less knowledgeable about the importance of feeding and ensuring infection prevention measures. For immediate newborn care, sepsis, and care of LBW babies, a higher percentage of health workers in Rwanda mentioned all relevant indicators (24%, 18% and 11%), although the percentages were still very low. Health workers in Kenya displayed greater overall knowledge in each category compared with other countries. Table 5 Individual components of health worker knowledge on newborn equipment and care Knowledge category and key equipment/practice Country (% health workers with number of correct answers) Basic equipment and supplies needed for immediate care after birth Ethiopia (N=76) Kenya (N=248) Mozambique (N=186) Rwanda (N=144) Tanzania (N=179) Two dry warm towels or cloths 55.0 74.0 78.0 36.0 52.0 Flat surface with warm cloth 17.0 75.0 22.0 29.0 12.9 Source of warmth—heat lamp 25.0 66.4 40.3 59.0 12.9 Self-inflating ventilation bag 49.0 28.4 29.0 55.0 29.1 Newborn face mask size 1 57.0 8.2 59.0 34.1 Newborn face mask size 0 38.0 43.0 21.8 Mucus extractor/suction/bulb syringe 88.0 41.8 43.5 76.0 44.7 Sterile or disinfected clamps, scissors, and cord ties 25.0 74.5 22.0 17.3 Sterile blade or scissors 66.7 Sterile or disposable cord ties/clamps 30.6 Clock or watch with seconds hand 13.0 26.4 20.4 14.0 3.4 Mean percentage score for category 41.0 49.3 38.0 44.0 25.3 Immediate newborn care Ethiopia (N=76) Kenya (N=248) Mozambique (N=186) Rwanda (N=144) Tanzania (N=250) Wipe face after birth 64.0 77.0 57.5 64.6 81.6 Conduct dry cord care (sterile cut, applying nothing to stump) 83.0 59.0 91 71.6 Cut cord with sterile blade/scissors 22.6 Conduct dry cord care (nothing applied to stump) 12.4 Ensure baby was breathing 36.0 49.0 38.2 50.7 47.6 Provide thermal protection 78.0 82.0 60.8 75 63.2 Initiate breast feeding within 1 hour 51.0 66.0 44.6 41.7 43.6 Assess/examine newborn within 1 hour 18.0 51.0 14.5 40.3 24.8 Provide eye prophylaxis 58.0 48.0 39.2 45.8 8.8 Mean percentage score for category 55.0 61.7 36.0 58.0 48.7 Signs of sepsis Ethiopia (N=78) Kenya (N=248) Mozambique (N=185) Rwanda (N=144) Tanzania (N=250) Poor/no breast feeding 67.0 70.7 48.6 57.6 62.4 Hypothermia/hyperthermia 81.0 88.5 88.9 69.6 Hypothermia 15.1 Hyperthermia 82.2 Restlessness/irritability 44.0 72.1 39.5 36.1 55.6 Breathing difficulties 45.0 29.8 41.1 49.3 49.2 Mean percentage score for category 59.0 65.3 27.0 58.0 59.2 Care of LBW newborns Ethiopia (N=78) Kenya (N=248) Mozambique (NA) Rwanda (N=144) Tanzania (N=248) Provide thermal protection 82.0 85.1 76.4 76.2 Provide extra support to mother to establish and maintain breast feeding 15.0 74.0 34.7 27.4 Monitor sucking capability 29.0 38.5 44.4 44.0 Monitor newborn closely for first 24 hours 15.0 33.2 36.1 33.5 Ensure infection prevention 17.0 48.2 31.3 24.2 Mean percentage score for category 32.0 54.0 45.0 41.0 The grey shaded boxes mean the question/item was not asked. LBW, low birthweight. In addition, due to limited numbers of actual asphyxiated newborn infants, NeoNatalie models were used to assess providers' skills in newborn resuscitation (n=514) in Madagascar, Rwanda and Tanzania. Demonstration of correct initial steps—including drying, rubbing of the back and extension of the neck—ranged from 52% to 72% of providers. Less than one-third (26–31%) of providers performed all steps without error to ventilate the NeoNatalie model using a bag and mask. Another 55–74% of providers readjusted the mask, checked airway positioning, or took further correct steps to improve ventilation. Discussion This assessment in six sub-Saharan African countries shows serious gaps in health facility readiness to provide quality newborn health services. The three main components of service delivery examined were (1) performance of key immediate ENC practices at hospitals and health centres, (2) health worker knowledge, and (3) the equipment and supplies to implement them. The study included an actual clinical observation of care provided to newborns during and immediately after birth. Almost all of the babies observed had their umbilical cord cut with clean or sterile scissors or blade, and about four out of five were dried immediately after birth. However, other essential care practices, including early initiation of breast feeding, skin-to-skin contact and delayed cord clamping, were performed for less than three-quarters of the observed deliveries. Studies have shown that early initiation of breast feeding and early skin-to-skin contact both contribute significantly to the maintenance of adequate thermal control and infection prevention in newborn infants.25 26 The earlier that breast feeding is initiated for a newborn, the lower the risk of newborn death.27 However, at an average of 45%, this was one of the least-performed newborn care practices by health workers in all the countries studied. A similar protective effect on mortality has been shown for hand washing by birth attendants.28 Unfortunately, hand washing in this study was also found to be low despite adequate availability of soap and water in delivery areas. Other studies have reported a much higher level of practice of some of these essential care practices in the study countries. The 2014–2015 Tanzania SPA reported skin-to-skin contact, early initiation of breast feeding, and drying and wrapping the baby as routine practice in over 90% of health facilities assessed.29 A similar finding was reported in Kenya in the 2010 SPA.30 However, the information was obtained through interviews with health facility staff and not actual clinical observation of their care practices. In addition to the routine newborn care practices reported by health facility staff in Tanzania and Kenya, the Kenya SPA reported on observed clinical newborn practices, information collected in collaboration with MCHIP reported in this paper. No SPA with information on maternal and newborn care services has been conducted for any of the remaining four study countries. The WHO's Service Availability and Readiness Assessment (SARA) tool does not include actual clinical observation for newborn services.31 Previous clinical observational studies in Ghana,15 the Philippines,8 Nepal32 and Bangladesh33 have also identified gaps in the quality of newborn healthcare. A health facility assessment in rural Ghana highlighted major gaps in newborn care equipment and quality of newborn care as well as the importance of quality improvement.15 A recent study in the Philippines also found that bathing of the baby in the first 24 hours was common and placing the baby skin-to-skin was infrequent.8 However, initiation of breast feeding was lower in this study (apart from in Kenya) than found in studies in the Philippines8 and Nepal.32 In addition, poor quality of neonatal care, specifically poor hygiene practices, lack of appropriate thermal care practices and substandard care of LBW babies were identified through a hospital survey in 18 hospitals in Bangladesh.33 Intrapartum-related complications is the second leading global cause of newborn deaths.1 Evidence shows that the majority of babies who experience difficulty in the initiation of breathing at birth can be saved with simple stimulation, such as drying and/or rubbing the baby's back.18 Of those newborns identified as requiring some form of resuscitation, 22% and 39% of them recovered through initial stimulation and bag and mask ventilation, respectively. Additionally, 28% recovered on their own without any intervention, and 11% died. Almost half of all of the observed babies who had difficulty initiating breathing at birth and required some form of resuscitation were not immediately dried, which could cause hypothermia and less responsiveness to resuscitation. When assessing the quality of resuscitation using the NeoNatalie model overall, less than a third of providers ventilated without error. The percentage of providers who correctly stimulated and made proper adjustments including repositioning the neck and the mask to facilitate breathing ranged from 50% to 75%. Every birth that occurs at a health facility that provides delivery services should have staff competent to resuscitate newborn infants who do not breathe at birth. Maintenance of resuscitation skills requires ongoing practice and periodic refresher training through on-site and off-site courses and mentorship. However, many service providers do not receive this type of support. In the Tanzania 2014–2015 SPA, only 33% of staff reported having received refresher training on newborn resuscitation in the 24 months prior to the assessment, and 42% have ever received training on newborn resuscitation.29 Data from the most recent SPA health facility surveys conducted in Ghana,34 Rwanda35 and Uganda36 indicate that only 2–12% of those responsible for conducting deliveries were trained in neonatal resuscitation, and only 8–22% of the facilities where deliveries were performed had the proper equipment for resuscitation. A study in the Philippines found that <50% of staff were trained in neonatal and paediatric resuscitation.8 With the exception of thermal care, health workers' knowledge related to immediate newborn care and associated equipment was poor and differed greatly among countries; mean knowledge per country for each category ranged roughly from about a quarter to two-thirds of providers. A study in 21 hospitals in 7 countries in sub-Saharan Africa and South Asia also found inadequate knowledge among doctors, nurses and medical assistants, particularly with regard to sepsis.9 Availability of essential newborn health equipment and other supplies, including bags and masks, was less of an issue in the assessed facilities. Most facilities had some supplies of newborn resuscitation equipment: suction, face mask and ventilation bag. The majority of facilities were equipped with sterile scissors or blades crucial to maintaining hygienic cord care and preventing infections, but, overall, the availability of all ENC supplies was poor. The largest gaps were found in the supply of towels and blankets necessary to dry and wrap newborns and promote thermoregulation. A study in Kenya14 that surveyed eight large district hospitals to assess the structural components of QoC found that equipment was not always available and that even large hospitals experienced poor QoC. Similarly, this study found that gaps were observed irrespective of the type of facility. Lack of infrastructural components and resources can negatively influence community perceptions of quality and usage of care, as found in Nigeria,37 and can further weaken linkages between the community and facility.14 Strengths and limitations of the study This paper describes directly observed quality of newborn care at hospitals and health centres through a facility assessment using a robust methodology in six countries in sub-Saharan Africa. It provides new information on the frequency with which evidence-based practices are performed according to international standards and guidelines. The use of direct observation in this study is a major strength. Direct observations of labour and delivery care are considered the gold standard in low-income and middle-income settings and more reliable than chart reviews, provider interviews or client exit interviews for assessing provider performance. However, due to cost and time requirements, observations are rarely conducted as part of health facility assessments, especially observation of labour and delivery care, and even more infrequently for management of newborn complications such as birth asphyxia. However, the study had limitations. Settings and facility samples differed greatly across countries, making it difficult to make comparisons. However, the purpose of the study was to inform country-specific quality improvement and policy development efforts, not to make cross-country comparisons. At times as the study evolved, survey questions differed slightly between countries, which limited samples for certain analyses. The availability of equipment was based on point prevalence at the time of observation and does not necessarily imply a constant supply, but this was the best possible assessment given the resources. Simulated resuscitation results may have been affected by previous exposure or non-exposure of health workers to the NeoNatalie model. Observations could have been biased as a result of the Hawthorne Effect, meaning that providers could have changed their behaviours because they knew that they were being watched. Additionally, observational data are based on the judgements of observers and could therefore result in some biases. The study did not include any tools to assess user perspectives on QoC, such as facility exit interviews or focus groups with recently delivered women in the community, as this was beyond the scope of the current study. Women's perceptions of QoC at facilities and their satisfaction with services are important determinants of whether or not women deliver at a facility, relevant for future investigation. Implications and next steps Many countries are implementing strategies to increase the percentage of pregnant women who deliver in health facilities as an intervention crucial to reducing maternal and newborn mortality. However, as this assessment shows, the care provided at most facilities is not optimal. These study findings will contribute to the evidence base on the quality of newborn facility care, calling attention to the issue and helping policymakers and stakeholders make informed decisions on the type of support needed for provision and maintenance of quality ENC services in health facilities. The hope is that such a study can help in closing the gap between policy and implementation. In order for countries to achieve the targets of the Sustainable Development Goals for ending preventable newborn and maternal deaths, it is essential that health workers at health facilities have the competencies required to provide quality ENC in a timely manner, with systems in place to facilitate implementation and maintain quality. Health facility assessments using tools such as the SPA and SARA are crucial first steps in identifying and then closing existing gaps in the quality of newborn care at health facilities. However, these assessments are usually conducted without observation of the actual provision of care by service providers, which might lead to some gaps being missed. Clinical observation data could be collected as part of such formal assessments or periodically as part of ongoing regular supportive supervision or at sentinel surveillance sites. Whether identified through interviews with health workers or observations of actual performance of ENC practices by health workers, gaps in the quality of newborn health services have to be closed to ensure that newborn infants receive optimal care during delivery. Comparison of the MDG 2010 and 2015 reports shows significant reduction in NMRs for all the six countries assessed (Ethiopia: 39–28 per 1000 live births, Kenya: 33–22 per 1000, Madagascar: 35–20 per 1000, Mozambique: 43–27 per 1000, Rwanda: 35–19 per 1000 and Tanzania: 33–19 per 1000).38 39 The contribution of actual improvement in newborn care practices by service providers in health facilities to the NMR reduction in these countries is unknown since the clinical observational assessments have not been repeated. Many countries already recognise the need to address the quality of maternal and newborn care in health facilities and are taking actions to address these concerns. There are also existing global initiatives, such as the ENAP,3 the Global Strategy for Women's, Children's and Adolescents' Health40 and the Mother-Baby Friendly Birthing Facilities Initiative,41 as well as new initiatives to encourage government accountability; these are not limited to increasing supply and demand for care but also focus on ensuring equitable care, an often neglected yet a critical objective of accountability.42 These initiatives provide frameworks for countries to assess gaps in their maternal, newborn and child health policies; service guidelines; and implementation plans, and to examine how they help health workers acquire, maintain and use their competencies. To further help countries improve the quality of their maternal and newborn health (MNH) services, the WHO recently released its Standards for improving quality of maternal and newborn care in health facilities.43 This document provides a framework with eight domains of QoC and associated performance indicators that countries could use to assess, improve and monitor the MNH services provided at health facilities. The implementation of this framework has been supported by a strong advocacy effort calling for quality, equity and dignity for all women and newborns.44 It is essential that quality improvement approaches be endorsed by national health authorities, integrated into existing systems and supported with strong policies in order to be effective, efficient and aligned with global initatives.45 46 The findings and lessons learnt from this study need to be taken forward as part of this WHO-led initiative and will be critical for advancing and supporting ministries of health to harmonise and lead quality improvement approaches across the health system. We hope that this paper will influence this important mechanism. Conclusion Results of the health facility assessments presented in this paper have highlighted major gaps in facility readiness and quality of immediate newborn care in six sub-Saharan African countries through observations of clinical care provided. There is a need for such studies to establish a better understanding of QoC practices for women and newborns at health facilities as countries seek to end preventable maternal and child deaths. Better understanding of the QoC will lead to improved supportive and enabling strategies and actions, such as on-site mentorship and skills updates, to close the gap in the delivery of optimal facility newborn care as women respond to the call to deliver in health facilities. It will require the continued leadership from ministries of health at all levels (national, regional and district) and collaboration with donors, international and local implementing partners, professional associations, preservice institutions, beneficiaries, and other stakeholders to ensure health workers are competent in knowledge and skills and that essential equipment and supplies are available to guarantee all newborns receive the essential care needed to enable their survival. The authors wish to thank all the interviewers, observers, service providers and patients in the six study countries for their work and contribution to making the data available. Contributors: BR designed the study, and BR, RK, AG, AT, JPR and GT conducted the study. The newborn component of the study was conceived by BR, JdG-J, HER, BR, SA and GM. The paper was conceptualised by LV, JdG-J, HER, BR, SA, NK, GM, RB and WM. LV drafted the paper. All authors reviewed the manuscript multiple times, provided feedback and gave final approval before submission. Funding: The study was funded by United States Agency for International Development, under the terms of the Leader with Associates Cooperative Agreement Number GHS-00-08-00002-00. The corresponding author had full access to all data and, together with the second author, the final responsibility to submit for publication. Competing interests: None declared. Ethics approval: The study was reviewed and approved by the Institutional Review Board (IRB) at Johns Hopkins Bloomberg School of Public Health (IRB number: 00002549) as well as by the ethical review boards in all study countries and by all partners. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Data are available on request. ==== Refs References 1 UNICEF . Child Mortality Estimates . Secondary Child Mortality Estimates 2015. http://www.childmortality.org/index.php?r=site/index . 2 Save the Children . Ending newborn deaths: ensuring every baby survives . London, UK : Save the Children , 2014 . 3 WHO, UNICEF . Every newborn: an action plan to end preventable deaths . Geneva : World Health Organization , 2014 . 4 Chou D , Daelmans B , Jolivet RR Ending preventable maternal and newborn mortality and stillbirths . BMJ 2015 ;351 :h4255 10.1136/bmj.h4255 26371222 5 Darmstadt GL , Bhutta ZA , Cousens S Evidence-based, cost-effective interventions: how many newborn babies can we save? Lancet 2005 ;365 :977 –88 . 10.1016/S0140-6736(05)71088-6 15767001 6 Wall SN , Lee AC , Niermeyer S Neonatal resuscitation in low-resource settings: what, who, and how to overcome challenges to scale up? Int J Gynaecol Obstet 2009 ;107 (Suppl 1) :S47 –62 , S63–4 10.1016/j.ijgo.2009.07.013 19815203 7 The Partnership for Maternal, Newborn and Child Health . A global review of the key interventions related to reproductive, maternal, newborn and child health . Geneva : PMNCH , 2011 . 8 Sobel HL , Silvestre MA , Mantaring JB III Immediate newborn care practices delay thermoregulation and breastfeeding initiation . Acta Paediatr 2011 ;100 :1127 –33 . 10.1111/j.1651-2227.2011.02215.x 21375583 9 Nolan T , Angos P , Cunha AJ Quality of hospital care for seriously ill children in less-developed countries . Lancet 2001 ;357 :106 –10 . 10.1016/S0140-6736(00)03542-X 11197397 10 Althabe F , Bergel E , Cafferata ML Strategies for improving the quality of health care in maternal and child health in low- and middle-income countries: an overview of systematic reviews . Paediatr Perinat Epidemiol 2008 ;22 (Suppl 1 ):42 –60 . 10.1111/j.1365-3016.2007.00912.x 18237352 11 Fauveau V , de Bernis L “Good obstetrics” revisited: too many evidence-based practices and devices are not used . Int J Gynaecol Obstet 2006 ;94 :179 –84 . 10.1016/j.ijgo.2006.05.020 16844125 12 van den Broek NR , Graham WJ Quality of care for maternal and newborn health: the neglected agenda . BJOG 2009 ;116 (Suppl 1 ):18 –21 . 10.1111/j.1471-0528.2009.02333.x 19740165 13 English M , Ntoburi S , Wagai J An intervention to improve paediatric and newborn care in Kenyan district hospitals: understanding the context . Implement Sci 2009 ;4 :42 10.1186/1748-5908-4-42 19627588 14 Opondo C , Ntoburi S , Wagai J Are hospitals prepared to support newborn survival? An evaluation of eight first-referral level hospitals in Kenya . Trop Med Int Health 2009 ;14 :1165 –72 . 10.1111/j.1365-3156.2009.02358.x 19695001 15 Vesel L , Manu A , Lohela TJ Quality of newborn care: a health facility assessment in rural Ghana using survey, vignette and surveillance data . BMJ Open 2013 ;3 :e002326 10.1136/bmjopen-2012-002326 16 Dickson KE , Simen-Kapeu A , Kinney MV Every newborn: health-systems bottlenecks and strategies to accelerate scale-up in countries . Lancet 2014 ;384 :438 –54 . 10.1016/S0140-6736(14)60582-1 24853600 17 Liu L , Oza S , Hogan D Global, regional, and national causes of under-5 mortality in 2000–15: an updated systemic analysis with implications for the Sustainable Development Goals . Lancet 2016 ;388 :3027 –35 . 10.1016/S0140-6736(16)31593-8 27839855 18 Lee AC , Cousens S , Wall SN Neonatal resuscitation and immediate newborn assessment and stimulation for the prevention of neonatal deaths: a systematic review, meta-analysis and Delphi estimation of mortality effect . BMC Public Health 2011 ;11 (Suppl 3 ):S12 10.1186/1471-2458-11-S3-S12 19 Newton O , English M Newborn resuscitation: defining best practice for low-income settings . Trans R Soc Trop Med Hyg 2006 ;100 :899 –908 . 10.1016/j.trstmh.2006.02.012 16757009 20 Duysburgh E , Zhang WH , Ye M Quality of antenatal and childbirth care in selected rural health facilities in Burkina Faso, Ghana and Tanzania: similar finding . Trop Med Int Health 2013 ;18 :534 –47 . 10.1111/tmi.12076 23398053 21 USAID/MCHIP . Introduction to the maternal and newborn quality of care surveys . Baltimore, MA : Jphieg , 2013 . 22 Columbia University . AMDD Needs assessment of emergency obstetric and newborn care. Obstetric facility assessment tool . New York 2010 . 23 WHO . Integrated management of pregnancy and childbirth. Managing complications in pregnancy and childbirth: a guide for midwives and doctors . Geneva, Switzerland : World Health Organization , 2007 . 24 ICF/Macro . Service Provision Assessment survey tool. http://www.measuredhs.com/What-We-Do/Survey-Types/SPA.cfm . 25 Conde-Agudelo A , Diaz-Rossello JL Kangaroo mother care to reduce morbidity and mortality in low birthweight infants . Cochrane Database Syst Rev 2014 ;(4) :CD002771 10.1002/14651858.CD002771.pub3 24752403 26 Edmond KM , Zandoh C , Quigley MA Delayed breastfeeding initiation increases risk of neonatal mortality . Pediatrics 2006 ;117 :e380 –6 . 10.1542/peds.2005-1496 16510618 27 NEOVITA Study Group . Timing of initiation, patterns of breastfeeding, and infant survival: prospective analysis of pooled data from three randomised trials . Lancet Glob Health 2016 ;4 :e266 –75 . 10.1016/S2214-109X(16)00040-1 27013313 28 Rhee V , Mullany LC , Khatry SK et al Impact of maternal and birth attendant hand-washing on neonatal mortality in Southern Nepal . Arch Pediatr Adolesc Med 2008 ;162 :603 –8 . 10.1001/archpedi.162.7.603 18606930 29 Ministry of Health and Social Welfare (MoHSW) . [Tanzania Mainland], Ministry of Health (MoH) [Zanzibar], National Bureau of Statistics (NBS), Office of the Chief Government Statistician (OCGS), and ICF International 2015. Tanzania Service Provision Assessment Survey 2014–15 . Rockville, MA, USA : MoHSW, MoH, NBS, OCGS, and ICF International http://dhsprogram.com/publications/publication-SPA22-SPA-Final-Reports.cfm . 30 National Coordinating Agency for Population and Development . (NCAPD) [Kenya], Ministry of Medical Services (MOMS) [Kenya], Ministry of Public Health and Sanitation (MOPHS) [Kenya], Kenya National Bureau of Statistics (KNBS) [Kenya], ICF Macro. 2011. Kenya Service Provision Assessment Survey 2010 . Nairobi, Kenya : National Coordinating Agency for Population and Development, Ministry of Medical Services, Ministry of Public Health and Sanitation, Kenya National Bureau of Statistics, and ICF Macro http://dhsprogram.com/publications/publication-SPA17-SPA-Final-Reports.cfm . 31 Service Availability and Readiness Assessment . (SARA): an annual monitoring system for service delivery reference manual. Version 2.2 2015 http://www.who.int/healthinfo/systems/sara_reference_manual/en/ . 32 Khanal V , Gavidia T , Adhikari M Poor thermal care practices among home births in Nepal: further analysis of Nepal Demographic and Health Survey 2011 . PLoS ONE 2014 ;9 :e89950 10.1371/journal.pone.0089950 24587145 33 Hoque DM , Rahman M , Billah SM An assessment of the quality of care for children in eighteen randomly selected district and sub-district hospitals in Bangladesh . BMC Pediatr 2012 ;12 :197 10.1186/1471-2431-12-197 23268650 34 Ghana Statistical Service (GSS), Health Research Unit, Ministry of Health, and ORC Macro . Ghana Service Provision Assessment Survey 2002 . Calverton, MA : Ghana Statistical Service and ORC Macro , 2003 http://dhsprogram.com/publications/publication-SPA6-SPA-Final-Reports.cfm 35 National Institute of Statistics . (NIS) [Rwanda], Ministry of Health (MOH) [Rwanda], and Macro International Inc. 2008. Rwanda Service Provision Assessment Survey 2007 . Calverton, MA, USA : NIS, MOH, and Macro International Inc http://dhsprogram.com/publications/publication-SPA15-SPA-Final-Reports.cfm 36 Ministry of Health . (MOH) [Uganda] and Macro International Inc. 2008. Uganda Service Provision Assessment Survey 2007 . Kampala, Uganda : Ministry of Health and Macro International Inc http://dhsprogram.com/publications/publication-SPA13-SPA-Final-Reports.cfm . 37 Uzochukwu BS , Onwujekwe OE , Akpala CO Community satisfaction with the quality of maternal and child health services in southeast Nigeria . East Afr Med J 2004 ;81 :293 –9 . 10.4314/eamj.v81i6.9178 16167676 38 World Health Organization and UNICEF . Countdown to 2015 decade report (2000–2010): taking stock of maternal, newborn and child survival . Geneva, Switzerland : World Health Organization , 2010 http://www.who.int/pmnch/topics/child/CountdownReportOnly.pdf?ua=1 . 39 World Health Organization and UNICEF . Countdown to 2015: a decade of tracking progress for maternal, newborn and child survival. The 2015 Report . Geneva, Switzerland: World Health Organization , 2015 http://countdown2015mnch.org/documents/2015Report/Countdown_to_2015_final_report.pdf . 40 United Nations . Survive, thrive, transform. The global strategy for women's, children's and adolescent's health . New York : United Nations , 2015 . 41 Lalonde AB , Miller S Mother-baby friendly birthing facilities initiative . Int J Gynecol Obstet 2015 ;128 :93 –4 . 10.1016/j.ijgo.2014.11.002 42 Barroso C , Lichuma W , Mason E Accountability for women's, children's and adolescents’ health in the Sustainable Development Goal era . BMC Public Health 2016 ;16 (Suppl 2 ):799 10.1186/s12889-016-3399-9 43 World Health Organization . Standards for improving quality of maternal and newborn care in health facilities 2016 http://www.who.int/maternal_child_adolescent/documents/improving-maternal-newborn-care-quality/en/ . 44 Kinney MV , Boldosser-Boesch A , McCallon B Quality, equity, and dignity for women and babies . Lancet 2016 ;388 :2066 –8 . 10.1016/S0140-6736(16)31525-2 .27642024 45 Milner KM , Duke T , Bucens I Reducing newborn mortality in the Asia-Pacific region: quality hospital services and community-based care . J Paediatr Child Health 2013 ;49 :511 –18 . 10.1111/jpc.12249 23713996 46 Tamburlini G , Yadgarova K , Kamilov A Improving the quality of maternal and neonatal care: the role of standard based participatory assessments . PLoS ONE 2013 ;8 :e78282 10.1371/journal.pone.0078282 24167616

PMC005xxxxxx/PMC5372101.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01395310.1136/bmjopen-2016-013953HIV/AIDSResearch15061842184316921724Incidence and prevalence of type 2 diabetes mellitus with HIV infection in Africa: a systematic review and meta-analysis Prioreschi A 1Munthali R J 1Soepnel L 12Goldstein J A 3Micklesfield L K 1Aronoff D M 4Norris S A 1 1 MRC/Wits Developmental Pathways for Health Research Unit, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa 2 Julius Global Health, Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht, The Netherlands 3 Department of Pathology, Microbiology and Immunology, Vanderbilt University Medical Center, Nashville, Tennessee, USA 4 Division of Infectious Disease, Department of Medicine, Vanderbilt University Medical Center, Nashville, Tennessee, USACorrespondence to Dr Alessandra Prioreschi; Alessandra.prioreschi@wits.ac.za2017 29 3 2017 7 3 e01395319 8 2016 1 12 2016 6 12 2016 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives This systematic review aims to investigate the incidence and prevalence of type 2 diabetes mellitus (T2DM) in patients with HIV infection in African populations. Setting Only studies reporting data from Africa were included. Participants A systematic search was conducted using four databases for articles referring to HIV infection and antiretroviral therapy, and T2DM in Africa. Articles were excluded if they reported data on children, animals or type 1 diabetes exclusively. Main outcome measures Incidence of T2DM and prevalence of T2DM. Risk ratios were generated for pooled data using random effects models. Bias was assessed using an adapted Cochrane Collaboration bias assessment tool. Results Of 1056 references that were screened, only 20 were selected for inclusion. Seven reported the incidence of T2DM in patients with HIV infection, eight reported the prevalence of T2DM in HIV-infected versus uninfected individuals and five reported prevalence of T2DM in HIV-treated versus untreated patients. Incidence rates ranged from 4 to 59 per 1000 person years. Meta-analysis showed no significant differences between T2DM prevalence in HIV-infected individuals versus uninfected individuals (risk ratio (RR) =1.61, 95% CI 0.62 to 4.21, p=0.33), or between HIV-treated patients versus untreated patients (RR=1.38, 95% CI 0.66 to 2.87, p=0.39), and heterogeneity was high in both meta-analyses (I2=87% and 52%, respectively). Conclusions Meta-analysis showed no association between T2DM prevalence and HIV infection or antiretroviral therapy; however, these results are limited by the high heterogeneity of the included studies and moderate-to-high risk of bias, as well as, the small number of studies included. There is a need for well-designed prospective longitudinal studies with larger population sizes to better assess incidence and prevalence of T2DM in African patients with HIV. Furthermore, screening for T2DM using gold standard methods in this population is necessary. Trial registration number PROSPERO42016038689. Type 2 diabetes mellitusHIVAfricacombination antiretroviral therapyincidenceprevalence ==== Body Strengths and limitations of this study This is the first systematic review of the literature examining associations between HIV infection and treatment with type 2 diabetes mellitus (T2DM) incidence and prevalence in Africa. The stringent inclusion criteria used is a strength of this systematic review. Differences in methods of T2DM diagnosis across studies is a limitation. Heterogeneity and moderate-to-high risk of bias across studies is a limitation. The small number of studies meeting the inclusion criteria is a limitation. Background The introduction of combination antiretroviral therapies (cARTs) in the treatment of HIV infection has resulted in significant extension of the predicted lifespan of patients with HIV infection.1 Consequently, patients with HIV are potentially at a greater risk of developing non-communicable diseases than due to the ageing process alone; as the disease itself,2 and treatments used to combat HIV, are associated with metabolic complications.3 Type 2 diabetes mellitus (T2DM) is one such disease that is becoming increasingly common, specifically in Africa due to rapidly transitioning lifestyles. An estimated 12.1 million people were living with T2DM in Africa in 20104 and it is predicted that this will increase to 23.9 million by 2030. Besides associations with age, obesity, sex and race,5 recent studies have associated T2DM with HIV infection, and with cART.1 3 5 The mechanisms underlying these associations are not fully elucidated, but may reflect chronic systemic inflammation in response to HIV infection despite treatment,6 7 antiretroviral drug-induced mitochondrial dysfunction, lipodystrophy and comorbidities.5 Conversely, some studies have shown a decreased incidence of T2DM in HIV-infected individuals compared with uninfected individuals.8 9 T2DM is associated with increased morbidity and mortality, an estimated 1.5 million deaths were attributed directly to T2DM in 2012,10 and the implications of HIV infection and treatment on the incidence of T2DM is therefore important to explore. The aim of this systematic review is to investigate the incidence of T2DM in patients with HIV infection in Africa, as well as, the prevalence of T2DM in patients with HIV infection treated with cART in comparison with non-infected and non-treated individuals. Methods The systematic review focused on the associations between HIV infection, antiretroviral therapy and T2DM. This review was registered in the PROSPERO registry for systematic reviews (registration number 42016038689),11 and was conducted in accordance with the PRISMA guidelines.12 Search strategy The search for this systematic review was conducted in May 2016 and included terms in the determinants of HIV infection and antiretroviral therapy, the domain of Africa and the outcome of T2DM. Restrictions included age (>13 years), date of publication (after 1 January 2008 due to the presence of an existing review examining prevalence of T2DM in HIV conducted in 200813). The title and abstracts of articles in PubMed, Scopus, the Cochrane library and Embase were searched; and a sample of the Embase search strategy is available online as a supplementary file. Keywords used included: ‘HIV’, ‘diabetes’, ‘Africa’ and ‘antiretroviral therapy’. 10.1136/bmjopen-2016-013953.supp1supplementary file Study selection All observational studies (cohort, case–control and cross sectional) that assessed the relationship between HIV seropositivity with or without cART therapy, and T2DM in Africa were included. Animal studies, biomolecular studies, studies not written in English or French, case reports and secondary analyses were excluded. Studies reporting outcomes in children or pregnant women, or reporting type 1 diabetes outcomes only, or not reporting T2DM incidence or prevalence (but hyperglycaemia or impaired glucose tolerance for example) were also excluded. Studies that did not report prevalence of T2DM in HIV-infected participants compared with HIV-uninfected participants; or prevalence of T2DM between cART exposure compared with untreated patients with HIV; or incidence of T2DM in patients with HIV infection were excluded. Authors of individual studies defined the criteria for T2DM diagnoses, and variant criteria were included provided diagnosis was made using a recognised score for a fasted blood glucose, or an oral glucose tolerance test (OGTT), or glycated haemoglobin (HbA1c) values.14 Screening and data extraction Two independent reviewers (AP, RJM) independently screened all articles retrieved by the search strategy by title and abstract for eligibility according to inclusion and exclusion criteria. Any discrepancies between the two reviewers were discussed and consensus was reached. The full text was accessed if necessary for further clarification. Full texts of eligible articles were then retrieved and divided among all reviewers. If no full text was available, one attempt was made to contact the author. Each full text was assessed for eligibility by one reviewer, and a second reviewer was available for consultation. Data extraction was then performed using a standardised data extraction form. One reviewer (AP) reassessed data extraction for all eligible full texts. Data of interest was study design, study setting and country, population, age, body mass index (BMI), number of patients included in each group, control population, cART treatment at the time of inclusion, duration of cART treatment, method of T2DM diagnosis, incidence of known risk factors for T2DM such as obesity, treatment provided for T2DM, incidence of T2DM in the control group and group with HIV and/or antiretroviral therapy, when applicable OR/risk ratio (RR) and follow-up duration. In cases of incomplete data, one attempt was made to contact the corresponding author by email and if no response was received the paper was excluded. Data synthesis Three separate analyses were performed for articles that examined incidence of T2DM; prevalence of T2DM in HIV-infected versus uninfected participants; and prevalence of T2DM in HIV-infected and treated versus untreated participants. Meta-analysis was conducted for articles with sufficiently homogenous outcome measures and study designs. The principle summary measure used was RR, and in cases of substantial heterogeneity (I2>50%), according to the Cochrane handbook,15 a binary random effects model (using the DerSimonian-Laird method) was applied. Analyses were conducted using OpenMetaAnalyst. A priori subgroup analyses based on geographical localisation, age, antiretroviral therapy (ART) medication/treatment strategy and duration, severity of HIV and method of T2DM diagnosis were not possible due to insignificant subcategorisation of data and insufficient number of included studies.15 Risk of bias assessment Studies were assessed for risk of bias using the Evidence Partner's risk of bias tool for cohort studies16 as ‘low risk’, ‘medium–low risk’, ‘medium–high risk’, ‘high risk’ or ‘not applicable’ for the categories of: similarity of intervention, adequacy of follow-up, assessment of outcome, assessment of prognostic factors, matching relevant variables between case and control, presence of outcome of interest at start of the study, assessment of exposure and selection of populations. This tool is available as an online supplementary file. 10.1136/bmjopen-2016-013953.supp2supplementary file Results The search provided 1056 results. After screening, 20 articles met the eligibility criteria17–36 and were included in the analysis (figure 1). Of these, seven17–23 articles reported incidence of T2DM in HIV-infected participants, eight24–31 reported prevalence of T2DM in HIV-infected participants compared with uninfected controls and five32–36 reported prevalence of T2DM in HIV-infected participants on treatment compared with untreated controls. In included studies, T2DM was diagnosed if participants were being treated for T2DM, or by OGTT. As summarised in table 1, four main criteria were used: WHO, American Diabetes Association (ADA), International Diabetes Federation (IDF) and National Cholesterol Education Programme (NCEP) criteria. Table 1 Overview of diagnostic criteria used in the included studies Criteria used Definitions WHO Fasting plasma glucose ≥7.0 mmol/L (126 mg/dL) or 2–h plasma glucose ≥11.1 mmol/L (200 mg/dL). ADA Fasting plasma glucose ≥126 mg/dL (7.0 mmol/L) or 2-hour plasma glucose ≥200 mg/dL (11.1 mmol/L) during OGTT (75 g) or A1C≥6.5% (48 mmol/mol) or Random plasma glucose ≥200 mg/dL (11.1 mmol/L) NCEP cut-offs Fasting plasma glucose ≥5.6 mmol/L IDF FPG≥100 mg/dl (5.6 mmol/L) A1C, glycated haemoglobin; ADA, American diabetes association; IDF, International Diabetes Federation; FPG, fasting plasma glucose; NCEP, National Cholesterol Education Programme; OGTT, oral glucose tolerance test. Figure 1 Flow diagram of article selection process and reasons for inclusion and exclusion. Risk of bias A summary of the risk of bias assessment is presented in figure 2. All included studies were observational, and 15 (71%) were case–control studies. In 5% of studies, there was a high risk of bias due to HIV treatment not being stated. In 25% of studies, there was a medium–high risk of bias due to confounding variables. Four (20%) of the studies had medium–low risk of bias due to T2DM diagnosis criteria. Three (14%) included studies were published conference proceedings. Figure 2 Risk of bias assessment for studies included in the analysis. Incidence of T2DM in HIV-infected participants Seven studies reported T2DM incidence in HIV-infected participants (n=57 006; table 2). One of the included studies compared incidence in treated versus untreated participants,21 and another compared incidence in infected versus uninfected participants.20 The rest of the studies assessed incidence in HIV-infected and treated participants with no control group. Most participants were on cART, except for participants in the Sagna et al22 study who were on first-line therapy, which was not clearly specified. Mean age of participants ranged from 33.5 years17 to 38 years19 23 (age was not stated by Magula et al20). Mean BMI ranged from 19.2 kg/m223 to 27.9 kg/m2 ,17 and was not stated in one of the studies.20 Mean duration of follow-up ranged from 1.56 years19 to 5.5 years,17 and the total number of participants followed up to completion was n=56 875. The majority of participants were women in all studies where sex was stated. Incidence of T2DM was reported as absolute incidence, cumulative incidence, incidence proportion and incidence rate per 1000 person years. Incidence rates ranged from 423 to 59,20 figure 3. The combined incidence rate for all the included studies over 89 640 person years of follow-up was 17.4. Table 2 Incidence data Author, year Setting Population Case Control ART Follow-up mean/median Diagnosis of T2DM Prevalence at baseline n (%) Prevalence at follow-up n (%) Cumulative incidence Incidence proportion Incidence rate (per 1000 person years) p Value Abrahams, 2015 South Africa 103 women Mean age=33.5 Mean BMI=27.9 NA NA Stavudine/lamivudine 5.5 years n=94 ADA criteria 1 (1.0) 7 (7.5) 6.5% 5.83% 11 0.07 George, 2009 South Africa 42 black participants, 65% women Mean age=34.4 Mean BMI=22.7 NA NA Stavudine/zidovudine 2 years n=42 NCEP cut-off 1 (2.4) 1(2.5) 0.1% 0.001% 5 >0.05 Karamchand, 2016 South Africa 56 298 participants, 64% women Mean age=38.14 Mean BMI=25.95 NA NA First line NNRTI regimen containing efavirenz or nevirapine 1.56 years n=56 298 Prescription of anti-diabetic medication 0 (0) 1500 (2.66) 2.66% 2.66% 13 Not reported Magula, 2014 South Africa 238 participants n=150 treated n=88 uninfected Initiated—tenofovir, lamivu- dine, efavirenz/nevirapine 2 years n=150 WHO criteria 0 (0) 13 (8.66) 8.66% 8.66% 59 Not reported Ndona, 2012 DRC 102 participants, 51% women, Mean age=43.4 Mean BMI=23.1 n=49 HIV+ treated n=53 HIV+ untreated Stavudine + lamivudine, zidovudine + lamivudine + nevirapine, or efavirenz 4 years n=102 WHO criteria Not stated 5 (4.9) 4.9% 4.9% 10 0.06 Sagna, 2013 Burkino Faso 144 participants, Mean age=37 NA NA Not stated (first-line therapy) 3 years n=128 Not stated Not stated 3 (2.3) 2.3% 2.1% 7 Not reported Zannou, 2009 Benin 79 participants, 59.5% women Mean age=38 Mean BMI=19.2 NA NA All started combination therapy. Lamivudine + stavudine+efavirenz 2 years n=61 WHO criteria 0 (0) 6 (7.6) 7.6% 7.6% 4 Not reported ADA, American Diabetes Association; ART, antiretroviral therapy; BMI, body mass index; DRC, Democratic Republic of Congo; NCEP, National Cholesterol Education Programme; NNRTI, non-nucleotide reverse transcriptase inhibitors; T2DM, type 2 diabetes mellitus. Figure 3 Incidence rates of T2DM in HIV-infected and treated participants in Africa. T2DM, type 2 diabetes mellitus. Prevalence of T2DM in HIV-infected compared with uninfected participants Table 3 shows the data for eight studies included in a meta-analysis comparing HIV-infected (n=1715) with uninfected participants (n=2853). The majority of included participants were women, except for the study conducted by Brand et al,27 who only included males, and by Becker et al,26 where the majority of participants were male. In four of the included studies, infected participants were not on treatment26 27 29 31 and in a further two,24 25 treatment was not stated. The remaining two studies examined participants on cART. Mean age ranged from 34.7 years25 to 62 years27 in uninfected participants and 37 years30 to 47 years27 in infected participants. Age was significantly different between the case and control groups in three studies.26–28 Mean BMI ranged from 20.6 kg/m225 to 28.1 kg/m231 in uninfected participants and from 21.1 kg/m225 to 25.1 kg/m231 in infected participants. BMI was significantly different between case and control groups in three studies.26 27 31 A meta-analysis using a random effects model (I2=84.79%) indicated no significant association between HIV infection and T2DM prevalence (RR=1.61, 95% CI 0.62 to 4.21, p=0.33), figure 4. Table 3 Prevalence data: HIV infected (treated and untreated) versus non-infected Author, year Setting Population Case HIV+ Control HIV ART Diagnosis Prevalence case % Prevalence control % p Value Amusa, 2015 Nigeria 200 adults n=150, 62.6% women Mean age=40.6 n=50, 60% women Mean age=40.2 Not stated FPG, criteria not stated 28.0 4.0 0.01 Anastos, 2010 Rwanda 824 women n=606 Mean age=42.4 Mean BMI=21.1 n=218 Mean age=34.7 Mean BMI=20.6 Not stated Self-report or WHO criteria 0.5 0.5 0.98 Becker, 2010 South Africa 60 adults n=30, 33% women Mean age=43 Mean BMI=25 n=30, 40% women Mean age=54 Mean BMI=28 Not on treatment Prescription of anti-diabetic medication or diagnosis on admission 3.0 23.0 0.05 Brand, 2014 South Africa 20 black males requiring amputation n=10 Mean age=47 Mean BMI=22.4 n=10 Mean age=62 Mean BMI=25.3 Not on treatment WHO criteria or prescription of anti-diabetic medication 50.0 0.0 <0.05 Edwards, 2015 Kenya 2206 adults n=210, 69% women Mean age=43 n=1996, 71% women Mean age=49 First-line ART used tenofovir/lamivudine/efavirenz; second line lopinavir/ritonavir instead of efavirenz WHO criteria 4.8 15.0 <0.01 Fourie, 2010 South Africa 600 adults n=300, 61% women Mean age=44 Mean BMI=22.9 n=300, 61% women Mean age=44 Mean BMI=22.8 Not on treatment IDF criteria 36.6  43.7 0.64 Maganga, 2015 Tanzania  454 adults n=301, 67.8% women Mean age=37 (untreated) and 40 (treated) Mean BMI=22.0 (untreated) and 23.7(treated) n=153, 61,4% women Mean age=38 Mean BMI=23.8 n=151 not on treatment n=150 on treatment—21% on protease inhibitors (lopinavir and ritonavir); rest on other ART: nevirapine, efavirenz, tenofovir, stavudine, zidovudine WHO criteria 9.3 5.2 0.04 (untreated vs control) and 0.001 (treated vs control) Ngatchou, 2013 Cameroon 204 adults n=108, 74% women Mean age=39 Mean BMI=25.1 n=96, 72% women Mean age=41 Mean BMI=28.1 Not on treatment WHO criteria 26.0 1.0 0.01 ART, antiretroviral therapy; BMI, Body mass index; FPG, fasting plasma glucose; IDF, International Diabetes Federation. Figure 4 Meta-analysis of studies comparing T2DM in HIV-infected and HIV-uninfected participants. T2DM, type 2 diabetes mellitus. Prevalence of T2DM in HIV-infected treated compared with untreated participants Table 4 shows the data for five studies included in the meta-analysis comparing HIV treated (n=1120) with untreated participants (n=828). The majority of included participants were women (ranged from 58%34 to 75% women36), and mean age ranged from 32.7 years33 to 44.2 years36 (age was not stated for Manuthu et al34). All treated participants were receiving cART (therapy not stated by Kagaruki et al32). Where stated, age was higher in treated compared with untreated participants,32 33 36 yet significance was not stated for these age differences. Mean BMI was only reported in two studies, and was in the WHO healthy weight category (22 kg/m2) for both groups in one study,33 and in the WHO overweight category (26.5 kg/m2) for both groups in the second study.36 A meta-analysis using a random effects model (I2=53.25%) indicated no significant association between HIV treatment and T2DM prevalence (RR=1.38, 95% CI 0.66 to 2.87, p=0.39), figure 5. Figure 5 Meta-analysis of studies comparing T2DM in HIV-infected treated and untreated participants. T2DM, type 2 diabetes mellitus. Discussion This systematic review and meta-analysis of African studies showed no statistically significant association between HIV infection or cART exposure, and T2DM prevalence. This is in contrary to study findings of international studies in Europe and North America that have shown a higher prevalence of T2DM in HIV-infected compared with uninfected participants,37 particularly when treated with cART.1 3 5 Incidence rates of T2DM in patients with HIV were described per 1000 person years of follow-up and ranged considerably among the included studies. Cumulative incidence rate for the included studies was 17.4. For comparison, the incidence rate of T2DM in a healthy American population in 2012 was lower at 7.8.38 Individually, three included papers reported lower incidence rates than the American population, and four reported higher incidence rates. There were no obvious differences between these studies in terms of age, sex and duration of follow-up, BMI or treatment; yet studies with larger sample sizes seemed to show higher incidence rates. A systematic review of T2DM in Sub-Saharan Africa4 found only one study reporting an incidence rate of 29 in healthy adults (> 40 years) in Kinshasa.39 In the present systematic review, only one included study on HIV-infected participants reported higher incidence rates than the healthy adults in Kinshasa.20 Therefore, from the limited data available and from the included studies in this systematic review, it does not seem that incidence is higher in populations infected with HIV in Africa than in a healthy ageing African population. T2DM incidence and prevalence rates have been reported internationally in patients with HIV infection and treated patients. De Wit et al40 reported T2DM incidence rates of 6 (and an incidence rate of 4 for definite cases of T2DM) from the Data Collection on Adverse Events of Anti-HIV Drugs (D:A:D) study. They examined 33 389 patients with HIV infection from 212 clinics in Europe, USA, Argentina and Australia, and found that treatment with stavudine increased the RR of T2DM by 1.19 per year of exposure (conversely, treatment with ritonavir and nevirapine decreased risk of T2DM). Interestingly, controlling for lipodystrophy did not modify this relationship and a direct effect of treatment on mitochondrial toxicity was thus suggested. Baseline prevalence of T2DM in this study was 2.9%. Findings from the Multicentre AIDS Cohort Study (MACS) showed a T2DM incidence of 47 in HIV-infected white males who were on cART versus 17 in those who were cART naïve; however, this study used only a single increased fasting plasma glucose as their diagnostic criteria.41 Nigatu et al,42 in 2013 conducted a systematic review looking at incidence of various comorbidities, including T2DM, with HIV infection, and found a combined T2DM incidence rate of 6 (with a range of 4.2–36) in a sample of 44 484 individuals. In the studies included in their systematic review, ART exposure increased incidence rates when compared with ART-naïve patients. Conversely, Tripathi et al8 found that 6816 patients with HIV infection (of which over 80% were treated with ART) had lower T2DM incidence rates than matched, non-infected individuals (11.4 vs 13.6). Similarly, Nix et al9 in 2014 stated that their summary of the literature found a similar decreased incidence of T2DM in HIV-infected individuals compared with controls. The present systematic review found a combined T2DM incidence rate of 17.4 in African patients with HIV infection who were cART treated, which is higher than incidence rates found in all of the above-mentioned studies except for the males in the MACS study. Therefore, although incidence does not seem to be higher in patients with HIV infection in Africa compared with a normal ageing population in Africa; T2DM incidence in HIV-infected people in Africa does appear to be higher than rates reported internationally for patients with HIV infection, and those reported for a healthy American population. It is possible that the higher incidence of T2DM in African HIV-infected individuals compared with international incidence data for HIV-infected individuals could be explained by the high presence of risk factors for T2DM in African populations, regardless of HIV status. Although prevalence of T2DM in Africa is lower than other regions in the world, the IDF Diabetes Atlas (7th edition) states that more than two-thirds of people with diabetes in Africa are undiagnosed, and that prevalence rates are expected to more than double in the next few years. This predicted increase is the highest of all regions worldwide, indicating the greatest increase in incidence rates. Furthermore, the IDF states that lack of prevalence data in Africa makes these estimates somewhat weak, and it is thus possible that prevalence in Africa is in fact higher than reported. As many African populations are undergoing rapid transitions, the toxic combination of early life undernutrition in utero and infancy, combined with excessive weight gain in later life may be contributing to increased T2DM susceptibility.43 In fact, in studies included in this systematic review where mean BMI was reported, a substantial proportion of participants infected with HIV were overweight or obese. This presents a different picture to the undernourished HIV-infected individual previously associated with Africa, and may explain a higher incidence of T2DM as an effect of lifestyle rather than an HIV disease-related risk. Although this systematic review has not shown a higher prevalence of T2DM in HIV-infected individuals compared with uninfected individuals, it does support the importance of screening for T2DM in African populations infected with HIV where T2DM incidence appears to be high. Furthermore, these findings reinforce the importance of managing and screening for metabolic disease, such as T2DM as part of routine clinical care of patients infected with HIV in order to support continuity of care.44 It is important to note that since none of the included studies were randomised, and there were too few studies for subgroup analyses, we cannot account for differences in disease course or lifestyle factors that confound exposure to cART or T2DM risk. Similarly, differences in cART exposure may be associated with regression or cure of illnesses in HIV, or with increased risk factors for T2DM. The mean age of included participants was generally <45 years, which may have influenced the cumulative incidence reported in this review, since age influences diabetic progression. There was also heterogeneity in the method of diagnosis of T2DM between studies, which could have confounded results. Although all of the diagnosis methods included in this systematic review were well recognised (see table 4), future T2DM screening programmes should strive to use gold standard diagnosis methods such as OGTT or HbA1c values.14 The findings of this systematic review are further limited by the high risk of bias of included studies, largely due to confounding factors and limited blinding. Furthermore, the small sample size of included studies as well as small number of studies available limit the conclusions that can be drawn. These limitations highlight the need for larger studies to be conducted examining T2DM incidence and prevalence in people with HIV in Africa, with focus on careful blinding and consideration of confounders. Table 4 Prevalence data: HIV-infected treated versus untreated Author, year Setting Population Case treated Control untreated ART Diagnosis Prevalence case % Prevalence control % p Value Kagaruki, 2014 Tanzania 671 participants, 70.5% women Mean age=38.7 n=354, 67.8% women Mean age=40.6 n=317, 73.5% women Mean age=36.7 Not stated WHO criteria 3.7 4.7 Not stated Manuthu, 2008 Kenya 295 participants, 58% women n=134 n=161 82.7% on d4t-based regimen, 51.1% on d4T+3TC+ nevirapine 31.6% on d4T+3TC+ efavirenz. 17.3% on AZT-based regimens 13.5% on AZT+3TC+ efavirenz 3.8% on AZT+3TC+ nevirapine; one PI-based regimen was AZT+3TC+ lopinavir. OGTT, criteria not stated 1.5 1.2 0.85 Mohammed, 2015 Ethiopia 393 adults, 66.9% women Mean age=37.9 n=284 n=109 32.1% used the drug combination zidovudine + lamivudine + nevirapine WHO criteria 8.5 0.9 <0.01 Nsagha, 2015 Cameroon 215 participants, 74.9% women Mean age 44.2 years Mean BMI=26.47 n=160, 77.5% women Mean age=44.7 Mean BMI=26.94 n=55, 67.3% women Mean age=38.6 Mean BMI=25.09 AZT+3TC+ efavirenx =1.3%, AZT+3TC+ nevirapine =50%, TDF+3TC+ efavirenz =27.5%, TDF+3TC+ nevirapine =13.1%, TDF+3TC+ lopinavir =8.1% WHO criteria 1.9 3.6 0.46 Tesfaye, 2014 Ethiopia 374 participants, 68% women Mean age=32.7 n=188, 63.8% women Mean age=32.7 Mean BMI=22.1 n=186, 68.8% women Mean age=32.6 Mean BMI=22.2 58% on regimen containing efavirenz and 42% on nevirapine as NNRTI IDF criteria 33.5 21.5 <0.05 3TC, lamivudine; ART, antiretroviral therapy; AZT, zidovudine; BMI, body mass index; d4t, stavudine, IDF, International Diabetes Federation; NNRTI, non-nucleotide reverse transcriptase; OGTT, oral glucose tolerance test; PI, protease inhibitor; TDF, tenofovir. In conclusion, this meta-analysis shows no significant association between HIV infection or treatment and T2DM prevalence in African population studies. Furthermore, incidence of T2DM in Africa in patients with HIV infection on cART is no greater than in a normal ageing population, yet is higher than incidence rates in HIV-infected individuals outside of Africa. Larger case–control studies with effective blinding and consideration of confounders need to be conducted in Africa in order to further elucidate these associations in comparison with international findings. Currently, HIV infection and cART do not seem to predispose patients in Africa to T2DM; however, high incidence rates warrant focus on screening and preventative programmes for HIV-infected people living in Africa. The support of the DST-NRF Centre of Excellence (CoE) in Human Development at the University of the Witwatersrand, Johannesburg, in the Republic of South Africa is acknowledged by AP (P2015004). The work of AP was also supported by the Claude Leon Foundation. Twitter: Follow Richard Munthali @https://twitter.com/RichardMunthali Contributors: AP contributed to conception and design; acquired, analysed and interpreted the data; drafted the article and approved the final version for publication. RJM acquired the data, revised the article and approved the final version for publication. LS, JAG, LKM and DMA acquired the data, revised the article and approved the final version for publication. SAN contributed to conception and design, acquired and interpreted the data, revised the article and approved the final version for publication. We have read and understood the BMJ policy on declaration of interests and declare that we have no competing interests. Funding: The work of DMA was supported by Vanderbilt Diabetes Research and Training Center funded by grant P30DK020593 from the National Institute of Diabetes and Digestive and Kidney Disease. Disclaimer: Opinions expressed and conclusions arrived at, are those of the author and are not to be attributed to the CoE in Human Development. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data available. ==== Refs References 1 Hadigan C , Kattakuzhy S Diabetes mellitus type 2 and abnormal glucose metabolism in the setting of human immunodeficiency virus . Endocrinol Metab Clin North Am 2014 ;43 :685 –96 . 10.1016/j.ecl.2014.05.003 25169561 2 Spollet G Hyperglycemia in HIV/AIDS . Diabetes Spectrum 2006 ;19 :163 –6 . 10.2337/diaspect.19.3.163 3 Samaras K The burden of diabetes and hyperlipidemia in treated HIV infection and approaches for cardiometabolic care . Curr HIV/AIDS Rep 2012 ;9 :206 –17 . 10.1007/s11904-012-0124-x 22752405 4 Hall V , Thomsen RW , Henriksen O Diabetes in Sub Saharan Africa 1999–2011: epidemiology and public health implications. A systematic review . BMC Public Health 2011 ;11 :564 10.1186/1471-2458-11-564 21756350 5 Paik IJ , Kotler DP The prevalence and pathogenesis of diabetes mellitus in treated HIV-infection . Best Pract Res Clin Endocrinol Metab 2011 ;25 :469 –78 . 10.1016/j.beem.2011.04.003 21663840 6 Brown TT , Tassiopoulos K , Bosch RJ Association between systemic inflammation and incident diabetes in HIV-infected patients after initiation of antiretroviral therapy . Diabetes Care 2010 ;33 :2244 –9 . 10.2337/dc10-0633 20664016 7 Ghislain M , Bastard JP , Meyer L Late antiretroviral therapy (ART) initiation is associated with long-term persistence of systemic inflammation and metabolic abnormalities . PLoS ONE 2015 ;10 :e0144317 10.1371/journal.pone.0144317 26636578 8 Tripathi A , Liese AD , Jerrell JM Incidence of diabetes mellitus in a population-based cohort of HIV-infected and non-HIV-infected persons: the impact of clinical and therapeutic factors over time . Diabet Med 2014 ;31 :1185 –93 . 10.1111/dme.12455 24673640 9 Nix LM , Tien PC Metabolic syndrome, diabetes, and cardiovascular risk in HIV . Curr HIV/AIDS Rep 2014 ;11 :271 –8 . 10.1007/s11904-014-0219-7 25027062 10 Global status report on diabetes 2012. Geneva: World Health Organization; 2012 . 11 Prioreschi A , Munthali R , Soepnel L The association between HIV, HIV treatment and type 2 diabetes in Africa: a systematic review. 2015 . http://www.crd.york.ac.uk/PROSPERO/myprospero.php 12 Liberati A , Altman DG , Tetzlaff J The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration . BMJ 2009 ;339 :b2700 10.1136/bmj.b2700 19622552 13 Samaras K Prevalence and pathogenesis of diabetes mellitus in HIV-1 infection treated with combined antiretroviral therapy . J Acquir Immune Defic Syndr 2009 ;50 :499 –505 . 10.1097/QAI.0b013e31819c291b 19223782 14 Sacks DB A1C versus glucose testing: a comparison . Diabetes Care 2011 ;34 :518 –23 . 10.2337/dc10-1546 21270207 15 Higgins JPT , Green S eds. Cochrane Handbook for Systematic Reviews of Interventions 4.2.6 [updated September 2006] . In: Chichester, UK : John Wiley & Sons, Ltd , 2006 :6 The Cochrane Library 16 Busse JW GG Tool to Assess Risk of Bias in Case-control Studies https://distillercercom/resources/ 2016 . 17 Abrahams Z , Dave JA , Maartens G Changes in blood pressure, glucose levels, insulin secretion and anthropometry after long term exposure to antiretroviral therapy in South African women . AIDS Res Ther 2015 ;12 :24 10.1186/s12981-015-0065-8 26251665 18 George JA , Venter WDF , Deventer HE A longitudinal study of the changes in body fat and metabolic parameters in a South African Population of HIV-positive patients receiving an antiretroviral therapeutic regimen containing stavudine . AIDS Res Hum Retroviruses 2009 ;25 :771 –81 . 10.1089/aid.2008.0308 19619010 19 Karamchand S , Leisegang R , Schomaker M Risk factors for incident diabetes in a cohort taking first-line Nonnucleoside reverse transcriptase inhibitor-based antiretroviral therapy . Medicine (Baltimore) 2016 ;95 :e2844 10.1097/MD.0000000000002844 26945366 20 Magula NP , Motala AA , Lalloo UG The prevalence and incidence of diabetes mellitus and other disorders of glycaemia in South African black patients on antiretroviral therapy . Int J Infect Dis 2014 ;21 :10 10.1016/j.ijid.2014.03.425 21 Mandina Ndona M , Longo-Mbenza B , Wumba R Nadir CD4+, religion, antiretroviral therapy, incidence of type 2 diabetes mellitus, and increasing rates of obesity among black Africans with HIV disease . Int J Gen Med 2012 ;5 :983 –90 . 10.2147/IJGM.S32167 23226071 22 Sagna Y , Tiéno H , Guira O Following of newly managed adult HIV infected patients in the National Referral Centre, Burkina Faso: three years outcomes study . Sex Transm Infect 2013 ;89 :A1 –A428 . 10.1136/sextrans-2012-050988 23862196 23 Zannou DM , Denoeud L , Lacombe K Original article Incidence of lipodystrophy and metabolic disorders in patients starting non-nucleoside reverse transcriptase inhibitors in Benin . Antivir Ther (Lond) 2009 ;14 :371 –80 .19474471 24 Amusa GA , Awokola BI , Akanbi MO Burden of cardiovascular disease risk factors in HIV-infected adults in North-Central Nigeria . Am J Respir Crit Care Med 2015 ;191 :A4521 . 25 Anastos K , Ndamage F , Lu D Lipoprotein levels and cardiovascular risk in HIV-infected and uninfected Rwandan women . AIDS Res Ther 2010 ;7 :2 –7 . 10.1186/1742-6405-7-34 20180959 26 Becker AC , Sliwa K , Stewart S Acute coronary syndromes in treatment-naive black South Africans with human immunodeficiency virus infection . J Interv Cardiol 2010 ;23 :70 –7 . 10.1111/j.1540-8183.2009.00520.x 20015160 27 Brand M , Woodiwiss AJ , Michel F Large vessel adventitial vasculitis characterizes patients with critical lower limb ischemia with as compared to without human immunodeficiency virus infection . PLoS ONE 2014 ;9 :e106205 10.1371/journal.pone.0106205 25170758 28 Edwards JK , Bygrave H , Van den Bergh R HIV with non-communicable diseases in primary care in Kibera, Nairobi, Kenya: characteristics and outcomes 2010-2013 . Trans R Soc Trop Med Hyg 2015 ;109 :440 –6 . 10.1093/trstmh/trv038 25997923 29 Fourie CM , Van Rooyen JM , Kruger A Lipid abnormalities in a never-treated HIV-1 subtype C-infected African population . Lipids 2010 ;45 :73 –80 . 10.1007/s11745-009-3369-4 19916038 30 Maganga E , Smart LR , Kalluvya S Glucose metabolism disorders, HIV and antiretroviral therapy among Tanzanian adults . PLoS ONE 2015 ;10 :e0134410 10.1371/journal.pone.0134410 26287742 31 Ngatchou W , Lemogoum D , Ndobo P Increased burden and severity of metabolic syndrome and arterial stiffness in treatment-naive HIV+ patients from Cameroon . Vasc Health Risk Manag 2013 ;9 :509 –16 . 10.2147/VHRM.S42350 24043942 32 Kagaruki GB , Mayige MT , Ngadaya ES Magnitude and risk factors of non-communicable diseases among people living with HIV in Tanzania : a cross sectional study from Mbeya and Dar es Salaam regions . BMC Public Health 2014 ;14 :1 –9 . 10.1186/1471-2458-14-904 24383435 33 Tesfaye DY , Kinde S , Medhin G Burden of metabolic syndrome among HIV-infected patients in Southern Ethiopia . Diabetes Metab Syndr 2014 ;8 :102 –7 . 10.1016/j.dsx.2014.04.008 24907175 34 Manuthu EM , Joshi MD , Lule GN Prevalence of dyslipidemia and dysglycaemia in HIV infected patients . East Afr Med J 2008 ;85 :10 –17 .18543521 35 Mohammed AE , Shenkute TY , Gebisa WC Diabetes mellitus and risk factors in human immunodeficiency virus-infected individuals at Jimma University Specialized Hospital, Southwest Ethiopia . Diabetes Metab Syndr Obes 2015 ;8 :197 –206 . 10.2147/DMSO.S80084 25926749 36 Nsagha DS , Clement J , Assob N Risk factors of cardiovascular diseases in HIV/AIDS patients on HAART . The Open AIDS Journla 2015 ;9 :51 –9 . 10.2174/1874613601509010051 37 Galli L , Salpietro S , Pellicciotta G Risk of type 2 diabetes among HIV-infected and healthy subjects in Italy . Eur J Epidemiol 2012 ;27 :657 –65 . 10.1007/s10654-012-9707-5 22722952 38 National Diabetes Statistics Report: Estimates of Diabetes and Its Burden in the United States, 2014 . Centers for Disease Control and Prevention . In: Atlanta, GA : U.S. Department of Health and Human Services , 2014 . 39 Longo-Mbenza B , On'kin JB , Okwe AN Metabolic syndrome, aging, physical inactivity, and incidence of type 2 diabetes in general African population . Diab Vasc Dis Res 2010 ;7 :28 –39 . 10.1177/1479164109346362 20368230 40 De Wit S , Sabin CA , Weber R Incidence and risk factors for new-onset diabetes in HIV-infected patients: the Data Collection on Adverse Events of Anti-HIV Drugs (D:A:D) study . Diabetes Care 2008 ;31 :1224 –9 . 10.2337/dc07-2013 18268071 41 Brown TT CS , Li X , Kingsley LA Antiretroviral therapy and the prevalence and incidence of diabetes mellitus in the multicenter AIDS cohort study . Arch Intern Med 2005 ;165 : 1179 –84 . 10.1001/archinte.165.10.1179 15911733 42 Nigatu T , Oldenburg B , Elliott J The incidence of cardiovascular disease, cancer and type 2 diabetes comorbidities in HIV infection: a systematic review . J Nurs EducPract 2013 ;3 : 58 –67 . 10.5430/jnep.v3n7p58 43 Victora CG AL , Fall C , Hallal PC Maternal and child undernutrition: consequences for adult health and human capital . Lancet 2008 ;2008 :340 –57 . 10.1016/S0140-6736(07)61692-4 44 Rabkin M , Melaku Z , Bruce K Strengthening health systems for chronic care: leveraging HIV programs to support diabetes services in Ethiopia and Swaziland . J Trop Med 2012 ;2012 : 137460 10.1155/2012/137460 23056058

PMC005xxxxxx/PMC5372102.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01331910.1136/bmjopen-2016-013319Renal MedicineProtocol15061728171917101723Randomised controlled trial comparing ofatumumab to rituximab in children with steroid-dependent and calcineurin inhibitor-dependent idiopathic nephrotic syndrome: study protocol Ravani Pietro 1Bonanni Alice 23Ghiggeri Gian Marco 23 1 Division of Nephrology, University of Calgary, Calgary, Alberta, Canada 2 Division of Nephrology, Dialysis, Transplantation, Giannina Gaslini Children's Hospital, Genoa, Italy 3 Laboratory on Pathophysiology of Uremia, Giannina Gaslini Children's Hospital, Genoa, ItalyCorrespondence to Dr Gian Marco Ghiggeri; gmarcoghiggeri@gaslini.org2017 17 3 2017 7 3 e0133194 7 2016 14 11 2016 18 11 2016 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction Oral steroids induce remission in about 90% of children with idiopathic nephrotic syndrome (INS), which is characterised by severe proteinuria and hypoalbuminaemia. Some children become steroid-dependent (SD) and require addition of calcineurin inhibitors (CNI) to maintain remission. Since these oral agents are toxic, alternative interventions are needed for long-term treatment. The anti-CD20 antibody rituximab has shown promising steroid-sparing properties in clinical trials, but benefits are less convincing in complicated forms of SD-INS. Ofatumumab, a new anti-CD20 antibody with stronger affinity to CD20, may be superior to rituximab in maintaining oral steroid-free and CNI-free disease remission in children with SD-INS. Methods and analysis This open-label, two-parallel-arm, controlled, phase II randomised clinical trial will enrol children with SD-INS maintained in remission with oral steroids and CNI. Children will be randomised to either ofatumumab or rituximab infusion. After infusion of either antibody, steroids will be maintained for 30 days and then tapered off by 0.3 mg/kg/week until complete withdrawal. 1 week after complete steroid withdrawal, CNI will be decreased by 50% and withdrawn within 2 additional weeks. We will enrol 140 children to detect as significant at the 2-sided p value of 0.01 with a power of >0.8, a reduction in the risk of 1-year relapse (primary end point) of at least 0.3 (ie, from 0.65 to 0.35; (risk ratio 0.54)) in the ofatumumab arm when compared with the rituximab arm. We will compare the amount of steroids required to maintain complete disease remission at 6 and 24 months, relapse-free period, relapse rate per year as secondary end points. Circulating cell populations will be studied as biomarkers or predictors of the anti-CD20 response. Ethics and dissemination The trial received ethics approval from the local ethics board. We will publish study results and present them at international scientific meetings. Trial registration numbers NCT02394119; 2015-000624-28; Pre-results. ==== Body Strengths and limitations of this study This study will be conducted as a multicentre randomised controlled trial coordinated by a large national referral centre for paediatric nephrology. It will provide patients and healthcare providers with important information on benefits and harms of two anti-CD20 antibodies in idiopathic nephrotic syndrome (INS). Disease relapse is an important outcome for children with INS for its social, psychological and clinical implications. More important outcomes such as end-stage kidney failure or death are difficult to include in randomised controlled trials in children as they occur relatively late in the disease course. Limitations related to lack of blinding are mitigated by the objective outcomes chosen and blinding of outcome assessors. Introduction Idiopathic nephrotic syndrome (INS) is a disease characterised by episodes of severe proteinuria and hypoalbuminaemia (serum albumin <2.5 g/dL), often associated with dyslipidaemia and hypercoagulability. In Western countries, INS affects 2–2.7 new children per 100 000 children per year and has a prevalence of 16 cases per 100 000.1 Oral corticosteroids are the cornerstone of therapy, inducing remission of INS in ∼90% of children.2 However, up to 85% of cases relapse within 5 years3 and many will develop steroid dependence (SD). SD-INS relapses within 2 weeks of steroid withdrawal and requires continuation of treatment. Clinical practice guidelines suggest using low-dose prednisone to maintain remission in SD-INS (evidence 2C-D), and use of steroid-sparing agents (ie, calcineurin inhibitors; CNI) for children who develop steroid-related adverse effects (evidence 1B).4 Given the toxicity of steroids and CNI, there is a need to investigate alternative treatment options. The interest in the anti-CD20 chimeric antibody rituximab followed the observation of a dramatic reduction in proteinuria in children with nephrotic syndrome who received rituximab to treat idiopathic thrombocytopenic purpura5 or a post-transplant lymphoproliferative disorder.6 7 More recent observational studies8–17 confirmed potential benefits in mixed populations with nephrotic syndrome, but clinical trials did not confirm expected benefits in steroid-resistant (SR) forms of INS.18 Although recent randomised controlled trials support the use of rituximab in SD-INS, benefits may be suboptimal, especially in complicated forms of the disease. While a single infusion of rituximab allowed steroid withdrawal and was non-inferior to steroids in maintaining remission in a trial of children with uncomplicated forms of SD-INS,19 in children with frequently relapsing and complicated forms of SD-INS requiring the use of CNI, benefits of rituximab are less convincing.20 21 Ofatumumab is a fully humanised anti-CD20 monoclonal IgG1(k) antibody of last generation. Ofatumumab binds with more affinity to CD20, potentially leading to more efficient complement-dependent cytotoxicity. In small series, ofatumumab induced remission in children with SR-INS who did not respond to rituximab.22 23 From an economic point of view, ofatumumab is not superior to rituximab as the cost of the two drugs for square metres of body surface is similar. Ofatumumab is currently in phase III clinical trials for the treatment of chronic lymphocytic leukaemia, relapsing lymphoma and rheumatoid arthritis.24 Owing to its stronger affinity for the CD20 and its fully humanised structure, ofatumumab may be used in larger doses with minimal risk of side effects leading to potentially larger benefits, that is, longer remission period, than the chimeric antibody rituximab. Some studies support this conclusion.25 26 The aim of this trial is to test whether ofatumumab is superior to rituximab in maintaining complete disease remission in children with steroid and CNI-dependent INS. Methods and design Rationale and justification of the active comparator Owing to its fully humanised structure and stronger affinity for the CD20 antigen, ofatumumab can be used in larger doses than the chimeric antibody rituximab with minimal risk of side effects. Knowledge generation to support the use of new steroid-sparing agents is an important unmet healthcare need in paediatric nephrology, especially for children with complicated forms of SD-INS requiring steroids and CNI (tacrolimus or ciclosporin). In the present study, CNI have been chosen among steroid-sparing agents because they are more effective than other drugs and therefore more commonly used.27 Head-to-head comparison between rituximab and ofatumumab is justified on the basis of available trials supporting the use of anti-CD20 antibodies to maintain remission of INS without steroids and CNI.19–21 Objectives This trial will test whether ofatumumab is superior to rituximab in maintaining oral drug-free disease remission (complete or partial remission) at 12 months (primary objective) and reduce the risk of relapse in a longer follow-up of 24 months (secondary objective) in children with SD and CNI-dependent INS. This study will also collect information on occurrence of side effects (ie, acute and long-term drug-related adverse events) and need to restart the use of steroids or CNI following the infusion of the trial interventions. Design This is an open-label, two-parallel-arm, controlled, phase II randomised clinical trial testing the superiority of ofatumumab over rituximab in maintaining steroid-free and CNI-free disease remission in children with SD-INS (figure 1). Eligible participants will enter a 1-month run-in period, during which instruction on urine collection and dipstick readings will be carefully reviewed, compliance assessed and CNI and prednisone doses will be reduced to the minimum doses required to maintain complete remission (usually the minimum doses used in the previous 6 months). After run-in, children will be randomised in a 1:1 ratio to either the intervention arm (ofatumumab) or the active comparator arm (rituximab). Figure 1 Schematic view of trial design. CNI, calcineurin inhibitors; ITT, intention to treat. After infusion of intervention or comparator, steroids will be maintained at the baseline (post-run-in) dose for 30 days and then tapered off by 0.3 mg/kg/week until complete withdrawal. One week after the steroid withdrawal, CNI will be decreased by 50% and withdrawn within 2 additional weeks. All children will be followed for up to 24 months. In the case of relapses during the study (see outcome section) children will be treated with oral prednisone (60 mg/m2), in order to achieve remission. At remission, children will be treated with another infusion of either ofatumumab or rituximab, according to the allocation arm. After infusion of intervention or comparator, steroids will be maintained at initial dose for 30 days and then tapered off by 0.3 mg/kg/week until complete withdrawal. One week after the steroid withdrawal, CNI will be decreased by 50% and withdrawn within 2 additional weeks. Inability to withdraw CNI or need to restart CNI (development of steroid resistance) will be considered a relapse (primary end point) and a reason to terminate the study. In this case, patients should be treated at the discretion of the investigator, according to the best clinical practice guidelines. Study enrolment started in June 2015 and is expected to be complete by the end of 2017. Setting Interventions will be administered during an inpatient setting at the Nephrology Unit of the Giannina Gaslini Children's Hospital, Genoa. Follow-up visits will be performed at the same institution or by local nephrologists if travel to Genoa is not possible. Eligibility criteria Patients aged 2–18 years, in complete disease remission and with a proven clinical history of SD and CNI-dependent INS are eligible for inclusion in the study. SD-INS will be defined by two consecutive relapses during corticosteroid therapy tapering or within 14 days of steroid withdrawal; CNI-dependent INS will be defined by relapse following ciclosporin or tacrolimus discontinuation. Exclusion criteria include: positivity to autoimmunity tests (ANA, nDNA, ANCA); reduction in C3 levels; steroid resistant forms of INS; eGFR<90/mL/min/1.73 m2 valuated according to revised Bedside Schwartz Formula for patients between 2 and 17 years and with CKD-EPI Creatinine 2009 Equation for 18 years old patients; pregnancy; neoplasm; previous or actual HBV (with HBeAb positivity) or HCV infection; CD20 B lymphocytes count <2.5% and treatment with rituximab or steroid-sparing agents other than CNI in the past 6 months. Participant identification process All patients affected by SD and CNI-dependent INS followed up in the nephrology outpatient clinic at the Giannina Gaslini Children's Hospital will be evaluated for recruitment as they return to clinic. A preliminary interview for clinical and pharmacological history will be performed in order to verify the eligibility criteria. A study coordinator will illustrate the project, deliver the information material (information sheet and informed consent form) and collect written informed consent in order to complete the screening. Parents'/guardian's written informed consent, and child's assent will be collected before any study-related procedure not part of the participant's normal medical care. Participants or their families will be able to withdraw consent at any time. Randomisation Participants will be randomised 1:1 to the intervention or active comparator arm. A distant site with no clinical involvement in the trial will generate two randomisation lists (for age ≥9 years and <9 years) using permuted blocks of variable size. Stratification by age is motivated by the need to maximise the likelihood of balancing factors potentially affecting the effects of the intervention on outcomes, which are associated with age (disease duration, age at onset, relapse history and disease severity).20 Assignments will be notified electronically after obtaining signed consent. A study coordinator responsible for recruitment will obtain signed consent (and assent for participants capable of assenting), assign a unique participant study number and request randomisation. An analyst from a distant site not involved in patient care, where the randomisation lists have been generated and kept concealed from the clinical investigators, will communicate the allocation arm to the study coordinator (based on the participant's study number). Blinding The following reasons justify the lack of participant and investigator blinding: Ofatumumab and rituximab require different methods for infusion; for example, the ofatumumab requires a filter that cannot be blinded. According to the producer indications, the same filter may alter/reduce the availability/stability rituximab; and Ofatumumab must be diluted in 1000 mL of normal saline and infused at a fixed high rate; the producer does not guarantee rituximab stability using the same dilution/infusion strategy. Treatment arms Intervention Children randomised to the intervention arm will receive the fully humanised monoclonal anti-CD20 antibody ofatumumab. Participants assigned to the intervention arm will receive a premedication with methylprednisolone (2 mg/kg infused in 30′ intravenous diluted in 100 mL of normal saline), oral cetirizine (0.2 mg/kg) and oral paracetamol (15 mg/kg), in order to reduce common reactions. Registered nurses will deliver the premedication and the intervention drug in the Nephrology Department of Giannina Gaslini Children's Hospital. Ofatumumab will be infused intravenous, at a dose of 1500 mg/1.73 m2, diluted in 1000 mL of normal saline, at a rate of 12 mL/hour in the first 30′. Thereafter, the infusion rate will be doubled every 30 min up to a maximum of 200 mL/hour. Active comparator Children randomised to the active comparator will receive the chimeric monoclonal anti-CD20 antibody rituximab. In order to reduce common reactions, children will receive the same premedication as described above for the ofatumumab arm. As for Ofatumumab, registered expert nurses will deliver the premedication and the active comparator in the Nephrology Department of Giannina Gaslini Children's Hospital. Rituximab will be administered at a dose of 375 mg/m2. For doses between 100 and 250 mg, rituximab will be diluted in 100 mL of normal saline and administered at 2 mL/hour for the first 30′; 3 mL/hour for the second 30′; 6 mL/hour for the third 30′; 15 mL/hour until the end of the infusion. For doses between 260 and 500 mg, rituximab will be diluted in 250 mL of normal saline and administered at 6 mL/hour for the first 30′; 9 mL/hour for the second 30′; 18 mL/hour for the third 30′; 36 mL/hour until the end. For doses between 510 and 1000 mg, rituximab will be diluted in 500 mL of normal saline and administered at 9 mL/hour for the first 30′; thereafter, the infusion rate will be doubled every 30 min up to a maximum of 72 mL/hour. Since ofatumumab and rituximab will be administered during a hospitalisation, the protocol does not include strategies to improve adherence to therapy. Relevant concomitant care and interventions that are permitted or prohibited during the trial Any medications not listed in the exclusion criteria may be given at the discretion of the investigator. The investigator will record all concomitant medications taken by the participant during the study from the date of informed consent, in the appropriate section of the case report form. In view of a possible role in the reduction in proteinuria and worsening of glomerular filtration rate, the use of ACE inhibitors/ARBs will not be allowed during the study. In hypertensive patients, these drugs will be replaced with other antihypertensive agents: α-adrenergic or β-adrenergic blockers (eg, carvedilol 0.2–1 mg/kg/day b.i.d., up to a maximum of 50 mg/day) will be administered as first-line drugs; in patients with intolerance or contraindication to β blockers (ie, children with asthma), calcium channel blockers (eg, amlodipine 0.06–0.3 mg/kg/day, up to a maximum of 10 mg/day) will be used. Outcomes The primary end point will be the occurrence of disease relapse or need of steroids or CNI to maintain complete remission within 12 months of randomisation. Relapse is defined by uPCR ≥2000 mg/g (≥200 mg/mmol) or >3+ protein on urine dipstick for 3 consecutive days.24 Failure to achieve full steroid and CNI withdrawal will be considered a relapse. The secondary end points will be relapse-free period, relapse rate per year and the amount of steroids required to maintain complete and partial disease remission at 6 and 24 months. Complete remission is defined by uPCR <200 mg/g (<20 mg/mmol) or 1+ protein on urine dipstick for 3 consecutive days.24 Partial remission is defined by proteinuria reduction of 50% or greater from the presenting value and absolute uPCR between 200 and 2000 mg/g (20–200 mg/mmol).28 Secondary aims will be to tests circulating cell populations as biomarkers or predictors of the anti-CD20 response. Safety end points will be frequency and severity of adverse events and abnormal values in biochemical tests and haematology assessments. Safety data We will collect any untoward medical occurrence in the form of signs, symptoms, abnormal laboratory findings or diseases that emerges or worsens relative to baseline (ie, present at the initial study visit). Policy and approach to define adverse events is reported in the online supplementary appendix. 10.1136/bmjopen-2016-013319.supp1supplementary appendix Data collection, management and analysis Data collection methods and adherence during follow-up Study visits will occur at baseline, after 1 month and every 3 months thereafter, unless complications or relapses occur. Determination of 24 hours proteinuria at baseline and after 6, 12 and 24 months will be performed at a central laboratory (in order to assess the primary and secondary outcome). Dipstick for proteinuria determination will be evaluated daily. In the case of dipstick positivity, the presence of proteinuria will be confirmed with 24-hour urine collection. Complete blood count, kidney function, plasma proteins, immunoglobulins, lipid status (cholesterol and triglycerides), albumin and lymphocyte subpopulations (for CD20 lymphocytes B count) will be obtained at 1, 3, 6, 9, 12, 15, 18, 21 and 24 months during protocol visits. Time schedule of enrolment, interventions assessments and visits for participants are shown in table 1. At discharge from Nephrology Unit, each patient will receive a clinical diary form, to be filled with proteinuria levels at dipstick, body weight, current treatment (steroid and CNI dosage) and monthly send to our centre via fax or email. A study coordinator will maintain ongoing contact with the children, their families and the family physician to collect clinical data including blood pressure and potential adverse events also in order to minimise loss to follow-up/dropout. Table 1 Participants timeline Study period Enrolment Allocation and treatment Follow-up Close-out Timepoint −1 0 t1 t3 t6 t9 t12 t15 t18 t24 ENROLMENT Relevant medical history X Eligibility screen X Project illustration X Informed consent X Instruction about immunosuppressive drugs tapering X X Allocation X INTERVENTIONS Ofatumumab/rituximab administration X ASSESSMENTS Dosage on 24-hours urine collection X X X X X X X X X X Physical examination and vital signs X X X X X X X X X X Haematology and biochemistry (complete blood count, kidney function, plasma proteins immunoglobulins, lipid status—cholesterol and triglycerides—, albumin, lymphocyte subpopulations—for CD20 lymphocytes B count) X X X X X X X X X X Adverse events data records X X X X X X X X X Data management The investigators, or designees, will be responsible for recording study data in the case report form and entering study data in the electronic database prepared by the study coordinating centre. Coordinating centre data management at the G. Gaslini Institute, Genoa, will be responsible for data processing, in accordance with G. Gaslini's data management procedures. Database will be locked once quality assurance procedures have been completed. Statistical methods We will use standard statistical methods to summarise the sample characteristics overall and by arm assignment, using statistics for quantitative and qualitative data as appropriate. For the analysis of the primary outcome, we will use logistic regression to compare the risk of relapse at 12 months with a two-sided p value of <0.05 as level of statistical significance. The comparison will be repeated at 6 months (secondary outcome). As in previous studies, we will use survival methods in secondary analyses to assess the risk of repeated relapses and estimate the average time to relapse by treatment group.29 In all analyses, we will use an intention-to-treat approach, whereby participants will be analysed as randomised regardless of protocol adherence. Missing data will be replaced in a way that reflects the worst-case scenario: missing data in the active comparator group will be considered as successes and missing data in the active intervention will be considered failures. Sample size We will enrol 140 children in this study, and follow them for at least 1 year. Assuming a baseline risk of relapse at 1 year of 0.65 among children assigned to rituximab, this sample size will allow to detect as significant at the two-sided p value of 0.01 with a power of >0.8, a reduction in the risk of 1 -year relapse of at least 0.3 (ie, from 0.65 to 0.35; risk reduction by 0.46). This sample size accounts for a total proportion of drop-out and drop-in <10%. Our centre is a national referral centre for paediatric nephrology. We treat 50 new cases of INS per year, of which 80% are SD-INS. Our records demonstrate that we were able to successfully recruit participants in similar studies.18–20 29 Study termination Participants will be informed that they have the right to withdraw from the study at any time, without prejudice to their medical care, and that they are not obliged to state their reasons. Any withdrawal must be fully documented in the case report form and source documents, and will be followed up by the investigator. Follow-up will be considered complete when the participant has completed all study procedures and assessments up to the month-12 visit (primary end point) and month-24 visit (secondary end point). The investigator may withdraw a participant at any time if this is considered to be in the participant's best interest. Study termination will be mandatory in the following situations: development of steroid resistance disease requiring other steroid-sparing agents, pregnancy, significant worsening of renal function, onset of malignancy, serious hypersensitivity or allergic reaction, any other serious adverse event, serious intercurrent illness, administrative reasons, or investigator's or participant's or parents'/legal tutor's request. We have no plans for interim analyses or prespecified stopping rules. The data safety and monitoring board will decide if the study is safe when 50% of the participants will be enrolled (without comparisons). The sponsor may temporarily or permanently discontinue the study for safety, ethical, compliance or other reasons. If the study is suspended or discontinued, the investigator will be responsible for promptly informing the Independent Ethics Committee. Where required by local regulations, the sponsor will be responsible for informing the Independent Ethics Committee of study discontinuation. Data safety and monitoring board, study monitoring and end point adjudication committee The members of DSMB will include experts in methodology and clinical nephrology; the sponsor will be responsible for study monitoring. The investigator or his/her delegate will ensure that the case report forms are completed in a timely manner to allow a study monitor to periodically access them and all study-related materials. The frequency of monitoring visits will be determined by the site enrolment rate. On study completion, the study monitor will visit the site to conduct a study termination visit. This will involve collection of any outstanding documentation. A blinded end point adjudication committee will ensure an ongoing review of all of the relevant data relating to each participant and to establish whether or not an end point has occurred. Ethics approval and trial status Before initiation of the study, we obtained written approval of the protocol, informed consent form and any information presented to potential participants from the local Independent Ethics Committee (Comitato Etico Regione Liguria). We also obtained approval from the Italian Drug Agency (Agenzia Italiana del FArmaco, AIFA). If any amendments to the protocol occur during the study, written approval must be obtained prior to its implementation. The Investigator is responsible for ensuring that these actions occur. Where required by local regulations, the Sponsor is responsible for ensuring Independent Ethics Committee approval of the study. The study was registered at https://clinicaltrials.gov (study number: NCT02394119) and https://eudract.ema.europa.eu (study number 2015-000624-28). The trial is currently recruiting: study enrolment started on 10 June 2015. Dissemination The completed randomised clinical trial study will be summarised in a final report that accurately and completely presents the study objectives, methods, results, limitations of the study and interpretation of findings. The authors of this study protocol will inform the contributing investigators (primary healthcare providers of study participants) in advance about any plans to publish or present data from this randomised controlled clinical trial. Any publications and presentations of the results (abstract in journals or newspapers, oral presentations, etc), either in whole or in part, by investigators or their representatives will require presubmission review by the authors of this study protocol and all coauthors. Discussion SD nephrotic syndrome carries a high risk of toxicity from steroids or standard steroid-sparing agents. Therefore, alternative treatment options are needed. Although recent studies support the use of rituximab as a steroid-sparing and CNI-sparing agent in SD-INS, benefits may be suboptimal, especially in complicated forms of the disease. The new anti-CD20 ofatumumab may be more effective than rituximab in controlling the disease, due to its stronger affinity for the CD20. Moreover, due to its fully humanised structure, it may be used in larger doses with minimal risk of side effects. These considerations motivated the use of an active comparator to test the effects of different agents blocking the CD20 antigen pathway. This is the first randomised controlled trial comparing the effects of two anti-CD20 antibodies on the risk of relapse of INS following steroid and CNI withdrawal. Strengths in the design of this trial include: objective and clinical important outcomes, identification of a high-risk population (SD-INS maintained in complete remission with steroids and CNI), methods to reduce bias (allocation concealment, objective outcomes, steps to ensure complete follow-up and blinding of outcome adjudicators to treatment assignment), careful collection of safety data and large catchment area of the study centre with clinical and research expertise in paediatric nephrology. The design of this trial has limitations. First, interventions are not blinded because ofatumumab and rituximab require different methods for infusion. Although strategies to mask the two interventions could have been designed, they would have been expensive. In addition and most importantly, the end point adjudication committee reviewing the clinical and laboratory data will be blinded to intervention assignments. Second, the primary end point is an intermediate rather than a final outcome. However, the laboratory-based measures we adopted to define disease relapse are objective and more distant outcomes, including cardiovascular or infectious complications of nephrotic syndrome, or progression of kidney disease to kidney failure are rarely observed during childhood. A very large multinational trial would be necessary to study these hard end points. On the other hand, relapse of nephrotic syndrome is an important outcome for children and their families, often requiring travel to the nephrology centre in order to access urgent care. Third, this trial will compare the risk of relapse at 1 year based on the first event that occurs. While we planned to maintain participants into the assigned arm for follow-up studies by treating repeated relapses with the same antibody they receive, comparison of the effect of these antibodies on the risk of repeated relapses is not included in the present trial. In summary, this study addresses an intervention question that is relevant to children with INS and their families. Results from this study may impact the management of paediatric SD-INS in the near future and provide information on safety of these agents more and more often used in clinical practice. In these children, ofatumumab may allow maintenance of long-term remission of INS without steroids and CNI therapy. Improvement in quality of life, reduction in hospitalisation rates and use of healthcare resources are other important expected benefits. The Institute Giannina Gaslini (trial sponsor) will provide logistic and financial support to the trial through grants from the ministry of health (‘Cinque per mille of IRPEF-Finanziamento della ricerca sanitaria’), the ‘Fondazione Malattie Renali del Bambino’ and the ‘Fondazione La Nuova Speranza’ (‘Progetto integrato per la definizione dei meccanismi implicati nella glomerulo sclerosi focale’). The sponsor will appoint a study monitor. The authors thank the members of the DSMB (Giovanni Candiano, PhD, Gianluca Caridi and Antonella Trivelli, MD) and end point adjudication committee (Drs Fabrizio Ginevri and Enrico Verrina). Contributors: PR, AB and GMG were involved in conception and trial design and in drafting of the article. Disclaimer: The study sponsor had no role in the study design and protocol development of this study. The sponsor will not have any role in the collection, analysis or interpretation of the data, or in the writing of report for publication. The researchers have complete independence from the sources of funding in all aspects of this study. Funding: Giannina Gaslini Children's Hospital will provide intervention and active comparator drugs (ofatumumab and rituximab, respectively) and the infrastructure for the trial. The ‘Fondazione La Nuova Speranza’ (‘Progetto integrato per la definizione dei meccanismi implicati nella glomerulo sclerosi focale’) founded the insurance coverage. Competing interests: None declared. Ethics approval: Local Independent Ethics Committee (Comitato Etico Regione Liguria). Provenance and peer review: Not commissioned; externally peer reviewed. ==== Refs References 1 Cameron JS , Turner DR , Ogg CS The nephrotic syndrome in adults with ‘minimal change’ glomerular lesions . Q J Med 1974 ;43 :461 –88 .4422336 2 McEnery PT , Strife CF Nephrotic syndrome in childhood. Management and treatment in patients with minimal change disease, mesangial proliferation, or focal glomerulosclerosis . Pediatr Clin North Am 1982 ;29 :875 –94 .7050865 3 Kyrieleis HA , Löwik MM , Pronk I Long-term outcome of biopsy-proven, frequently relapsing minimal-change nephrotic syndrome in children. Clin J Am Soc Nephrol 2009 ;4 :1593 –600 . 10.2215/CJN.05691108 19808243 4 Hahn D , Hodson EM , Willis NS Corticosteroid therapy for nephrotic syndrome in children . Cochrane Database Syst Rev 2015 ;18 :CD001533 10.1002/14651858.CD001533.pub5 5 Benz K , Dötsch J , Rascher W Change of the course of steroid-dependent nephrotic syndrome after rituximab therapy . Pediatr Nephrol 2004 ;19 :794 –7 . 10.1007/s00467-004-1434-z 15071769 6 Peskovitz MD , Book BK , Sidner RA Resolution of recurrent focal segmental glomerulosclerosis proteinuria after rituximab treatment . N Engl J Med 2006 ;354 :1961 –3 . 10.1056/NEJMc055495 16672715 7 Nozu K , Iijima K , Fujisawa M Rituximab treatment for posttransplant lymphoproliferative disorder (PTLD) induces complete remission of recurrent nephrotic syndrome . Pediatr Nephrol 2005 ;20 :1660 –3 . 10.1007/s00467-005-2013-7 16133051 8 Guigonis V , Dallocchio A , Baudouin V Rituximab treatment for severe steroid- or cyclosporine-dependent nephroticsyndrome: a multicentric series of 22 cases . Pediatr Nephrol 2008 ;23 :1269 –79 . 10.1007/s00467-008-0814-1 18465150 9 Peters HP , van de Kar NC , Wetzels JF Rituximab in minimal change nephropathy and focal segmental glomerulosclerosis: report of four cases and review of the literature . Neth J Med 2008 ;66 : 408 –15 .19011266 10 Kamei K , Ito S , Nozu K Single dose of rituximab for refractory steroid-dependent nephrotic syndrome in children . Pediatr Nephrol 2009 ;24 :1321 –8 . 10.1007/s00467-009-1191-0 19421786 11 Fernandez-Fresnedo G , Segarra A , Gonza′lez E Trabajo de Enfermedades Glomerulares de la Sociedad Espa∼nola de Nefrologı′a (GLOSEN): rituximab treatment of adult patients with steroid-resistant focal segmental glomerulosclerosis . Clin J Am Soc Nephrol 2009 ;4 :1317 –23 .19578004 12 Fujinaga S , Hirano D , Nishizaki N Single infusion of rituximab for persistent steroid-dependent minimal-change nephrotic syndrome after long-term cyclosporine . Pediatr Nephrol 2010 ;25 : 539 –44 . 10.1007/s00467-009-1377-5 20049616 13 Gulati A , Sinha A , Jordan SC Efficacy and safety of treatment with rituximab for difficult steroid-resistant and -dependent nephrotic syndrome: multicentric report . Clin J Am Soc Nephrol 2010 ;5 : 2207 –12 . 10.2215/CJN.03470410 20798255 14 Prytuła A , Iijima K , Kamei K Rituximab in refractory nephrotic syndrome . Pediatr Nephrol 2010 ;25 :461 –8 . 10.1007/s00467-009-1376-6 20033225 15 Kemper MJ , Gellermann J , Habbig S Long-term follow-up after rituximab for steroid-dependent idiopathic nephrotic syndrome . Nephrol Dial Transplant 2012 ;27 :1910 –5 . 10.1093/ndt/gfr548 22076431 16 Sellier-Leclerc AL , Baudouin V , Kwon T Rituximab in steroid dependent idiopathic nephrotic syndrome in childhood—follow-up after CD19 recovery . Nephrol Dial Transplant 2012 ;27 : 1083 –9 . 10.1093/ndt/gfr405 21810762 17 Ito S , Kamei K , Ogura M Survey of rituximab treatment for childhood-onset refractory nephrotic syndrome . Pediatr Nephrol 2013 ;28 :257 –64 . 10.1007/s00467-012-2319-1 23052656 18 Magnasco A , Ravani P , Edefonti A Rituximab in children with resistant idiopathic nephrotic syndrome . J Am Soc Nephrol 2012 ;23 :1117 –24 . 10.1681/ASN.2011080775 22581994 19 Ravani P , Rossi R , Bonanni A Rituximab in children with steroid-dependent nephrotic syndrome: a multicentre, open-label, non inferiority, randomised controlled trial . J Am Soc Nephrol 2015 ;26 :2259 –66 . 10.1681/ASN.2014080799 25592855 20 Ravani P , Magnasco A , Edefonti A Short-term effects of rituximab in children with steroid- and calcineurin-dependent nephrotic syndrome: a randomised controlled trial . Clin J Am Soc Nephrol 2011 ;6 :1308 –15 . 10.2215/CJN.09421010 21566104 21 Iijima K , Sako M , Nozu K Rituximab for childhood-onset, complicated, frequently relapsing nephrotic syndrome or steroid-dependent nephrotic syndrome: a multicentre, double-blind, randomised, placebo controlled trial . Lancet 2014 ;4 :1273 –81 . 10.1016/S0140-6736(14)60541-9 22 Basu B Ofatumumab for rituximab-resistant nephrotic syndrome . N Engl J Med 2014 ;370 :1268 –70 . 10.1056/NEJMc1308488 24670185 23 Bonanni A , Rossi R , Murtas C Low-dose ofatumumab for rituximab-resistant nephrotic syndrome . BMJ Case Rep 2015 ;2015 :pii: bcr2015210208 10.1136/bcr-2015-210208 24 Robak T Ofatumumab, a human monoclonal antibody for lymphoid malignancies and autoimmune disorders . Curr Opin Mol Ther 2008 ;10 :292 –309 . 25 Vivarelli M , Serafinelli J , Colucci M Ofatumumab in two pediatric nephrotic syndrome patients allergic to Rituximab . Ped Nephrol 2017 ;32 :181 –4 . 26 Bonanni A , Bertelli E , Moscatelli A Ofatumumab-associated acute respiratory manifestations: clinical characteristics and treatment . Br J Clin Pharmacol 2016 ;82 :1146 –8 . 10.1111/bcp.13029 27353034 27 Gellermann J , Weber L , Pape L Mycophenolate mofetil versus cyclosporine A in children with frequently relapsing nephrotic syndrome . J Am Soc Nephrol 2013 ;24 :1689 –97 . 10.1681/ASN.2012121200 23813218 28 Kidney Disease: Improving Global Outcomes (KDIGO) Glomerulonephritis Work Group. KDIGO Clinical Practice Guideline for Glomerulonephritis. Kidney Int 2012;2(Suppl.):139–274. 29 Ravani P , Ponticelli A , Siciliano C Rituximab is a safe and effective long-term treatment for children with steroid and calcineurin inhibitor-dependent idiopathic nephrotic syndrome . Kidney Int 2013 ;84 :1025 –33 . 10.1038/ki.2013.211 23739238

PMC005xxxxxx/PMC5372103.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01463110.1136/bmjopen-2016-014631General practice / Family practiceProtocol1506169617041730Protocol for determining primary healthcare practice characteristics, models of practice and patient accessibility using an exploratory census survey with linkage to administrative data in Nova Scotia, Canada Marshall Emily Gard 1Gibson Richard J 2Lawson Beverley 1Burge Frederick 1 1 Department of Family Medicine, Dalhousie University, Halifax, Nova Scotia, Canada 2 Department of Family Practice, Nova Scotia Health Authority, Halifax, Nova Scotia, CanadaCorrespondence to Dr Emily Gard Marshall; emily.marshall@dal.ca2017 16 3 2017 7 3 e01463111 10 2016 24 1 2017 20 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction There is little evidence on how primary care providers (PCPs) model their practices in Nova Scotia (NS), Canada, what services they offer or what accessibility is like for the average patient. This study will create a database of all family physicians and primary healthcare nurse practitioners in NS, including information about accessibility and the model of care in which they practice, and will link the survey data to administrative health databases. Methods and analysis 3 census surveys of all family physicians, primary care nurse practitioners (ie, PCPs) and their practices in NS will be conducted. The first will be a telephone survey conducted during typical daytime business hours. At each practice, the person answering the telephone will be asked questions about the practice's accessibility and model of care. The second will be a telephone survey conducted after typical daytime business hours to determine what out-of-office services PCP practices offer their patients. The final will be a tailored fax survey that will collect information that could not be obtained in the first 2 surveys plus new information on scope of practice, practice model and willingness to participate in research. Survey data will be linked with billing data from administrative health databases. Multivariate regression analysis will be employed to assess whether access and availability outcome variables are associated with PCP and model of practice characteristics. Negative binomial regression analysis will be employed to assess the association between independent variables from the survey data and health system use outcomes from administrative data. Ethics and dissemination This study has received ethical approval from the Nova Scotia Health Authority and the Health Data Nova Scotia Data Access Committee. Dissemination approached will include stakeholder engagement at local and national levels, conference presentations, peer-reviewed publications and a public website. PRIMARY CARENova Scotia Health Research Foundationhttp://dx.doi.org/10.13039/501100000194PSO-Research Programs- 2012-8293 ==== Body Strengths and limitations of this study This is the first census study in Canada to systematically capture practice and provider characteristics and link them to administrative data to answer questions about models of primary healthcare and healthcare access, comprehensiveness and utilisation. This study was developed in collaboration with local knowledge users in response to an identified priority area for data and research, with contributions to tool development and ensuring the relevance of the research to policy priorities. This study will use multiple methods to collect data from practices and providers, including sources that do not only rely on directly surveying physicians to lessen the burden on providers' time, and when required, using tailored methods to improve physician response rates. Multiple methods allow the establishment of reliability between data sources and to triangulate study findings. Limited coding structures for primary care providers and inconsistencies in data entry may present limitations in the data available in the administrative databases. Response rates for surveys of physicians are historically low; this study will mitigate this limitation with robust recruitment strategies and data collection plan, minimising physician and nurse practitioner time. Introduction Access to and quality of primary healthcare (PHC), including models of collaborative and team-based care, are key areas of interest for policy and research. Accessible, well-coordinated team-based PHC is associated with better health outcomes and lower healthcare system costs.1–6 In Nova Scotia (NS), improved access to PHC and the development of interprofessional teams are major strategic goals of the provincial Department of Health and Wellness and the Nova Scotia Health Authority (NSHA).7 8 However, to date, there is very little empirical evidence on how primary care providers (PCPs) model their primary care practices in NS, what services they offer or what accessibility is like for the average patient. This innovative Models and Access Atlas of Primary Care in Nova Scotia (MAAP-NS) study will create a database, or atlas, of all family physicians and PHC nurse practitioners (ie, PCPs) in NS, accessibility to their services, models of care and PCP characteristics. The second phase of MAAP-NS will link the survey data ‘atlas’ of all PCPs to administrative health databases. This will allow for exploration of research issues such as: the association of models of primary care with chronic disease prevention and management; accessibility of primary care with patient use of emergency departments; and walk-in clinic utilisation and preventive and screening services. This research will: (1) identify areas of PHC requiring enhancement, such as where gaps exist in accessibility (eg, geographical location, time of day and availability of specific primary care services); (2) provide a baseline for evaluating future innovations and (3) create new knowledge about the current landscape of PHC access, models of care being offered and their impacts on utilisation. Access to PHC The vast majority of medical care in NS is primary care provided by family physicians and nurse practitioners. Of the ∼4.1 million medical encounters each year in NS (primary care visits, emergency room visits, surgeries and hospital admissions), 80% are to family physicians and nurse practitioners.9 It is well documented that accessible and well-coordinated primary care is associated with better health outcomes and lower costs to the healthcare system.1 3 5 Adults with an accessible source of primary care are more likely to receive preventive care and experience fewer disparities in care when compared with those without a source of care.10 However, having a PCP does not ensure access; in one study, the majority of participants reported having a PCP but most of those reported having difficulties accessing care.11 In large surveys of Canadians in 2007 and 2008, 84–91% of participants reported having a PCP.3 12 The most recent data indicate that 85.1% of Canadians have a regular PCP.13 Yet, the percentage of Canadians reporting unmet healthcare needs has increased from 1995 (4.2%) to 2014 (11.2%).13 14 Reasons for unmet healthcare needs include wait times, the geographic location of physicians' offices, the limited hours of operation at family physicians' offices and, broadly, that services were not available at the time they were required.12 14 15 Models of primary care practice Increasingly, new models of primary care have been implemented across Canada and in NS. An inventory and synthesis of new models of health services delivery reported on new, emerging and existing models of health services delivery in Canada and found that in NS, there are eight different models (primary and non-PHC) in practice, while in Ontario, there are 32.16 Similarly, the Health Council of Canada has outlined types of primary care teams currently operating in NS17 and in 2008, the Nova Scotia Department of Health and Wellness used the work of Pineault et al18 to describe key features of various models of PHC practice, focusing on elements of accessibility, continuity of care, number of PCPs and other allied health professionals involved in the practice, and collaboration and integration. Currently, the NSHA is using the following classification of primary care models in their work: (1) Solo Family Physician Practice; (2) Group Practice (an organisation of two or more family physicians who work together, share client/patient records, office space, staff, technology and on-call coverage); (3) Interdisciplinary Healthcare Team or Network (includes a group of individuals with diverse training who work together to deliver patient care, such as family physicians, nurses, social workers, dieticians and other healthcare providers) and (4) Other. However, these model frameworks are not a comprehensive list of what models of care are present in NS and there are still several unknowns regarding models of and access to primary care in NS, including: if the listed classifications are sensitive enough to capture the reality of practice offerings in NS; how many physicians are practicing in each mode and what components of access are correlated with each model.17–19 Study context and objectives To ensure the feasibility of obtaining response rates from PCPs, we first conducted a pilot-feasibility study. In the pilot phase, we built an interprofessional research team, including healthcare researchers, family physicians, nurse practitioners and primary care representatives from the provincial Department of Health and Wellness and NSHA; built relationships with our stakeholder groups; and refined our data collection tools, which will be used in the full study. The survey tools were developed from an extensive search of existing tools, including the Canadian Institute of Health Information Primary Health Care Indicators Chartbook.15 Our team strategised methods to obtain data with minimal reliance on PCP responses and with maximised PCP response rates, where required. This included maximising data collection from non-PCP sources. These sources include information publicly available from the College of Physician and Surgeons website; a list of PCPs' names, contact information, training and specialty from the Nova Scotia Department of Health and Wellness; and conducing the first surveys by calling the PCP office and asking the person who answers the phone a number of questions. During this iterative pilot study phase, item responses were examined to determine if they could be shorter and if any required responses from the PCP directly. For example, forms of remuneration were often not known by office phone respondents. Team review of the survey tools and pilot responses led to the development of shorter, targeted surveys that make the most efficient use of PCP time. The full study will use these novel survey methods to investigate the current availability of and access to all PCPs in NS from the patient perspective plus provider and practice characteristics. The objectives of this study are: To create an atlas of all known actively practicing family physicians and nurse practitioners in NS with key characteristics (ie, location, gender, provider identification number, scope of practice and willingness to participate in future research). To determine the models of primary care practice in which all family physicians and nurse practitioners in NS work (ie, solo physician, group practice or interdisciplinary team; disciplines included on primary care teams). To assess access to and availability of all family physicians and nurse practitioners in NS (ie, telephone and email access, message service, office hours, out-of-hours arrangements, accepting new patients, next available appointment, same-day appointments) and examine the relationship to PCP key characteristics and the model of primary care in which they practice. To examine the association of key PCP and practice characteristics (such as location, gender, models of primary care practice, accessibility) and patient health service utilisation outcomes (such as emergency department and walk-in clinic use; prevention and screening services), using information captured in provincial administrative health data. Methods and analysis Survey data collection The survey data collection phase of the study will involve three population surveys: two telephone surveys and one fax survey. For all three surveys, information supplied by the Department of Health and Wellness of NS will be used to contact all primary care practices and providers in NS (projected population size=828 PCPs). Since all three surveys will be complete censuses of the population of PCPs, no sample size calculations are required. The first survey is a telephone survey of all primary care practices and will be conducted during typical daytime business hours. PCP offices will be notified about the study with a letter of information that demonstrates the significance of the research and the team's direct links to policymakers. Each office phone number will be called by a trained research assistant a maximum of five times on different days of the week at different times of the day to maximise the opportunity to reach the practice. If offices are unreachable after five attempts, a personalised letter will be sent to the PCP explaining the study, that their office was unreachable despite five attempts on different days of the week and that they are invited to contact the study by email if they want to arrange participation. At each practice, the person answering the telephone will be asked about whether the practice is accepting new patients, the hours of operation, the next available patient appointment for routine care or for urgent care and whether same-day appointments are offered. Information about the structure of the practice in terms of number of providers, presence of interdisciplinary professionals, sharing of resources and the presence of on-call and after-hours care will also be collected (see online supplementary appendix A). For group practices, data will be collected for each individual PCP working in the practice. Colocation and collaboration indicators will be used to determine group practices versus colocated independent PCPs. 10.1136/bmjopen-2016-014631.supp1supplementary appendix Similarly, the second survey will be a telephone survey of all primary care practices but will be conducted after typical business hours. It will determine if practices provide, via recorded message or on-call service, information on out-of-hours arrangements for patients and what they are (see online supplementary appendix B). 10.1136/bmjopen-2016-014631.supp2supplementary appendix The third survey is a fax survey of all PCPs in NS. We will use fax communication to reach providers directly as it reflects the communication technologies currently in use by PCPs in NS; we heard from PCPs in our pilot study that this was their preferred medium for the surveys as it fits directly into their work flow. The information sought will concern scope of practice and aspects of each provider's model of primary care and will include a request for consent to be contacted for future research (see online supplementary appendix C). This survey will be tailored to each PCP to remove items that ask about information that was obtained in the first two surveys. 10.1136/bmjopen-2016-014631.supp3supplementary appendix Detailed protocols and instructions for the provider and practice surveys will be used to train and guide research assistants (see online supplementary appendix D). Data collection began January 2014 and is ongoing, with an anticipated end date of June 2017. All data will be entered into spreadsheets and analysed with SPSS, V.22 (IBM Corp. Released 2013. IBM SPSS Statistics for Windows, V.22.0. Armonk, New York, USA: IBM Corp). 10.1136/bmjopen-2016-014631.supp4supplementary appendix Data linkage In Canada, PCPs deliver healthcare free of cost to the general public. To be remunerated, they bill the provincial government for their work using unique billing codes for each service provided to each patient. Some family physicians and nurse practitioners are paid using alternative methods, such as salaries, but they still submit ‘shadow’ billing information to the provincial government. These administrative billings databases document which medical services (including referrals to specialists and use of emergency departments) are provided to which patient by which provider. These databases also contain some patient information, such as diagnostic codes for chronic diseases and procedures delivered to patients. This study will not access the individual patient data, but will have access to patient data aggregated at the provider level. Following the survey data collection, survey data will be linked with billing data from the administrative databases held by Health Data Nova Scotia (HDNS). The PCPs' individual billing numbers will be sent to Medavie Blue Cross, which maintains the Medical Services Insurance (MSI) database for NS, for encryption. Accompanying the billing numbers will be a unique identifier generated within the MAAP-NS data set. The encrypted billing numbers and MAAP-NS ID numbers will be sent to the Maritime SPOR Support Unit (MSSU). The encrypted billing numbers will allow extraction of billing data for the PCPs in the study and the MAAP-NS ID will allow linkage of the PCPs' survey data to their administrative billing data (see figure 1). Figure 1 Data linkage in Models and Access Atlas of Primary Care—Nova Scotia. Analysis plan Initial analyses to develop the atlas will be primarily descriptive. Frequencies, proportions, measures of central tendency and variance will be applied to all data. For example, mean age (with SD), proportion of men and women, mean time since graduation, proportion rural/urban, proportion of PCPs practicing in each model of primary care will be calculated, as well as access and availability frequencies, means and distributions. Next, bivariate analyses will be used to examine the relationship between key characteristics of actively practicing PCPs, model of primary care practice and accessibility (eg, mean time to next available non-urgent/urgent appointment in days by age of PCP, rural/urban characteristics or by model of primary care practice). Finally, multivariate regression analysis will be employed to assess whether access and availability outcome variables (eg, accepting new patients, wait times) are associated with physician and model of practice characteristics. For example, a logistic regression will be used to assess if primary care practice structures are associated with accepting new patients (yes vs no) after controlling for geographical location, number of hours available for patients per week, physician gender, age and years since graduation. The development of indices of collaboration will involve factor analysis of the survey items related to collaborative elements. These include: the sharing of physical resources (waiting rooms, examination room and equipment); sharing of human resources (staff, on-call, coverage during absences); sharing of patients and records; and formal or informal practice protocols. Development of understanding of models of care will include examining combinations of constituent elements related to the descriptions of primary care models in the literature, including types and numbers of interprofessional care providers colocated, methods of collaboration, methods of remuneration and clinic governance. The various models and their constituent components will all be analysed to see if there is a relationship with key outcomes of interest, including availability, accessibility and comprehensiveness of services offered. Billing data linkage will allow for analysis to see if provider characteristics, practice characteristics, aspects of collaboration and models of care predict healthcare utilisation patterns of patients, rates of using walk-in clinics and emergency department use, and whether patients are more likely to receive appropriate chronic disease prevention and management. See supplementary appendix E for the objectives and methods for the billing data linkage analysis. 10.1136/bmjopen-2016-014631.supp5supplementary appendix Negative binomial regression analysis will be employed to assess the association between independent variables of interest such as models of primary care and accessibility and ‘count’ data, including the total number of visits to an emergency department, walk-in clinic, total ambulatory visits and total targeted services. Unadjusted regression will be followed by multivariate techniques to allow for the control of various patient, physician and practice factors that may modify the effect of the association. If a disproportionately large number of patients did not experience any emergency or walk-in visits (ie, count=0), these analyses will be repeated using logistic regression techniques to evaluate the probability of having made at least one visit versus none. To address the possibility of non-independence among cases (ie, that patients of one provider may be more similar to one another than they are to patients of another provider), we will attempt multilevel regression modelling. This procedure adds a variable to each regression model for which each provider has a value. This allows the inclusion of individuals and groups of individuals in the same model to avoid flouting the assumption of independent cases. Dissemination All information received from participants will be treated in a confidential manner and stored in password-protected files on a secure research drive at Dalhousie University Department of Family Medicine, Primary Care Research Unit. The linked data set will be held and analysed in the high-security facilities of the MSSU. Knowledge translation and exchange activities will be facilitated by numerous committees and working groups in which the research team members are actively engaged, including (1) the Canadian Association of Health Services and Policy Research (CAHSPR) Primary Healthcare Theme Working Group, which provides opportunities for engagement and sharing of knowledge with PHC decision makers and researchers from across Canada, (2) the Transdisciplinary Understandings and Training on Research—Primary Health Care (TUTOR-PHC) programme, which provides interdisciplinary PHC research training and mentorship to a cohort of 12 graduate students from multiple disciplines from across the country annually and (3) the Canada Working Group on Primary Healthcare Improvement. As well as presentations at national and international meetings such as CAHSPR and North American Primary Care Research Group (NAPCRG), MAAP-NS findings will be made public on a website developed, especially for this study. The website will be maintained on the Dalhousie University server (http://www.dal.ca/maapstudy) with up-to-date findings of the present study, as they become available, and will encourage visitors to provide feedback. The overarching goal of this study is to provide evidence for models of primary care practice that lead to the highest level of access for the right care at the right time from the right provider. A provincial assessment of access is inline with a population health approach and will facilitate improving access to primary care to reduce growing health disparities. This study will generate a database (atlas) of PCPs in NS, accessibility to their services and models of care and will link this information to administrative databases. The atlas will be a valuable resource for answering future research and health policy questions. By establishing a database that contains relevant indicators of access to primary healthcare province-wide, it will be possible to determine what factors contribute to a more accessible primary care system, identify where gaps exist and where improvements can be made. Funding: This study is supported by an Establishment Grant from the Nova Scotia Health Research Foundation. Competing interests: None declared. Ethics approval: Research Ethics Board of the Nova Scotia Health Authority and from the Health Data Nova Scotia Data Access Committee. Provenance and peer review: Not commissioned; externally peer reviewed. ==== Refs References 1 Starfield B , Shi L , Macinko J Contribution of primary care to health systems and health . Milbank Q 2005 ;83 :457 –502 . 10.1111/j.1468-0009.2005.00409.x 16202000 2 Khan E , McIntosh C , Sanmartin C Working paper series: primary health care teams and their impact on process and outcomes for care (catalogue number 82-622-X no. 002) Ottawa, ON: Statistics Canada , 2008 . 3 Schoen C , Osborn R , Doty MM Toward higher-performance health systems: adults’ health care experiences in seven countries, 2007 . Health Aff (Millwood) 2007 ;26 :w717 –34 . 10.1377/hlthaff.26.6.w717 17978360 4 Russell GM , Dahrouge S , Hogg W Managing chronic disease in Ontario primary care: the impact of organizational factors . Ann Fam Med 2009 ;7 :309 –17 . 10.1370/afm.982 19597168 5 Baicker K , Chandra A Medicare spending, the physician workforce, and beneficiaries’ quality of care . Health Aff (Millwood) 2004 ;4 :w184 –97 . 10.1377/hlthaff.w4.184 6 Hollander MJ , Kadlec H , Hamdi R Increasing value for money in the Canadian healthcare system: new findings on the contribution of primary care services . Healthc Q 2009 ;12 :32 –44 . 10.12927/hcq.2013.21050 7 Nova Scotia Health Authority . Nova Scotia Health Authority Business Plan 2015 http://www.nshealth.ca/sites/nshealth.ca/files/2015-16_nsha_business_plan_-_master_final_2015-06-03.pdf 8 Capital District Health Authority . The Promise of Primary Health Care: The Community Master Plan Working Document. (n.d. ). http://www.cdha.nshealth.ca/system/files/sites/123/documents/community-master-plan.pdf 9 Nova Scotia Department of Health and Wellness, Business Intelligence Analytics and Privacy Division . Retrieved 30 September 2016 http://www.nshealth.ca/about-us 10 Beal AC , Doty MM , Hernandez SE Closing the divide: how medical homes promote equity in health care. Results from the Commonwealth Fund 2006 Health Care Quality Survey , 2007 http://www.commonwealthfund.org/usr_doc/1035_Beal_closing_divide_medical_homes.pdf?section=4039 11 Wellstood K , Wilson K , Eyles J ‘Reasonable access’ to primary care: assessing the role of individual and system characteristics . Health Place 2006 ;12 :121 –30 . 10.1016/j.healthplace.2004.10.010 16338628 12 Canadian Institute for Health Information . Analysis in Brief. Experiences with primary health care in Canada, July 2009 http://secure.cihi.ca/cihiweb/products/cse_phc_aib_en.pdf 13 Statistics Canada . Unmet health care needs, 2014. Health Fact Sheets. Statistics catalogue no. 82-625-X 2016 Retrieved 30 September 2016 http://www.statcan.gc.ca/pub/82-625-x/2016001/article/14310-eng.htm 14 Sanmartin C , Houle C , Tremblay S Changes in unmet health care needs . Health Rep 2002 ;13 :15 –21 .12743957 15 Canadian Institute for Health Information (CIHI) . Primary health care (PHC) Indicators chartbook: an illustrative example of using PHC data for indicator reporting . Ottawa, ON : CIHI , 2008 https://secure.cihi.ca/estore/productSeries.htm?pc=PCC426 16 Hollander Analytical Services Ltd . Assessing new models for the delivery of medical services: inventory and synthesis. August 2003 http://www.physicianhr.ca/reports/hollander-e.pdf 17 Health Council of Canada . Primary health care teams—a Nova Scotia perspective 2009 http://www.healthcouncilcanada.ca/docs/rpts/2009/TeamsInAction_NovaScotia_Summary.pdf 18 Pineault R , Levesque JF , Roberge D Primary care services organizational models and the population's care experience, Government du Quebec, Centre de recherché de l'hopital Charles LeMoyne. September 2008 http://www.greas.ca/evolution/pdf/843_modeles_organisation_ang.pdf 19 Health Council of Canada . Teams in action: primary health care teams for Canadians 2009 http://www.dsrf.org/media/teamsinaction.pdf

PMC005xxxxxx/PMC5372104.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01401810.1136/bmjopen-2016-014018Occupational and Environmental MedicineResearch1506171617311692Association of respiratory symptoms and asthma with occupational exposures: findings from a population-based cross-sectional survey in Telemark, Norway Abrahamsen R 12Fell A K M 1Svendsen M V 1Andersson E 3Torén K 3Henneberger P K 4Kongerud J 25 1 Department of Occupational and Environmental Medicine, Telemark Hospital, Skien, Norway 2 Institute of Clinical Medicine, Faculty of Medicine, University of Oslo, Oslo, Norway 3 Department of Occupational and Environmental Medicine, Institute of Medicine, The Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden 4 Respiratory Health Division, National Institute for Occupational Safety and Health (NIOSH), Morgantown, West Virginia, USA 5 Department of Respiratory Medicine, Oslo University Hospital, Oslo, NorwayCorrespondence to Dr Regine Abrahamsen; regine.abrahamsen@sthf.no2017 22 3 2017 7 3 e01401824 8 2016 15 2 2017 24 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives The aim of this study was to estimate the prevalence of respiratory symptoms and physician-diagnosed asthma and assess the impact of current occupational exposure. Design Cross-sectional analyses of the prevalence of self-reported respiratory health and association with current occupational exposure in a random sample of the general population in Telemark County, Norway. Settings In 2013, a self-administered questionnaire was mailed to a random sample of the general population, aged 16–50, in Telemark, Norway. The overall response rate was 33%, comprising 16 099 responders. Outcome measures The prevalence for respiratory symptoms and asthma, and OR of respiratory symptoms and asthma for occupational groups and exposures were calculated. Occupational exposures were assessed using self-reported exposure and an asthma-specific job-exposure matrix (JEM). Results The prevalence of physician-diagnosed asthma was 11.5%. For the occupational groups, the category with agriculture/fishery workers and craft/related trade workers was associated with wheezing and asthma attack in the past 12 months, showing OR 1.3 (1.1 to 1.6) and 1.9 (1.2 to 2.8), respectively. The group including technicians and associated professionals was also associated with wheezing OR 1.2 (1.0 to 1.3) and asthma attack OR 1.4 (1.1 to 1.9). The JEM data show that exposure to flour was associated with wheezing OR 3.2 (1.4 to 7.3) and woken with dyspnoea OR 3.5 (1.3 to 9.5), whereas exposures to diisocyanates, welding/soldering fumes and exposure to vehicle/motor exhaust were associated with dyspnoea OR 2.9 (1.5 to 5.7), 3.2 (1.6 to 6.4) and 1.4 (1.0 to 1.8), respectively. Conclusions The observed prevalence of physician-diagnosed asthma was 11.5%. The ‘manual’ occupations were associated with respiratory symptoms. Occupational exposure to flour, diisocyanates, welding/soldering fumes and vehicle/motor exhaust was associated with respiratory symptoms in the past 12 months and use of asthma medication. However, prospective data are needed to confirm the observed associations. respiratory diseaseEPIDEMIOLOGYQuestionaireOccupationalPopulation-based studyJob exposure matrix ==== Body Strengths and limitations of this study The study included 16 099 subjects from the general population in Telemark, Norway, which is a county with a variety of occupational exposures including large industrial clusters, craft, trade and agricultural activities. The study used cross-sectional data to investigate possible associations of self-reported respiratory symptoms and asthma with occupation and occupational exposures. The study is strengthened by the use of an asthma-specific job-exposure matrix on a large random sample from the general population. A potential study limitation is selection bias due to the cross-sectional design and non-participation; however, inverse probability of participation weights was used to minimise selection bias from non-participation. Introduction Studies have shown that at least one out of three adults in Norway will experience long-term respiratory symptoms during their lifetime.1 2 In Europe, a substantial proportion of adults report symptoms that are initiated or exacerbated by environmental and/or occupational exposures.3–9 Globally, 15–20% of all asthma cases are considered to be related to occupation with population attributable risks (PAR) as high as 50% among non-smokers.4 A Swedish study in 2011 reported a cumulative incidence for adult-onset asthma of 2.3% and PAR of occupational asthma of 17.3% among men and 5.1% among women.10 In southern Norway, an area which also includes Telemark County, it was reported in 2013 that the prevalence of wheezing and chest tightness during the past month was slightly above the country average with 5.9% and 5.1%, respectively.11 No information regarding respiratory symptoms by occupation was available in that study, but the subjects with the lowest educational level had the highest prevalence of wheezing and chest tightness. The incidence of asthma in Telemark County has so far not been studied, but it is known that the use of medication for chronic respiratory disease in this region is above the national average. In 2014, the Norwegian medication prescription database reported that the number of users of medication for chronic respiratory diseases was 84 per 1000 inhabitants in Norway, while in Telemark the number was 93 per 1000 inhabitants.12 Historically, Telemark has been one of the main onshore industrialised centres in Norway. The region still has a high proportion of industrial workers employed in companies producing nitrate fertiliser, organic chlorine and chlor-alkali derivatives, polyvinylchlorides, cement and refined manganese alloys, among others. Whereas these companies have reduced the number of workers during the past decades, several smaller industrial service providers have been established. These employ many of the previous industrial workers which perform similar tasks as they did in their past jobs. Thus, inclusion of the large companies and the whole working population from the area is of importance when assessing prevalence of respiratory symptoms and impact of occupational exposure. In Norway, there has been an increase in cases of chronic respiratory diseases referred to all Departments of Occupational and Environmental Medicine. However, studies regarding the importance of occupational exposures are still reporting conflicting results and many have a lack of specific exposure information. Moreover, there is a need for updated information regarding the impact of occupational exposures on respiratory health because production processes, raw-material and additives of products, and organisation of work change. This makes the identification of occupational groups at risk for chronic respiratory diseases difficult and hence there is uncertainty where to implement preventative measures. The aim of this study was to estimate the prevalence of respiratory symptoms and physician-diagnosed asthma and assess the impact of current occupational exposure in a large random sample from Telemark County, Norway. Methods Study population A cross-sectional survey was carried out from February to August 2013 in Telemark County, which is located in the south-eastern part of Norway and has a population of around 170 000. Grenland is the most densely populated part of the County, includes several industrial sites and has a population of 100 000. A sample of 50 000 male and female subjects from the ∼80 000 residents in Telemark, aged 16–50 years, was drawn randomly using the services of the Norwegian national population registry. The subjects between 16 and 18 years of age were included to take into account those who have to complete a mandatory training programme as apprentices attending school and work simultaneously or those entering the workforce without further education. Design of the study The participants were asked to complete a postal questionnaire (see online supplementary file 1) and mail it back in a prepaid envelope. Two reminders were sent by mail. The questionnaire included questions regarding occupational history and exposure, physician-diagnosed asthma, respiratory symptoms, allergy, use of asthma medication and possible confounders. The questionnaire was based on the European Community Respiratory Health Survey (ECRHS) questionnaire and a questionnaire from a similar study from west Sweden.10 The questions regarding physician-diagnosed asthma and respiratory symptoms are consistent with other studies addressing asthma and asthma-like symptoms.10 13–16 10.1136/bmjopen-2016-014018.supp1supplementary file Outcomes Occupational history and exposure In the postal questionnaire, the participants were asked to list their occupational history in free text including occupational title, industry or sector, some key words indicating main tasks and period of employment. The participants with sufficient information regarding occupational history were then classified according to their self-reported occupation in 2013 (current occupation) with the use of the International Standard Classification of Occupations (ISCO-88) coding system17 by a trained research assistant. Each occupation was given a code according to the ISCO-88 which classifies occupations into 10 major groups that are related to formal education/qualifications. The participants were first classified into these groups as shown in table 1, and certain groups were then combined for the analyses: (1) legislators, senior officials, managers and professionals (groups I and II); (2) technicians and associated professionals (group III); (3) clerks and service workers and shop and market sales workers (groups IV and V); (4) skilled agriculture, fishery workers and craft and related trade workers (groups VI and VII) and (5) plant, machine operators and assemblers and elementary occupations (groups VIII and IX). Group 1 was used as an unexposed reference group. All occupations with at least 6 months employment were classified. Table 1 Population characteristics for responders and non-responders Register based Responders (n=16 099) Non-responders (n=33 901) n Per cent n Per cent Population size  Urban areas/Grenland 10 296 64.0 21 348 63.0  Rural areas/outside Grenland 5803 36.0 12 553 37.0 Gender  Men 7159 44.5 18 358 54.2  Women 8940 55.5 15 543 45.8 Age (years)  16–30 5282 32.8 14 626 43.1  31–40 4126 25.6 9085 26.8  41–50 6691 41.6 10 190 30.1 Responders (n=16 099) Non-responders (n=260) Questionnaire based n Per cent n Per cent Smoking habits  Never-smokers 8950 55.6 119 46  Past smokers 3271 20.3 77 30  Current smokers 3749 23.3 64 25  Missing 129 0.8 0 0 Home/housing conditions  Detached 11 181 69.5 NA  Row 1565 9.7  Apartment 2879 17.9  Other 325 2.0  Missing 149 0.9 Current occupation (ISCO-88)    Armed forces (Group 0) 43 0.3 0 0  Legislators, senior officials and managers (Group I) 498 3.1 5 2  Professionals (Group II) 1748 10.9 35 13  Technicians and associated professionals (Group III) 3000 18.6 52 20  Clerks (Group IV) 674 4.2 12 5  Service workers and shop and market sales workers (Group V) 1717 10.7 46 18  Skilled agriculture and fishery workers (Group VI) 158 1.0 2 1  Craft and related trade workers (Group VII) 1240 7.7 35 13  Plant and machine operators and assemblers (Group VIII) 779 4.8 21 8  Elementary occupations (Group IX) 438 2.7 12 5  Missing* 5804 36.1 40 15 *The 5804 missing are the responders without data regarding current occupation. Of these, 58.9% state that they have been employed some time during the past 12 months, but have not presented enough occupational data to be classified according to ISCO-88. The remaining 41.1% did not state that they had been employed during the past 12 months. ISCO-88, International Standard Classification of Occupations; NA, not applicable. We also used an asthma-specific job-exposure matrix (JEM) developed for the northern European countries (N-JEM) to assess the participants' occupational exposures based on their self-reported jobs in 2013.8 9 A few job titles (n=39) specific to industries in Telemark County were added to the N-JEM. The N-JEM defines an occupation as exposed based on the assumption that at least half of the subjects with this specific code would have a high probability of being exposed to irritants or allergens. As an example, the category ‘flour-associated antigens’ consists of bakers, pastry makers, grain millers and operators of baked goods. The N-JEM data can be collapsed into six main exposure groups: high molecular weight (HMW) agents, low molecular weight (LMW) agents, irritating agents, accidental peak exposures to irritants, uncertain or low exposures and an unexposed reference group. Respiratory symptoms Asthma, use of asthma medication, respiratory symptoms and nasal allergies were defined as positive responses to the following questions: ever asthma: ‘Have you ever had asthma?’, physician-diagnosed asthma: ‘Have you been diagnosed by a physician as having asthma?’, asthma attack: ‘Have you had an attack of asthma in the past 12 months?’, use of asthma medication: ‘Do you currently use asthma medication?’, ever wheezing: ‘Have you ever had whistling or wheezing in the chest?’, wheezing past 12 months: ‘Have you had whistling or wheezing in the chest at any occasion during the past 12 months?’, woken with dyspnoea: ‘Have you been woken by an attack of shortness of breath at any time in the past 12 months?’, chronic cough: ‘Have you had a persistent cough during the last years?’, productive cough: ‘Do you usually bring up any phlegm or do you have phlegm in your chest, that is difficult to cough up?’, nasal allergy as a marker for allergic status: ‘Do you have an allergy with nasal symptoms, including hay-fever?’. These symptoms were used to describe the population, while only the symptoms addressing the past 12 months, current asthma medication and physician-diagnosed asthma were used to assess associations with current occupation or exposure. A symptom score was then calculated based on the number of affirmative answers to the four questions addressing symptoms, and use of asthma medication in the past 12 months. Covariates and confounders The participant's age, gender and residential address was census-derived, whereas housing condition, smoking habits, household smoking and exposure to damp and mould at home was based on self-report from the questionnaires. For housing condition, the following question was used: ‘What type of home do you live in?’ The alternative categories were: detached house, row house/semidetached house, apartment or other. Smoking was defined as follows: Current smokers answered affirmative to the question: ‘Do you smoke every day (also applies if you only smoke a few cigarettes, cigars or light a pipe each day)?’ or the question: ‘Do you only smoke occasionally (not each day, but weekends, parties or similar)?’ Past smoker were those who answered affirmative to the question: ‘Did you use to smoke?’ Those not answering any of the three questions were defined as missing, and the remaining as never-smokers. Passive smoking was defined as an affirmative answer to ‘Does anyone smoke inside your current home?’ Damp and mould exposure at home was defined as positive answer to at least one of the following questions: ‘Have you ever had the following in your property? Water damage inside in the dwelling on walls, floors or ceilings? Dented plastic mats, stained plastic or hardwood flooring that has been darkened by the moisture?’ or ‘Visible mould on walls, floors or ceilings?’ Ethics The REC only allowed the sending of two reminders to the study sample. The participants signed and returned a consent form. Statistical analysis The subjects with missing or insufficient information regarding current occupation were excluded from the analyses addressing occupational exposure. When respondents left an answer blank, these were handled as ‘no’ (not having the symptom). Sensitivity analyses were performed with and without missing. The selection of individual jobs for the analysis assessing occupation/exposure and respiratory outcomes was based on the work codes with at least 100 respondents in the major groups 5–9 (1–4 are those with non-exposure or low exposure) supplemented with the occupational groups previously identified as having changed work because exposures at work affected their breathing.18 χ2 and Fisher’s exact tests were used to compare groups with regard to the prevalence of respiratory symptoms. To estimate the effects of occupations, occupational groups and occupational exposures, the ORs and corresponding 95% CIs were computed by logistical regression. Four different analyses were performed separately; for each occupation, for the occupational groups, for large exposure groups and for the single exposures. Unadjusted and adjusted estimates were calculated. The estimates were adjusted for age, gender, area of residence, smoking status, household smoking, damp and mould at home and housing conditions. To adjust for possible participation bias, all analyses were performed on weighted data sets. The weighting was based on inverse probability of participation among all 50 000 invited and calculated for each combination of gender, 5-year age intervals and area of residence by grouping of postal codes.19 Each combination of these background variables had at least 19 responders when calculating weights. Additionally, weights based on inverse probability of participation by use of asthma medication and chronic cough derived from 260 participants in the non-responder study were performed. The final weights used in the current analysis were the product of these two weights. As part of the sensitivity analysis, the statistical analyses were performed with and without these weights to evaluate how adjusting for non-response affected the study results.19 PAR was calculated for some exposure and symptoms. The statistical significance level was set at p<0.05 and results reaching significance are marked in bold in the tables. Statistical analysis was performed with IBM SPSS Statistics for Windows (V.23, IBM SPSS, Armonk, New York, USA) using the SPSS complex sample module where appropriate. Results Participation The overall response rate to the Telemark study was 33%, comprising 16 099 responders. Out of the 16 099 responders, 10 004 (20%) responded to the first mailed questionnaire, while 3123 (6%) responded to the first reminder and 2972 (6%) to the last reminder. The inclusion of participants stopped 3 months after the second reminder was mailed. Eighty-three per cent of the responders (13 302) reported being employed in the past 12 months. Among all 16 099 responders, 10 295 (63.9%) reported their occupation in 2013 and were classified by the ISCO-88 (table 1). Repeating the analyses without responses with missing values or without adjustment for non-participation did not significantly influence the results presented. Study population The population characteristics of responders and non-responders are presented in table 1. The whole sample including 16 099 responders contained more women than men (55.5% vs 44.5%, respectively). Approximately twice as many lived in the urban area of Grenland (64.0%) as outside this area (36.0%). Seventy per cent of the subjects lived in a detached house and 18% in apartments. By age, 41.6% of the responders were between 41 and 50 years, while 32.8% and 25.6% were between 16 and 30 years and 31 and 40 years, respectively. A little more than half (55.6%) of the responders were never-smokers while the current smokers (23.3%) slightly outnumbered the past smokers (20.3%). Prevalence of respiratory symptoms, smoking habits and occupation The prevalence of respiratory symptoms and asthma are shown in figure 1. The most common respiratory outcomes ever experienced were nasal allergies (29.9%), wheezing (26.7%) and chronic cough (20.6%). The most common outcome in the past 12 months was wheezing (20.0%). The prevalence of a history of physician-diagnosed asthma was 11.5%, and 4.1% reported an asthma attack in the past 12 months. Among the 2137 responders (13.3%) who reported ‘ever asthma’, 1727 reported their age at which they first experienced asthma symptoms and 71.3% of those reported to have had their first symptoms before they turned 18 years (n=1231). The prevalence of physician-diagnosed asthma declined with advancing age: 12.4% if 16–30 years, 11.4% if 31–40 years and 10.9% if 41–50 years. Out of the 1839 responders reporting physician-diagnosed asthma, 893 (48.1%) reported current use of asthma medication and 531 (28.6%) reported an asthma attack in the past 12 months. Figure 1 The prevalence of respiratory symptoms and asthma ever and for the past 12 months in the Telemark study. The prevalence of respiratory symptoms and diseases by the possible confounders, age, gender, area of residence, smoking habits and housing conditions, is shown in table 2. Table 2 Prevalence of asthma, respiratory symptoms ever and symptoms during the past 12 months by possible confounders: area of residence, smoking habits, housing conditions, age and gender Symptoms or disease Area of residence Smoking habits Age (years) Gender Housing conditions Number Yes/no Grenland Outside Grenland p Value* Never- smokers Past smokers Current smokers p Value† 16–30 31–40 41–50 p Value† Men Women p Value* Detached Row/ semidetached Apartment Other p Value† Ever  Ever asthma 2137/13962 1348 (13.1) 789 (13.6) 0.371 1128 (12.6) 471 (14.4) 526 (14.0) 0.012 759 (14.4) 553 (13.4) 825 (12.3) 0.005 871 (12.2) 1266 (14.2) <0.001 1416 (12.7) 232 (14.8) 432 (15.0) 45 (13.8) 0.002  Physician-diagnosed asthma 1857/14042 1178 (11.4) 679 (11.7) 0.626 983 (11.0) 407 (12.4) 457 (12.2) 0.032 656 (12.4) 471 (11.4) 730 (10.9) 0.036 757 (10.6) 1100 (12.3) 0.001 1226 (11.0) 207 (13.2) 376 (13.1) 39 (12.0) 0.002  Ever wheezing 4294/11805 2788 (27.1) 1 506 (26.0) 0.124 1893 (21.2) 988 (30.2) 1399 (37.3) <0.001 1323 (25.0) 1125 (27.3) 1846 (27.6) 0.005 1858 (26.0) 2436 (27.2) 0.067 2919 (26.1) 436 (27.9) 837 (29.1) 81 (24.9) 0.008  Chronic cough 3319/12780 2210 (21.5) 1109 (19.1) <0.001 1663 (18.6) 636 (19.4) 1001 (26.7) <0.001 1082 (20.5) 853 (20.7) 1384 (20.7) 0.959 1334 (18.6) 1985 (22.2) <0.001 2234 (20.0) 347 (22.2) 653 (22.7) 68 (20.9) 0.006  Productive cough 2363/13736 1589 (15.4) 774 (13.3) <0.001 1006 (11.2) 425 (13.0) 921 (24.6) <0.001 830 (15.7) 533 (12.9) 1000 (14.9) 0.001 1018 (14.2) 1345 (15.0) 0.145 1494 (13.4) 252 (16.1) 539 (18.7) 64 (19.7) <0.001  Nasal allergies 4814/12085 3148 (30.6) 1666 (28.7) 0.013 2682 (30.0) 1030 (31.5) 1087 (29.0) 0.073 1461 (27.7) 1298 (31.5) 2055 (30.7) <0.001 2122 (29.6) 2692 (30.1) 0.522 3388 (30.3) 504 (32.2) 832 (28.9) 68 (20.9) <0.001 Past 12 months  Wheezing 3226/12963 2093 (20.3) 1133 (19.5) 0.227 1378 (15.4) 620 (19.0) 1213 (32.4) <0.001 1021 (19.3) 792 (19.2) 1413 (21.1) 0.015 1371 (19.2) 1855 (20.7) 0.012 2164 (19.4) 333 (21.3) 645 (22.4) 65 (20.0) 0.002  Asthma attack 664/15 035 447 (4.3) 217 (3.7) 0.070 345 (3.9) 138 (4.2) 174 (4.6) 0.119 222 (4.2) 136 (3.3) 306 (4.6) 0.005 211 (2.9) 453 (5.1) <0.001 412 (3.7) 71 (4.5) 162 (5.6) 13 (4.0) <0.001  Woken by dyspnoea 1179/14920 790 (7.7) 389 (6.7) 0.024 533 (6.0) 243 (7.4) 396 (10.6) <0.001 318 (6.0) 264 (6.4) 597 (8.9) <0.001 503 (7.0) 676 (7.6) 0.201 766 (6.9) 126 (8.1) 258 (9.0) 22 (6.8) 0.001  Current use of asthma medication 1171/14928 730 (7.1) 441 (7.6) 0.242 610 (6.8) 262 (8.0) 291 (7.8) 0.035 376 (7.1) 262 (6.3) 533 (8.0) 0.006 435 (6.1) 736 (8.2) <0.001 762 (6.8) 138 (8.8) 243 (8.4) 23 (7.1) 0.002 Results reaching significance, p<0.05, are marked in bold. *Fisher’s exact test. †Pearson χ2 test. Ever having respiratory symptoms and woken by dyspnoea attack in the past 12 months, but not the asthma outcomes are more common in Grenland as compared with outside Grenland. The prevalence of physician-diagnosed asthma, use of asthma medication and respiratory symptoms in the past 12 months were associated with smoking and housing condition. The prevalence of physician-diagnosed asthma decreased with increasing age and was also more common among women than men. The prevalence of respiratory symptoms in the past 12 months and nasal allergies were associated with age category. Association of self-reported current job categories and specific occupation with respiratory symptoms and disease The relationship between respiratory symptoms/diseases and the participants' occupations is shown in table 3. Table 3 Adjusted* OR (ORadj) for respiratory symptoms past 12 months, current use of asthma medication and ever physician-diagnosed asthma by current occupation (ISCO-88)† Major groups (ISCO-88 first digit) Lower respiratory tract Total N Wheezing ORadj Woken with dyspnoea ORadj Asthma attack ORadj Use of asthma medication ORadj At least one symptom last 12 months ORadj Physician- diagnosed asthma ORadj 1. Legislators, senior officials and managers and professionals and armed forces (Groups 0, I and II only)—Reference 2204 1 1 1 1 1 1 2. Technicians and associated professionals (Group III) 3091 1.2 (1.0 to 1.3) 0.95 (0.76 to 1.2) 1.4 (1.1 to 1.9) 1.2 (0.99 to 1.5) 1.1 (0.99 to 1.3) 0.95 (0.79 to 1.1) 3. Clerks and Service workers and shop and market sales workers (Groups IV and V) 2547 1.1 (0.95 to 1.3) 1.0 (0.82 to 1.3) 1.3 (0.95 to 1.8) 1.0 (0.82 to 1.3) 1.1 (0.93 to 1.2) 0.89 (0.74 to 1.1) 4. Skilled agriculture and fishery workers and craft and related trade workers (Groups VI and VII) 1467 1.3 (1.1 to 1.6) 1.2 (0.91 to 1.6) 1.9 (1.2 to 2.8) 1.1 (0.83 to 1.6) 1.2 (1.0 to 1.5) 0.80 (0.62 to 1.0) 5. Plant and machine operators and assemblers and Elementary occupations(Groups VIII and IX) 1294 1.2 (1.0 to 1.5) 1.2 (0.90 to 1.6) 1.1 (0.73 to 1.7) 1.1 (0.81 to 1.5) 1.2 (1.0 to 1.5) 0.80 (0.63 to 1.0) Specific occupations (ISCO-88 four digits) Legislators, senior officials, managers, professionals and armed forces (Groups 0, I and II only)—Reference 2204 1 1 1 1 1 1 Healthcare professionals  Health-associated professionals not classified elsewhere (3229) 630 1.4 (1.1 to 1.7) 1.2 (0.82 to 1.8) 1.4 (0.87 to 2.4) 2.1 (1.5 to 3.1) 1.4 (1.1 to 1.7) 0.79 (0.56 to 1.1)  Nursing (3230–3232) 435 1.2 (0.86 to 1.5) 0.79 (0.49 to 1.3) 1.2 (0.68 to 2.1) 0.78 (0.48 to 1.3) 1.0 (0.77 to 1.4) 0.87 (0.60 to 1.2) Cooks (5122) 168 1.4 (0.90 to 2.1) 1.2 (0.65 to 2.1) 1.5 (0.65 to 3.3) 1.2 (0.57 to 2.5) 1.3 (0.84 to 1.9) 0.91 (0.54 to 1.5) Waiters, waitresses and bartenders (5123) 163 1.1 (0.75 to 1.7) 1.5 (0.84 to 2.7) 1.8 (0.89 to 3.7) 1.3 (0.67 to 2.4) 1.4 (0.93 to 2.0) 1.3 (0.81 to 2.1) Hairdressers (5141) 81 1.5 (0.82 to 2.6) 0.96 (0.34 to 2.7) 1.8 (0.57 to 5.8) 0.94 (0.29 to 3.1) 1.5 (0.85 to 2.5) 0.82 (0.37 to 1.8) Shops salesperson (5220) 1084 1.1 (0.86 to 1.3) 1.0 (0.73 to 1.4) 1.2 (0.77 to 1.9) 1.1 (0.81 to 1.6) 1.1 (0.87 to 1.3) 0.78 (0.60 to 1.0) Gardeners (6113) 47 0.88 (0.37 to 2.1) 1.4 (0.42 to 4.8) 1.2 (0.16 to 9.2) 0.94 (0.20 to 4.4) 0.95 (0.43 to 2.1) 0.72 (0.24 to 2.2) Carpenters and jointers (7124) 178 1.3 (0.87 to 2.0) 1.0 (0.52 to 2.1) 1.7 (0.60 to 4.5) 0.82 (0.35 to 1.9) 1.3 (0.84 to 1.9) 0.79 (0.43 to 1.4) Building workers (7120–7123, 7129, 7133, 9313) 213 0.71 (0.45 to 1.1) 0.94 (0.49 to 1.8) 0.60 (0.19 to 1.9) 0.45 (0.19 to 1.1) 0.73 (0.48 to 1.1) 0.39 (0.21 to 0.73) Electricians (7137, 7241, 7245) 203 1.5 (0.97 to 2.2) 1.4 (0.74 to 2.6) 1.5 (0.63 to 3.5) 0.89 (0.44 to 1.8) 1.3 (0.90 to 1.9) 0.69 (0.41 to 1.2) Welders (7212) 35 1.7 (0.79 to 3.8) 5.5 (2.1 to 14) 0.60 (0.072 to 5.0) 1.3 (0.44 to 3.9) 1.9 (0.86 to 4.1) 0.66 (0.23 to 2.0) Sheet metal workers (7213) 38 1.6 (0.73 to 3.6) 0.76 (0.15 to 3.9) 1.1 (0.14 to 9.4) 1.3 (0.39 to 4.2) 1.4 (0.63 to 2.9) 0.38 (0.091 to 1.6) Machinery mechanics (7230–7233) 255 1.2 (0.82 to 1.7) 0.54 (0.27 to 1.1) 1.0 (0.47 to 2.4) 0.70 (0.36 to 1.3) 1.1 (0.76 to 1.5) 0.63 (0.39 to 1.0) Motor vehicle drivers (8320–8324) 205 1.2 (0.81 to 1.7) 1.2 (0.66 to 2.0) 0.62 (0.21 to 1.8) 1.1 (0.61 to 1.9) 1.2 (0.85 to 1.7) 0.66 (0.39 to 1.1) Cleaners (9131, 9132) 185 0.76 (0.47 to 1.2) 0.63 (0.27 to 1.4) 1.4 (0.61 to 3.2) 1.6 (0.80 to 3.1) 0.90 (0.58 to 1.4) 0.92 (0.51 to 1.6) Agricultural labour (9211) 53 1.3 (0.58 to 2.9) 0.28 (0.038 to 2.1) 1.2 (0.28 to 5.2) 0.24 (0.032 to 1.8) 1.0 (0.45 to 2.3) 1.1 (0.46 to 2.7) Results reaching significance, p<0.05, are marked in bold. *Adjusted for age, gender, area of residence, smoking, passive smoking, housing conditions, damp and mould at home and nasal allergy. †Self-reported occupation categorised by the researchers using ISCO-88. ISCO-88, International Standard Classification of Occupations. Using the main occupational groups, the category with the combination agriculture/fishery workers and craft/related trade workers showed the highest OR for wheezing and asthma attack (table 3). The PARs for asthma attack during the past 12 months for the main occupational groups 2 and 4 were both 11%. The strongest associations between respiratory symptoms and current job category were for health-associated professionals (OR between 1.4 and 2.1) and welders (OR 5.5) (table 3). Association between occupational exposure (N-JEM) and respiratory symptoms and asthma To further explore the relationship between occupational exposure and respiratory symptoms and diseases, we used the N-JEM to assess the exposures (table 4). Table 4 Adjusted* OR (ORadj) for respiratory symptoms past 12 months, current use of asthma medication and ever physician-diagnosed asthma by exposure generated from ISCO-88 work codes and the use of N-JEM† Lower respiratory tract Exposure category Total number Wheezing last 12 months ORadj Woken with dyspnoea ORadj Asthma attack ORadj Use of asthma medication ORadj At least one symptom last 12 months ORadj Physician- diagnosed asthma ORadj Main categories  HMW agents 1022 1.1 (0.93 to 1.3) 1.0 (0.75 to 1.4) 1.2 (0.80 to 1.7) 0.98 (0.73 to 1.3) 1.1 (0.91 to 1.3) 0.90 (0.71 to 1.1)  LMW agents 584 1.2 (0.97 to 1.5) 1.2 (0.86 to 1.7) 1.1 (0.69 to 1.9) 0.94 (0.63 to 1.4) 1.2 (0.94 to 1.5) 0.99 (0.73 to 1.3)  Irritating agents 2543 1.0 (0.91 to 1.2) 1.2 (1.0 to 1.5) 1.2 (0.92 to 1.6) 1.2 (0.95 to 1.5) 1.1 (0.97 to 1.3) 1.0 (0.86 to 1.2)  Peak exposure to irritants 239 1.0 (0.70 to 1.5) 1.2 (0.77 to 2.0) 0.82 (0.37 to 1.8) 0.82 (0.46 to 1.5) 0.90 (0.64 to 1.3) 0.71 (0.43 to 1.2)  Uncertain or low exposure 759 1.1 (0.85 to 1.3) 0.86 (0.59 to 1.2) 0.93 (0.56 to 1.6) 0.86 (0.59 to 1.3) 1.0 (0.81 to 1.2) 0.96 (0.72 to 1.3)  Reference group/unexposed‡ 5597 1 1 1 1 1 1 Detailed categories  Animal-derived agents 165 0.97 (0.58 to 1.6) 0.94 (0.41 to 2.2) 1.1 (0.43 to 2.9) 0.61 (0.21 to 1.8) 0.91 (0.55 to 1.5) 1.1 (0.61 to 2.0)  Flour 33 3.2 (1.4 to 7.3) 3.5 (1.3 to 9.5) 2.6 (0.47 to 15) 3.3 (1.1 to 10) 3.4 (1.6 to 7.6) 1.6 (0.53 to 4.7)  Mixed agricultural work (not animal) 113 1.1 (0.68 to 1.8) 1.2 (0.52 to 2.7) 1.6 (0.45 to 5.4) 1.1 (0.45 to 2.6) 1.1 (0.65 to 1.7) 1.1 (0.53 to 2.1)  Moulds and other bioaerosols 141 1.1 (0.65 to 2.0) 1.2 (0.53 to 2.8) 0.46 (0.12 to 1.8) 1.0 (0.32 to 3.3) 1.2 (0.68 to 2.0) 0.49 (0.24 to 1.0)  Latex proteins 639 1.2 (0.89 to 1.7) 1.3 (0.79 to 2.1) 1.9 (0.97 to 3.6) 1.6 (0.89 to 2.7) 1.3 (0.97 to 1.8) 1.1 (0.72 to 1.7)  Pharmaceutical product agents 474 0.87 (0.60 to 1.3) 0.60 (0.33 to 1.1) 0.54 (0.25 to 1.2) 0.44 (0.22 to 0.89) 0.74 (0.51 to 1.1) 0.83 (0.50 to 1.4)  Acrylates 219 1.1 (0.72 to 1.7) 1.8 (1.0 to 3.4) 1.2 (0.47 to 3.1) 1.1 (0.56 to 2.2) 0.92 (0.63 to 1.4) 1.0 (0.61 to 1.8)  Epoxy chemicals 187 0.63 (0.31 to 1.3) 0.11 (0.034 to 0.35) 0.15 (0.024 to 0.93) 0.55 (0.13 to 2.4) 0.54 (0.27 to 1.1) 0.89 (0.30 to 2.7)  Diisocyanates 268 1.5 (0.90 to 2.6) 2.9 (1.5 to 5.7) 3.5 (0.91 to 13) 1.5 (0.50 to 4.5) 1.7 (1.0 to 2.9) 1.4 (0.59 to 3.4)  Other reactive agents 195 1.2 (0.83 to 1.8) 0.93 (0.48 to 1.8) 1.2 (0.57 to 2.6) 0.73 (0.36 to 1.5) 1.1 (0.77 to 1.6) 0.69 (0.39 to 1.2)  Cleaning agents 223 0.91 (0.60 to 1.4) 1.1 (0.61 to 2.0) 1.2 (0.56 to 2.6) 1.3 (0.73 to 2.3) 1.0 (0.71 to 1.5) 0.96 (0.58 to 1.6)  Wood and paper dust 288 1.0 (0.73 to 1.5) 0.95 (0.56 to 1.6) 1.8 (0.84 to 4.0) 1.2 (0.68 to 2.3) 1.1 (0.77 to 1.5) 1.3 (0.81 to 2.0)  Inorganic dust and fumes 789 1.0 (0.78 to 1.3) 1.3 (0.90 to 2.0) 1.0 (0.54 to 1.9) 0.84 (0.53 to 1.3) 1.0 (0.82 to 1.3) 0.77 (0.54 to 1.1)  Welding and soldering fumes and other metal dust 134 1.6 (0.88 to 3.0) 3.2 (1.6 to 6.4) 0.90 (0.24 to 3.4) 1.2 (0.51 to 3.0) 1.8 (1.0 to 3.2) 1.6 (0.73 to 3.4)  Metal working fluids 191 0.83 (0.52 to 1.3) 0.36 (0.16 to 0.84) 1.4 (0.56 to 3.4) 0.81 (0.38 to 1.7) 0.72 (0.46 to 1.1) 0.72 (0.38 to 1.4)  Vehicle/motor exhaust 1124 1.1 (0.95 to 1.4) 1.4 (1.0 to 1.8) 1.3 (0.88 to 1.9) 1.4 (1.0 to 1.8) 1.2 (1.0 to 1.4) 1.2 (0.93 to 1.5)  Environmental tobacco smoke 210 0.92 (0.63 to 1.4) 1.5 (0.92 to 2.6) 1.4 (0.68 to 2.8) 0.72 (0.37 to 1.4) 1.1 (0.75 to 1.5) 1.7 (1.1 to 2.6)  Peak exposure to irritants 239 0.76 (0.44 to 1.3) 0.47 (0.23 to 0.95) 0.57 (0.19 to 1.7) 0.69 (0.31 to 1.5) 0.61 (0.37 to 1.0) 0.45 (0.24 to 0.84)  Uncertain or low exposure 759 1.1 (0.85 to 1.3) 0.85 (0.59 to 1.2) 0.96 (0.58 to 1.6) 0.85 (0.58 to 1.2) 0.99 (0.81 to 1.2) 0.97 (0.73 to 1.3)  Reference group/unexposed† 5597 1 1 1 1 1 1 Results reaching significance, p<0.05, are marked in bold. *Adjusted for age, gender, area of residence, smoking, passive smoking, housing conditions, damp and mould at home and nasal allergy. †Exposure was assessed by using N-JEM on the self-reported occupations categorised by the researchers using ISCO-88. ‡The reference group are those with current occupations classified as unexposed only by N-JEM. HMW, high molecular weight; ISCO-88, International Standard Classification of Occupations; LMW, low molecular weight; N-JEM, job-exposure matrix (JEM) developed for the northern European countries. As shown in table 4, irritating agents showed a statistically significant elevated OR for woken with dyspnoea. When analysing specific exposure subgroup from the N-JEM (table 4), flour exposure had the highest ORs for respiratory symptoms in the past 12 months. Other specific exposures with statistically significant elevated ORs were diisocyanates, welding/soldering fumes and vehicle/motor exhaust. Low risks were observed for pharmaceutical product agents, epoxy chemicals, metal working fluids and peak exposures to irritants. The PAR for having at least one symptom during the past 12 months for the exposures flour, diisocyanates, welding/soldering fumes and vehicle/motor exhaust was 0.8%, 1.8%, 1.1% and 2.0%, respectively. Discussion In this population-based study, the prevalence of physician-diagnosed asthma was 11.5%. The study identified occupations and occupational exposures associated with respiratory symptoms in the past 12 months, use of asthma medication and respiratory diseases by the use of an asthma-specific JEM. The high-risk occupations associated with several respiratory outcomes were health-associated professionals and welders. The specific occupational exposures that had associations with respiratory symptoms were flour, diisocyanates, welding/soldering fumes and vehicle/motor exhaust. The study also identified negative associations between some respiratory symptoms and exposures, which may indicate the presence of healthy worker effects. The prevalence of physician-diagnosed asthma was higher in Telemark than in a similar study conducted in Gothenburg in 2008 which found a prevalence of 8.3%.20 Also, the prevalence of symptoms in the past 12 months and use of asthma medication was slightly more common in our study compared with that study. Globally, the asthma prevalence among adults is highest in high-income regions and varies between 1% and 21%, with estimates of ∼6–20% for different parts of western Europe and 11% for Norway.21 While there is no previous estimates on physician-diagnosed asthma from Telemark County, the prevalence of wheezing in the past 12 months was 5.9% in southeast Norway in 2013.11 In our study, a larger proportion of responders and non-responders reported wheezing in the past 12 months, with 20.0% and 21.3%, respectively. This finding indicates that further assessment of risk factors for respiratory disease is needed in Telemark. The prevalence of physician-diagnosed asthma and respiratory symptoms in the past 12 months decreased with increasing age and was more common among women as compared with men (12.3% vs 10.6%, respectively). This finding is in contrast with the Hordaland study from west Norway where 6.1% of men and 5.7% of women reported physician-diagnosed asthma.6 In that study, few associations between occupational exposures assessed by the ALOHA JEM were detected by the use of three categories of exposure, biological dust, mineral dust and gas or fumes, and further division into non-exposed, low or high exposed within these categories. There are some important differences between our study and the Hordaland study. The latter was published in 2009 but includes data collected two decades ago. The applied JEM was not adjusted to the geographical area, and used wide exposure categories which may have introduced misclassification of exposure. Further, the authors stated that their population sample was likely to be representative for the whole country, whereas our study includes an area with high levels of industrial activity (Grenland). In terms of smoking habits, the prevalence of physician-diagnosed asthma did not differ between smoking categories in our study. This is in line with the observation in the study from west Sweden where asthma incidence rates during the never-smoking years were the same as during the smoking years.20 Our results also showed that respiratory symptoms were associated with living in an urban area and housing conditions. Similar observations were reported by Montnémery et al22 in a study from Sweden, where living by the sea and urban living were associated with asthma. That study also reported an association between asthma and unskilled and semiskilled workers, although the results were not statistically significant. We have adjusted our results for smoking, housing condition, damp and mould, and area of living. These variables may be considered as indicators of socioeconomic status, although income is more commonly used. The Telemark study does not include income data. As differences in workers' salaries are relatively small in Norway, income may not be well suited to describe socioeconomic differences.23 Our study identified two job categories that were associated with different respiratory symptoms in the past 12 months and use of asthma medication: health-associated professionals and welders. These results support earlier findings from cross-sectional studies from specific industries or sectors showing that working as a healthcare professional is associated with respiratory symptoms and asthma.24–26 Healthcare professionals consist of several occupational groups with a variety of potential exposures to occupational hazards that may have asthmagenic and/or allergenic properties. Especially, exposures to cleaning and disinfecting products may explain the increased risk of occupational asthma.26 In our study, we observed that health-associated professionals were associated with several respiratory symptoms in the past 12 months and the use of current asthma medication. This is in line with findings from Lillienberg et al8 showing an increased risk for new-onset asthma in female nurses (HR 1.5, 95% CI 0.95 to 2.1). Similar findings were seen in the ECRHS II cohort where elevated relative risks of asthma (RR 2.2, 95% CI 1.0 to 4.5) were observed among nurses who reported using ammonia and/or bleach cleaning products more than once per week.24 In Norway, one might speculate that this group of healthcare workers including nursing assistants are more exposed to asthmagenic and/or irritating agents as they perform more activities related to cleaning and use of disinfectants compared with other healthcare workers, and also work more frequently in private homes where exposure to passive smoking may occur. Latex proteins were not identified as a risk factor by the use of N-JEM in our study. One reason may be that there has been considerable focus on the allergic properties of these proteins, leading to the use of alternatives. Our findings also showed positive associations for respiratory symptoms and exposure to welding/soldering fumes, and for woken with dyspnoea and the occupational group welders. Welders are exposed to a variety of agents like LMW agents (metals), irritating gases/fumes and potentially high irritant peaks. These exposures have previously been associated with an increased risk of asthma.8 27 28 A longitudinal study of Norwegian smelters also showed that low to moderate levels of dust exposure were associated with increased incidence of work-related asthma-like symptoms.29 However, as the CI for this finding was wide, this result must be interpreted with caution. As shown in table 4, the study also identified negative associations between respiratory symptoms and exposures. For instance, a negative association was identified between physician-diagnosed asthma (OR 0.45, CI 0.24 to 0.84) and woken with dyspnoea (OR 0.47, CI 0.23 to 0.95) in regard to peak exposure to irritants. Some negative associations may be explained by residual confounding from socioeconomic status as it has been shown that airflow limitation may be related to education or income.30 Such effects must be considered when interpreting these results. Time-dependent analyses of the onset of symptoms and asthma in regard to occupational exposures will be considered in future data analyses. The PAR for having at least one symptom during the past 12 months attributable to workplace factors was between 1% and 2%, whereas the PAR for asthma attacks was 11% for the main occupational groups 2 and 4. A similar study from Sweden reported that proportion of asthma attributable to exposure groups varied from 3% to 14%.8 As that study assessed only new-onset asthma, the results are difficult to compare. Occupational exposures were assessed by the use of a JEM. The occupational exposures associated with an increased risk of respiratory symptoms in the past 12 months were flour, diisocyanates, welding/soldering fumes and vehicle/motor exhaust. Among these exposures, flour exposure had the highest risk of respiratory symptoms. Flour is a well-known asthmagen.31 32 Exposure to vehicle/motor exhaust was also associated with self-reported respiratory symptoms. Vehicle/motor exhaust is a complex mixture of different agent(s)/particles and therefore it is difficult to identify specific causal agents. Dumas et al27 showed that exposure to combustion particles/fumes was associated with physician-diagnosed asthma with an OR of 2.04 (95% CI 1.55 to 2.68). In addition, Henneberger et al33 showed an association between severe exacerbation of asthma and exposure to combustion particles/fumes. Other studies have also shown associations between occupational exposures to diisocyanates and respiratory health outcomes that are consistent with our study findings.9 34 The positive association between respiratory symptoms and exposure to welding/soldering fumes in the N-JEM supported the findings regarding this exposure from the ISCO-88 data. The N-JEM findings confirm that certain occupational groups are still at high risk of being exposed to respiratory hazards in the workplace and that information regarding specific exposure is crucial to achieve targeted prevention of respiratory symptoms and diseases. The low response rate of the Telemark study may be considered as an important limitation. To address non-participation, adjustments by inverse probability weighting were made to account for non-response bias. In a separate study in 2016,19 we reported demographics characteristics, respiratory symptoms and use of asthma medication for the non-responders and identified possible selection bias. A total of 260 out of 700 randomly selected non-responders (37%) participated. No statistically significant differences were detected for asthma and several respiratory symptoms between those that responded to the Telemark study and those that did not. However, we adjusted the prevalence estimates for chronic cough and current use of asthma medication as these were over-represented among the responders. Even though inverse probability of non-response weighting aims to adjust for non-response, results may still not be representative of the initial population. All the statistical analyses in this study were performed with and without weighted data sets to evaluate how adjusting for non-response affected the study results. We observed that the weights had little impact on the study outcomes. Regarding the external validity of the study, it may be argued that Telemark County includes an industrialised region (Grenland) and thus our results are not generalisable to the whole country. However, this would mainly affect the prevalences and to a lesser extent the associations between exposure and outcomes. Also, the similarity of some current findings to reports from other countries suggests our findings are relevant beyond Norway. Another limitation may be that some of the occupational groups were small. In population-based studies, this will often be the case and implicates that effects have to be large to be detected. Thus, respiratory effects from some exposures and for some occupations may have been underestimated. On the other hand, the grouping of occupations by exposures may increase the study power to detect associations with respiratory outcomes. Further, JEM-derived data are considered to be less susceptible to recall bias leading to non-random misclassification of exposure than self-report because a JEM uses occupation and not exposures/measurements. We used an asthma-specific JEM that probably will identify well-known exposures associated with respiratory outcomes, but less likely to identify associations that are not well documented. Another limitation may be that the ISCO-88 classification is based on education and not on individual exposure information or measurements, which may lead to misclassification of exposure and underestimation of effects. Further studies should be performed to confirm these results including also exposure from previous jobs as our study addressed only current occupation. Using a sample from the general population may reduce the probability of healthy worker effect. However, in cross-sectional analyses, selection out of the work force may occur. Negative associations between the occupational groups 4 and 5 (table 3) and physician-diagnosed asthma may indicate this type of selection bias. However, our analyses of non-responders showed that their use of asthma medication was somewhat lower than among study participants, indicating that more severe asthmatics attended the study. The inclusion of subjects from the age of 16 may also to some extent have reduced the probability of selection bias in this cross-sectional analysis as some adolescents start their working life without further education or as apprentices. Apprentices may experience respiratory symptoms or develop occupational asthma leading to early selection out of certain jobs. Such effects may be difficult to capture in cross-sectional analyses which do not include this age group. However, the drawback is that many of these subjects do not have an occupational history resulting in a larger proportion of missing data. Efforts were made to reduce the likelihood of selection, confounding and information bias as explained above. Also, the sample size was considered to ensure adequate statistical power for the reported associations. Thus, it can be argued that multiple testing procedures should not necessarily be used.35 Nevertheless, further studies should be performed to confirm our results regarding the impact of occupational exposure for specific jobs and exposures. As in most cross-sectional studies based on questionnaire data, we cannot exclude the presence of common method bias, but using current occupation instead of self-reported exposure may have reduced the possibility. In conclusion, the prevalence of physician-diagnosed asthma was 11.5% in this sample from the general population. Analysis of risk factors contribute to the evidence that occupational exposure to flour, diisocyanates, welding/soldering fumes and vehicle/motor exhaust is associated with respiratory symptoms in the past 12 months and use of asthma medication. Selection bias may have been present although inverse probability of non-participation weighting was performed to counter this bias. Until prospective data are available, we recommend reduction of occupational exposures and early identification of workers with respiratory symptoms should remain an important priority. Contributors: RA, MVS, PKH, KT, JK and AKMF were involved in the design of the study. RA was involved in the data collection, data analyses, data management, data interpretation and primary manuscript preparation. AKMF was involved in the data collection, data management, data interpretation and critical revision of the manuscript. MVS was involved in the data collection, data analyses, data management, data interpretation and critical revision of the manuscript. EA contributed to the data interpretation and critical revision of the manuscript. KT was involved in data interpretation and critical revision of the manuscript. PKH was involved in data interpretation and critical revision of the manuscript. JK was involved in the data interpretation and critical revision of the manuscript. All authors provided input on the manuscript, and all authors read and approved the final manuscript. Funding: This work was supported by funding from Telemark Hospital, Norway. Disclaimer: The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the National Institute for Occupational Safety and Health. Competing interests: None declared. Ethics approval: This study was conducted with the approval of The Regional Committee for Medical and Health Research Ethics in Norway (REC 2012/1665). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Frostad A , Søyseth V , Haldorsen T Respiratory symptoms and long-term cardiovascular mortality . Respir Med 2007 ;101 :2289 –96 . 10.1016/j.rmed.2007.06.023 17681463 2 Voll-Aanerud M , Eagan TM , Wentzel-Larsen T Respiratory symptoms, COPD severity, and health related quality of life in a general population sample . Respir Med 2008 ;102 :399 –406 . 10.1016/j.rmed.2007.10.012 18061422 3 Kogevinas M , Zock JP , Jarvis D Exposure to substances in the workplace and new-onset asthma; an international prospective population-based study (ECRHS-II) . Lancet 2007 ;370 :336 –41 . 10.1016/S0140-6736(07)61164-7 17662882 4 Torén K , Blanc PD Asthma caused by occupational exposures is common—a systematic analysis of estimates of the population-attributable fraction . BMC Pulm Med 2009 ;9 :7 10.1186/1471-2466-9-7 19178702 5 Jakkola MS , Jakkola JJ Assessment of public health impact of work-related asthma . BMC Med Res Methodol 2012 ;5 :12 –22 . 6 Skorge TD , Eagan TM , Eide GE Occupational exposure and incidence of respiratory disorders in a general population . Scand J Work Environ Health 2009 ;35 :454 –61 . 10.5271/sjweh.1352 19806271 7 Gundersen H , Magerøy N , Moen BE Low traffic and respiratory symptoms among smoking females: the Hordaland Health Study . Arch Environ Occup Health 2012 ;67 :189 –98 . 10.1080/19338244.2011.619214 23074976 8 Lillienberg L , Andersson E , Janson C Occupational exposure and new-onset asthma in a population-based study in Northern Europe (RHINE) . Ann Occup Hyg 2013 ;57 :482 –92 . 10.1093/annhyg/mes083 23204511 9 Lillienberg L , Dahlman-Höglund A , Schöler L Exposures and asthma outcomes using two different job exposure matrices in a general population study in northern Europe . Ann Occup Hyg 2014 ;58 :469 –81 . 10.1093/annhyg/meu002 24504176 10 Torén K , Ekerljung L , Kim JL Adult-onset asthma in west Sweden—Incidence, sex differences and impact of occupational exposures . Respir Med 2011 ;105 :1622 –8 . 10.1016/j.rmed.2011.06.003 21757331 11 The National Institute of Occupational Health (STAMI) . http://noa.stami.no/arbeidsmiljoindikatorer/helseutfallarbeidsskader/helseplager/luftveisplager/ (accessed 8 Dec 2016 ). 12 The Norwegian Institute of Public Health . http://www.reseptregisteret.no/Prevalens.aspx (accessed 11 Mar 2016 ). 13 European Community Respiratory Health Survey II Steering Committee . The European community respiratory health survey II . Eur Respir J 2002 ;20 :1071 –9 . 10.1183/09031936.02.00046802 12449157 14 Burney PG , Laitinen LA , Perdrizet S Validity and repeatability of the IUATLD (1984) Bronchial Symptoms Questionnaire: an international comparison . Eur Respir J 1989 ;2 :940 –5 .2606194 15 Kongerud J , Vale JR , Aalen OO Questionnaire reliability and validity for aluminum potroom workers . Scand J Work Environ Health 1989 ;15 :364 –70 . 10.5271/sjweh.1838 2799323 16 Torén K , Brisman J , Järvholm B Asthma and asthma-like symptoms in adults assessed by questionnaires. A literature review . Chest 1993 ;104 :600 –8 .7802735 17 ILO . International Standard Classification of Occupations ISCO 88 . Geneva, Switzerland : ILO , 1991 ISBN 92-2-106438-7 . 18 Fell AKM , Abrahamsen R , Henneberger PK Breath-taking jobs: a case-control study of respiratory work disability by occupation . Occup Environ Med 2016 ;73 :600 –6 . 10.1136/oemed-2015-103488 27365181 19 Abrahamsen R , Svendsen MV , Henneberger PK Nonresponse in a cross-sectional study of respiratory health in Norway . BMJ Open 2016 ;6 :e009912 10.1136/bmjopen-2015-009912 20 Lötvall J , Ekerljung L , Rönmark EP West Sweden Asthma Study: prevalence trends over the last 18 years argues no recent increase in asthma . Respir Res 2009 ;10 :94 10.1186/1465-9921-10-94 19821983 21 To T , Stanojevic S , Moores G Global asthma prevalence in adults: findings from the cross-sectional world health survey . BMC Public Health 2012 ;12 :204 10.1186/1471-2458-12-204 22429515 22 Montnémery P , Bengtsson P , Elliot A Prevalence of obstructive lung diseases and respiratory symptoms in relation to living environment and socio-economic group . Respir Med 2001 ;95 :744 –52 . 10.1053/rmed.2001.1129 11575896 23 Statistics Norway . https://www.ssb.no/statistikkbanken/selectout/print.asp?FileformatId=2&Queryfile=20161216105842489815693Lonn AnsMedKjNar&PLanguage=1&MainTable=LonnAnsMedKjNar &potsize=100 (accessed 16 Dec 2016 ). 24 Mirabelli MC , Zock JP , Plana E Occupational risk factors for asthma among nurses and related healthcare professionals in an international study . Occup Environ Med 2007 ;64 :474 –9 . 10.1136/oem.2006.031203 17332135 25 Arif AA , Delclos GL , Serra C Occupational exposures and asthma among nursing professionals . Occup Environ Med 2009 ;66 :274 –8 . 10.1136/oem.2008.042382 19164328 26 Dumas O , Varraso R , Zock JP Asthma history, job type and job changes among US nurses . Occup Environ Med 2015 ;72 :482 –8 . 10.1136/oemed-2014-102547 25713153 27 Dumas O , Laurent E , Bousquet J Occupational irritants and asthma: an Estonian cross-sectional study of 34,000 adults . Eur Respir J 2014 ;44 :647 –56 . 10.1183/09031936.00172213 24743968 28 Hedmer M , Karlsson JE , Andersson U Exposure to respirable dust and manganese and prevalence of airways symptoms, among Swedish mild steel welders in the manufacturing industry . Int Arch Occup Environ Health 2014 ;87 :623 –34 . 10.1007/s00420-013-0896-3 23979145 29 Søyseth V , Johnsen HL , Henneberger PK The incidence of work-related asthma-like symptoms and dust exposure in Norwegian smelters . Am J Respir Crit Care Med 2012 ;185 :1280 –5 . 10.1164/rccm.201110-1809OC 22517789 30 Smith M , Li L , Augustyn M Prevalence and correlates of airflow obstruction in ∼317,000 never-smokers in China . Eur Respir J 2014 ;44 :66 –77 . 10.1183/09031936.00152413 24603814 31 Baur X , Degens PO , Sander I Baker's asthma: still among the most frequent occupational respiratory disorders . J Allergy Clin Immunol 1998 ;102 (Pt 1 ):984 –97 . 10.1016/S0091-6749(98)70337-9 9847440 32 Meijster T , Warren N , Heederik D What is the best strategy to reduce the burden of occupational asthma and allergy in bakers? Occup Environ Med 2011 ;68 :176 –82 . 10.1136/oem.2009.053611 20798001 33 Henneberger PK , Liang X , Lillienberg L Occupational exposures associated with severe exacerbation of asthma . Int J Tuberc Lung Dis 2015 ;19 :244 –50 . 10.5588/ijtld.14.0132 25574926 34 Torén K , Järvholm B , Brisman J Adult-onset asthma and occupational exposures . Scand J Work Environ Health 1999 ;25 :430 –5 . 10.5271/sjweh.456 10569463 35 Goldman M , Silbergeld E Assuring access to data for chemical evaluations . Environ Helath Perspect 2013 ;121 :149 –52 . 10.1289/ehp.1206101

PMC005xxxxxx/PMC5372105.txt

==== Front RMD OpenRMD OpenrmdopenrmdopenRMD Open2056-5933BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR 28405475rmdopen-2016-00042610.1136/rmdopen-2016-000426Vasculitis1506Original articleModified-release prednisone for polymyalgia rheumatica: a multicentre, randomised, active-controlled, double-blind, parallel-group study Cutolo Maurizio 1Hopp Michael 2Liebscher Stefan 2Dasgupta Bhaskar 3Buttgereit Frank 4 1 Research Laboratories and Academic Division of Clinical Rheumatology, University of Genova, Genova, Italy 2 Mundipharma Research GmbH & Co. KG, Limburg, Hessen, Germany 3 Department of Rheumatology, Southend University Hospital, Westcliff-on-Sea, UK 4 Department of Rheumatology and Clinical Immunology, Charité University Medicine, Berlin, GermanyCorrespondence to Professor Maurizio Cutolo; mcutolo@unige.itBD and FB share senior authorship. 2017 17 3 2017 3 1 e00042622 12 2016 22 2 2017 24 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective To assess the efficacy and safety of modified-release (MR) versus immediate-release (IR) prednisone in newly diagnosed glucocorticoid (GC)-naïve patients with polymyalgia rheumatica (PMR). Methods Patients were randomised to double-blind MR prednisone (taken at approximately 22:00) or IR prednisone (taken in the morning), 15 mg/day for 4 weeks. The primary end point was complete response rate (≥70% reduction in PMR visual analogue scale, duration of morning stiffness and C reactive protein (CRP) (or CRP <2× upper limit of normal (ULN))) at week 4. Non-inferiority was decided if the lower 95% confidence limit (MR vs IR prednisone) was above −15%. 400 patients were planned but only 62 were enrolled due to difficulties in recruiting GC-naïve patients with PMR with CRP ≥2×ULN. Results The percentage of complete responders at week 4 was numerically greater for MR prednisone (53.8%) than for IR prednisone (40.9%). Non-inferiority of MR versus IR prednisone was not proven in the primary analysis on the per protocol population (N=48; treatment difference: 12.22%; 95% CI −15.82% to 40.25%). However, sensitivity analysis on the full analysis population showed an evident trend favouring MR prednisone (N=62; treatment difference: 15.56%; 95% CI −9.16% to 40.28%). Adverse events were generally mild and transient with no unexpected safety observations. Conclusions The study showed a clear trend for favourable short-term efficacy of MR prednisone versus IR prednisone in early treatment of PMR. Further studies are warranted. Trial registration number EudraCT number 2011-002353-57; Results. Polymyalgia RheumaticaCorticosteroidsAutoimmune DiseasesInflammationTreatment ==== Body Key messages What is already known about this subject? Optimising the timing of glucocorticoid (GC) administration in relation to endogenous cortisol rhythms and symptom severity using modified-release (MR) formulations can improve therapeutic potential in inflammatory conditions, as already demonstrated in rheumatoid arthritis. Preliminary studies suggest that MR prednisone may also have benefits in polymyalgia rheumatica (PMR), but data from randomised controlled trials are lacking. What does this study add? This randomised controlled study showed a clear trend for a larger reduction in key PMR symptoms and levels of pro-inflammatory cytokine interleukin (IL)-6 with MR prednisone compared with immediate-release (IR) prednisone, although only 62 of a planned 400 patients were enrolled due to difficulties in recruiting GC-naïve patients with PMR and the study did not meet its primary objective of showing non-inferiority. How might this impact on clinical practice? Results from this study, though limited, suggest favourable efficacy of MR prednisone over IR prednisone in patients with PMR, and further confirmation should be sought from larger clinical trials. Experience from this study demonstrates that careful consideration of PMR inclusion and response criteria is required. Introduction Polymyalgia rheumatica (PMR) is a common inflammatory rheumatic disease of older people1 characterised by new-onset bilateral shoulder and/or hip girdle pain with pronounced stiffness and an acute phase response.2 3 Glucocorticoids (GCs) are the mainstay of treatment and rapidly improve PMR symptoms;4 an initial flexible minimum effective GC dose of 12.5–25 mg prednisone (recognising demographics, comorbidities, comedications, GC risk factors and disease severity) is recommended by the 2015 European League Against Rheumatism (EULAR) / American College of Rheumatology (ACR) guidelines.5 6 Patients with inflammatory conditions such as PMR typically show circadian variations in clinical symptoms related to altered concentrations of inflammatory cytokines, melatonin and cortisol, with key symptoms usually most severe in the early morning.7–10 Modified-release (MR) prednisone has been developed to optimise oral GC treatment strategies with respect to circadian rhythms of inflammation by releasing prednisone ∼4 hours after the administration of the tablet in the late evening. The CAPRA (Circadian Administration of Prednisone in Rheumatoid Arthritis) studies confirmed that optimising the timing of GC administration improves the benefit:risk ratio of long-term, low-dose GC treatment in patients with rheumatoid arthritis (RA).10–12 Preliminary studies in PMR suggest that MR prednisone may also have benefits in this inflammatory condition.13 This randomised, double-blind, active-controlled, parallel-group, non-inferiority phase III clinical study aimed to assess the efficacy and safety of evening MR prednisone compared with morning administration of immediate-release (IR) prednisone in newly diagnosed patients initiating GC treatment for PMR. Methods Setting, patients and treatments The study (EudraCT number 2011-002353-57) was conducted between March 2013 and March 2014 at 41 secondary care centres across nine European countries. Patients aged ≥50 years, newly diagnosed with PMR and previously untreated with GCs for PMR, were eligible for inclusion. According to the 2012 EULAR/ACR provisional classification criteria for PMR,14 15 the diagnosis had to be confirmed by all of the following at screening: (1) new-onset bilateral shoulder pain with/without hip girdle pain; (2) a PMR visual analogue scale (VAS) score ≥50 (0–100 scale); (3) duration of morning stiffness >45 min and (4) acute phase response shown by elevated C reactive protein (CRP; ≥2 times the upper limit of normal (ULN)). Patients were randomised in a 1:1 ratio to MR or IR prednisone for 4 weeks, and received 15 mg IR prednisone/placebo between 5:00 and 9:00 and 15 mg MR prednisone/placebo at 22:00±30 min. No rescue medication was used, and other medications for the treatment of PMR, including analgesics and coanalgesics, were prohibited during the study. Study assessments Every morning and evening, patients completed an electronic diary (Log Pad, PHT Corporation), recording how their PMR had affected them (PMR VAS), their overall pain (global pain VAS), pain in their arms and shoulders (shoulder pain VAS), their overall level of fatigue/tiredness (fatigue VAS), duration of morning stiffness and time of medication intake. Interleukin-6 (IL-6), CRP and erythrocyte sedimentation rate (ESR) were measured at baseline, week 1 (CRP/ESR only) and week 4. Adverse events (AEs) were recorded throughout the study. Statistical analyses The primary objective was to demonstrate non-inferiority of MR prednisone administered in the evening versus IR prednisone administered in the morning, with regard to the percentage of complete responders at week 4 (primary end point). Complete response was defined as ≥70% improvement from baseline in PMR VAS, duration of morning stiffness and CRP (or CRP <2×ULN)).16 The primary end point was analysed using logistic regression with treatment as a factor and baseline PMR VAS score, duration of morning stiffness and CRP as covariates. Non-inferiority was concluded if the lower limit of the two-sided 95% CI was above −15%. The primary analysis was performed using the per protocol population (PPP) with sensitivity analysis performed using the full analysis population (FAP; all randomised patients who received ≥1 dose of study treatment). Assuming a complete response rate of 69%16 in the comparator arm at week 4, an expected treatment difference of 0%, a non-inferiority bound of −15%, 80% power and a two-sided α of 0.05, a sample size of 300 patients in the PPP (∼400 randomised patients) was required. Further methodology can be found in the online supplementary material. 10.1136/rmdopen-2016-000426.supp1supplementary data Results Study patients The study enrolled, randomised and treated 62 patients (21 male patients, 41 female patients), all Caucasian, with a mean age of 69 years (see online supplementary table S1). All patients were included in the FAP and safety population, while 48 patients (77.4%) were included in the PPP. The study experienced a high screen failure rate (62/124 patients screened), primarily due to an insufficiently high CRP value (77% of screen failures did not have CRP ≥2×ULN). The difficulties in recruiting GC-naïve patients with PMR fulfilling the inclusion criteria, and cessation of production of the comparator Decortin 1 mg tablets, led to the premature termination of the study after 11 months' recruitment. Participant flow is shown in figure 1. Figure 1 Participant flow. Primary end point The percentage of complete responders at week 4 was numerically higher for MR prednisone (53.8%) than for IR prednisone (40.9%, table 1), although non-inferiority of MR versus IR prednisone was not proven in the primary PPP analysis (treatment difference: 12.22% in favour of MR prednisone; 95% CI −15.82% to 40.25%). Sensitivity analysis on the FAP showed a trend in favour of MR prednisone (treatment difference: 15.56%; 95% CI −9.16% to 40.28%). The study was relevantly underpowered (N=48 vs planned 300 patients in the PPP) due to recruitment difficulties and early study termination. Table 1 Response rates Response Week Modified-release prednisone Immediate-release prednisone Per protocol population (N=26) Full analysis population (N=32) Per protocol population (N=22) Full analysis population (N=30) Complete response* (n (%)) 1 3 (11.5) 5 (15.6) 4 (18.2) 4 (13.3) 2 7 (26.9) 9 (28.1) 7 (31.8) 7 (23.3) 4 14 (53.8) 17 (53.1) 9 (40.9) 10 (33.3) Partial response† (n (%)) 1 8 (30.8) 8 (25.0) 1 (4.5) 1 (3.3) 2 11 (42.3) 13 (40.6) 2 (9.1) 2 (6.7) 4 6 (23.1) 8 (25.0) 3 (13.6) 4 (13.3) *Complete response was defined as all three of the following: (1) ≥70% improvement from baseline in the polymyalgia rheumatica visual analogue scale, (2) ≥70% reduction in the duration of morning stiffness and (3) ≥70% reduction in the C reactive protein (CRP) value (or CRP <2× upper limit of normal). †Partial response was defined as two of the above three criteria being met. Secondary end points A clear consistent trend for a larger favourable effect of MR prednisone compared with IR prednisone was observed for all secondary end points (except CRP and ESR) at weeks 1 and 4 (table 2). The percentage of responders (patients with ≥70% improvement from baseline) was also greater for MR prednisone than for IR prednisone at weeks 1 and 4 for all secondary end points (see online supplementary table S2). Clinically significant mean reductions and treatment differences of more than 10 points in favour of MR prednisone were observed for PMR VAS, global pain VAS and shoulder pain VAS (table 2). MR prednisone was markedly more effective than IR prednisone in reducing morning stiffness duration from as early as week 1 (mean reduction of 326 vs 160 min, table 2), which was supported by the percentage of responders (44% vs 17%, online supplementary table S2). Table 2 Secondary efficacy results (full analysis population) Parameter Visit Modified-release prednisone (N=32) Immediate-release prednisone (N=30) Estimate (95% confidence limit)* p Value for treatment difference n Mean (SD)† n Mean (SD)† PMR VAS (0–100 scale) Baseline 32 80.7 (12.88) 30 81.0 (11.70) Week 1 31 −37.1 (25.70) 30 −28.2 (27.55) −9.3 (−22.97 to 4.30) 0.176 Week 4 27 −70.4 (20.81) 23 −59.8 (24.02) −12.8 (−22.58 to −3.05) 0.011 PMR VAS at awakening (0–100 scale) Baseline 32 81.7 (17.26) 30 85.9 (9.68) Week 1 32 −36.6 (29.40) 30 −29.4 (29.78) −10.9 (−25.45 to 3.61) 0.138 Week 4 27 −70.7 (18.72) 23 −63.7 (23.25) −11.6 (−21.33 to −1.79) 0.021 Duration of morning stiffness (minutes) Baseline 32 530 (531.0) 30 616 (591.0) Week 1 31 −326 (435.3) 30 −160 (412.4) 134.5 (18.50 to 250.50)‡ 0.021 Week 4 27 −457 (517.9) 23 −417 (574.7) 46.9 (−110.00 to 203.83)‡ 0.592 Global pain VAS (0–100 scale) Baseline 32 80.0 (13.14) 30 78.2 (13.96) Week 1 31 −36.2 (26.73) 30 −27.1 (27.96) −7.5 (−21.10 to 6.11) 0.275 Week 4 27 −68.7 (21.73) 23 −55.5 (25.55) −13.6 (−23.23 to −3.26) 0.011 Global pain VAS at awakening (0–100 scale) Baseline 32 80.6 (18.94) 30 83.4 (13.13) Week 1 32 −35.2 (30.03) 30 −28.5 (32.46) −9.1 (−23.78 to 5.57) 0.219 Week 4 27 −69.3 (20.67) 23 −60.9 (28.28) −12.3 (−23.23 to −1.42) 0.028 Shoulder pain VAS (0–100 scale) Baseline 32 81.0 (13.31) 30 79.9 (13.01) Week 1 31 −36.9 (26.85) 30 −28.4 (28.67) −7.4 (−21.55 to 6.66) 0.295 Week 4 27 −68.4 (21.46) 23 −57.7 (25.92) −11.1 (−21.30 to −0.91) 0.033 Fatigue VAS (0–100 scale) Baseline 32 72.9 (19.31) 30 75.7 (14.31) Week 1 31 −29.8 (24.94) 30 −24.6 (27.66) −7.8 (−20.67 to 5.07) 0.230 Week 4 27 −59.4 (27.34) 23 −57.6 (23.16) −6.4 (−16.76 to 4.05) 0.225 C reactive protein (mg/L) Baseline 32 50.6 (32.34) 30 69.6 (49.38) Week 1 30 −40.8 (27.46) 30 −48.9 (44.27) −6.2 (−13.61 to 1.22) 0.100 Week 4 27 −44.0 (32.24) 23 −52.9 (48.33) −8.0 (−16.37 to 0.37) 0.060 Erythrocyte sedimentation rate (mm/hour) Baseline 32 66.5 (21.62) 30 68.3 (22.84) Week 1 29 −25.8 (17.03) 35 −23.9 (15.29) Week 4 31 −38.8 (24.14) 25 −40.1 (23.65) Not estimable§ Interleukin-6 (pg/mL) Screening 28 41.4 (34.97) 30 40.9 (35.18) Week 4 19 −37.4 (41.27) 22 −29.8 (32.61) −6.5 (−11.84 to −1.23) 0.017 Larger decreases from baseline represent a favourable treatment effect for all secondary efficacy end points. *Estimates are least square means with confidence limits from a repeated measures analysis of covariance model, if not stated otherwise. †Raw mean at screening/baseline and mean change from baseline at weeks 1 and 4. ‡Estimates and confidence limits stem from the Hodges-Lehmann method. §Convergence criteria not met. PMR, polymyalgia rheumatica; VAS, visual analogue scale. MR prednisone indicated good efficacy in reducing IL-6 levels compared with IR prednisone at week 4 (decrease from baseline of −37.4 vs −29.8 pg/mL, table 2). Notable reductions in CRP and ESR were observed in both treatment groups, with a larger decrease for IR prednisone (although we noted that baseline CRP values were higher in the IR prednisone group, and mean CRP values at week 4 were lower for MR prednisone (8.4 mg/L) than for IR prednisone (17.9 mg/L)). Safety The AEs reported during the study are presented in online supplementary table S3. While more patients experienced AEs in the MR prednisone group (19 (59%); 8 (25%) related) than the IR prednisone group (9 (30%); 3 (10%) related), this was not driven by any specific AEs and the majority of subjects in both treatment groups experienced non-related AEs. Two patients experienced serious AEs (pancytopenia and temporal arteritis), which were not treatment-related. Three patients prematurely discontinued due to AEs: upper abdominal pain in the MR prednisone group and temporal arteritis and burning sensation in the IR prednisone group (figure 1). Discussion In this study, the complete response rate at week 4 was numerically greater with MR prednisone (53.8%) than with IR prednisone (40.9%). Non-inferiority of MR prednisone versus IR prednisone was not proven for the primary end point; however, even with only 48 patients in the PPP (vs the 300 planned), the point estimate was clearly in favour of MR prednisone and the lower 95% confidence limit was only marginally below (−15.82%) the decided non-inferiority threshold of −15%. Sensitivity analysis on the FAP supported the trend in favour of MR prednisone, and secondary efficacy results were encouraging, showing a clear consistent trend for a stronger effect of MR prednisone compared with IR prednisone. Of note, evening MR prednisone was associated with significantly greater reductions in IL-6 levels than morning IR prednisone, suggesting a more effective downregulation of night cytokine synthesis.17 These results are similar to findings in the CAPRA studies in RA.10 12 Consistent with the CAPRA-1 study, we observed no significant difference between MR prednisone and IR prednisone on other inflammatory markers (CRP and ESR). The study experienced a higher than expected screen failure rate (50%), primarily due to patients demonstrating lower CRP values not fulfilling the strict inclusion criteria when referred. It appears that in suspected PMR, primary care practitioners are willing to refer GC-naïve only patients with mild elevation or normal levels of inflammatory markers. However, it is likely that this population may contain predominantly non-inflammatory PMR mimics.18 The CAPRA studies have already confirmed that optimising the timing of GC administration improves the benefit:risk ratio of low-dose GC treatment in RA, and the open-label extension of the CAPRA-1 study11 and other publications19 20 confirmed positive long-term effects. A recent clinical experimental study further supports the use of chronotherapy in PMR.9 Although the formal criteria of non-inferiority were not met, our study showed a clear trend for a favourable clinical effect of MR prednisone over IR prednisone in patients with PMR. Further confirmation should be sought from a large clinical trial in patients with PMR, with careful consideration of inclusion and response criteria, based on the present study experience. The authors would like to thank all the study investigators and participants and members of the study team from Mundipharma Research GmbH & Co. KG. Karen Mower (Scientific Editor) provided medical writing assistance that was funded by Mundipharma Research Limited. Contributors: MC, MH, BD and FB were involved in the development of study protocol. MC, BD and FB were involved in patient recruitment and data acquisition. SL was involved in data analysis. MC, MH, SL, BD and FB were involved in manuscript preparation and approved the final version of the manuscript. Funding: The investigation was supported by Mundipharma Research Limited. Competing interests: MC received grant/research support from Horizon Pharma, is a consultant for Mundipharma International and is on the speaker’s bureau for Mundipharma International. MH is an employee of Mundipharma Research GmbH & Co. KG. SL is an employee of Mundipharma Research GmbH & Co. KG. BD received grant/research support from NAPP Pharmaceuticals, and is a consultant for Servier, Roche, Merck, GSK & Mundipharma. FB received grant/research support from Horizon Pharma, is a consultant for Mundipharma International and Horizon Pharma, and is on the speaker’s bureau for Mundipharma International and Horizon Pharma. Ethics approval: The Independent Ethics Committee of each centre reviewed and approved the protocol, and written informed consent was obtained from all patients before enrolment. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Lawrence RC , Felson DT , Helmick CG Estimates of the prevalence of arthritis and other rheumatic conditions in the United States, Part II . Arthritis Rheum 2008 ;58 :26 –35 . 10.1002/art.23176 18163497 2 Salvarani C , Cantini F , Boiardi L Polymyalgia rheumatica and giant-cell arteritis . N Engl J Med 2002 ;347 :261 –71 . 10.1056/NEJMra011913 12140303 3 Weyand CM , Goronzy JJ Giant-cell arteritis and polymyalgia rheumatica . Ann Intern Med 2003 ;139 :505 –15 . 10.7326/0003-4819-139-6-200309160-00015 13679329 4 Hernández-Rodríguez J , Cid MC , López-Soto A Treatment of polymyalgia rheumatica: a systematic review . Arch Intern Med 2009 ;169 :1839 –50 . 10.1001/archinternmed.2009.352 19901135 5 Dejaco C , Singh YP , Perel P 2015 recommendations for the management of polymyalgia rheumatica: a European League Against Rheumatism/American College of Rheumatology collaborative initiative . Ann Rheum Dis 2015 ;74 :1799 –807 . 10.1136/annrheumdis-2015-207492 26359488 6 Dejaco C , Singh YP , Perel P 2015 recommendations for the management of polymyalgia rheumatica: a European League Against Rheumatism/American College of Rheumatology collaborative initiative . Arthritis Rheumatol 2015 ;67 :2569 –80 . 10.1002/art.39333 26352874 7 Spies CM , Cutolo M , Straub RH More night than day—circadian rhythms in polymyalgia rheumatica and ankylosing spondylitis . J Rheumatol 2010 ;37 :894 –9 . 10.3899/jrheum.091283 20360185 8 Buttgereit F How should impaired morning function in rheumatoid arthritis be treated? Scand J Rheumatol Suppl 2011 ;125 :28 –39 . 10.3109/03009742.2011.566438 21529308 9 Galbo H , Kall L Circadian variations in clinical symptoms and concentrations of inflammatory cytokines, melatonin, and cortisol in polymyalgia rheumatica before and during prednisolone treatment: a controlled, observational, clinical experimental study . Arthritis Res Ther 2016 ;18 :174 10.1186/s13075-016-1072-4 27455959 10 Buttgereit F , Doering G , Schaeffler A Efficacy of modified-release versus standard prednisone to reduce duration of morning stiffness of the joints in rheumatoid arthritis (CAPRA-1): a double-blind, randomised, controlled trial . Lancet 2008 ;371 :205 –14 . 10.1016/S0140-6736(08)60132-4 18207016 11 Buttgereit F , Doering G , Schaeffler A Targeting pathophysiological rhythms: prednisone chronotherapy shows sustained efficacy in rheumatoid arthritis . Ann Rheum Dis 2010 ;69 :1275 –80 . 10.1136/ard.2009.126888 20542963 12 Buttgereit F , Mehta D , Kirwan J Low-dose prednisone chronotherapy for rheumatoid arthritis: a randomised clinical trial (CAPRA-2) . Ann Rheum Dis 2013 ;72 :204 –10 . 10.1136/annrheumdis-2011-201067 22562974 13 Zakout S , Kirwan JR Polymyalgia rheumatica has a nocturnal rise in serum interleukin-6 which is almost completely suppressed by nighttime prednisone . Arthritis Rheum 2011 ;63 (Suppl 10 ):79 10.1002/acr.20347 14 Dasgupta B , Cimmino MA , Maradit-Kremers H 2012 provisional classification criteria for polymyalgia rheumatica: a European League Against Rheumatism/American College of Rheumatology collaborative initiative . Ann Rheum Dis 2012 ;71 :484 –92 . 10.1136/annrheumdis-2011-200329 22388996 15 Dasgupta B , Cimmino MA , Maradit Kremers H 2012 provisional classification criteria for polymyalgia rheumatica: a European League Against Rheumatism/American College of Rheumatology collaborative initiative . Arthritis Rheum 2012 ;64 :943 –54 . 10.1002/art.34356 22389040 16 Matteson EL , Maradit-Kremers H , Cimmino MA Patient-reported outcomes in polymyalgia rheumatica . J Rheumatol 2012 ;39 :795 –803 . 10.3899/jrheum.110977 22422492 17 Straub RH , Glück T , Cutolo M The adrenal steroid status in relation to inflammatory cytokines (interleukin-6 and tumour necrosis factor) in polymyalgia rheumatica . Rheumatology (Oxford) 2000 ;39 :624 –31 . 10.1093/rheumatology/39.6.624 10888707 18 Dasgupta B , Hutchings A , Matteson EL Polymyalgia rheumatica: the mess we are now in and what we need to do about it . Arthritis Rheum 2006 ;55 :518 –20 . 10.1002/art.22106 16874793 19 Pfeiffer BM , Krenzer S , Dockhorn R Impact of modified-release prednisone on functional ability in patients with rheumatoid arthritis . Rheumatol Int 2013 ;33 :1447 –54 . 10.1007/s00296-012-2583-1 23179262 20 Cutolo M , Iaccarino L , Doria A Efficacy of the switch to modified-release prednisone in rheumatoid arthritis patients treated with standard glucocorticoids . Clin Exp Rheumatol 2013 ;31 :498 –505 .23415134

PMC005xxxxxx/PMC5372107.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01329810.1136/bmjopen-2016-013298SurgeryResearch150617371737Surgery for Type B Ankle Fracture Treatment: a Combined Randomised and Observational Study (CROSSBAT) http://orcid.org/0000-0002-0852-8868Mittal Rajat Harris Ian A http://orcid.org/0000-0003-0780-4338Adie Sam Naylor Justine M for the CROSSBAT Study GroupAllcock Paul Alttahir Mustafa Balogh Zsolt Bhimani Aziz Bourne Russell Boyle Richard Broe David Bucknill Andrew Chehade Mellick Chin Raymond Clout Andrew Connon Frank Cooke Cameron Dave Chandra Dave Jaykar Dekkers Mark Drobetz Herwig Farrugia Richard Fritsch Brett Gupta Sanjeev Hoffman Chris Holt Michael Horseley Mark Hutchinson Stephen Jeyaprakash Prajith Kent Steven King Doug Ko Victoria Kulkarni Vinay Laird Martin Lieu David Loefler Andreas Lunz David Lynch Genni Malisano Lawrie Meakin Ian Molnar Robert Morrey Chris Mulford Jonathan Munsif Ashish Muscio Paul Narayan Alok Nicholls Alexander Nouh Fred O’Carrigan Tim O'Leary Ed Lorentzos Peter Pant Sushil Perriman Diana Petchell Jeffrey Pirpiris Marinis Pohl Tony Punjabi Vaibhav Randhawa Sunil Rau Matthias Scarvell Jennie Schuetz Michael Scwartz Ben Smith Paul Sivakumar Brahman Solomon Bogdan Spencer Jonathan Stalley Paul Steele Mitchell Suthersan Mayuran Szomor Zoltan Tamblyn Peter Tarrant Seth Tetsworth Kevin Viswanathan Sameer Walker Richard Orthopaedic Department, Whitlam Orthopaedic Research Centre, Ingham Institute for Applied Medical Research, South Western Sydney Clinical School, UNSW Australia, Liverpool BC, New South Wales, AustraliaCorrespondence to Dr Rajat Mittal; rajatmittal.syd@gmail.com2017 27 3 2017 7 3 e01329813 7 2016 10 1 2017 20 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Background Isolated type B ankle fractures with no injury to the medial side are the most common type of ankle fracture. Objective This study aimed to determine if surgery is superior to non-surgical management for the treatment of these fractures. Methods A pragmatic, multicentre, single-blinded, combined randomised controlled trial and observational study. Setting Participants between 18 and 65 years with a type B ankle fracture and minimal talar shift were recruited from 22 hospitals in Australia and New Zealand. Participants willing to be randomised were randomly allocated to undergo surgical fixation followed by mobilisation in a walking boot for 6 weeks. Those treated non-surgically were managed in a walking boot for 6 weeks. Participants not willing to be randomised formed the observational cohort. Randomisation stratified by site and using permuted variable blocks was administered centrally. Outcome assessors were blinded for the primary outcomes. Primary outcomes Patient-reported ankle function using the American Academy of Orthopaedic Surgeons Foot and Ankle Outcomes Questionnaire (FAOQ) and the physical component score (PCS) of the SF-12v2 General Health Survey at 12 months postinjury. Primary analysis was intention to treat; the randomised and observational cohorts were analysed separately. Results From August 2010 to October 2013, 160 people were randomised (80 surgical and 80 non-surgical); 139 (71 surgical and 68 non-surgical) were analysed as intention to treat. 276 formed the observational cohort (19 surgical and 257 non-surgical); 220 (18 surgical and 202 non-surgical) were analysed. The randomised cohort demonstrated that surgery was not superior to non-surgery for the FAOQ (49.8 vs 53.0; mean difference 3.2 (95% CI 0.4 to 5.9), p=0.028), or the PCS (53.7 vs 53.2; mean difference 0.6 (−2.9 to 1.8), p=0.63). 23 (32%) and 10 (14%) participants had an adverse event in the surgical and non-surgical groups, respectively. Similar results were found in the observational cohort. Conclusions Surgery is not superior to non-surgical management for 44-B1 ankle fractures in the short term, and is associated with increased adverse events. Trial registration number NCT01134094. ORTHOPAEDIC & TRAUMA SURGERY ==== Body Strengths and limitations of this study The strengths of CROSSBAT (Combined Randomised and Observational Study of Surgery for type B Ankle fracture Treatment) include allocation concealment. In the randomised cohort, loss to follow-up and cross-over rates were low, and the as-treated analysis supported the findings of the intention-to-treat analysis. Outcome tools were validated and relevant, and assessors were blinded. The addition of the observational arm added to generalisability of the findings and addressed selection bias. Limitations include the lack of blinding of the surgeons and participants which is unavoidable with this trial design and the use of subjective scoring only. Ankle fractures are common, with 1 in 800 people fracturing their ankle every year.1–3 The most common pattern involves a fracture of the distal fibula (lateral malleolus) at the level of the tibiofibular syndesmosis, otherwise known as an Association for the Study of Internal Fixation (AO) or Orthopaedic Trauma Association (OTA) type B ankle fracture.4–7 If combined with displacement of the ankle mortise or a fracture of the medial malleolus, surgical fixation is the preferred treatment. However, the most common type of ankle fracture involves a type B lateral malleolus fracture without fracture of the medial malleolus or displacement of the talus (AO/OTA-type 44-B1).8 Management options for these AO 44-B1 ankle fractures include surgical stabilisation by internal fixation using a plate and screws or non-surgical management using a cast or a walking boot.1 Advocates for surgical management emphasise the importance of achieving an anatomic reduction with internal fixation, thereby limiting the potential for displacement and instability.9 Advocates for non-surgical management argue that functional outcomes are not superior to surgical stabilisation and surgery is associated with significant costs and possible adverse events.8 10–12 These include the general risks of anaesthesia and surgery, such as death, venous thromboembolism, infection, failure of fixation and the need for revision surgery.12 Slobogean et al13 showed that the average costs of non-surgical and surgical management of an unstable, isolated, lateral malleolar fracture were US$1892 and US$6404, respectively. A national survey of 358 orthopaedic surgeons in Australia revealed that surgical management of this common fracture is preferred by ∼40% of surgeons, despite a lack of evidence to support this approach.14 Recognising the costs and risks associated with surgery, the lack of evidence supporting the benefit of surgery and the considerable practice variation, we designed a randomised trial to determine the comparative effectiveness of surgical and non-surgical management. In this study involving participants with a 44-B1 ankle fracture, we sought to determine whether surgical management provided superior ankle function and quality of life at 12 months postinjury when compared with non-surgical management. A concurrent observational cohort study was included to provide further evidence regarding the outcomes obtained in routine practice and to improve the generalisability of the results. Methods Study design CROSSBAT (Combined Randomised and Observational Study of Surgery for type B Ankle fracture Treatment) was a pragmatic, multicentre, parallel-group, superiority, randomised controlled trial with an observational cohort that recruited participants from August 2010 to October 2013. It involved 22 hospitals in Australia and New Zealand that were a mix of rural, regional and metropolitan centres (a list of recruiting hospitals is provided in the online supplementary appendix). The main study was the randomised group, and participants declining randomisation were invited to participate in the observational cohort. The protocol was approved by the ethics committees relevant for each site. The full protocol can be accessed as an online supplementary material on the BMJ website. 10.1136/bmjopen-2016-013298.supp1supplementary appendix Participants Consecutive adult patients presenting to a recruiting hospital during the study period with an isolated, closed AO-type 44-B1 distal fibula fracture without significant talar shift presenting within 10 days of injury were screened for eligibility. Significant talar shift was defined as medial clear space being at least 2 mm wider than the superior clear space on a mortise X-ray view of the ankle. Further inclusion criteria were patients aged between 18 and 65 years inclusive with no other concomitant fractures/dislocations; mobilising unaided/independently preinjury; and willing to be followed up for 12 months. Exclusion criteria were participants who were medically unfit for anaesthesia/surgery; skeletally immature; previous trauma or surgery to the fractured ankle; inability to consent; pregnancy; the presence of comorbidities that impede mobilisation; and non-English speaking. Written informed consent was obtained from all patients willing to participate. Randomisation and blinding Eligible participants willing to be randomised were randomly allocated in a 1:1 ratio to either the surgical or non-surgical intervention. The National Health and Medical Research Council Clinical Trials Centre (not otherwise involved in the study) generated the randomisation schedule using a permuted block approach with variable block size and stratified by site. Randomisation was administered using an automated telephone-based system that provided allocation concealment. Owing to the nature of the interventions, neither the investigators nor the participants were blinded. Outcome assessors were independent of the treating teams, and collected data using a standardised telephone interview. As part of the opening conversation, patients were advised not to disclose their treatment so that the assessor could remain blind to treatment. After randomisation, the surgical group received surgery within 10 days of injury. Eligible participants who declined randomisation were invited to enter the observational cohort. Treatment for the observational cohort was determined by participant and surgeon preference. Procedures During protocol development, members of the Australian Orthopaedic Trauma Society were consulted regarding the best practice for the surgical and non-surgical management of 44-B1 ankle fractures as well as the primary and secondary outcomes. Patient eligibility centred on the presence of the fracture of interest. An external rotation stress test to assess the stability of the ankle was not performed as it was not routine practice in Australia owing to uncertainty about its validity and clinical utility.15 16 The focus for the effectiveness of the interventions was patient-reported outcomes. Radiological measures beyond 6 weeks were not required as they were unlikely to demonstrate any osteoarthritic changes and because late malalignment was considered rare (with both methods of treatment) and unlikely to influence management without clinical symptoms. One recruiting site declined to randomise participants due to lack of equipoise within the orthopaedic department and contributed to the observational cohort only. The technique for surgical management was surgical fixation using a plate and screws. Surgeries were performed by orthopaedic surgeons or by orthopaedic trainees under the supervision of consultant orthopaedic surgeons following the AO principles of fracture fixation. Plate placement and reduction techniques were left to the discretion of the surgeon. Adverse intraoperative or postoperative events were recorded. Postoperatively, all participants were non-weight bearing and placed in a below-knee plaster cast or walking boot. Discharge from hospital was determined by the participant's ability to walk 25 m unaided with standby assistance as determined by a physiotherapist (usual discharge criteria). The treating surgeon reviewed the participant after 10–14 days for wound assessment and change of cast to a walking cast or a walking boot (cam walker). The participant was then allowed full weight bearing. The treating surgeon reviewed the participant 6 weeks postinjury with ankle radiographs and removed the cast or walking boot. Participants who were treated non-surgically were managed with a walking boot and allowed full weight bearing. Discharge from hospital was determined as for the surgical group. All participants were examined within 10–14 days postinjury by the treating surgeon who assessed the patient with new ankle radiographs. The treating surgeon reviewed the participant 6 weeks postinjury with repeat ankle radiographs and removed the cast or walking boot. Referral to physiotherapy for all participants was at the discretion of the treating surgeon. Outcomes The primary outcome measures were patient-reported ankle function using the American Academy of Orthopaedic Surgeons Foot and Ankle Outcomes Questionnaire (FAOQ) and the health-related quality of life using the physical component score (PCS) of the SF-12v2 General Health Survey at 12 months postinjury. The FAOQ is a validated, patient-reported outcome measure that assesses ankle function with a higher value indicating better function.17 18 Normative FAOQ scores were used, with a score of 50 representing the mean in the general population and an SD of 10.19 Similarly, the SF-12v2 is a validated patient-reported outcome measure that has been used for the assessment of people with ankle fractures, with a higher value indicating better health.20–22 Both the SF-12v2 and the FAOQ have been used previously for patients with ankle fractures.22 23 Secondary end points included any adverse events in the 12 months postinjury; return to work at 6 weeks and 3, 6 and 12 months postinjury; the PCS and FAOQ at 3 and 6 months postinjury; and the mental component score of the SF-12v2 at 3, 6 and 12 months postinjury. Adverse events were classified as major (unplanned/repeat surgery; infection requiring admission to hospital; pulmonary embolus or death) or minor (neurological injury not requiring further intervention; infections not requiring hospital admission; deep vein thrombosis or other adverse events not requiring hospital admission or surgery).24 The adverse events were collected at 6 weeks and 3, 6 and 12 months postinjury. Follow-up assessments were conducted by telephone. Physiotherapy use (number of visits) was measured. Statistical analysis The PCS has an SD of 10 points and a 5-point difference (equivalent to a 0.5 SD) is considered to be the minimum clinically important difference.20 21 25 A sample size of 160 in the randomised cohort was used to provide 80% power to detect a five-point difference in the PCS between the two groups at a significance level of 0.05, allowing for a 20% loss to follow-up. The normative FAOQ score has an SD of 10, with a 5-point difference (0.5 SD) regarded as the minimum clinically important difference.19 The same sample size (160) would provide the same power to detect a 0.5 SD difference in the FAOQ. There was no sample size target for the observational cohort as this cohort was to provide online supplementary information for the randomised cohort. The randomised and observational cohorts were analysed separately. The primary analysis, conducted using intention-to-treat principles, was performed on the randomised cohort; an as-treated analysis was also performed on the randomised cohort for sensitivity testing. Normality was assessed and Student's t-test was used to compare continuous variables between groups. Missing data were not imputed. The χ2 or Fisher's exact test was used for categorical data analysis as appropriate. Statistical analysis was conducted using SAS V.9.4 (Cary, North Carolina, USA). Both primary outcomes were required to be significantly better in the surgical arm in order for surgery to be regarded as superior. The trial was registered with clinicaltrials.gov (NCT01134094). 10.1136/bmjopen-2016-013298.supp2supplementary information Patient involvement Patients were involved in the development of the outcome measures.17 19–21 They were not involved in the development or conduct of the study. Publication details will be disseminated to study participants who expressed an interest in knowing the results of this study. All participants were thanked in Acknowledgements statement for participating in this study. The burden of intervention on patients was assessed and considered to be low by the ethics committee that assessed the research project (given that both the intervention and control arms are routine practice); no patients were involved in that assessment. This was done as part of a survey of patient factors influencing participation in surgical randomised trials embedded within CROSSBAT.26 Role of the funding source This trial was supported in part by a grant from the Australian Orthopaedic Association Research Foundation. RM was supported with: a postgraduate scholarship from the National Health and Medical Research Council, Avant Doctors-in-training research scholarship and the Foundation for Surgery John Loewenthal Research Fellowship from the Royal Australasian College of Surgeons. The funding organisations of the study had no role in the study design, data collection, data analysis, data interpretation or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication. Results From 15 August 2010 to 3 October 2013, 436 participants who presented with an isolated, closed AO-type 44-B1 distal fibula fracture with minimal talar shift were screened and all were recruited; 160 participants were randomised to the randomised cohort and all 276 participants who declined randomisation were included in the observational cohort. The cohort ascertainment and retention flow chart is presented in figure 1. Figure 1 Cohort ascertainment and retention. DVT, deep vein thrombosis. In the randomised cohort, 80 participants were randomised to non-surgical management and 80 were randomised to surgical management. At 12 months, 68 (85%) and 71 (89%) participants were followed up in the non-surgical and surgical groups, respectively. The intention-to-treat analysis kept participants in the groups to which they were randomised, but the numbers are incomplete due to missing data. In the observational cohort, 257 participants were treated non-surgically and 19 were treated surgically as most patients declined surgery when informed of equipoise regarding the two treatment arms. At 12 months, 202 (79%) participants were followed up in the non-surgical group, and 18 (95%) participants were followed up in the surgical group. Baseline participant characteristics were similar between the two groups in the randomised cohort. In the observational cohort, the surgical group was significantly younger than the non-surgical group (mean difference 8.3, 95% CI 2.6 to 14.0; p=0.007). There were no other significant differences in baseline demographics between the two groups in the observational cohort. Baseline characteristics are shown in table 1. Comparison of baseline data between the randomised and observational cohorts showed that both cohorts had similar demographic profiles. Table 1 Baseline demographics for CROSSBAT Randomised cohort Observational cohort Variable Surgical (n=80) Non-surgical (n=80) Surgical (n=19) Non-surgical (n=257) Age, mean (SD), years 38.1 (13.0) 39.8 (13.7) 31.1 (11.5)* 39.4 (13.7)* Female, n (%) 42 (53) 41 (51) 5 (26) 115 (45) BMI, mean (SD), kg/m2 27.7 (5.2) 28.4 (6.6) 26.2 (2.9) 27.6 (5.5) Left side, n (%) 41 (51) 46 (58) 11 (58) 120 (47) Mechanism, n (%)  Fall <1 m 70 (90) 67 (84) 17 (90) 232 (92)  Fall >1 m 6 (8) 8 (10) 0 (0) 9 (4)  Motor vehicle accident 2 (3) 5 (6) 2 (11) 11 (5) Education, n (%)  High school or lower 31 (39) 44 (55) 11 (58) 100 (39)  TAFE/diploma 30 (38) 23 (29) 4 (21) 78 (30)  University or above 17 (21) 12 (15) 4 (21) 73 (29) Diabetes mellitus, n (%) 3 (4) 4 (5) 0 (0) 10 (4) Peripheral vascular disease, n (%) 1 (1) 0 (0) 0 (0) 1 (1) Alcohol, n (%)† 60 (78) 63 (79) 15 (79) 177 (69) Smoker, n (%)‡ 29 (36) 28 (35) 9 (47) 74 (29) Working, n (%) 64 (80) 65 (81) 15 (79) 197 (77) Insurance status, n (%)  Public 50 (63) 57 (71) 7 (37) 160 (63)  Private 18 (23) 19 (24) 9 (47) 75 (30)  Compensation 10 (13) 3 (4) 3 (16) 18 (7) *Surgical group was significantly younger than non-surgical group in the observational cohort (p=0.007). †A patient was described as consuming alcohol if they were drinking one or more standard drinks per month. ‡A patient was described as a smoker if they were smoking one or more cigarettes per month. BMI, body mass index; CROSSBAT, Combined Randomised and Observational Study of Surgery for type B Ankle fracture Treatment; TAFE, Technical and Further Education. For the randomised cohort, at 12 months, intention-to-treat analysis demonstrated that the surgical group was not superior to the non-surgical group. With respect to the FAOQ, there was a statistically significant difference favouring the non-surgical group (mean difference 3.2; 95% CI 0.4 to 5.9; p=0.028), but this difference was not clinically meaningful. The minimum and maximum values of FAOQ scores were 5.8–55.6 and 32.6–55.6 for the surgical and non-surgical groups, respectively. The surgical group was not superior to the non-surgical group with respect to the PCS (mean difference 0.6, favouring the non-surgical group; 95% CI −2.9 to 1.8; p=0.63). The surgical group had a significantly higher proportion of participants with overall adverse events (risk ratio (RR)=2.3; 95% CI 1.2 to 5.4; p=0.01) and minor adverse events (RR=2.9; 95% CI 1.3 to 6.4; p=0.009). No significant differences in the proportion of participants with major adverse events were found (RR=2.0; 95% CI 0.5 to 7.8; p=0.30). A breakdown of the adverse events is provided in the online supplementary appendix. There was one death in the non-surgical group. This participant was an intravenous drug user who overdosed and died between 6 and 12 months postinjury. The length of hospital stay was shorter in the non-surgical group (mean difference 1.5 days; 95% CI 0.9 to 2.0; p<0.001). A significantly higher proportion of participants from the surgical group used outpatient physiotherapy (RR=1.5; 95% CI 1.1 to 2.2; p=0.01). There was no significant difference between the surgical and non-surgical groups with respect to the proportion of participants (of those who were working preinjury) returning to work at 6 weeks (RR=0.87; 95% CI 0.62 to 1.2; p=0.41). A summary of the outcomes is presented in table 2 and figure 2. Table 2 Results for the intention-to-treat analysis Variable Randomised cohort (intention-to-treat analysis) Surgical Non-surgical Difference (95% CI) p Value 3 months n=72 n=69 FAOQ, mean (SD) 43.8 (12.0) 44.7 (12.2) 0.9 (−3.1 to 5.0)* 0.65 PCS, mean (SD) 47.1 (10.5) 46.8 (11.6) 0.24 (−3.9 to 3.5)* 0.90 MCS, mean (SD) 55.0 (10.3) 56.4 (7.4) 1.4 (−1.6 to 4.4)* 0.37 Working, n (%)† 55/64 (86%) 57/61 (93%) 0.47 (0.15 to 1.4)‡ 0.17 6 months n=72 n=69 FAOQ, mean (SD) 49.1 (8.4) 51.9 (5.6) 2.7 (0.4 to 5.1)* 0.025 PCS, mean (SD) 50.4 (8.9) 52.3 (7.4) 1.9 (−0.90 to 4.6)* 0.18 MCS, mean (SD) 56.6 (7.2) 57.2 (7.9) 0.6 (−2.0 to 3.1)* 0.66 Working, n (%)† 62/63 (98) 61/61 (100) NA 1.00 12 months n=71 n=68 FAOQ, mean (SD) 49.8 (10.6) 53.0 (5.2) 3.2 (0.4 to 5.9)* 0.028 PCS, mean (SD) 53.7 (7.1) 53.2 (6.7) 0.6 (−1.8 to 2.9)* 0.63 MCS, mean (SD) 55.2 (11.1) 56.5 (9.7) 1.3 (−2.2 to 4.8)* 0.47 Working, n (%)† 62/63 (98) 60/60 (100) NA 1.00 Any adverse event, n (%) 23/73 (32) 10/74 (14) 2.3 (1.2 to 4.5)‡ 0.009 Major adverse event, n (%) 6/73 (8) 3/74 (4) 2.0 (0.5 to 7.8)‡ 0.33 Minor adverse event, n (%) 20/73 (27) 7/74 (10) 2.8 (1.3 to 6.4)‡ 0.006 Physiotherapy use, n (%) 44/73 (60) 28/72 (39) 1.5 (1.1 to 2.2)‡ 0.010 *Mean difference (95% CI). †Based on the number of participants working preinjury. ‡Risk ratio (95% CI). FAOQ, American Academy of Orthopaedic Surgeons Foot and Ankle Outcomes Questionnaire; MCS, mental component scores; NA, not available; PCS, physical component scores. Figure 2 Differences between surgical and non-surgical groups with respect to ankle function and general health for the randomised cohort. FAOQ, PCS and MCS are the scores of the SF-12v2 general health survey for the randomised and cohort. Higher value represents better function. Error bars represent 95% CI. Solid black line represents the non-surgical group while the dashed grey line represents the surgical group. FAOQ, American Academy of Orthopaedic Surgeons Foot and Ankle Outcomes Questionnaire; MCS, mental component scores; PCS, physical component scores. There were 10 protocol violations; 8 patients randomised to the surgical group were treated non-surgically (7 later declined surgery; 1 was diagnosed with a deep vein thrombosis presurgery) and 2 patients randomised to the non-surgical group were treated surgically due to protocol violations by treating surgeons. An as-treated analysis of the randomised cohort was also conducted. It also showed that the surgical group was not superior to the non-surgical group for any outcomes. These results are presented in the online supplementary appendix. Results for the observational cohort are presented in the online supplementary appendix as well. Discussion Principal findings In adult patients aged from 18 to 65 years with an isolated type B ankle fracture with minimal talar shift, surgical management was not superior to non-surgical management in terms of ankle function and health-related quality of life at 12 months postinjury. Furthermore, surgical management was not superior to non-surgical management for any secondary outcomes and it was associated with a longer length of hospital stay and a higher rate of adverse events. CROSSBAT was a randomised controlled trial with a parallel observational cohort. The randomised cohort provides a robust comparison of effectiveness between the two treatment groups while the observational cohort provides a concurrent cohort subjected to routine clinical practice. The two cohorts had largely similar baseline characteristics indicating that the results of the randomised trial are generalisable to similar patients who decline randomisation. Further details of baseline comparisons are provided in the appendix. For these reasons, we believe that dissemination of the results of CROSSBAT will help address the practice variation that exists in this area.14 27 Comparison with other studies A recent systematic review conducted by Donken et al1 showed that there was insufficient evidence to justify surgical management of type B ankle fractures. This is because the prevailing RCTs identified by the review included patients with either different patterns of ankle fractures and/or with significant talar shift that potentially confounds the need for surgery.7 28–32 A recent study consented 81 patients to either surgical or non-surgical management for potentially unstable type B ankle fractures (type B ankle fractures that had a positive external rotation stress test indicating a significant lateral talar shift).33 Despite the presence of slight talar misalignment in 20% of the non-surgical group at 1 year, patients managed surgically did not have superior functional outcomes to those managed non-surgically.33 It is possible that a minority of patients in the non-surgical group studied within CROSSBAT also had some misalignment at 1 year, but it was likely to have been subclinical given the good clinical scores. To assess the longer term implications of surgical and non-surgical management of these ankle fractures, we plan to conduct longer term follow-up of the participants using both radiographic and functional measures. Strengths and limitations The strengths of CROSSBAT include allocation concealment, which was assured through employment of a third party overseeing randomisation and allocation. In the randomised cohort, loss to follow-up and cross-over rates were low, and the as-treated analysis supported the findings of the intention-to-treat analysis. Outcome tools were validated and relevant, and assessors were blinded. The addition of the observational arm added to generalisability of the findings and addressed selection bias. Limitations include the lack of blinding of the surgeons and participants which is unavoidable with this trial design. It is also possible that some eligible participants were missed, as recruitment fluctuated over time and between sites, given that dedicated research officers were not present at the sites due to funding constraints. However, all participants who were approached were willing to be recruited to either the randomised or observational cohort. The physiotherapy practices postinjury were not controlled, as participants were free to access physiotherapy services as desired. It was noted that a higher proportion of participants managed surgically sought physiotherapy. This, however, did not result in improved patient-reported outcomes for the surgical group. Further, a recent review by Lin Chung-Wei et al34 showed no evidence of improved outcomes with physiotherapy-based rehabilitation following ankle fractures. Some may consider the use of subjective scoring to be a limitation; however, both the SF-12v2 and the FAOQ have been validated and used previously for patients with ankle fractures.22 23 It can also be argued that clinical decisions about treating patients should be based on symptoms rather than radiographs. Although this study presents 1-year results, future research would include further follow-up of this cohort to assess the longer term effect of surgical and non-surgical management of these 44-B1 ankle fractures. Conclusion The results of this study demonstrate that surgical management is not superior to non-surgical management in type B ankle (fibula) fractures with minimal talar shift in the short term and is associated with increased adverse events. Further follow-up is needed to assess the difference between the two groups in the longer term. The authors would like to thank all the participants who participated in this study. Collaborators: CROSSBAT Study Group: SA, Liverpool Hospital, New South Wales, Australia, sam.adie@gmail.com; Paul Allcock, Lyell McEwin Hospital, South Australia, Australia, Paul.Allcock@health.sa.gov.au; Mustafa Alttahir, Campbelltown Hospital, New South Wales, Australia, Mustafa.alttahir@gmail.com; Zsolt Balogh, John Hunter Hospital, New South Wales, Australia, Zsolt.Balogh@hnehealth.nsw.gov.au; Aziz Bhimani, Wollongong Hospital, New South Wales, Australia, azizb@mac.com; Russell Bourne, Nambour Hospital, Queensland, Australia, bourner@bigpond.com; Richard Boyle, Royal Prince Alfred Hospital, New South Wales, Australia, rich_boyle@hotmail.com; David Broe, Prince of Wales Hospital, New South Wales, Australia, davebroe@lycos.com; Andrew Bucknill, Royal Melbourne Hospital, Victoria, Australia, andrew.bucknill@mh.org.au; Mellick Chehade, Royal Adelaide Hospital, South Australia, Australia, mellick.chehade@adelaide.edu.au; Raymond Chin, Liverpool Hospital, New South Wales, Australia, ray.chin@iinet.net.au; Andrew Clout, Campbelltown Hospital, New South Wales, Australia, a.j.clout@gmail.com; Frank Connon, St George Hospital, New South Wales, Australia, francisconnon@gmail.com; Cameron Cooke, Princess Alexandra Hospital, Queensland, Australia, cameroncooke@lowerlimb.com.au; Chandra Dave, Liverpool Hospital, New South Wales, Australia, chandradave61@gmail.com; Jaykar Dave, Liverpool Hospital, New South Wales, Australia, orthopod@optusnet.com.au; Mark Dekkers, Princess Alexandra Hospital, Queensland, Australia, mark_dekkers@health.qld.gov.au; Herwig Drobetz, Mackay Base Hospital, Queensland, Australia, drobo@aon.at; Richard Farrugia, Royal Melbourne Hospital, Victoria, Australia, richard.farrugia@mh.org.au; Brett Fritsch, Royal Prince Alfred Hospital, New South Wales, Australia, brett@brettfritsch.com.au; Sanjeev Gupta, Royal Prince Alfred Hospital, New South Wales, Australia, sanjeev.gupta@bigpond.com; IAH, Liverpool Hospital, New South Wales, Australia, iaharris1@gmail.com; Chris Hoffman, Wellington Hospital, Wellington, New Zealand, cwhoffman@xtra.co.nz; Michael Holt, Royal Brisbane Hospital, Queensland, Australia, mjholt@wesportsmed.com.au; Mark Horseley, Royal Prince Alfred Hospital, New South Wales, Australia, drmhorsley@gmail.com; Stephen Hutchinson, Prince of Wales Hospital, New South Wales, Australia, shutchin1978@gmail.com; Prajith Jeyaprakash, University of New South Wales, New South Wales, Australia, prajithjeyaprakash@gmail.com; Steven Kent, Wollongong Hospital, New South Wales, Australia, skent1985@gmail.com; Doug King, Princess Alexandra Hospital, Queenslnd, Australia, lloyd_king@health.qld.gov.au; Victoria Ko, Liverpool Hospital, New South Wales, Australia, vwmko@yahoo.com; Vinay Kulkarni, Campbelltown Hospital, New South Wales, Australia, vinay@doctor.id.au; Martin Laird, Liverpool Hospital, New South Wales, Australia, mlaird@gmp.usyd.edu.au; David Lieu, Liverpool Hospital, New South Wales, Australia, david.t.lieu@bigpond.com; Andreas Loefler, Prince of Wales Hospital, New South Wales, Australia, andreasloefler@optusnet.com.au; David Lunz, Prince of Wales Hospital, New South Wales, Australia, footandanklesurgery@yahoo.com; Genni Lynch, Princess Alexandra Hospital, Queensland, Australia, genni.lynch@health.qld.gov.au; Lawrie Malisano, Princess Alexandra Hospital, Queensland, Australia, l.malisano@uq.edu.au; Ian Meakin, Bankstown Hospital, New South Wales, Australia, imeakin@bigpond.net.au; Robert Molnar, Westmead Hospital, New South Wales, Australia, rmolnar@optusnet.com.au; Chris Morrey, Cairns Base Hospital, Queensland, Australia, cmorrey@coc.net.au; Jonathan Mulford, Prince of Wales Hospital, New South Wales, Australia, jonathanmulford1971@gmail.com; Ashish Munsif, University of New South Wales, New South Wales, Australia, z3288258@zmail.unsw.edu.au; Paul Muscio, Cairns Base Hospital, Queensland, Australia, paul_muscio@health.qld.gov.au; Alok Narayan, Wollongong Hospital, New South Wales, Australia, rahintern@yahoo.com.au; Alexander Nicholls, Canberra Hospital, Australian Capital Territory, Australia, alexandernicholls@gmail.com; JMN, University of New South Wales, New South Wales, Australia, Justine.Naylor@sswahs.nsw.gov.au; Fred Nouh, Campbelltown Hospital, New South Wales, Australia, frednouh@yahoo.com.au; Tim O’Carrigan, Campbelltown Hospital, New South Wales, Australia, Dr.Ocarrigan@sbjc.com.au; Ed O'Leary, St George Hospital, New South Wales, Australia, edmundoleary@hotmail.com; Peter Lorentzos, Westmead Hospital, New South Wales, Australia, peter.lorentzos@gmail.com; Sushil Pant, St George Hospital, New South Wales, Australia, sushpant@gmail.com; Diana Perriman, Canberra Hospital, Australian Capital Territory, Australia, Diana.perriman@act.gov.au; Jeffrey Petchell, Royal Prince Alfred Hospital, New South Wales, Australia, jfpetchell@aapt.net.au; Marinis Pirpiris, Royal Melbourne Hospital, Victoria, Australia, surgeons@orthopaedicsvictoria.com.au; Tony Pohl, Royal Adelaide Hospital, South Australia, Australia, anthony.pohl@adelaide.edu.au; Vaibhav Punjabi, Liverpool Hospital, New South Wales, Australia, vaibhav_punjabi@yahoo.com; Sunil Randhawa, Liverpool Hospital, New South Wales, Australia, sunil.randhawa@ausdoctors.net; Matthias Rau, Rockhampton Hospital, Queensland, Australia, gainsgorn@yahoo.com.au; Jennie Scarvell, Canberra Hospital, Australian Capital Territory, Australia, Jennie.Scarvell@canberra.edu.au; Michael Schuetz, Princess Alexandra Hospital, Queensland, Australia, m.schuetz@qut.edu.au; Ben Scwartz, Wollongong Hospital, New South Wales, Australia, drbenschwarz@gmail.com; Paul Smith, Canberra Hospital, Australian Capital Territory, Australia, psmith.admin@orthoact.com.au; Brahman Sivakumar, Campbelltown Hospital, New South Wales, Australia, brahman.sivakumar@gmail.com; Bogdan Solomon, Royal Adelaide Hospital, South Australia, Australia, bogdan.solomon@health.sa.gov.au; Jonathan Spencer, Sir Charles Gairdner Hospital, Western Australia, Australia, jonmfspencer@gmail.com; Paul Stalley, Royal Prince Alfred Hospital, New South Wales, Australia, pdstalley@optusnet.com.au; Mitchell Steele, Wollongong Hospital, New South Wales, Australia, mitchelljsteele@gmail.com; Mayuran Suthersan, Campbelltown Hospital, New South Wales, Australia, msuthersan@gmail.com; Zoltan Szomor, St George Hospital, New South Wales, Australia, zszomor@hotmail.com; Peter Tamblyn, Flinders Medical Centre, South Australia, Australia, Peter.Tamblyn@health.sa.gov.au; Seth Tarrant, John Hunter Hospital, New South Wales, Australia, sethtarrant@gmail.com; Kevin Tetsworth, Royal Brisbane Hospital, Queensland, Australia, kevin_tetsworth@health.qld.gov.au; Sameer Viswanathan, Campbelltown Hospital, New South Wales, Australia, sameerv@me.com; Richard Walker, Liverpool Hospital, New South Wales, Australia, drrmwalker@hotmail.com. Contributors: RM and IAH had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis and act as guarantors. RM, IAH, SA, JMN and CROSSBAT Study Group were involved in study, concept, design, acquisition, critical revision of the manuscript for important intellectual content. RM and IAH conducted statistical analysis and are responsible for the data analysis. Funding: This trial was supported in part by a grant from the Australian Orthopaedic Association Research Foundation. RM was supported by: a postgraduate scholarship from the National Health and Medical Research Council, Avant Doctors-in-training research scholarship and the Foundation for Surgery John Loewenthal Research Fellowship from the Royal Australasian College of Surgeons. None of the organisations had a direct role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review or approval of the manuscript; and decision to submit the manuscript for publication. Competing interests: None declared. Ethics approval: This study was approved by the following ethics committees: Central Regional Ethics Committee. Reference Number: CEN/12/06/030 Ethics of Human Research Committee (TQEH and LMH). Reference Number: 2010130 Flinders Clinical Research Ethics Committee. Reference Number: 358.10 Hunter New England Human Research Ethics Committee. Reference Number: 09/12/16/5.02 Melbourne Health Human Research Ethics Committee. Reference Number: 2010.027 Metro South Human Research Ethics Committee. Reference Number: HREC/11/QPAH/145 Royal Adelaide Hospital Research Ethics Committee. Reference Number: 100705 Sir Charles Gairdner Group Human Research Ethics Committee. Reference Number: 2010-092 University of New South Wales Human Research Ethics Committee. Reference Number: HC12187 Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Additional data from the study can be obtained from the corresponding author at rajatmittal.syd@gmail.com ==== Refs References 1 Donken CC , Al-Khateeb H , Verhofstad Michael HJ Surgical versus conservative interventions for treating ankle fractures in adults . Cochrane Database Syst Rev 2012 ;(8) :CD008470 10.1002/14651858.CD008470.pub2 22895975 2 Court-Brown CM , Caesar B Epidemiology of adult fractures: a review . Injury 2006 ;37 :691 –7 . 10.1016/j.injury.2006.04.130 16814787 3 Melton LJ III, Crowson CS , O'Fallon WM Fracture incidence in Olmsted County, Minnesota: comparison of urban with rural rates and changes in urban rates over time . Osteoporos Int 1999 ;9 :29 –37 . 10.1007/s001980050113 10367027 4 Court-Brown CM , McBirnie J , Wilson G Adult ankle fractures–an increasing problem? Acta Orthop Scand 1998 ;69 :43 –7 . 10.3109/17453679809002355 9524517 5 Martin AG Weber B ankle fracture: an unnecessary fracture clinic burden . Injury 2004 ;35 :805 –8 . 10.1016/j.injury.2003.12.013 15246805 6 Clare MP A rational approach to ankle fractures . Foot Ankle Clin 2008 ;13 :593 –610 . 10.1016/j.fcl.2008.09.003 19013398 7 Phillips WA , Schwartz HS , Keller CS A prospective, randomized study of the management of severe ankle fractures . J Bone Joint Surg Am 1985 ;67 :67 –78 . 10.2106/00004623-198567010-00010 3881447 8 Michelson JD Fractures about the ankle . J Bone Joint Surg Am 1995 ;77 :142 –52 . 10.2106/00004623-199501000-00020 7822349 9 Michelson JD , Magid D , McHale K Clinical utility of a stability-based ankle fracture classification system . J Orthop Trauma 2007 ;21 :307 –15 .17485995 10 Van Schie-Van der Weert EM , Van Lieshout EMM , De Vries MR Determinants of outcome in operatively and non-operatively treated Weber-B ankle fractures . Arch Orthop Trauma Surg 2012 ;132 :257 –63 . 10.1007/s00402-011-1397-z 21959696 11 Murray AM , McDonald SE , Archbold P Cost description of inpatient treatment for ankle fracture . Injury 2011 ;42 :1226 –9 . 10.1016/j.injury.2010.08.023 20869055 12 SooHoo NF , Krenek L , Eagan MJ Complication rates following open reduction and internal fixation of ankle fractures . J Bone Joint Surg Am 2009 ;91 :1042 –9 . 10.2106/JBJS.H.00653 19411451 13 Slobogean GP , Marra CA , Sadatsafavi M on behalf of the Canadian Orthopedic Trauma S . Is surgical fixation for stress-positive unstable ankle fractures cost effective? Results of a multicenter randomized control trial . J Orthop Trauma 2012 ;26 :652 –8 .22473067 14 Ansari U , Adie S , Harris IA Practice variation in common fracture presentations: a survey of orthopaedic surgeons . Injury 2011 ;42 :403 –7 . 10.1016/j.injury.2010.11.011 21163480 15 Koval KJ , Egol KA , Cheung Y Does a positive ankle stress test indicate the need for operative treatment after lateral malleolus fracture? A preliminary report . J Orthop Trauma 2007 ;21 :449 –55 . 10.1097/BOT.0b013e31812eed25 17762475 16 Egol KA , Amirtharajah M , Tejwani NC Ankle stress test for predicting the need for surgical fixation of isolated fibular fractures . J Bone Joint Surg Am 2004 ;86 :2393 –8 . 10.2106/00004623-200411000-00005 15523008 17 Johanson NA , Liang MH , Daltroy L American Academy of Orthopaedic Surgeons lower limb outcomes assessment instruments. Reliability, validity, and sensitivity to change . J Bone Joint Surg Am 2004 ;86-A :902 –9 .15118030 18 Suk M Musculoskeletal outcomes measures and instruments . Thieme , 2009 :1 . 19 Hunsaker FG , Cioffi DA , Amadio Peter C The American Academy of Orthopaedic Surgeons outcomes instruments: normative values from the general population . J Bone Joint Surg Am 2002 ;84-A :208 –15 .11861726 20 Ware JJ , Kosinski M , Keller SD A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity . Med Care 1996 ;34 :220 –33 .8628042 21 Ware JE Jr, Kosinski M , Turner-Bowker DM User's manual for the SF-12v2 health survey with a supplement documenting SF-12 health survey . Lincoln, RI : QualityMetric Incorporated , 2002 . 22 Sultan MJ , Zhing T , Morris J Compression stockings in the management of fractures of the ankle: a randomised controlled trial . Bone Joint J 2014 ;96-B :1062 –9 .25086122 23 So S , Harris IA , Naylor JM Correlation between metal allergy and treatment outcomes after ankle fracture fixation . J Orthop Surg (Hong Kong) 2011 ;19 :309 –13 . 10.1177/230949901101900309 22184160 24 Dindo D , Demartines N , Clavien P-A Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey . Ann Surg 2004 ;240 :205 –13 . 10.1097/01.sla.0000133083.54934.ae 15273542 25 Resnick B , Nahm ES Reliability and validity testing of the revised 12-item Short-Form Health Survey in older adults . J Nurs Meas 2001 ;9 :151 –61 .11696939 26 Mittal R , Harris IA , Adie S Factors affecting patient participation in orthopaedic trials comparing surgery to non-surgical interventions . Contemp Clin Trials 2016 ;3 :153 –7 . 27 Timmermans S , Mauck A The promises and pitfalls of evidence-based medicine . Health Affairs 2005 ;24 :18 –28 . 10.1377/hlthaff.24.1.18 15647212 28 Pakarinen HJ , Flinkkilä TE , Ohtonen PP Syndesmotic fixation in supination-external rotation ankle fractures: a prospective randomized study . Foot Ankle Int 2011 ;32 :1103 –9 . 10.3113/FAI.2011.1103 22381193 29 Kortekangas TH , Pakarinen HJ , Savola O Syndesmotic fixation in supination-external rotation ankle fractures: a prospective randomized study . Foot Ankle Int 2014 ;35 :988 –95 . 10.1177/1071100714540894 24962527 30 Bauer M , Bergstrom B , Hemborg A Malleolar fractures: nonoperative versus operative treatment. A controlled study . Clin Orthop 1985 ;199 :17 –27 . 31 Makwana NK , Bhowal B , Harper WM Conservative versus operative treatment for displaced ankle fractures in patients over 55 years of age. A prospective, randomised study.[see comment] . J Bone Joint Surg Br 2001 ;83 :525 –9 . 10.1302/0301-620X.83B4.11522 11380123 32 Rowley DI , Norris SH , Duckworth T A prospective trial comparing operative and manipulative treatment of ankle fractures . J Bone Joint Surg Br 1986 ;68 :610 –13 . 10.2106/00004623-198668040-00021 3090049 33 Sanders DW , Tieszer C , Corbett B , Society OBOTCOT . Operative versus nonoperative treatment of unstable lateral malleolar fractures: a randomized multicenter trial . J Orthop Trauma 2012 ;26 :129 –34 . 10.1097/BOT.0b013e3182460837 22330975 34 Lin Chung-Wei C , Donkers Nicole AJ , Refshauge Kathryn M Rehabilitation for ankle fractures in adults . Cochrane Database Syst Rev 2012 [cited 2015 Oct 29]. http://onlinelibrary.wiley.com/doi/10.1002/14651858.CD005595.pub3/abstract

PMC005xxxxxx/PMC5372108.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01298510.1136/bmjopen-2016-012985Intensive CareResearch150617071719Investigating skin-to-skin care patterns with extremely preterm infants in the NICU and their effect on early cognitive and communication performance: a retrospective cohort study Gonya Jenn 1Ray William C 2Rumpf R Wolfgang 2Brock Guy 3 1 The Research Institute at Nationwide Children's Hospital, Center for Perinatal Research, Columbus, Ohio, USA 2 The Research Institute at Nationwide Children's Hospital, Battelle Center for Mathematical Medicine, Columbus, Ohio, USA 3 Department of Biomedical Informatics, Center for Biostatistics, The Ohio State University, Columbus, Ohio, USACorrespondence to Dr Jenn Gonya; Jenn.Gonya@nationwidechildrens.org2017 20 3 2017 7 3 e0129858 6 2016 9 12 2016 19 12 2016 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives The primary objective of the study was to investigate how patterns of skin-to-skin care might impact infant early cognitive and communication performance. Design This was a retrospective cohort study. Setting This study took place in a level-IV all-referral neonatal intensive care unit in the Midwest USA specialising in the care of extremely preterm infants. Participants Data were collected from the electronic medical records of all extremely preterm infants (gestational age <27 weeks) admitted to the unit during 2010–2011 and who completed 6-month and 12-month developmental assessments in the follow-up clinic (n=97). Outcome measures Outcome measures included the cognitive and communication subscales of the Bayley Scales of Infant Development, Third Edition (Bayley-III); and skin-to-skin patterns including: total hours of maternal and paternal participation throughout hospitalisation, total duration in weeks and frequency (hours per week). Analysis Extracted data were analysed through a multistep process of logistic regressions, t-tests, χ2 tests and Fisher's exact tests followed with exploratory network analysis using novel visual analytic software. Results Infants who received above the sample median in total hours, weekly frequency and total hours from mothers and fathers of skin-to-skin care were more likely to score ≥80 on the cognitive and communication scales of the Bayley-III. However, the results were not statistically significant (p>0.05). Mothers provided the majority of skin-to-skin care with a sharp decline at 30 weeks corrected age, regardless of when extremely preterm infants were admitted. Additional exploratory network analysis suggests that medical and skin-to-skin factors play a parallel, non-synergistic role in contributing to early cognitive and communication performance as assessed through the Bayley-III. Conclusions This study suggests an association between early and frequent skin-to-skin care with extremely preterm infants and early cognitive and communication performance. extreme prematurityNEONATOLOGYNational Center for Advancing Translational Scienceshttp://dx.doi.org/10.13039/100006108UL1TR001070 ==== Body Strengths and limitations of this study Identifies natural, emergent patterns of skin-to-skin care with extremely preterm infants to reflect authentic human engagement experiences. Uses the evidence to suggest ways to target specific intervention areas for increasing skin-to-skin care. Supports current literature on the longer term benefits of skin-to-skin care. Uses one instrument to assess early cognitive and communication performance. Uses retrospective data which exclude variables known to impact neonatal neurodevelopment. Introduction The birth and subsequent hospitalisation of an extremely preterm infant is a trauma event. Unlike term infants, extremely preterm infants (infants born at <27 weeks) spend the last trimester of their gestation ex utero, in an artificial, technology-laden neonatal intensive care unit (NICU) that places them at a developmental disadvantage. Monitors, tubing and wires often create an environment that makes it difficult for authentic positive human interaction. In response, skin-to-skin care (SSC) has been incorporated into many NICUs across the world to re-establish this positive human contact. SSC is the practice of holding an infant upright on a parent's chest in a manner that provides maximum bare skin ventral contact. The practice impacts infant physiological stability, stress and sleep as well as maternal stress and parenting behaviour. SSC studies over the last 25 years1–15 have collectively translated into a global acknowledgement that SSC is medically safe and significantly affects longer term neurodevelopmental cognitive, social and emotional outcomes.16–22 Despite the benefits of SSC, it is often difficult to engage some families in the practice. Findings from one of the most recent and comprehensive systematic reviews of the barriers and promoters of SSC (included in the complete package known as Kangaroo Mother Care)23 identified over 35 factors involved in integrating SSC into the NICU environment. The top three barriers to SSC were issues with the NICU physical facility, negative impressions by the staff about the practice and fear of injuring the infant during SSC. In contrast, SSC increased when mothers felt attached to their infants, felt confident in their parenting role and received support from family, friends or other mothers. While current studies, such as those found in the systematic review, can help in the design of new interventions promoting SSC, many are a reflection of participating in a highly supported and scrutinised form of SSC rather than parent practice as it naturally occurs in the NICU. What remains unknown is how parents are actually engaging in the practice of SSC in an all-referral NICU setting in the USA when they are not involved in an SSC study. A rigorous study of routine SSC across a cohort of extremely preterm infants could identify specific strategies and intervention points for care providers to target their efforts at increasing parental engagement in SSC. Therefore, the purpose of the current study was to identify the naturalistic patterns of SSC that parents engage in with their extremely preterm infants in an all-referral NICU and investigate how these patterns impact early infant cognitive and communicative performance. A secondary aim was to compare the relative effects of amount and intensity of SSC on these outcomes. Patients and methods This study was a retrospective cohort study of all infants admitted to the Small Baby Intensive Care Unit (SBICU) at Nationwide Children's Hospital (NCH) between 1 January 2010 and 30 November 2011. The SBICU is a specialised level-IV all-referral unit staffed by a centralised team of nurses who provide protocol-driven care23–25 to neonates born at a gestational age (GA) <27 completed weeks. These protocols, organised within the Small Baby Guidelines, outline how to specifically address the medical and developmental needs of extremely preterm infants. SSC is specifically designated as a critical practice for medical stability and neurodevelopmental outcomes and is described as a care piece that should be strongly encouraged whenever possible, as long as possible. All patients cared for in this unit are outborn and are transported to the SBICU for care of complications of prematurity including necrotising enterocolitis, sepsis, surgical issues, brain injury and so on. Data Retrospective data were extracted from the electronic medical record within three categories: (a) medical (b) SSC and (c) cognitive and communication outcomes at follow-up. Medical record information extracted for each patient included gender, GA, birth weight (BW), length of hospital stay (LOS), occurrence or absence of intraventricular haemorrhage in the brain (IVH), number of days on a ventilator (IPPV days), days until first full feed by mouth (PO DOL), whether the patient was a twin, triplet, etc (multiple births), and whether the patient received antenatal steroids. These variables were selected based on the outcome trajectories calculator developed by the Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network.26 Total hours of SSC for each parent were recorded for each day after the baby was admitted to the NICU until discharge. Hours were documented by the nursing staff in the patient medical record. (Audits performed comparing parental report and nurse report of SSC time indicated 89% consistency.) Summary measures of SSC use included total hours of SSC, the number of days between the day of admission and the first onset of SSC, total hours of SSC performed by the mother and father, intensity of SSC (average days of SSC per week) and whether the family participated in SSC after their child reached 33 weeks corrected age, a critical period of auditory development.27 To reduce the number of tested associations and aid in clinical interpretation, families were further classified as having a ‘high’ level of SSC participation if they were above the median in total hours, total hours for mother and total hours for father, and frequency of SSC (ie, above the median for each of the four variables). The remaining families were classified as having a ‘low’ level of SSC participation. Cognitive and communication early performance outcomes were determined through the Bayley Scales of Infant Development, Third Edition (Bayley-III), a valid and reliable developmental assessment tool that is widely used in neonatal follow-up. Assessments were performed at 6 and 12 months by licensed professionals certified and trained in the tool28 and scores were adjusted for prematurity. Descriptive classifications were used according to the protocol outlined by Pearson Clinical with infants scoring <80 being described as ‘Borderline’ for developmental disability.29 Consequently, scores were treated as continuous variables and as dichotomised variables of scores <80 and scores ≥80. Statistical analysis Statistical analysis was divided into three parts to address the clinical questions of interest. First, since the study is observational in nature, patterns of SSC participation (‘high’ vs ‘low’, as defined in the ‘Data’ section) were investigated graphically and associations between SSC measures and medical factors were tested. These associations were considered to explore for potential factors associated with SSC participation and to account for potential confounding of SSC with these other clinical/medical variables. A logistic regression model was fit to contrast the probability of being a high versus low SSC participant (as defined in the ‘Data’ section) as a function of gender, GA, BW, IVH, IPPV days, PO DOL, multiple births and receipt of antenatal steroids. Backwards elimination was used to select a final explanatory model based on minimising the Akaike's information criteria. Second, the association between SSC participation (high vs low) and Bayley-III scores was evaluated. Strength of association between raw Bayley scores and SSC participation (high vs low) was quantified and tested using point biserial correlations (rpb, tested against a null that the correlation was zero). Bayley scores dichotomised at the borderline disability level (<80 vs ≥80) were tested for association with SSC participation using the χ2 test or Fisher's exact test. Associations between dichotomised Bayley scores and SSC participation were additionally adjusted for confounding based on the medical factors found to be associated with SSC participation. Patterns between medical factors and SSC became evident and patterns of SSC and the Bayley-III scores became evident. Consequently, for the final analysis, we used our StickWRLD visual analytic software30 31 to investigate potential triangulations among specific aspects of medical factors, SSC and the Bayley-III scores. All factors were loaded into the StickWRLD visual framework and initial two-node association patterns were set with an initial residual value32 of 0.2. Subsequent analyses were performed incrementally at lower residual values to identify and compare associative relationships and to search for significant emerging triangular data patterns. Analyses concluded when the model reached a threshold residual value corresponding to visual associative overload. Results A total of 97 NICU patients were included in the study. The GA ranged from 22 to 26 weeks with an overall median of 25 weeks. Summary statistics of SSC usage (overall participation, participation by parent, SSC intensity and onset of SSC) are given in table 1. Mothers represented the majority of overall SSC participation, as evidenced by figure 1. Nine families were missing information on some aspect of SSC involvement. Among the remaining 88 families, 30 (34%) were classified as ‘high’ participants in SSC (above the median for total SSC hours, hours per parent and SSC intensity) while the other 58 (66%) were classified as ‘low’ participation in SSC. Table 1 Summary statistics of skin-to-skin care (SSC) from admission to 40 weeks postmenstrual age SSC Metric Mean (SD) Median (IQR) Min, max Total SSC (hours) 27.4 (29.8) 17.2 (5.1, 36.6) 0, 129.8 Mother SSC (hours) 22.8 (22.4) 17.2 (4.6, 30.9) 0, 97 Father SSC (hours) 5.8 (10.4) 1 (0, 7.5) 0, 58 SSC frequency (days/week) 2.3 (1.2) 2.2 (1.3, 3.2) 0, 5 SSC onset (days) 6.2 (7.4) 4 (1.8, 8) 0, 45 IQR (25th centile, 75th centile). Figure 1 Overall and parent-specific SSC participation: boxplots displaying the distribution of overall SSC participation and by parent. Thick horizontal lines give medians while boxes display the middle 50% of the data (25th and 75th centiles). Whiskers extend to no more than 1.5 times the IQR (IQR= difference between 75th and 25th centiles) from the edge of the box. Points beyond the whiskers represent outliers. SSC, skin-to-skin care. Patterns of intensity and total hours of SSC participation between the postmenstrual ages of 23 and 40 weeks are displayed on a study-wide (total person-hours per week, figure 2) and family (hours per family per week, figure 3) basis. There was a steady increase in total hours and hours per family until about 30 weeks, after which there was a corresponding precipitous decline until 40 weeks. Differences in medical factors between families with high versus low SSC participation are given in table 2. Receipt of antenatal steroids was the only significant (p<0.05) finding, with 71% of children from families with high SSC participation receiving antenatal steroids and 91% of children with low family SSC participation receiving them. Table 2 Medical factors influencing skin-to-skin patterns Medical factors (categorical) SSC low SSC high p Value Gender  Female 23 (0.4) 8 (0.27) 0.25  Male 35 (0.6) 22 (0.73) Multiple births  No 39 (0.67) 18 (0.60) 0.64  Yes 19 (0.33) 12 (0.40) Antenatal steroids  No 5 (0.09) 8 (0.29) 0.02  Yes 53 (0.91) 20 (0.71) IVH  No 28 (0.48) 10 (0.33) 0.26  Yes 30 (0.52) 20 (0.67) Medical factors (continuous) Gestational age (weeks) 24.9 (1) 24.4 (1.1) 0.06 Birth weight (g) 748.2 (164.5) 719.6 (188.5) 0.48 Length of hospital stay (days) 117.6 (46.1) 127.8 (40) 0.28 Days on ventilator 41.6 (28.6) 45 (33.2) 0.63 PO DOL (days) 106.1 (35.9) 117.7 (44.6) 0.26 Numbers in each cell are mean (SD) for continuous and N (%) for categorical. p Value for categorical based on χ2, for continuous based on t-test. IVH, intraventricular haemorrhage; PO DOL, days until first full feed by mouth; SSC, skin-to-skin care. Figure 2 Total SSC hours per week. Blue line displays the total number of SSC hours per week for all families in the study, by postmenstrual age. Red line gives the total number of babies in the NICU for the given week. NICU, neonatal intensive care unit; SSC, skin-to-skin care. Figure 3 SSC intensity (hours per week per family). Blue line displays the average SSC intensity (hours per family per week) by postmenstrual age. Red line gives the total number of babies in the NICU for the given week. NICU, neonatal intensive care unit; SSC, skin-to-skin care. These factors (minus LOS, which was omitted because infants with longer LOS might be expected to have longer total SSC duration) were subsequently used to build a model to analyse the variance in SSC participation based on logistic regression with backwards elimination. The resulting model included antenatal steroids, BW and IVH as predictors (table 3). We investigated various cut-points for dichotomising BW and found the 75th centile to provide the best fit. Receipt of antenatal steroids (OR=0.136) and BW in the top quartile (OR=0.152) were associated with reduced odds of high SSC participation, while presence of IVH was associated with increased odds (OR=1.92). Table 3 ORs for high participation in skin-to-skin care (SSC) based on medical factors Univariable Multivariable Factor Levels High SSC OR 95% CI p Value OR 95% CI p Value Antenatal steroids Yes 20/73 (27%) 4.16 1.06 to 18.2 0.024 7.36 1.67 to 32.53 0.008 No 8/13 (62%) Birth weight 844+ 5/24 (21%) 2.41 0.74 to 9.35 0.13 6.59 1.46 to 29.84 0.014 <844 25/64 (39%) IVH Yes 20/50 (40%) 0.54 0.19 to 1.46 0.26 0.52 0.19 to 1.43 0.2 No 10/38 (26%) IVH, intraventricular haemorrhage. Next, association between SSC participation and cognitive and communication outcomes (Bayley scores) at follow-up were investigated. Figure 4 displays boxplots of the Bayley scores stratified by high and low SSC participation, while table 4 gives mean values for each examination by participation group. Communication scores at 12 months were somewhat higher in the high SSC group (p=0.05). We then dichotomised the Bayley score at the borderline disability level (<80 vs ≥80). Table 5 displays the number and percentage of patients that fall below this borderline disability level along with univariate ORs. To account for potential confounding, ORs were further adjusted by fitting a multivariable model including the factors identified to be associated with SSC participation (BW, antenatal steroids and IVH, table 5). None of the associations (except for cognitive examination at 6 months) reached statistical significance. However, there was a relatively consistent OR of 2 for associations between each dichotomised Bayley score and SSC participation in the multivariable models. Adjusted ORs were higher than the unadjusted because SSC participation was associated with factors that were also generally associated with lower Bayley scores. Table 4 Point biserial correlations (rpb) between low and high participation in skin-to-skin care (SSC) and Bayley-III cognitive and communication outcomes Bayley-III assessment SSC low* SSC High* rpb p Value† Cognitive 6 months 92.7 (15.7) 96.3 (15.1) 0.11 0.30 Cognitive 12 months 93.1 (14.6) 93.9 (19.2) 0.03 0.82 Communication 6 months 93.1 (12.9) 96.9 (16.6) 0.13 0.24 Communication 12 months 90.7 (15.4) 98.2 (16.4) 0.23 0.05 Composite (cog/comm) 6 months 92.9 (12.6) 96.6 (14.4) 0.13 0.21 Composite (cog/comm) 12 months 91.9 (13.6) 96.1 (16.1) 0.14 0.22 *Numbers in each cell are mean (SD). †p Values are from test that point biserial correlations are different from zero. Cog/comm, cognitive/communication. Table 5 Associations between low and high participation in skin-to-skin care (SSC) and borderline disability (<80 vs ≥80) Bayley-III cognitive and communication outcomes Per cent borderline developmental disability Univariable Multivariable Bayley examination Low SSC High SSC OR 95% CI p Value OR 95% CI p Value Cognitive 6 months 11/58 (19%) 2/30 (7%) 3.28 (0.92 to 11.67) 0.07 4.46 (1.08 to 18.41) 0.04 Cognitive 12 months 8/49 (16%) 2/28 (7%) 2.54 (0.66 to 9.77) 0.18 2.87 (0.62, 13.26) 0.18 Communication 6 months 9/58 (16%) 4/30 (13%) 1.19 (0.52 to 2.72) 0.67 1.72 (0.64 to 4.61) 0.28 Communication 12 months 13/49 (27%) 5/29 (17%) 1.73 (0.88 to 3.42) 0.11 2.00 (0.87 to 4.57) 0.10 Composite (cog/comm) 6 months 8/58 (14%) 3/30 (10%) 1.44 (0.52 to 3.95) 0.48 2.22 (0.68 to 7.28) 0.19 Composite (cog/comm) 12 months 8/49 (16%) 3/28 (11%) 1.63 (0.58 to 4.53) 0.35 2.22 (0.66 to 7.44) 0.20 Multivariable models include antenatal steroids, birth weight and intraventricular haemorrhage. Cog/comm, cognitive/communication. Figure 4 Bayley Cognitive and Communication Scores by SSC participation: boxplots displaying the distribution of Bayley cognitive, communication and combined cognitive–communication (Cog–Comm) scores at 6 and 12 months by SSC participation (high vs low). High SSC participation was defined as having above the median participation for total SSC hours, mother SSC hours, father SSC hours and SSC intensity (see text for details). Thick horizontal lines give medians while boxes display the middle 50% of the data (25th and 75th centiles). Whiskers extend to no more than 1.5 times the IQR (IQR= difference between 75th and 25th centiles) from the edge of the box. Points beyond the whiskers represent outliers. A final exploratory analysis of the data set was performed using StickWRLD software to identify possible emergent interactive network associations among SSC measures, medical factors and Bayley-III scores. Network displays (figure 5) indicated separate, but convergent significant associations between SSC measures and medical factors and the 12-month Bayley-III cognitive scores. Figure 5 Visual analytical display from StickWRLD software. The visual space is divided into three main regions: medical factors, skin-to-skin factors and Bayley-III Scores. Each line represents a significant correlation between factors with stronger correlations represented by lines that are thicker in diameter. Line colours are assigned randomly and are used only to aid in visual comparisons of associations. Within-region correlations are evident and expected. Two blue arrows indicate two unexpected strong correlations (one from each region) that converge on the Bayley-III 12-Month Cognitive Score, suggesting skin-to-skin Care frequency and presence or absence of intraventricular haemorrhage (IVH) as parallel, but non-interactive factors impacting the score. Discussion Engagement in SSC with extremely preterm infants in the NICU varies among families. However, SSC patterns are evident in this population and potentially have an impact on early cognitive and communication performance. These findings are not new in that numerous studies11–22 have been devoted to the short-term and long-term developmental benefits of SSC. What is initially novel about our findings is that we have found a strong indication that SSC before 30 weeks postmenstrual age may play a role in the cognitive and communication development of extremely preterm infants. The period of time before 32 weeks is often considered a developmentally marginal time period in communication and language development. Underlying brain structure and auditory/visual development are not at full capacity33 and preterm infants are not yet prepared to vocalise or discern formal speech sounds.34 35 However, studies in neurobiology indicate that social development does occur during this time36–39 and could possibly represent an early foundational stage in the developmental continuum of communication and language. A second novel finding is that extremely preterm infants, who had higher BWs, had received antenatal steroids and who did not have IVH, were at decreased odds of receiving a ‘high’ level of SSC (where high level was defined as above the median for total hours, frequency and hours for each parent). One possible explanation is that infants who were perceived as being ‘less sick’ were at reduced odds of receiving a high level of SSC. This poses questions about how medical caregivers and parents perceive the practice of SSC and how developmental information is being communicated between parents and the medical team. Finally, our findings identify a concerning gap in SSC from 30 weeks corrected age to term age. Numerous studies highlight the essential nature of this time period for appropriate neurodevelopment.40–44 However, extremely preterm infants, who represent one of the highest risk categories for neurodevelopmental disability, are not receiving an SSC intervention shown to improve neurodevelopment at a fundamental time point of their developmental trajectory. This elicits additional questions about why parents choose to stop at this corrected age, the underlying mechanisms of communication development in this population and the potential added dimensional role of SSC as a communication intervention in the NICU. One important limitation of our study is that it is retrospective in nature within an all-referral hospital system in the USA, which impedes our ability to capture and analyse additional factors shown to impact neonatal neurodevelopment. Consequently, we could not explore the effects of maternal health, socioeconomic status or education, which are shown to significantly influence neonatal developmental outcome. However, the intent of our study was to investigate if patterns of parental skin-to-skin behaviour should be considered in the overall discussion of mesosystemic variables contributing to longer term outcomes. The study is also limited by sample size and possible confounding, which is inherently an issue when investigating associations. However as noted previously, NICU patients receiving SSC were actually associated with medical factors that were in turn associated with lower early-stage cognitive and communication scores. The study is also based on patients from a single hospital, and may not generalise to other neonatal units, especially because our unit is a level-4 all-referral unit with high-acuity infants and geographically distant parents. While we did see a relatively consistent pattern of association (c.f. table 5) between high/low SSC participation and the Bayley-III Cognitive and Communication Outcomes dichotomised at borderline disability (<80 vs ≥80), we reiterate that none of these associations achieved statistical significance and thus can only be viewed as suggestive results in need of confirmatory analysis. If the observed associations were to hold in the population along with the same level of SSC participation and prevalence of borderline disability, then roughly 460 total subjects would be needed to achieve 80% power to detect the association in a larger study. Our research suggests that developmental investigation into very early time points in the life of extremely preterm infants that incorporates medical and behavioural components is warranted and critical to understanding how to fully optimise future developmental social and cognitive processes. Further prospective studies, involving more comprehensive measures and analyses of the early developmental NICU environment (22–40 weeks postmenstrual age), could help inform new designs for developmental caregiving and promotion of SSC throughout the duration of hospitalisation. We would like to thank Dr Leif Nelin and the Small Baby Program for their continual advocacy and advancement towards the highest standard of developmentally appropriate care of extremely preterm infants. Contributors: All authors contributed equally to this manuscript. JG and GB designed and conducted the study, analysed the biostatistical portion, and wrote and reviewed manuscript drafts. JG, WCR and RWR performed visual analytics, wrote sections of the manuscript, and reviewed and refined manuscript drafts. Funding: This project was supported by award number UL1TR001070 from the National Center for Advancing Translational Sciences. Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Advancing Translational Sciences or the National Institutes of Health. Competing interests: None declared. Ethics approval: This study was approved by the Institutional Review Board of Nationwide Children's Hospital (IRB#13-00042) as an expedited study that meets the criteria for waiver of authorisation. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Data are stored on our internal, high-security server. De-identified data set is available on request. ==== Refs References 1 Colonna F , Uxa F , da Graca AM The “kangaroo-mother” method: evaluation of an alternative model for care of low birth weight newborns in developing countries . Int J Gynaecol Obstet 1990 ;31 :335 –9 . 10.1016/0020-7292(90)90911-4 1969818 2 de Leeuw R , Colin EM , Dunnebier EA Physiological effects of kangaroo care in very small preterm infants . Biol Neonate 1991 ;59 :149 –55 . 10.1159/000243337 2054424 3 Ludington-Hoe SM , Thompson C , Swinth J Kangaroo care: research results, and practice implications and guidelines . Neonatal Netw 1994 ;13 :19 –27 . 4 Bauer K , Uhrig C , Sperling P Body temperatures and oxygen consumption during skin-to-skin (kangaroo) care in stable preterm infants weighing less than 1500 grams . J Pediatric 1997 ;130 :240 –4 . 10.1016/S0022-3476(97)70349-4 5 Feldman R , Eidelman AI Intervention programs for premature infants. How and do they affect development? Clin Perinatol 1998 ;25 : 613 –26 .9779337 6 Tornhage CJ , Stuge E , Lindberg T First week kangaroo care in sick very preterm infants . Acta Paediatr 1999 ;88 :1402 –4 . 10.1111/j.1651-2227.1999.tb01059.x 10626530 7 Ohgi S , Fukuda M , Moriuchi H Comparison of kangaroo care and standard care: behavioral organization, development, and temperament in healthy, low-birth-weight infants through 1 year . J Perinatol 2002 ;22 :374 –9 . 10.1038/sj.jp.7210749 12082472 8 Ludington-Hoe SM , Ferreira C , Swinth J Safe criteria and procedure for kangaroo care with intubated preterm infants . J Obstet Gynecol Neonatal Nurs 2003 ;32 :579 –88 . 10.1177/0884217503257618 9 Begum EA , Bonno M , Ohtani N Cerebral oxygenation responses during kangaroo care in low birth weight infants . BMC Pediatr 2008 ;8 :51 10.1186/1471-2431-8-51 18990243 10 Neu M , Robinson J Maternal holding of preterm infants during the early weeks after birth and dyad interaction at six months . J Obstet Gynecol Neonatal Nurs 2010 ;39 :401 –14 . 10.1111/j.1552-6909.2010.01152.x 11 Conde-Agudelo A , Belizan JM , Diaz-Rossello J Kangaroo mother care to reduce morbidity and mortality in low birthweight infants . Cochrane Database Syst Rev 2011 ;3 :CD002771 . 12 Ludington-Hoe SM Thirty years of kangaroo care science and practice . Neonatal Netw 2011 ;30 :357 –62 . 10.1891/0730-0832.30.5.357 21846634 13 Kaffashi F , Scher MS , Ludington-Hoe SM An analysis of the kangaroo care intervention using neonatal EEG complexity: a preliminary study . Clin Neurophysiol 2013 ;124 :238 –46 . 10.1016/j.clinph.2012.06.021 22959415 14 Conde-Agudelo A , Diaz-Rossello J Kangaroo mother care to reduce morbidity and mortality in low birthweight infants . Cochrane Database Syst Rev 2014 ;4 :CD002771 . 15 Mörelius E , Örtenstrand A , Theodorsson E A randomized trial of continuous skin-to-skin contact after preterm birth and the effects on salivary cortisol, parental stress, depression, and breastfeeding . Early Hum Dev 2015 ;91 :63 –70 . 10.1016/j.earlhumdev.2014.12.005 25545453 16 Boundy EO , Dastjerdi R , Spegelman D Kangaroo mother care and neonatal outcomes: a meta-analysis . Pediatrics 2016 ;137 http://dx.doi.org/10.1542/peds.2015-2238 17 Flacking R , Lehtnonen L , Thomson G Closeness and separation in neonatal intensive care . Acta Paediatr 2012 ;101 :1032 –7 . 10.1111/j.1651-2227.2012.02787.x 22812674 18 Feldman R , Rosenthal Z , Eidelman AI Maternal-preterm skin-to-skin contact enhances child physiologic organization and cognitive control across the first 10 years of life . Biol Psychiatry 2014 ;75 :56 –64 . 10.1016/j.biopsych.2013.08.012 24094511 19 Roberts KL , Paynter C , McEwan B A comparison of kangaroo mother care and conventional cuddling care . Neonatal Netw 2000 ;19 :31 –5 . 10.1891/0730-0832.19.4.31 20 Rojas MA , Kaplan M , Quevedo M Somatic growth of preterm infants during skin-to-skin care versus traditional holding: a randomized, controlled trial . J Dev Behav Pediatr 2003 ;24 :163 –8 . 10.1097/00004703-200306000-00006 12806228 21 Schneider C , Charpak N , Ruiz-Peláez JG Cerebral motor function in very premature-at-birth adolescents: a brain stimulation exploration of kangaroo mother care effects . Acta Paediatr 2012 ;101 :1045 –53 . 10.1111/j.1651-2227.2012.02770.x 22734793 22 Scher MS , Ludington-Hoe S , Kaffashi F Neurophysiologic assessment of brain maturation after an 8-week trial of skin-to-skin contact on preterm infants . Clin Neurophysiol 2009 ;120 :1812 –18 . 10.1016/j.clinph.2009.08.004 19766056 23 Seidman G , Unnikrishnan S , Kenny E Barriers and enablers of kangaroo mother care practice: a systematic review . PLoS ONE 2015 ;10 :e0125643 10.1371/journal.pone.0125643 25993306 24 Nankervis CA , Martin EM , Crane ML Implementation of a multidisciplinary guideline-driven approach to the care of the extremely premature infant improved hospital outcomes . Acta Paediatr 2010 ;99 :188 –93 . 10.1111/j.1651-2227.2009.01563.x 19863632 25 Shepherd E , Calvert T , Martin E Outcomes of extremely premature infants admitted to a children's hospital depends on referring hospital . J Neonat Perinat Med 2011 ;4 :45 –53 . 26 Ambalavanan N , Carlo WA , Tyson JE Outcome trajectories in extremely preterm infants . Pediatrics 2012 ;130 : e115 –25 . 10.1542/peds.2011-3693 22689874 27 McMahon E , Wintermark P , Lahav A Auditory brain development in premature infants: the importance of early Experience . Ann N Y Acad Sci 2012 ;1252 :17 –24 . 10.1111/j.1749-6632.2012.06445.x 22524335 28 National Institute of Child Health and Human Development Neonatal Research Network . Follow-up program. 2013; Retrieved from https://neonatal.rti.org/about/fu_background.cfm 29 Pearson Education . Bayley scales of infant development .3rd edn Pearson Clinical 2008 Training . 30 Ray WC MAVL and StickWRLD: visually exploring relationships in nucleic acid sequence alignments . Nucleic Acids Res 2004 ;32 :59 –63 . 31 Rumpf RW , Gonya J , Ray W Visual hypothesis and correlation discovery for precision medicine. AMIA workshop on visual analytics in healthcare conference paper 2014 . 32 Ray WC , Wolock SL , Li N StickWRLD: Interactive visualization of massive parallel contingency data for personalized analysis to facilitate precision medicine. Proceedings of the 3rd annual Workshop on Visual Analytics in Healthcare, in conjunction with the American Medical Informatics Symposium . 2013 ; November . 33 Key APF , Lambert EW , Aschner JL Influence of gestational age and postnatal age on speech sound processing in NICU infants . Psychophys 2012 ;49 :720 –31 . 10.1111/j.1469-8986.2011.01353.x 34 Caskey M , Stephens B , Tucker R Importance of parent talk on the development of preterm infant vocalizations . Pediatrics 2011 ;128 :910 –16 . 10.1542/peds.2011-0609 22007020 35 Caskey M , Stephens B , Tucker R Adult talk in the NICU with preterm infants and developmental outcomes . Pediatrics 2014 ;133 :578 –84 . 10.1542/peds.2013-0104 36 Silva MG , Barros MC , Pessoa UM Kangaroo-mother care method and neurobehavior of preterm infants . Early Hum Dev 2016 ;95 :55 –9 . 10.1016/j.earlhumdev.2016.02.004 26962849 37 Feldman R , Eidelman AI Skin-to-skin contact (kangaroo care) accelerates autonomic and neurobehavioural maturation in preterm infants . Dev Med Child Neurol 2003 ;45 :274 –81 . 10.1111/j.1469-8749.2003.tb00343.x 12647930 38 Feldman R , Eidelman AI , Sirota Comparison of skin-to-skin (kangaroo) and traditional care: parenting outcomes and preterm infant development . Pediatrics 2002 ;110 :16 –26 . 10.1542/peds.110.1.16 12093942 39 Feldman R From biological rhythms to social rhythms: physiological precursors of mother-infant synchrony . Dev Psychol 2006 ;42 :175 –88 .16420127 40 Reissland N , Kisilevsky BS , eds. Fetal development: research on brain and behavior, environmental influences, and emerging technologies . Switzerland : Springer International , 2016 . 41 Thomason ME , Grove L , Lozon TA Age related increases in long-range connectivity in fetal functional neural connectivity networks in utero . Developmental Cog Neurosci 2015 ;11 :96 –104 . 10.1016/j.dcn.2014.09.001 42 Krueger C , Garvan C Emergence and retention of learning in early fetal development . Infant Behavior and Dev 2014 ;37 :162 –73 . 10.1016/j.infbeh.2013.12.007 43 Pearson J , Tarabulsy GM , Bussieres E Foetal programming and cortisol secretion in early childhood: a metal-analysis of different programming variables . Infant Behavior and Dev 2015 ;40 :204 –15 . 10.1016/j.infbeh.2015.04.004 44 Harding R , Bocking AD , eds. Fetal growth and development . UK : Cambridge University Press, 2001 .

PMC005xxxxxx/PMC5372110.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01436010.1136/bmjopen-2016-014360Cardiovascular MedicineResearch150616831730Design and reliability of a specific instrument to evaluate patient safety for patients with acute myocardial infarction treated in a predefined care track: a retrospective patient record review study in a single tertiary hospital in the Netherlands Eindhoven Daniëlle C 1Borleffs C Jan Willem 1Dietz Marlieke F 1Schalij Martin J 1Brouwers Corline 2de Bruijne Martine C 2 1 Department of Cardiology, Leiden University Medical Center, Leiden, The Netherlands 2 Department of Public and Occupational Health, EMGO Institute for Health and Care Research, VU University Medical Center, Amsterdam, The NetherlandsCorrespondence to Dr Martin J Schalij; m.j.schalij@lumc.nl2017 20 3 2017 7 3 e01436021 9 2016 24 1 2017 14 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective Numerous studies have shown that a substantial number of patients suffer from adverse events (AEs) as a result of hospital care. However, specific data on AEs in acute cardiac care are scarce. The current manuscript describes the development and validation of a specific instrument to evaluate patient safety of a predefined care track for patients with acute myocardial infarction (AMI). Design Retrospective patient record review study. Setting and participants A total of 879 hospital admissions treated in a tertiary care centre for an AMI (age 64±12 years, 71% male). Main outcome measure In the first phase, the medical records of patients with AMI warranting coronary angiography or coronary intervention were analysed for process deviations. In the second phase, the medical records of these patients were checked for any harm that had occurred which was caused by the healthcare provider or the healthcare organisation (AE) and whether the harm that occurred was preventable. Results Of all 879 patients included in the analysis, 40% (n=354) had 1 or more process deviation. Of these 354 patients, 116 (33%) had an AE. Patients with AE experienced more process deviations compared with patients without AE (2±1.7 vs 1.5±0.9 process deviations per patient, p=0.005). Inter-rater reliability in assessing a causal relation of healthcare with the origin of an AE showed a κ of 0.67 (95% CI 0.51 to 0.83). Conclusions This study shows that it is possible to develop a reliable method, which can objectively assess process deviations and the occurrence of AEs in a specified population. This method could be a starting point for developing an electronic tracking system for continuous monitoring in strictly predefined care tracks. Patient SafetyMedical ErrorPatient HarmProcess Assessment ==== Body Strengths and limitations of this study This study describes the development of a new valid instrument to objectively assess and monitor the occurrence of process deviations and adverse events in patients with acute myocardial infarction treated in a predefined care track. This study focuses on identifying care track-specific process deviations, assessing the type and preventability of adverse events and the correlation between process deviations and adverse events. This specific instrument can be used as a template for developing a quality instrument to objectify patient safety in a predefined care track. This specific instrument has a substantial inter-rater reliability and is tested using a large number of patient records. Limitations include the limited availability of information sources (ie, medical and nursing records only) and the absence of information regarding adverse events after discharge. Introduction Patient safety, defined by the National Academy of Medicine, formerly called the American Institute of Medicine, as the prevention of harm to patients, is the minimum pre-requisite for a good quality of care.1 In 1999, they published a report called ‘To Err is Human’, which drew attention to the fact that a significant number of patients suffered from injuries or even had died as a result of care delivered in hospitals.2 Subsequently, various studies in different countries reported that 2.9–16.6% of in-hospital patients experienced one or more adverse events (AEs) and that in 5–13% of the AEs the patients died.3–13 In these studies, an AE was defined as an unintended injury that results in disability at the time of discharge, death or prolonged hospital stay and is caused by healthcare management rather than by the patient's underlying disease process. In various studies, a large variation in the incidence of AEs among the different hospital departments was shown (0.5–29.9%).4 14–16 However, these studies had a general hospital-wide approach and provided hardly any insight into causal relations on a departmental level. To develop specific interventions at each department in order to improve patient safety appeared even more challenging, particularly cardiology, which is a department with a high intervention rate and a large number of patients with life-threatening illnesses. A subset of studies contain results on the occurrence of AE among general cardiac patients, based on small numbers of patients and showing a substantial variety in the incidence of AE (13.3–29.9%).15 16 Therefore, sufficiently powered studies are needed for specific patient groups to gain more insight into the incidence of (preventable) AEs and to define ‘how safe is our care’. Inspired by high-risk industries and best-practice hospitals, the aim of this study is to provide a system to review our work and to define if our work process is sufficiently safe. High-risk industries, such as the aviation and chemical industry, are required to perform structured assessments of all processes that contribute to a particular activity, which allows them to make a reasoned claim regarding safety. In the healthcare sector, best-practice hospitals such as the Intermountain Healthcare Group provide a framework such as the Quality and Patient Safety Plan on which an integrated and comprehensive programme to monitor, assess and improve the quality and safety of patient care is delivered.17 This study attempts to analyse process deviations and potential correlation with AEs in hospitalised patients who are treated according to a protocolised care pathway, in this case patients who suffer from an acute myocardial infarction (AMI). The method developed to assess AEs in patients with AMI is based on the Harvard Medical Practice Study (HMPS), which is a structured patient record review that has also been used in other AE studies.4 16 18 19 This manuscript describes how this commonly used method is adapted for our specific patient population. In addition, it was examined whether the linkage between process deviations and AEs will increase the uniformity of the assessments of AEs and creates the possibility to develop better improvement strategies. Methods Patient population Patients who were admitted in 2012 and 2013 to the Leiden University Medical Center (LUMC) with an AMI warranting coronary angiography (CAG) or percutaneous coronary intervention (PCI) and treated according to the MISSION!-protocol were included.20 The LUMC functions as a tertiary referral centre performing PCI procedures on a 24/7 basis and serves an area of ∼750 000 inhabitants. The MISSION!-protocol contains a prehospital, in-hospital and outpatient framework for clinical decision-making and treatment for the different diagnosis in AMIs (unstable angina (UA), ST segment elevation myocardial infarction (STEMI) and non-STEMI (NSTEMI)). This study focused on the in-hospital programme (early reperfusion, same diagnostic trajectory like two-dimensional-echocardiography, structured medical therapy and disease education). Generally, patients are planned for discharge 12 hours after a CAG, or 48 hours after a PCI. The MISSION!-protocol is based on the evolving guidelines of the European Society of Cardiology.20–22 The MISSION!-patient records were extracted from the electronic patient file system (EPD-Vision, LUMC, Leiden, The Netherlands) by selecting the diagnose coding of a diagnosis–treatment combination for UA (11.203), STEMI (11.204) and NSTEMI (11.205). Patients with any of these three diagnosis codes were linked with the clinical database to select the patients who received an intervention procedure (CAG or PCI) within 24 hours after admission. This was applied to all patients admitted to the LUMC between 1 January 2012 and 31 December 2013. Patients with an urgent coronary artery bypass grafting after CAG were excluded because they underwent a different treatment path. Review process The method used in this study was based on a protocol originally developed by the HMPS. A modified version of this protocol was used in studies in Australia, Canada, Denmark, France, New Zealand, the UK and the USA.3 5 10 11 13 14 A Dutch protocol, based on the Canadian Adverse Events study, was used in studies in 2004, 2008 and 2011/2012.15 16 19 To identify high-risk patient records, the HMPS developed 18 triggers (ie, unplanned return to the operating room or hospital-incurred patient injury). The presence of one or more of these 18 triggers was established in phase 1. In case a trigger was found, the patient record entered a second phase, which focused on identifying whether harm was done to the patient, whether the harm was due to the care that the patient received, and whether the harm was preventable. To increase the uniformity of the assessment of AE and to gain more insight into patterns of AEs, the triggers of the HMPS were specified for acute cardiac care, thereby creating the opportunity to identify specific process deviations. In the first phase of the review process, we focused on identifying these process deviations in the patient records, and in the second phase AEs were identified. In case an AE was identified, it was also scored on preventability.3 Phase 1: specification of process deviations for AEs In the MISSION!-protocol, the patient can have different workflows according to their electrocardiographic diagnosis at admission (STEMI, NSTEMI or UA). In the search for AEs, all process deviations from the MISSION!-protocol were identified. During this review phase, a process deviation was defined as every operation or treatment that differed from the MISSION!-protocol, such as additional procedures (a pacemaker implantation or second PCI), prescription of extra medication other than described in the protocol (use of antiaritmica, anticoagulation, inotropics or diuretics) or omission of a procedure (no diagnostics performed). If a patient had a transient heart rhythm disorder without treatment consequences, the process deviation was noted as an observation of the heart rhythm. Figure 1 shows all the defined process deviations. The review process of all medical records (nursing and medical records) was performed by a physician with work experience at the clinical department and who is familiar with reviewing electronic files. In contrast to the original triggers of the HMPS and other process deviation frameworks like the Process Deviations Analysis Framework, no judgements were made during phase 1 on whether the process deviation was ‘unexpected’ or ‘unplanned’.23 After identifying all process deviations, the process deviations were categorised into main categories and translated back to one (or more) of the original 18 triggers of the HMPS. Owing to the defined inclusion criteria stated in the MISSION!-protocol and restrictions concerning the availability of data from the patient records in the peripheral hospitals up to 30 days after discharge, it was decided beforehand that events experienced prior to (ie, unplanned admission before index admission), or after the index admission (ie, readmission) were excluded from the review process. Figure 1 Newly defined process deviations in relation to the original Harvard Medical Practice Study. Phase 2: determination of an AE and preventability During phase 2, the nursing and medical records of the admission were reviewed for AEs by a clinical physician. If applicable, the records of patients who were transferred to another hospital during their admission were also traced. In cases where the physician found more than one AE in a patient, they were separately registered. Event classification First, it was assessed whether the event resulted in harm to the patient. The definitions are mentioned in online supplementary file 1. If the patient experienced harm that resulted in any disadvantage for the patient, such as prolonged admission, temporary or permanent (physical and/or mental) impairment or death, it was rated whether the harm was caused by healthcare (ie, an AE) and, if so, whether it was preventable (ie, caused by an error). Both the causation and the preventability were scored on a six-point Likert scale. Preventability of the AE was assessed by indicating a score between 4 and 6 on the Likert scale. This is in accordance with other AE studies (see online supplementary files 2 and 3). 10.1136/bmjopen-2016-014360.supp1supplementary files To support the reviewer in the rather implicit judgement of determining the causation in healthcare of the AE, the causation score was preceded by structuring questions to direct if the injury was indeed caused by medical care rather than the underlying acute coronary syndrome (see online supplementary files 2 and 3). For example: ‘Does the timing of the (adverse) event suggest that the injury is related to the treatment?’ and ‘Is the lack of treatment or delayed treatment a recognised cause of this injury?’ Analogously, preceding questions were used to judge if an AE was preventable. An AE was found to be preventable when the performance of the practitioner fell short of the expected level of competence based on the professional standard. Appropriate management of the myocardial infarction was outlined in the previously mentioned MISSION!-protocol. Also, local hospital guidelines on precautionary measures to prevent common events, such as measures to prevent a delirium, were taken into account by the reviewer and the expert panel. The questions preceding the preventability score were also used to evaluate the complexity of the medical history and comorbidity of the patient. In addition, it was of importance to consider the potential benefit of the procedure, the calculated risk and the degree of emergency in treating a patient with a myocardial infarction. Therefore, the original preceding questions were augmented by two extra questions on whether the management of the AMI was appropriate, and on the estimated risk of an AE associated with the management (see online supplementary file 3). Expert panel In case of doubt regarding the causality and/or the preventability of an event, an advisory opinion of an expert panel was requested. The expert panel consisted of two consultants, cardiologists with a wide range of experience in interventional cardiology or electrophysiology. Both cardiologists were either involved in managerial tasks or departmental incident analysis. In addition, two cardiologists in their final year of training and involved in the daily clinical practice formed part of the expert panel. In case there was a discrepancy about the causation or preventability between the reviewer and the expert panel, the structuring questions were used to guide the discussion and reach a final decision (see online supplementary files 2 and 3). The expert panel was also involved in reaching consensus on the causality and preventability of events, which occurred frequently such as a groin haematoma. Privacy Guarding privacy and anonymity was considered to be a high priority. To warrant the privacy and anonymity, all people involved in the study signed a confidentiality agreement to maintain the confidentiality of the information. Study results were stored in a Microsoft Access 2010 Database on a safety disc which can only be accessed by individuals who are involved in the study. Ethical approval The LUMC gave a declaration of ‘medical–ethical permittance not necessary’ for this retrospective records study (reference number P15.133). The peripheral hospitals had formally consented to obtain data from outpatient clinical records, in accordance with their local medical–ethical committee. Analysis and statistics Patient characteristics and process deviations For all patients, baseline characteristics such as age, sex, medical history and admission characteristics like length of stay, comorbidities, cardiac diagnosis and procedural characteristics (stenting or not) were retrieved. Continuous variables are presented as the mean with SD or median with 25th and 75th centile, where appropriate. Dichotomous variables are presented as numbers and percentages. The number of process deviations within each main category were calculated by summing the number of process deviations of each subcategory, after which they were plotted in a pie chart. Baseline comparability between patients with or without process deviations were evaluated by descriptive statistics and independent t-test or χ2 test, when appropriate. A p value smaller than 0.05 was considered to be statistically significant. Furthermore, the process deviation:AE ratio was determined to assess the effectivity of the new method based on identifying process deviations, and a χ2 analysis was performed to assess differences between patients with and without process deviations. Inter-rater reliability To assess the reliability of the assessment of the presence and preventability of an AE of the first reviewer, 10% (n=87) of the patient records were independently screened by an experienced cardiologist from another centre. This second reviewer was blinded to the outcome of the first review. To maximise efficiency, stratified sampling was performed. AE positive patient records were oversampled in relation to AE negative patient records, with a ratio of 2:1 (AE:non-AE). These patients were randomly selected from the whole MI study population. Consecutively, all patient records that contained an AE according to both reviewers were also reviewed on preventability. The percentage of agreement on causality and preventability is determined on a patient level and expressed separately for positive and negative rating. In addition, Cohen's κ statistics was calculated.24 To avoid any potential bias in the κ's coefficient, caused by the stratified sampling, the κ statistics were also calculated separately for the patient records with and without an AE. Data accuracy Data accuracy (missing data, inconsistent data) was checked on a regular basis and analysed using Microsoft Access 2010 and IBM SPSS Statistics for Windows, V.23.0, Armonk, New York, USA: IBM Corp. Results Process deviations and AEs In total, 879 patients (age 64±12 years, 71% male) were reviewed, including the follow-up records of 274 patients who were transferred to three affiliated hospitals after a cardiac procedure in the tertiary care centre (table 1). In 347 patients (39%), one or more process deviations during admission (587 deviations in total) were found. The process deviations (n=587) were categorised into four main categories: observation, diagnostic, therapy or transfer (figure 2). Most process deviations were found during observations (especially observation of mental and physical symptoms or rhythm disorders) and therapy (especially anticoagulation therapy or return to catheterisation room for a second CAG or PCI procedure; figure 1). Patients with one or more process deviations were significantly older than patients without a process deviation (67±12 vs 61±11, p<0.001). In addition, female patients (66% vs 75% male patients, p=0.006) and patients with a lower renal function had a significantly higher risk for process deviations (table 1). Table 1 Baseline characteristics All patients N=879 With process deviation N=347 No process deviation N=532 p Value Demographic characteristics  Age (years) 64±12 67±12 61±11 ≤0.001  Male sex 626 (71%) 229 (66%) 397 (75%) 0.006  BMI (kg/m2) 27±4 26±4 27±4 0.070  Length of stay (days) (median, IQR) 3 (2–4) 3 (2–5) 2 (2–3) ≤0.001 Comorbidities  Hypertension 352 (40%) 155 (45%) 200 (38%) 0.066  Hyperlipidaemia 198 (23%) 75 (22%) 123 (23%) 0.625  Diabetes mellitus 115 (13%) 52 (15%) 63 (12%) 0.177  Known coronary disease 145 (16%) 65 (19%) 80 (15%) 0.147  Known pulmonary disease 88 (10%) 43 (12%) 45 (8%) 0.055  Renal clearance (mL/min/1.73 m2) 75±23 72±25 77±22 0.002 Infarct characteristics  Diagnosis 0.052   STEMI 594 (68%) 234 (67%) 360 (68%)   NSTEMI 135 (15%) 63 (18%) 72 (14%)   UAP 114 (13%) 34 (10%) 80 (15%)   Other 36 (4%) 16 (5%) 20 (4%)  Treated with PCI 747 (85%) 297 (86%) 450 (85%) 0.684 BMI, body mass index; NSTEMI, non-ST-elevated myocardial infarction; PCI, percutaneous coronary intervention; STEMI, ST-elevated myocardial infarction; UAP, unstable angina pectoris. Figure 2 Five hundred and eighty-seven process deviations of all 879 patients. Translation of triggers to process deviations After categorising the process deviations, the categories were translated back to one (or more) of the 18 original triggers of the HMPS (figure 1). Owing to the inclusion criteria of the study population, original triggers like ‘unplanned admission before index admission’, ‘unplanned transfer to another acute care hospital’, ‘injury related to abortion or delivery or neonatal complications’ were not applicable to this population. Moreover, ‘inappropriate discharge to home’ and ‘unplanned readmission after discharge’ were not included because the medical records were only reviewed while the patient was admitted to the hospital. Triggers like ‘dissatisfaction with care documented in the medical record’ and ‘documentation or correspondence indicating litigation’ were not used if they did not result in an alteration of the workflow. Eleven original triggers remained. Triggers used for diagnostic or therapy procedures especially became more specified by using process deviations (unplanned return to the operating room; unplanned removal, injury or repair of organ during surgery’ and other patient complication (no natural consequence of disease)). Likewise, the trigger ‘unexpected death’ was more specified in process deviations as ‘resuscitation’ or ‘positioning of an intra-aortic balloon pump’. In 116 patients, 33% of all patients with a process deviation, an AE was found. The majority of patients with an AE had more than 1 process deviation (64 of 116 patients (55%), average of 2.0±1.7 process deviations per patient). In the group of patients with no AE, 150 of 231 patients (65%) had more than 1 process deviation, average of 1.5±0.9 process deviations per patient (p=0.005). No significant differences were found in patients with and without an AE in the distribution of the type of process deviations (table 2). Likewise, no differences in mortality rate (5 of 116 patients with AE died (4.3%) vs 13 of 763 patients without AE (1.7%), p=0.065). All patients who died (n=18, 2.0%) during their hospital stay experienced a process deviation (n=23 in 18 patients) and 5 patients an AE. The process deviations that were found in deceased patients were mainly related with therapy (21) or transfer (2) and no differences were seen between deceased patients with and without an AE in the type of process deviation (4 of 5 therapy deviation in patients with AE vs 17 of 18 therapy deviation in patients without AE, p=0.395). In patients without a process deviations, 2 patients (2 of 532, 0.4%) experienced a non-preventable AE. Table 2 Comparison of process deviations Process deviations in all patients Process deviations in all patients without an adverse event Process deviations in all patients with an adverse event p Value Process deviations N=587 N=354 N=233 Number of patients 879 231 116 Observation 199 (34%) 124 (35%) 75 (32%) 0.477 Diagnostic 79 (13%) 46 (13%) 33 (14%) 0.685 Therapy 280 (48%) 163 (46%) 117 (50%) 0.322 Transfer 29 (5%) 21 (6%) 8 (3%) 0.172 None 532 n/a n/a n/a Inter-rater reliability The inter-rater reliability assessment was carried out for 87 patients (10% of n=879) by a second independent cardiologist with experience in the assessment of the presence of preventable AEs by means of medical record reviewing. On a patient level, there was an agreement in 73 patient records (agreement level of 84%). The positive agreement on the presence of an AE in the patient records was 87%, and the negative agreement on the absence of an AE in the patient records was 80%. No differences were found when performing an independent analysis of the agreement in patient records with AE and without AE (independent analysis AE positive agreement 86%, independent analysis non-AE: negative agreement 98%). In 45 patients, preventability was assessed on which there was an agreement of 84%. The κ statistics of the preventability was substantial (κ 0.67 (95% CI 0.45 to 0.90); Table 3). Of note, a common type of AE was a groin complication (groin haematoma), on which consensus was established by the expert panel. In principle, a groin haematoma was judged to be preventable. The expert panel decided that despite adequate action, a groin haematoma is less preventable in high-risk frailty groups like the obese, restless, elderly and patients with known peripheral vascular disease when using dual antiplatelet therapy. Table 3 Assessment of the inter-rater reliability between two physicians Causality 87 patients Agreement 84% Positive agreement 87% Negative agreement 80% κ statistic (95% CI) 0.67 (0.51 to 0.83) Preventability 45 patients Agreement 84% Positive agreement 87% Negative agreement 80% κ statistic (95% CI) 0.67 (0.45 to 0.90) To assess the reliability of the preventability, the sample size population contained an over-representation of AEs compared with the whole MI study population (AE:non-AE ratio is 2:1). AE, adverse event; MI, myocardial infarction. Discussion This study describes the development of a new valid screening tool for identifying process deviations and AEs in a specified patient population: patients with an AMI, treated according to a strict defined protocol. First of all, process deviations were identified. We reviewed 879 patient records in which 587 process deviations in 347 patients were found. In 116 patients, 33% of all patients with a process deviation, an AE was found, leading to an AE:process deviation ratio of 1:3. This ratio seems higher in comparison to other studies where one AE was found in every sixth or seventh patient record with a trigger.3 4 18 However, it is important to take into account that previous studies were focused on a wider range of specialties, which can explain a lower prevalence of AEs. Interestingly, the type of process deviation is not associated with experiencing an AE; no differences in the type of process deviations were found between patients with an AE compared with patients without an AE. This finding could be explained by the fact that, despite process deviations being present in a patient record, they may not be the root cause of the AE. Various preventive actions may have taken place after the occurrence of the process deviation which could have averted patient harm. In addition, AEs are more likely to be caused by a combination of factors.25 This makes it difficult to develop a targeted approach to improve patient safety. More research, for instance based on incident analysis methods using more sources of information than the patient record, may help to identify other variables (ie, patient characteristics) that predict an AE. The majority of process deviations were found during clinical observation and therapy, which is in line with prior studies using the related original HMPS triggers.17 18 On the basis of the inter-rater reliability that was found in this study (causality and preventability agreement are both 84%), this adapted method appears to be a reliable and suitable instrument to use in this well-defined patient population. In previous studies, the reliability of the occurrence of AEs in general hospitals was moderate to substantial (κ ranged from 0.42 to 0.83, agreement ranged from 76% to 92%) and, if measured, it was moderate for determining the preventability of AEs (κ ranged from 0.33 to 0.69, agreement ranged from 68% to 91%; see online supplementary file 4).3 4 8 11 13 18 26–29 Recently, a study focused on hip fractures in elderly patients showed an agreement of 85% in the presence of an AE, with a κ value of 0.52.30 The substantial reliability in this study suggests that an assessment procedure shows reduced inter-rater variation when being performed in a specified population with a strictly predefined protocol. Nevertheless, a 100% agreement score was not reached. This is probably due to the different perspectives of physicians and limitations in our medical knowledge in causal relations; therefore, the assessment of preventability remains under debate until new scientific evidence is discovered. Previous studies on AEs were performed in general hospital populations and encountered a large variation among specialties with regard to the risk of the procedures employed and the severity of illness of the patients. As a result, heterogeneous numbers and causes of (preventable) AEs were reported among the different specialties.4 8 18 26 Focusing on one specific illness leads to a decrease in the workload for the hospital staff, makes it easier to plan the collection of data, and limits the interruption of the work flow. In the end, this specific instrument is likely to provide more valuable insights into the specific cause of an AE in patients with myocardial infarction and, consequently, possibilities to improve patient safety. Similar, AEs can show recurrent wrong patterns in the management within a clinic department. Some limitations of the proposed method have to be considered. First, the validation method of the new process deviations tool can be considered incomplete because there was no direct comparison with an alternative screening tool, such as the original HMPS trigger tool. Therefore, we cannot exclude the possibility that other numbers of process deviations or triggers were found if a comparison using another tool was performed. In addition, tools for identifying process deviations, such as the Process Deviations Analysis Framework or the (Lean) Six Sigma Model, were not used for the design of the current method because they are not (yet) suitable for assessing whether process deviations are associated with undesirable (healthcare) outcomes.23 31 Besides, there are still considerable challenges when it comes to implementing process deviations frameworks in a healthcare setting. Current process deviation frameworks are highly measurement and data driven while a healthcare setting is mostly dependent on human behaviour, which is difficult to quantify. Another limitation is that this study depended exclusively on documentation in medical and nursing records. However, the likelihood that this has affected the quality of our study is low since previous studies showed that a record review method is sensitive for identifying AEs.32 Also, direct comparison of these results with different hospitals is difficult, as record reviewing highly depends on the level of record completeness and the use of a (electronic) patient record. To lower the possibility of hindsight bias, a prospective design with a weekly review process of a researcher could be considered.8 33 Future implications In the future, it could be of interest to explore whether application of this new method makes it possible to assign AEs to uniform groups after which action can be taken. Furthermore, the eventual goal is to monitor patient safety with real-time process deviations and AEs measurements or, even better, to predict which patients are at risk for AEs to prevent harm. For this next step, existing approaches for measuring process deviations used in other fields can be helpful. An integration of HMPS and Lean Six Sigma, for example, may be beneficial for healthcare. HMPS's strength is to structure implicit relations and to define harm; on the other hand, Lean Six Sigma is a data-driven approach, which may facilitate detection of process deviations as indicators of AEs. It should be explored whether it is possible to develop an electronic tracking system, as part of an electronic patient records system, which enables the continuous monitoring of care. This will shift the emphasis away from focusing solely on medical errors, and more on real-time performance and measures that relate to future risks and resilience of organisations. This could be the starting point for the development of a hospital benchmark quality instrument to objectify patient safety as part of quality of care in a specific patient population.34 Bottom–up, this could serve as an incentive to improve safety, and top–down it will give insight to redesign patient work flows which can improve efficacy and quality of care. Structural process deviations seem more useful for educational purposes compared with individual preventable incidents. Although this method is focused on one illness, this approach may also be applied to other patient populations, or to evaluate other care tracks like the MISSION!-protocol, for example, the total hip arthroplasty procedures.17 Conclusion In conclusion, this paper describes the development of a reliable method to objectively assess the process deviations and the occurrence of AEs in a specified population. The authors gratefully thank L Relik-van Wely, cardiologist for screening the records for the assessment of the reliability. Furthermore, they also would like to thank the member of the expert panel, consisting of M Bootsma, electrophysiologist; F van der Kley, interventional cardiologist; WTJ Aanhaanen, cardiologist in training at the Leiden University Medical Center; M van Vliet, cardiologist in training at the Leiden University Medical Center. Contributors: MCdB and DCE designed the study. DCE drafted the manuscript, as well as collected and analysed the data. MCdB and CJWB contributed to the conception and design of the study, and provided daily supervision. CJWB helped in drafting the article and revising the content of the article. MJS supervised the conduct of the study and provided comments on subsequent versions of the manuscript. MFD supported the data collection. All authors read and approved the final manuscript. Funding: This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Competing interests: None declared. Ethics approval: Medical Ethics Committee of the Leiden University Medical Center (reference number P15.133). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Institute of Medicine Committee on Data Standards for Patient Safety . In: Aspden P , Corrigan JM , Wolcott J , Erickson SM eds. Patient Safety: Achieving a New Standard for Care . Washington DC : National Academies Press (US) , 2004 . 2 Institute of Medicine . To Err Is Human: Building a Safer Health System , 1999 . 3 Baker GR , Norton PG , Flintoft V The Canadian Adverse Events Study: the incidence of adverse events among hospital patients in Canada . CMAJ 2004 ;170 :1678 –86 . 10.1503/cmaj.1040498 15159366 4 Brennan TA , Leape LL , Laird NM Incidence of adverse events and negligence in hospitalized patients. Results of the Harvard Medical Practice Study I. N Engl J Med 1991 ;324 :370 –6 .1987460 5 Davis P , Lay-Yee R , Briant R Adverse events in New Zealand public hospitals I: occurrence and impact . N Z Med J 2002 ;115 :U271 .12552260 6 Hayward RA , Hofer TP Estimating hospital deaths due to medical errors: preventability is in the eye of the reviewer . JAMA 2001 ;286 :415 –20 . 10.1001/jama.286.4.415 11466119 7 Jarman B , Gault S , Alves B Explaining differences in English hospital death rates using routinely collected data . BMJ 1999 ;318 :1515 –20 . 10.1136/bmj.318.7197.1515 10356004 8 Michel P , Quenon JL , de Sarasqueta AM Comparison of three methods for estimating rates of adverse events and rates of preventable adverse events in acute care hospitals . BMJ 2004 ;328 :199 10.1136/bmj.328.7433.199 14739187 9 Park RE , Brook RH , Kosecoff J Explaining variations in hospital death rates. Randomness, severity of illness, quality of care . JAMA 1990 ;264 :484 –90 .2195173 10 Schioler T , Lipczak H , Pedersen BL [Incidence of adverse events in hospitals. A retrospective study of medical records] . Ugeskr Laeger 2001 ;163 :5370 –8 .11590953 11 Thomas EJ , Studdert DM , Burstin HR Incidence and types of adverse events and negligent care in Utah and Colorado . Med Care 2000 ;38 :261 –71 . 10.1097/00005650-200003000-00003 10718351 12 Vincent C , Neale G , Woloshynowych M Adverse events in British hospitals: preliminary retrospective record review . BMJ 2001 ;322 :517 –19 . 10.1136/bmj.322.7285.517 11230064 13 Wilson RM , Runciman WB , Gibberd RW The quality in Australian Health Care Study . Med J Aust 1995 ;163 :458 –71 .7476634 14 Michel P , Quenon JL , Djihoud A French national survey of inpatient adverse events prospectively assessed with ward staff . Qual Saf Health Care 2007 ;16 :369 –77 . 10.1136/qshc.2005.016964 17913779 15 Langelaan M , Baines R , Broekens M Monitor zorggerelateerde schade 2008: dossieronderzoek in Nederlandse ziekenhuizen . Utrecht : NIVEL , 2010 . 16 Langelaan M , De Bruijne M , Baines R Monitor zorggerelateerde schade 2011/2012: Dossieronderzoek in Nederlandse ziekenhuizen . Amsterdam : NIVEL , 2013 . 17 James BC , Savitz LA How intermountain trimmed health care costs through robust quality improvement efforts . Health Aff (Millwood) 2011 ;30 :1185 –91 . 10.1377/hlthaff.2011.0358 21596758 18 Baines RJ , Langelaan M , de Bruijne MC Changes in adverse event rates in hospitals over time: a longitudinal retrospective patient record review study . BMJ Qual Saf 2013 ;22 :290 –8 . 10.1136/bmjqs-2012-001126 19 de Bruijne M , Zegers M , Hoonhout L Onbedoelde schade in Nederlandse ziekenhuizen: dossieronderzoek van ziekenhuisopnames in 2004 . Amsterdam : Instituut voor Extramuraal Geneeskundig Onderzoek , 2007 . 20 Liem SS , van der Hoeven BL , Oemrawsingh PV MISSION!: optimization of acute and chronic care for patients with acute myocardial infarction . Am Heart J 2007 ;153 :14.e1 –11 . 10.1016/j.ahj.2006.10.002 21 Steg PG , James SK , Atar D ESC guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation . Eur Heart J 2012 ;33 :2569 –619 . 10.1093/eurheartj/ehs215 22922416 22 Roffi M , Patrono C , Collet JP 2015 ESC guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation: Task Force for the Management of Acute Coronary Syndromes in patients presenting without persistent ST-segment elevation of the European Society of Cardiology (ESC) . Eur Heart J 2016 ;37 :267 –315 . 10.1093/eurheartj/ehv320 26320110 23 Depaire B , Swinnen J , Jans M A Process Deviation Analysis Framework . (Proceedings of the 1st Joint Int. Workshop on Security in Business Processes) 2013 :701 –06 . 24 de Vet HC , Mokkink LB , Terwee CB Clinicians are right not to like Cohen's kappa . BMJ 2013 ;346 :f2125 .23585065 25 Reason J Human error: models and management . BMJ 2000 ;320 :768 –70 .10720363 26 Zegers M , de Bruijne MC , Wagner C Design of a retrospective patient record study on the occurrence of adverse events among patients in Dutch hospitals . BMC Health Serv Res 2007 ;7 :27 10.1186/1472-6963-7-27 17319971 27 Soop M , Fryksmark U , Koster M The incidence of adverse events in Swedish hospitals: a retrospective medical record review study . Int J Qual Health Care 2009 ;21 :285 –91 . 10.1093/intqhc/mzp025 19556405 28 Hogan H , Healey F , Neale G Preventable deaths due to problems in care in English acute hospitals: a retrospective case record review study . BMJ Qual Saf 2012 ;21 :737 –45 . 10.1136/bmjqs-2011-001159 29 Sari AB , Sheldon TA , Cracknell A Extent, nature and consequences of adverse events: results of a retrospective casenote review in a large NHS hospital . Qual Saf Health Care 2007 ;16 :434 –9 . 10.1136/qshc.2006.021154 18055887 30 Merten H , Johannesma PC , Lubberding S High risk of adverse events in hospitalised hip fracture patients of 65 years and older: results of a retrospective record review study . BMJ Open 2015 ;5 :e006663 10.1136/bmjopen-2014-006663 31 Bandyopadhyay J , Coppens K Six Sigma approach to healthcare quality and productivity management . Int J Qual Product Manag 2005 ;5 :1 –12 . 32 Thomas EJ , Petersen LA Measuring errors and adverse events in health care . J Gen Intern Med 2003 ;18 :61 –7 .12534766 33 Wong BM , Dyal S , Etchells EE Application of a trigger tool in near real time to inform quality improvement activities: a prospective study in a general medicine ward . BMJ Qual Saf 2015 ;24 :272 –81 . 10.1136/bmjqs-2014-003432 34 Vincent C , Burnett S , Carthey J The measurement and monitoring of safety: The Health Foundation 2013 http://www.health.org.uk/measuresafety

PMC005xxxxxx/PMC5372111.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01503610.1136/bmjopen-2016-015036Public HealthProtocol1506172416921706Protocol for a cluster randomised stepped wedge trial assessing the impact of a community-level hygiene intervention and a water intervention using riverbank filtration technology on diarrhoeal prevalence in India McGuinness Sarah L 1O'Toole Joanne E 1Boving Thomas B 2Forbes Andrew B 1Sinclair Martha 1Gautam Sumit K 3Leder Karin 1 1 Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Victoria, Australia 2 Department of Civil and Environmental Engineering and Geosciences, University of Rhode Island, Kingston, Rhode Island, USA 3 The Energy and Resources Institute (TERI), Bangalore, Karnataka, IndiaCorrespondence to Dr Sarah L McGuinness; sarah.mcguinness@monash.com2017 17 3 2017 7 3 e0150364 11 2016 2 2 2017 23 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction Diarrhoea is a leading cause of death globally, mostly occurring as a result of insufficient or unsafe water supplies, inadequate sanitation and poor hygiene. Our study aims to investigate the impact of a community-level hygiene education program and a water quality intervention using riverbank filtration (RBF) technology on diarrhoeal prevalence. Methods and analysis We have designed a stepped wedge cluster randomised trial to estimate the health impacts of our intervention in 4 rural villages in Karnataka, India. At baseline, surveys will be conducted in all villages, and householders will receive hygiene education. New pipelines, water storage tanks and taps will then be installed at accessible locations in each village and untreated piped river water will be supplied. A subsequent survey will evaluate the impact of hygiene education combined with improved access to greater water volumes for hygiene and drinking purposes (improved water quantity). Villages will then be randomly ordered and RBF-treated water (improved water quality) will be sequentially introduced into the 4 villages in a stepwise manner, with administration of surveys at each time point. The primary outcome is a 7-day period prevalence of self-reported diarrhoea. Secondary outcomes include self-reported respiratory and skin infections, and reported changes in hygiene practices, household water usage and water supply preference. River, tank and tap water from each village, and stored water from a subset of households, will be sampled to assess microbial and chemical quality. Ethics and dissemination Ethics approval was obtained from the Monash University Human Research Ethics Committee in Australia and The Energy and Resources Institute Institutional Ethics Committee in India. The results of the trial will be presented at conferences, published in peer-reviewed journals and disseminated to relevant stakeholders. This study is funded by an Australian National Health and Medical Research Council (NHMRC) project grant. Trial registration number ACTRN12616001286437; pre-results. PUBLIC HEALTHINFECTIOUS DISEASESEPIDEMIOLOGYNational Health and Medical Research Councilhttp://dx.doi.org/10.13039/501100000925Career Development Fellowship APP1084351 (KL)Postgraduate Scholarship APP1115196 (SLM)Project Grant APP1083408 ==== Body Strengths and limitations of this study This is the first randomised control trial (RCT) to assess the health impact of a community-level water intervention using riverbank filtration technology. This community-level study is novel as it considers the impact of interventions that target water access/quantity and improved water quality. The stepped wedge RCT design allows all participants to eventually receive the intervention, improving equity and acceptability; intervention effects are estimated from within-village differences while controlling for time trends. Collection of household stored water samples prior to health survey administration enables assessment of the relationship between microbiological quality of stored drinking water and diarrhoea. Incorporation of only four clusters (villages) in the trial may limit generalisability. Introduction Access to a safe, reliable and continuous supply of water for drinking, cooking and personal hygiene is an essential requirement for good health. The WHO estimates that around 10% of the total burden of disease worldwide could be prevented by improvements related to water, sanitation and hygiene.1 A leading cause of death and disease globally is diarrhoea, and an estimated 90% of diarrhoea-related deaths occur as a result of unsafe drinking water and sanitation.2 3 Global initiatives such as the Sustainable Development Goals have led to significant advances in access to improved water sources and improved sanitation;4 however, progress has been uneven and large disparities remain between rural and urban areas. An estimated 663 million people worldwide still use unimproved water sources, and one-fifth of these people live in South Asia.4 India has the highest burden of child mortality and morbidity related to diarrhoea in Asia.2 5 Despite recent improvements in water sources and sanitation facilities for some of the population, many people in rural areas still cannot access improved water sources and gross disparities in coverage exist across the country.3 6 Technical water treatment solutions are out of the financial reach of most communities. Affordable water treatment systems that have beneficial human health effects and that can be maintained and sustained using local resources are needed. Most previous studies of water quality interventions in developing settings have focused on household based ‘point-of-use’ (POU) interventions, which can be delivered to and randomised at the level of single households.7 Systematic reviews of these studies have demonstrated a reduction in diarrhoeal disease associated with the use of such devices; however, effect sizes are heterogeneous.8–10 Additionally, POU interventions are frequently limited by poor uptake and lack of sustained use, and may have low acceptability to the population.7–9 Community-level interventions are an alternative approach to water quality improvement, where improved water supplies (eg, filtered water) are piped to communal tanks or directly to households. These interventions may have broader uptake and coverage, and may have additional health benefits due to increased water availability for personal hygiene and household use. Unfortunately, such interventions are often limited by cost, and there is currently insufficient evidence to know if they consistently reduce diarrhoeal rates.8 An inexpensive community-level intervention approach is riverbank filtration (RBF), a water treatment technique used in Europe for over 100 years that can remove contaminants from water to improve water quality.11 12 RBF wells can be installed in the immediate proximity of polluted rivers (figure 1). When an RBF well is pumped, river water is induced to flow through porous riverbed sediments and in the process bacterial pathogen removal efficiencies of three to four log can be achieved13 and heavy metal and organic pollutant concentrations are reduced. Pilot studies have demonstrated the feasibility of using RBF technology for water treatment in rural communities in India, and the ability of such technology to supply quantities of water sufficient to ensure all personal hygiene, food hygiene, laundry and bathing needs are met.12 14 RBF uses auto-regenerative, natural treatment processes, so properly engineered systems can remain effective indefinitely. Unlike other community-level interventions, such as high-capacity reverse osmosis filtration systems, RBF systems are technologically non-complex and comparatively inexpensive; required maintenance is limited to disinfection of the well(s) and/or pipeline and storage tanks, if indicated by the presence of faecal indicator bacteria in routinely collected water samples. An RBF system consists of one or more RBF wells, electric submersible pump(s), water storage tank(s) and pipeline(s) fitted with taps at designated water distribution points. While RBF operators require minimal training and can therefore be recruited from local communities, the placing and design of RBF wells and water supply infrastructure requires input from hydrogeologists and engineers. Figure 1 Schematic of a riverbank filtration system. We have designed a stepped wedge cluster randomised control trial (RCT) of a community-level hygiene education intervention and a source-based improvement in community water supply using RBF in four rural Indian communities. We plan to address the following objectives: Primary objective: to measure the effect of hygiene education and provision of RBF-treated water on the prevalence of diarrhoea in four rural India communities. Specific research question: in communities receiving hygiene education, does supply of piped RBF-treated water reduce the (7-day period) prevalence of reported diarrhoeal disease in comparison to piped unfiltered river water? Secondary objectives To measure the effect of hygiene education and provision of RBF-treated water on the (7-day period) prevalence of other hygiene-related illnesses, including respiratory and skin infections; To determine whether hygiene education and provision of RBF-treated water leads to changes in hygiene practices, preferred domestic water supply or household water consumption; To determine whether installation of RBF systems leads to effective and sustainable removal of bacterial faecal indicators, turbidity reduction and chemical removal; To evaluate the cost-effectiveness of installing RBF systems in India; To confirm the feasibility of transferring RBF system maintenance to local communities; To document household water storage practices and to compare the microbiological quality of piped and stored household water; To investigate the temporal association between the microbiological quality of stored household water and the prevalence of reported diarrhoea. Methods and analysis Study design This study will use a closed cohort stepped wedge cluster randomised design (figure 2), which involves a sequential crossover of clusters from the control to the intervention arm, so that every cluster begins in the control condition and eventually receives the intervention, with the order of crossover randomly determined. Figure 2 Stepped wedge schematic for the study. Each cell represents a data collection point. Each ‘step’ from T1 to T5 will be ∼12 weeks in length. Shaded red cells represent the prepipe augmentation stage and the baseline (T0) survey will be performed during this stage. Once the augmented pipework has been completed in all four villages, piped untreated river water will be delivered to all villages (unshaded cells, pre-riverbank filtration (RBF) intervention). The RBF intervention (shaded blue cells) will then be sequentially introduced to the villages in random order. The T1–T5 household surveys will start in week 5 of each step. The study will be conducted in four rural villages, currently using untreated river water as their primary source of water for drinking and hygiene. At the start of the study period, baseline (T0) demographic and health data will be collected from each consenting household and baseline hygiene education will be provided. Following this, new pipes, water storage tanks and taps will be installed in accessible locations throughout all villages. Once completed, simultaneous supply of piped untreated river water to all villages (control arm) will be provided (start of T1). This will increase the quantity of water available and ensure that all households in the villages have improved spatial and temporal (seasonal) access to greater volumes of untreated river water for hygiene and drinking purposes (improved water quantity). The second (T1) health survey will start 4 weeks after the initiation of piped untreated river water supply to evaluate the impact of hygiene education combined with improved water quantity compared with baseline (T0). RBF-treated water (intervention arm) will then be sequentially introduced to each village in random order at 12-week intervals (T2–T5), with health surveys performed 4 weeks after the implementation of the intervention to assess the additional effects of improved water quality. RBF systems will be built during the construction phase but will remain disconnected until the relevant village is randomised to receive the intervention. River, tank and tap water from each village will be sampled to assess microbial and chemical quality, and microbiological analysis will be performed on stored water from a subset of households. The stepped wedge cluster RCT is a pragmatic study design in which all villages within the study eventually receive the intervention, thereby improving equity and acceptability. Comparisons between intervention and control are made within each cluster, maximising statistical power, particularly when cluster sizes are large. Between-cluster variation and seasonal effects are accommodated in the statistical analysis model. This study is registered with the Australian New Zealand Clinical Trials Registry (ACTRN12616001286437) and the WHO Trial Registration Data Set for this study is included in online supplementary appendix A. Study registration was delayed until after recruitment due to human resource constraints; however, protocols developed prior to data collection were strictly followed. At the time of registration, only baseline demographic and health data had been collected; water supply systems, representing the control condition for the stepped wedge study design, had not yet been implemented. The study is expected to end in January 2018. 10.1136/bmjopen-2016-015036.supp1supplementary appendix Setting The study will be conducted in four similarly sized villages of ∼3000 people situated along the Krishna River in the Belgaum district of the state of Karnataka, India. The Krishna River is India's fourth largest river basin, receiving effluent and wastewater from large cities, including Hyderabad, Pune, Kolhapur, Satara and Kurnool.15 Local sewage treatment plants are typically working beyond capacity, so most wastewater is discharged into the river untreated. This region is appropriate for this study for several reasons: (1) prior pilot work has established excellent working relationships with local NGOs and community representatives, (2) many of the villages in the area do not currently have access to improved water sources, and (3) since the Krishna River is one of India's most polluted rivers, the capacity of RBF to reduce turbidity, remove heavy metals and remove Escherichia coli (the latter measure indicative of bacterial faecal pathogen removal) will be quantifiable. Site selection More than 20 villages along the Krishna River were visited by study investigators and evaluated for inclusion in the study. Features assessed included village demographics, current water usage and sources, hydrogeological conditions, availability of suitable land for RBF infrastructure, community receptiveness to installation of RBF systems and commitment to ongoing operation and maintenance of RBF infrastructure. Villages were excluded if: (1) there were an insufficient number of households, (2) untreated river water was not the primary source of drinking water for most households and/or (3) necessary water supply works were unlikely to be feasible due to cost, logistics or community participation. Following application of exclusion criteria, four villages remained that fulfilled the study requirements. Among selected villages, there is substantial seasonal variation in water availability and accessibility. During the wet season, river levels are high and villagers can collect water from nearby sections of the river. During the dry season, some villagers travel up to 8 km to collect water from the river. A small subset of households have access to alternative water supplies, such as privately piped unfiltered river water, or private bore water. Engagement Relevant local information was obtained via meetings between researchers, members of the local government body (Gram Panchayat) and village elders in each of the candidate villages. A participatory community model was used to gather information about existing water supply systems, sanitation facilities, electricity supply, up-to-date population and household statistics and to gauge interest in the project. The planned study design was discussed and village representatives were advised that the RBF intervention would be allocated in a random order. The Gram Panchayat were asked about barriers, if any, to village participation. Discussions were conducted in the local language, Kannada. Bore wells, pipework, community tanks and village standpipes were marked on a Google map image of each village. Stickers representing pumps, valves, storage tanks and other water supply features (sourced from an Australian Community Water Planner field guide)16 were then employed to assist in mapping out proposed water supply works with village representatives. Following identification of four villages that met inclusion criteria and agreed to participation, relevant approvals were sought from the Indian government at a ministerial level (Belgaum District Chief Engineer) and from each Gram Panchayat. Intervention The intervention will incorporate a hygiene education program primarily focusing on hand hygiene and safe storage of household water supplies, and a community-level water quality intervention using RBF technology. Educational materials have been developed from open content resources produced by the Centre for Affordable Water and Sanitation Technology (CAWST) and HydrAid, licensed under the Creative Commons Attribution 2.5 Canada Licence. Education will be structured around two posters (see online supplementary figures 1 and 2) which have been translated into the local language, Kannada. Following enrolment, members of each consenting household will be provided with hygiene education by trained fieldworkers, and laminated copies of educational materials will be supplied to each household. A poster focusing on reducing contamination of food and utensils by improving hand hygiene will be provided for display near each household's handwashing station, and a poster focusing on reducing the risk of contamination of stored water by promoting covering drinking water containers and safely accessing drinking water will be provided for display near water storage containers. At each subsequent household visit, the presence and location of these posters will be assessed, and any missing posters replaced. Posters will also be displayed on project tanks and in public spaces in each village. 10.1136/bmjopen-2016-015036.supp3supplementary figures The water quality intervention will use RBF technology. At each RBF site, one or more wells will be installed in a location at least 50 m away from the river. PVC pipes with a diameter of 15 cm will be used, with vertical slots of ∼15 cm length cut into the pipe. If possible, an outer steel casing and a bentonite seal will be installed to protect the PVC well casing. If this is not possible, the uppermost 1 m around the PVC pipe will be excavated and back-filled with concrete. Above the surface, the well heads will be encased in a 1 by 1 m concrete slab for additional protection. The well head will be covered by a steel cap and further protected by a locked steel cage (figure 3). A pump submerged into each well will induce water to flow through porous riverbed (alluvial) sediments, facilitating removal of pathogens and dissolved/suspended contaminants via a combination of physical, chemical and biological processes.17 The pump in the RBF well will be sized to provide sufficient pressure to move water from the well to storage tanks. Prior to use, all wells will be sanitised with a solution of one part 5% sodium hypochlorite (NaOCl) bleach to three parts water. Each RBF well will be fitted with a calibrated water meter to gauge volumes of water produced, and a pressure head logger will be installed at each well field to automatically monitor the well's water level. Figure 3 Schematic of a riverbank filtration well. Covered storage tanks (2000–3000 L), pipes and taps will be installed in all villages during the project's baseline (T0) stage. At least one storage tank will be placed at the topographic highest point in the village to enable gravity flow to tanks and taps, but if gravitational pressure is insufficient, transfer pumps will be installed. If needed, sediment filters will be installed at the RBF well head to remove suspended solids. During the T1 stage, unfiltered river water will be pumped from the river to the storage tanks. Since each village is allocated to RBF, the pipework for that village will be connected to the RBF field by a buried 5 cm plastic pipe. Depending on the distance from the river, the size of the village and the exact routing of the water distribution system in each village, the length of the pipeline and the number of storage tanks and taps will vary from site to site, that is, the pipeline length will range from about 2000 to 3900 m and the number of storage tanks will be ∼12 to 17 per village. For cost reasons, individual tap connections will be installed in appropriate communal locations rather than to each household. RBF system maintenance will be conducted as required throughout the study. Maintenance will consist of disinfection of the well(s) and/or the pipeline and storage tanks, if indicated by the presence of faecal indicator bacteria in routinely collected water samples. A sodium hypochlorite solution will be prepared, added to the well(s) and pumped through the distribution system. During these maintenance periods, lasting ∼24 hours, access to the water distribution system will be prohibited. Recruitment All households from each village will be approached. Each household approached will be allocated a unique number irrespective of whether the household consents to participation. Twelve pairs of surveyors (three for each village) who are fluent in English and Kannada will provide an explanatory statement describing the study (see online supplementary appendix B) to a self-nominated adult household representative. The household representative will then be invited to participate in the study and provide consent on behalf of their household. All household members will be eligible for inclusion in the study, regardless of age. At the time of recruitment, householders will also be given the option to provide written informed consent to: (1) participation in future interviews, (2) collection of stored water samples and/or (3) photography of household water storage containers and handwashing area(s). Each household will have the option to participate in up to five subsequent surveys. 10.1136/bmjopen-2016-015036.supp2supplementary appendix Data collection Health surveys Twelve pairs of surveyors will be trained to conduct household surveys. In consenting households, each household member will be allocated a person identifier number and identifiable information will be collected (age, first name, gender, relationship to head of household, marital status, educational attainment and main occupation/activity) and entered onto a standardised paper registration form, which will be stored in a secure location (locked filing cabinet with restricted access). All subsequent survey data will be deidentified (using unique household and person identifier numbers only) and collected using an electronic data collection form on handheld electronic tablet devices, designed using Magpi's data collection platform (http://home.magpi.com/). Magpi is a real-time, configurable, cloud-based mobile collection and communication tool with global positioning system coordinates and time stamping. Questionnaires will be written in English, and imported onto tablet devices for onsite data entry. Validation rules and range limits will be used to ensure precise data entry. Following survey completion, deidentified data will be transmitted to a central database via the cloud. During each survey period, study investigators will inspect data on at least a weekly basis to verify data upload, assess completeness of data entry and identify missing data fields and duplicate records. Where required, appropriate remedial action (eg, refresher training of health surveyors) will be undertaken. Surveys will be administered in Kannada and will include questions on household water sources, storage, treatment and usage, household handwashing facilities and practices, and sanitation facilities and practices. Surveys will also collect information on the health of each household member and on socioeconomic indicators. Specific questions will focus on diarrhoeal, skin and respiratory symptoms, health-seeking behaviour and consumption of water outside of the household setting. Specimen collection and analysis River water, RBF tank water and tap water from each village will be sampled at each stage of the study to assess bacteria (E. coli), organoleptic (smell, taste) and physico-chemical (pH, temperature, electrical conductivity, turbidity and dissolved oxygen) parameters using calibrated digital field meters and specialised laboratory equipment. Approximately 6000 water samples will be collected over the entire study period (up to 184 samples per week). In a subset of consenting households (∼160 per village), household stored water samples will be collected alongside the T1–T5 health surveys using sterile sample containers and aseptic technique. Results will be compared with samples taken from piped water systems to assess for contamination of drinking water associated with unsafe storage or poor hygienic practices. Stored water samples will be collected 1 week prior to administration of health surveys to enable assessment of temporal associations between contaminated stored household water and reported diarrhoea. Samples for microbiological testing will be analysed within hours of sample collection by a trained local technician at a purpose-built local laboratory. Escherichia coli bacteria will be analysed as most probable number using the US Environmental Protection Agency (US EPA) approved Colilert method (IDEXX Laboratories). Following the Bureau of Indian Standards (BIS) water quality guidelines,18 a subset of samples will be analysed by an independent laboratory for chemical quality (anions and cations, heavy metals and organic pollutants). RBF system electricity consumption and water output will be recorded in a field journal and/or tablet device. Water quality, microbiological and chemical testing results for RBF-treated water samples will be assessed according to Indian drinking water quality guidelines.18 If RBF-treated samples do not meet national standards, the RBF operation will be reviewed and optimised, and/or additional water treatment measures (eg, chlorination) will be considered. Study outcomes The primary outcome of the study is a 7-day period prevalence of diarrhoea among villagers of all ages. Secondary outcomes include a 7-day period prevalence of other hygiene-related illnesses (respiratory and skin infections), reported changes in hygiene practices, household water usage and water supply preference. Outcomes will be measured at each of the six survey visits. Sample size Sample size estimations for the primary outcome were based on detecting a relative reduction in the prevalence of diarrhoea in the past 7 days with 80% power and α=0.05. Calculations were based on a two-level hierarchy stepped wedge design structure of individuals nested within villages. We obtained estimates of diarrhoeal prevalence and household size from pilot work conducted in Karnataka, India, that included a survey of 7 villages (110 households).19 We used a within-village correlation of a 7-day diarrhoea period prevalence of 0.02, which is conservative relative to the value of 0.01 obtained from a large cluster randomised trial in India20 and estimated a 50% increase in prevalence during the wet season (June–October). On the basis of results of systematic reviews of other water and hygiene interventions,9 10 21 we assumed that our intervention would lead to a ∼35% relative reduction in self-reported diarrhoea. The required sample size was obtained using numerical simulation. Hypothetical trial data were simulated from a binary random effects model with a logarithmic link and with the desired within-cluster correlation,22 followed by an analysis aggregating the data to prevalences for each village at each time point and fitting a linear regression model with the logarithm of the prevalence as the dependent variable, controlling for seasonal effects and with fixed effects for villages. Using this approach, we estimated that 430 households (with an average household size of 5 persons) will need to be recruited from each of the 4 villages to detect a 35% reduction from a baseline diarrhoeal prevalence of 3% with 80% power. Estimated power with a within-village correlation of 0.01 (rather than 0.02) is 87%. We also verified that the type I error rate of 5% was preserved with this procedure with only four villages (clusters). To allow for attrition across the five repeat surveys, we plan to recruit all available consenting households to ensure that adequate power is retained for analysis of the primary outcome. Randomisation and blinding Randomisation will be performed by the study statistician (ABF); random ordering of the allocation sequence will be determined from computer-generated uniform random numbers. The allocation sequence will only be made available to two study investigators (ABF and MS). Indian study investigators will be blinded to the allocation sequence with only the next village randomised for rollout being revealed at each intervention implementation time point. Study participants will be blinded to the allocation sequence and those not yet receiving the intervention will not be aware of the time at which they will have the intervention implemented. Blinding to the intervention (ie, the type of water being received) is not possible due to potential differences in turbidity of untreated and RBF-treated river water. Statistical analysis The primary outcome (diarrhoeal prevalence) will be calculated for each cell in the stepped wedge design by aggregating over all individuals surveyed in each village during each time period. Estimation of intervention effects will be obtained from a linear regression of the logarithm of the village-aggregated prevalence adjusting for seasonal effects and incorporating village as a fixed effect. The intervention effect coefficient will be exponentiated to produce an estimated relative reduction (with 95% CIs) in the overall prevalence of diarrhoea in the intervention periods (post-RBF) compared with control periods (piped but unfiltered water). This analysis model controls for both clustering of individuals within villages and for repeated assessments of villages over time. Analyses of secondary outcomes will be made using the same methods as above for prevalence outcomes, and using linear regression models of village-time averages without log transformation for continuous outcomes. We will use multiple imputation to impute missing outcomes at the individual person level which will then be aggregated for the village-level analyses. We will perform post hoc sensitivity analyses adjusting for aggregates at the village level of variables that are observed to change over time. We will perform limited subgroup analyses assessing intervention effects in children under 5 years of age and in adults aged over 65 years, although such analyses will have limited power. Cost-effectiveness and cost–benefit analysis Economic evaluation of RBF will be performed at the conclusion of the study. Cost-effectiveness of RBF for prevention of illness, especially diarrhoea, will be measured and reported as cost per event prevented and cost per disability-adjusted life year averted. Benefits including productivity gains will be quantified, cost-effectiveness ratios and a cost-benefit analysis will be evaluated and extensive sensitivity analysis will be performed to assess the impact of cost drivers and uncertain variables. Strengths This is the first randomised controlled study to assess the health impact of a community-level water intervention using RBF technology. This community-level study is also novel as it considers the impact of interventions that target increased water access and improved water quality. The stepped wedge RCT design allows all participants to eventually receive the intervention, improving equity and acceptability. While waterborne and hygiene-related illnesses vary with temporal factors such as season, we are able to control for temporal effects in the model because these factors apply to all villages together and the stepped wedge model includes a time factor variable. Factors that remain constant over time (eg, demographic factors such as age and gender) will be controlled for in the statistical analyses involving within-village comparisons. Limitations The incorporation of only four clusters (villages) in the trial may limit the generalisability of results. While characteristics of study villages (eg, literacy, employment and household sizes) are comparable to Karnataka state averages, they may not be generalisable to the rest of India. Baseline imbalances (eg, the proportion of persons from Scheduled Castes, the proportion of children aged under 5 years and open defecation rates) will be dealt with using fixed village effects. Our analyses are purely within-village comparisons; hence, village-level factors that differ across villages at baseline but remain constant over time are automatically controlled for in these analyses. While the allocation sequence will not be revealed to participants, we expect that participants will become aware that they are receiving the intervention (RBF-treated water) due to potential differences in turbidity and taste of untreated river water and RBF-treated water. Since water usage and health symptoms will be self-reported and blinding is difficult, any preconceived ideas about the value of increased water quantity or water treatment could bias symptom reporting. This so-called ‘courtesy bias’ (a degree of minimisation of symptoms) might occur when piped water is initially delivered (T1), or when RBF is supplied (T2–5). To detect courtesy bias and assess the validity of self-reporting, we have included falsification symptoms in health surveys (such as the presence of scrapes or bruises or urinary tract symptoms). These fairly common symptoms have been deliberately chosen as they might be prone to courtesy bias, but would not be expected to be affected by the RBF intervention. Contamination of stored water within the household may bias the effect of consuming RBF-treated water; however, collection of household stored water samples prior to health surveys in a subset of households will allow an assessment of the likelihood of contamination and the relationship between the microbiological quality of stored drinking water and diarrhoea. Our primary research question is based on the provision of RBF-treated water rather than adherence to the intervention, as indicated by our use of an intention-to-treat analysis. However, we expect that adherence to RBF-treated water will be imperfect, as alternative drinking water sources are available to study participants. Each survey will collect information on preferred household drinking water sources, and this information will be reported, as the use of alternative water sources may be a limitation of the intervention. Safety and data monitoring This is a low-risk study, and there are no foreseeable risks to participants from the intervention. Therefore, a data monitoring committee is not required and no interim analysis is planned. Dissemination Surveyors will provide householders with a written explanatory statement and will verbally explain the broad aims of the study before providing householders with an opportunity to ask questions. Participants will be advised at the time of recruitment that participation in the study is voluntary and that they may withdraw from the study at any time. Written signed consent will be obtained from an adult householder (aged ≥18 years). For illiterate participants, an ink thumb-print will be used in lieu of a signature. Any modifications to the protocol which may impact on the conduct of the study (eg, changes to eligibility criteria, outcomes, analyses) will be approved by ethics committees in Australia and India prior to implementation, and communicated to participants and other relevant stakeholders via written correspondence. Protocol changes will also be formally documented on the ANZCTR website. Results of the study will be published in peer-reviewed journals, presented at national and international conferences, and discussed with project partners. Participants will have access to results via access to The Energy and Resources Institute (TERI) and Monash University websites, which will contain a summary of the study findings. In addition, summary results will be sent to village elders for their dissemination to the community. The authors gratefully acknowledge SP Luby for his input into the study design. Contributors: SLM revised the study protocol and drafted the manuscript. JEOT codrafted the manuscript. JEOT, MS, TBB, SKG, ABF and KL conceived and designed the study. JEOT designed the data collection tools and developed laboratory protocols. ABF determined the sample size considerations and statistical analysis plan, and will conduct the primary statistical analysis. All authors contributed to refinement of the study protocol and approved the final manuscript. Funding: This work was supported by an Australian National Health and Medical Research Council (NHMRC) project grant (APP1083408). KL holds an NHMRC Career Development Fellowship (APP1084351). SLM holds an NHMRC Postgraduate Scholarship (APP1115196). Competing interests: None declared. Ethics approval: Ethics approval was obtained from the Monash University Human Research Ethics Committee (project number CF15/522-2015000248) in Australia and The Energy and Resources Institute Institutional Ethics Committee in India. Provenance and peer review: Not commissioned; externally peer reviewed. ==== Refs References 1 Pruss-Ustun A , Bos R , Gore F Safer water, better health: costs, benefits and sustainability of interventions to protect and promote health . Geneva : WHO Press , 2008 . 2 United Nations Children's Fund (UNICEF), World Health Organisation (WHO) . Diarrhoea: why children are still dying and what can be done . Geneva : WHO Press , 2009 . 3 Black RE , Morris SS , Bryce J Where and why are 10 million children dying every year? Lancet 2003 ;361 :2226 –34 . 10.1016/S0140-6736(03)13779-8 12842379 4 United Nations . The Millennium Development Goals Report 2015 . New York http://www.un.org/millenniumgoals/reports.shtml (accessed 9 May 2016 ). 5 Kumar S , Vollmer S Does access to improved sanitation reduce childhood diarrhea in rural India? Health Econ 2013 ;22 :410 –27 . 10.1002/hec.2809 22438282 6 Fan VYM , Mahal A What prevents child diarrhoea? The impacts of water supply, toilets and hand-washing in rural India . J Dev Effectiveness 2011 ;3 :340 –70 . 10.1080/19439342.2011.596941 7 Schmidt WP The elusive effect of water and sanitation on the global burden of disease . Trop Med Int Health 2014 ;19 :522 –7 . 10.1111/tmi.12286 24576060 8 Clasen TF , Alexander KT , Sinclair D Interventions to improve water quality for preventing diarrhoea . Cochrane Database Syst Rev 2015 ;(10) :CD004794 10.1002/14651858.CD004794.pub3 26488938 9 Wolf J , Pruss-Ustun A , Cumming O Assessing the impact of drinking water and sanitation on diarrhoeal disease in low- and middle-income settings: systematic review and meta-regression . Trop Med Int Health 2014 ;19 :928 –42 . 10.1111/tmi.12331 24811732 10 Waddington H , Snilstveit B Effectiveness and sustainability of water, sanitation, and hygiene interventions in combating diarrhoea . J Dev Effectiveness 2009 ;1 :295 –335 . 10.1080/19439340903141175 11 Schubert J Hydraulic aspects of riverbank filtration—field studies . J Hydrol 2002 ;266 :145 –61 . 10.1016/S0022-1694(02)00159-2 12 Boving TB , Choudri BS , Cady P Hydraulic and hydrogeochemical characteristics of a riverbank filtration site in rural India . Water Environ Res 2014 ;86 :636 –48 . 25112031 13 Cady P , Boving TB , Choudri BS Attenuation of bacteria at a riverbank filtration site in rural India . Water Environ Res 2013 ;85 :2164 –74 . 10.2175/106143013X13736496909428 14 Howard G , Bartram J Domestic water quantity, service level and health . Geneva : WHO Press , 2003 . 15 Maharashtra State Plan Division PC, Government of India . Maharashtra State Development Report . Chapter 15: Water and environment New Delhi : Planning Commission, Government of India , 2005 . 16 Grey-Gardner R , Taylor K Community water planner field guide . Canberra, Australia : Australian Government National Water Commission , 2009 . 17 Keuhn W , Mueller U Riverbank filtration: an overview . J Am Water Works Assoc 2000 ;92 :60 –9 . 18 Bureau of Indian Standards (BIS) . Indian standard: drinking water—specification (Second Revision) IS 10500 . New Delhi : Agriculture DoFa , 2012 . 19 Boving T , Choudri BS , Cady P Clean water by riverbank filtration . Final report prepared for World Bank Development Marketplace Program (DM 2007 Grant 2447) , 2010 . 20 Clasen T , Boisson S , Routray P Effectiveness of a rural sanitation programme on diarrhoea, soil-transmitted helminth infection, and child malnutrition in Odisha, India: a cluster-randomised trial . Lancet Glob Health 2014 ;2 :e645 –53 . 10.1016/S2214-109X(14)70307-9 25442689 21 Fewtrell L , Kaufmann RB , Kay D Water, sanitation, and hygiene interventions to reduce diarrhoea in less developed countries: a systematic review and meta-analysis . Lancet Infect Dis 2005 ;5 :42 –52 . 10.1016/S1473-3099(04)01253-8 15620560 22 Yelland LN , Salter AB , Ryan P Adjusted intraclass correlation coefficients for binary data: methods and estimates from a cluster-randomized trial in primary care . Clin Trials 2011 ;8 :48 –58 . 10.1177/1740774510392256 21335589

PMC005xxxxxx/PMC5372112.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01293510.1136/bmjopen-2016-012935OncologyResearch150617171684Use of a computerised decision aid (DA) to inform the decision process on adjuvant chemotherapy in patients with stage II colorectal cancer: development and preliminary evaluation Miles A 1Chronakis I 2Fox J 23Mayer A 4 1 Birkbeck, University of London, London, UK 2 University College London, London, UK 3 Oxford University, Oxford, UK 4 Royal Free London NHS Trust, London, UK▸ Additional material is available. To view please visit the journal online (http://dx.doi.org/10.1136/bmjopen-2016-012935). Correspondence to Dr A Miles; ae.miles@bbk.ac.uk2017 24 3 2017 7 3 e0129353 6 2016 26 9 2016 8 11 2016 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives To develop a computerised decision aid (DA) to inform the decision process on adjuvant chemotherapy in patients with stage II colorectal cancer, and examine perceived usefulness, acceptability and areas for improvement of the DA. Design Mixed methods. Setting Single outpatient oncology department in central London. Participants Consecutive recruitment of 13 patients with stage II colorectal cancer, 12 of whom completed the study. Inclusion criteria were: age >18 years; complete resection for stage II adenocarcinoma of the colon or rectum; patients within 14–56 days after surgery; no contraindication to adjuvant chemotherapy; able to give written informed consent. Exclusion criterion: previous chemotherapy. Primary outcomes Patient perceived usefulness (assessed by the PrepDM questionnaire) and acceptability of the DA. Results PrepDM scores, measuring the perceived usefulness of the DA in preparing the patient to communicate with their doctor and make a health decision, were above those reported in other patient groups. Patient acceptability scores were also high; however, interviews showed that there was evidence of a lack of understanding of key information among some patients, in particular their baseline risk of recurrence, the net benefit of combination chemotherapy and the rationale for having chemotherapy when cancer had apparently gone. Conclusions Patients found the DA acceptable and useful in supporting their decision about whether or not to have adjuvant chemotherapy. Suggested improvements for the DA include: sequential presentation of treatment options (eg, no treatment vs 1 drug, 1 drug vs 2 drugs) to enhance patient understanding of the difference between combination and single therapy, diagrams to help patients understand the rationale for chemotherapy to prevent a recurrence and inbuilt checks on patient understanding of baseline risk of recurrence and net benefit of chemotherapy. CHEMOTHERAPYQUALITATIVE RESEARCHdecision aidadjuvantpatient communication ==== Body Strengths and limitations of this study This study evaluated patient acceptability, perceived usefulness and understanding of a decision aid (DA) using mixed-methods at the point patients made their decision about whether or not to have adjuvant chemotherapy. We recruited men and women with a range of different ages and levels of education, including people with no educational qualifications, enabling us to capture a range of different responses to the DA. People educated to degree level or above were over-represented in the study, and this may have led to an overestimate of perceived usefulness and acceptability of the DA. Introduction Ethical considerations and recent policy changes have put patients at the centre of health decisions, aiming to make shared-decision-making ‘the norm’.1 2 One key issue for people who have undergone surgical resection of colorectal cancer is the decision about whether or not to have adjuvant chemotherapy, which is typically either 5-FU or capecitabine, given alone or in combination with oxaliplatin.3 4 While surgery is highly effective for localised disease, up to 85% of patients with lymph node involvement (stage III) relapse within 5 years. Since chemotherapy can prevent recurrence in up to 25% of patients with stage III disease,5 it is offered routinely. For patients without lymph node involvement (stage II), the risk of relapse is lower (20–40%) and a smaller number of patients (3–7%) are expected to benefit.6 7 Since there is a risk of side effects following chemotherapy, some of which can be life-threatening or permanent and the balance of harms and benefits is marginal in patients with stage II, the use of adjuvant chemotherapy in this patient group has remained controversial.8 Patients with colorectal cancer have expressed a desire for more information, notably about their cancer, prognosis and treatment options,9 10 and effective communication of diagnostic and prognostic information has been shown to enhance patient well-being and patients' perceptions of the quality of doctor–patient interactions.11 However, patients can hold more positive views on adjuvant therapy than physicians12 and may want chemotherapy even when there is little evidence that it will help prevent a recurrence;13 14 for example, women with breast cancer with a low risk of recurrence are more likely to decline chemotherapy if they are aware of the (small) impact,15 underlining the need for effective communication. People have problems understanding risk information,16 and while there is an emphasis on patients making decisions that are consistent with their values, evidence points to the instability of preferences, with some researchers arguing that true preferences may not always exist, but are constructed ‘on the spot’ using the information available at the time of decision-making.17 Decision aids To facilitate informed decision-making, a wide range of decision aids (DAs) have been developed. DAs are used alongside patient–physician interactions and contain detailed information about clinical options and outcomes, are explicit about the choices facing the patient and encourage patients to express their preference. DAs typically result in greater knowledge, a higher proportion of people with accurate risk perceptions, lower decisional conflict (eg, feeling better informed and clearer about personal values) and more active roles in decision-making compared with usual care.18 Although DAs have been developed to aid treatment decisions for a number of different cancers,19–22 little research has been done in the context of colorectal cancer. One exception is a booklet to help patients make decisions about adjuvant chemotherapy or radiotherapy, although this was only evaluated among patients who had opted to have adjuvant therapy, and was developed to be used by the patient at home rather than as part of a clinical consultation.10 23 The aim of the present study was to develop a DA for chemotherapy for stage II that included the information identified as important to patients in previous research9 10 and examine its usefulness and acceptability to patients making a decision about whether or not to have adjuvant chemotherapy, as well as to identify any areas where patient understanding of key information could be improved. Methods Development of the DA PROforma decision support technology (a formal language for automating clinical processes and protocols) and the Tallis tool set (used to design the decision support, check the syntactic integrity of the programme and run the application in a test environment) were used to develop the DA;24 25 http://openclinical.net/index.php?id=390. A variety of decision support applications based on PROforma technology have demonstrated a significant positive impact on a number of outcome measures such as reducing prescription errors and adherence to evidence-based guidelines.26–28 The DA captured patient demographic and clinical information (eg, histology type, evidence of extramural vascular invasion), as well as patient fitness to undergo chemotherapy (assessed via the presence of comorbidity and ECOG performance status29 ranging from fit and well (0) to bed bound (4)). Risk of recurrence with and without chemotherapy was mathematically derived based on: (1) the baseline risk of recurrence without chemotherapy, which was calculated for each patient using information from the published SEER database (http://seer.cancer.gov/data/) using the adjuvant!online risk calculator (http://www.adjuvantonline.com);30 and (2) the effectiveness of single and combination chemotherapy, taken from published randomised controlled trials and meta-analysis.6 7 This calculator matches baseline risk of recurrence data to the individual patient's clinicopathological characteristics (T stage, number of lymph nodes examined, histology type and histopathological grading). The age-specific natural mortality was derived from published national mortality data. In line with the presentation of information in Adjuvant!online, and consistent with recommendations, we used a simple percent format, and absolute risk when conveying information about the net benefit of chemotherapy along with a bar chart showing the likelihood of different outcomes, thereby using visual and numeric information to convey risk of recurrence information, keeping the denominator consistent and stating the reference class and time frame.31 32 Patients were first shown one graph detailing their baseline risk of recurrence in the absence of chemotherapy (see figure 1) and graphs showing the net benefit of single and combination chemotherapy on the final page of the DA (see figure 2). Since the risk prediction of Adjuvant!online does not include clinical presentation (obstruction or perforation), or the risk factors vascular, lymphovascular and perineural invasion, these were included under headings for ‘indicators for good prognosis’ (eg, tumour stage T3) and ‘indicators for bad prognosis’ (eg, extramural vascular invasion), presented on the final page of the DA (see figure 2). Figure 1 Graph showing baseline risk of recurrence. Figure 2 Final page of decision aid showing risk of recurrence and net benefit of treatment options alongside pros and cons of different treatment options. Information about the method of administration for the different drugs, the need for the insertion of a central line, the number of times the patient would need to come to hospital and the duration of treatment was given. The list of side effects were put into three different sections: common (>30% of patients affected), less common but serious (potentially life-threatening) and potentially permanent side effects. It was made clear that the first two sections related to all chemotherapy options, but that the long-term side effects only related to the two-drug chemotherapy option (oxaliplatin carries a high risk of long-term numbness of the hands and feet).4 The values clarification exercise was adapted from one developed for use in colorectal cancer screening among elderly people33 to capture the main issues involved in the decision, including features identified as important to patients with colorectal cancer9 (see online supplementary table S1). Patients were asked to complete the questionnaire on the computer, selecting 1 of the 2 options presented for all 10 questions. Patients were able to review and change their responses to the questions at any time during or after the consultation. 10.1136/bmjopen-2016-012935.supp1supplementary data The final page of the DA combined graphical presentation of risk of recurrence information for the three treatment options alongside arguments laying out the pros and cons for each option, and the results of the values clarification exercise (see figure 2). Design A mixed methods design was employed. Participants Potential patients were identified at the colorectal multidisciplinary team meeting at the Royal Free NHS Trust. Inclusion criteria were: male or female patients aged >18 years; patients who had undergone complete resection for stage II adenocarcinoma of the colon or rectum; patients within 14–56 days after surgery; no contraindication to adjuvant chemotherapy; able to give written informed consent. The exclusion criterion was previous chemotherapy. Target N was 12 in line with similar research on DAs.34 Patients were recruited consecutively. Procedures After informed consent, patients were randomised to receive the DA at either the first or second consultation. All patients were given written information about the side effects of chemotherapy, as per current practice. When seen back in clinic, after 1–2 weeks, patients were given the opportunity to ask questions, at which point the patient's treatment decision was recorded. Patients were invited to complete the values clarification exercise on the computer, but received help from relatives or the clinician if requested. All consultations took place in the Outpatients Oncology Department at the Royal Free London NHS Trust. Patients were given a questionnaire assessing responses to the DA after the consultation at which the DA had been used. Questionnaire Information about age, gender, educational level, ethnicity and employment status was collected. Perceived usefulness of the DA in preparing the patient to communicate with their doctor and make a health decision was assessed with the PrepDM questionnaire, a 10-item scale designed to measure the usefulness of DAs, developed across a number of different patient groups, including patients with cancer, with good internal reliability and validity.35 Cronbach's α in this study was 0.935. Example items include: ‘Did the decision aid…Help you think about the pros and cons of each option? … Help you identify questions you want to ask your doctor?’ Response options are on a five-point Likert scale with higher scores indicating higher agreement. Acceptability of the DA was assessed using seven items; three were from Bennett et al,35 while the remaining items were developed for this study (see table 1 for questions and response options). Data were analysed using SPSS V.22. Table 1 Perceived usefulness and acceptability of the decision aid Number Usefulness of information in preparing patient for decision (PrepDM) (1–5) mean (SD) 4.28 (0.90) Acceptability of DA35 Helpful  Very 10  Somewhat 1  A little 1  Not helpful 0 Recommend to others  Definitely yes 6  Probably yes 5  Probably not 1  Definitely not 0 Information clear  Everything clear 8  Mostly clear 4  Some clear 0  Mostly unclear 0 Did the DA add anything to the consultation  An extreme amount 1  Very much 7  A moderate amount 4  A little bit 0  Nothing 0 Graphical representation in the DA helpful  Extremely 4  Very 6  Moderately 2  A little bit 0  Not at all 0 Understood risk of recurrence info. in DA  Yes 10  No 0  Not sure 1 Helpfulness of risk of recurrence info. in DA  Extremely 3  Very 6  Moderately 1  A little bit 1  Not at all 0 Interview Face-to-face interviews were used to gain an in-depth understanding of patients’ responses to the DA, and their understanding of key information. These were conducted face to face at the hospital following the second consultation, between March 2012 and July 2014, by AMi, who has experience of conducting interviews with patients with colorectal cancer,36 and was not involved in the patients' care. The interviews lasted an average of 33 min (range: 18–54) and were recorded and transcribed verbatim by a transcription company, and reviewed for accuracy by AMi using the original recordings. The initial Topic Guide is provided as online supplementary data, but during the interviews, it became clear that patients were confused about what risk of recurrence meant and how chemotherapy could act on the body to prevent this. Subsequent patients were therefore asked about these issues in more detail. Data were analysed using thematic analysis,37 facilitated by the software package NVIVO V.10. AMi developed the codes by examining all instances in the data set that related to the DA, risk of recurrence information and the rationale for having chemotherapy. Themes were derived and developed via an iterative process, with constant comparisons of the data that identified similarities and differences within and across individual interviews. The codes and supporting quotes were reviewed for appropriateness and accuracy by AMa. Patients were not asked to verify the thematic analysis as we did not wish to increase participation burden. Results Fifteen patients were approached by clinic staff, 13 were recruited, with 12 completing the study. The age of participants ranged from 33 to 82, with a median age of 67. The ratio of men to women was 1:2. Over half (n=7) were educated to degree level or equivalent and the majority were of white ethnicity (British or Irish) (n=9). Approximately half were employed (n=6), one was unemployed and the remainder were retired. One patient reported having had chemotherapy prior to surgery (s6). This person was included since it was not felt to impact on the reported outcomes. Eleven of the 12 patients declined chemotherapy (all except s8). Patients’ perceptions of the usefulness of the DA in helping them communicate with their doctor and make a decision were good, with scores (mean 4.28, SD 0.9) comparing favourably with those reported in other clinical populations, where scores ranged from 3.3 to 3.9.35 Patients also reported finding the risk of recurrence information in the DA and graphical representation of information in the DA helpful, with all except one reporting that they were happy they understood the meaning of the information in the DA about risk of recurrence (see table 1). Graphical representation of information on risk A number of patients felt the graphs increased the clarity of the risk of recurrence information, “I liked the graphs, I suppose they're accurate really… I worked with statistics enough to know that you can kind of hide the numbers…. it made me consider chemotherapy more than I possibly would if I was just given words” (s10). Another participant described how the graphical presentation looked more scientific and hence more credible than being told the same information verbally: “Really it's no different to what the surgeon had already told us yesterday… it's a very much more scientific, you know, sort of presentation to it…” (s4). For some patients, the graphs provided information they would otherwise have overlooked: for one patient, it was the additional benefit conferred by two, compared with one, chemotherapy drug “I didn't realise that the one was quite like a significant difference there, between the two (chemotherapy drugs)” (s3); for another patient, it was the risk of death from other causes: “When we're there we only talk about coming back and not coming back. But we do know that death is something that can come anytime… So I just put them in the two categories but on the graph I saw the three there” (s9). Some patients found it helpful to see the portion of the graph that illustrated the net benefit of chemotherapy, whether they wanted chemotherapy or not, and used it to help them cope with their decision: “that made me feel quite positive… even though there was only 5%… I'm going into the chemotherapy, now, thinking about that line” (s8), “the graph…the bit that she was talking about (net benefit of chemotherapy) is just tiny is quite good” (s3). Some patients liked the positive information about their general prognosis—“I was quite pleased to see that. That was nice. There was lots of green (baseline risk of recurrence)” (s8). One person liked it because they described themselves as ‘a visual person’ (s3). However, not everyone was able to articulate whether or why they had found the graphical information helpful, particularly if they had taken a younger relative into the consultation with them and felt confident their relative understood what was being said. “I didn't really understand it very much…I'm a bit of a dinosaur… I know my niece would have understood it all” (s11), “I just can't say…My son was with me now and he was very informative afterwards, he was, talked about it a lot…I think he found it very helpful” (s12). Values clarification The belief that they could not predict whether they would benefit from chemotherapy, the small risk of cancer coming back and the long-term side effects of chemotherapy were the main barriers to chemotherapy (see online supplementary table S2). Of least concern were the short-term side effects, insertion of a central line and repeat attendance at hospital, although around half were still concerned about these things. Reactions to the values clarification exercise were more mixed than to the graphical display. Some patients found it helpful, while others reported conflicting feelings about the information. Positive reactions included the feeling that it was an accurate reflection of their (often mixed) views about chemotherapy, was clearer, more explicit, more rational and less emotional, and empowering for the patient: “it's good in terms of sometimes you struggle to articulate why you make one decision versus another” (s10) “getting me to kind of click on things, and say yes… puts the onus on me to make a decision and own my decision and feel empowered to make a decision independent of what the doctor thinks…” (s8). “Rather than having an emotional response, there's an actual figure there to say, ‘This is what you decided.’ …it's good to be logical” (s3). Others commented on the fact that not all the items were equally important and that numbers for and against chemotherapy were a bit simplistic, or having to select a choice yes/no without saying why did not tell the whole story. One patient reported feelings of guilt when answering the values clarification exercise, because although they viewed their health as important they were opting to do something that did not necessarily maximise their chances of maintaining their health: “my health comes first above everything. And so a lot of questions there made me feel when I opted for a different answer to that, to really putting that first, I felt very guilty about, which was quite interesting. And I did pause and stress about that a little bit… But just because I answered the other way doesn't mean I still don't have my health, you know, at the forefront of my thoughts… However, there are other issues at stake here” (s4). Another felt it highlighted the importance of life over everything else: “I'm worrying about hair loss and this and that…. And what about that? Is that more important than life itself? Well, no, it isn't… But that tool…did think, you know, all of that is trivial in comparison with life itself. It's very precious” (s7). However, one woman felt she only needed the risk of recurrence statistics to make a decision, and did not need the values clarification exercise: “It didn't seem to be difficult to answer really…. the biggest thing that sticks out in my mind is the statistics and the fact that taking the chemo again made really very little difference, that my mind was pretty well made up straight away” (s6). Others forgot whether or not they had performed the exercise: “In all honesty, I'd forgotten that, you know. I don't remember I did or not do” (s11). Patient understanding of risk of recurrence statistics The majority of patients understood that the aim of chemotherapy was to reduce the chances of cancer coming back, although people were more likely to mention the figure for net benefit of one chemotherapy drug (2–3%) rather than two (5%): “if I had chemotherapy, my chances of living more than five years would be about one or two percent increased, compared with not doing anything” (s1). “The chance of the cancer coming back was low—without the chemo, was low, anyway… Yes, the chemo might improve it by a couple of percent” (s4). Two patients (both in their 30s) reported higher net benefit of chemotherapy figures (of 5%) and a clearer understanding of the difference between one and two chemotherapy drugs. “I think off the top of my head there was 86% chance of it not recurring within five years if I stopped it now… then for every additional drug chemo that I had was approximately a 2.5% additional chance of survival” (s10). One patient confused net benefit (additional 2–5% avoid recurrence) with overall recurrence rates (20–40%). “(Oncologist's name) said to me, ‘There's just 2% chance.’ When I'd said, ‘No, I do not want chemo,’…. I took for granted that sort of the chances of it coming back is extremely low” (s11). Patient understanding of how and why cancer recurs and how chemotherapy may act to reduce risk of recurrence There was confusion over whether the figures about recurrence referred to another primary cancer or progression of the current episode, or both. “Was it a recurrence of the same cancer or was it from somewhere else?” (s6). While some patients understood that any recurrence would potentially be more aggressive and harder to treat than their colorectal cancer had been, others assumed the recurrence would be in the gut: “When you say it's a recurrence, you presume it will be expected to be from the same site where the primary is… somewhere in the gut…” (s6). “Actually, in the beginning I was thinking that they were talking about that, the cancer will affect my bowel again. But from what she's (the oncologist) saying…it's not only the bowel, but you can get other type of cancer” (s9). “Obviously if it comes back it's not going to be as straightforward… It wouldn't be as simple as just having an operation and removing it, cos it might go to another part of the body” (s7). An additional area of confusion was whether they were currently cancer free or not, and the rationale for giving chemotherapy when no cancer was apparently present or if you were a person ‘prone’ to cancer. “People say that you're ‘cured’ now because the cancer's been removed surgically. But then you've got this risk of recurrence. I found that quite difficult to wrestle in my head because it doesn't feel like I'm cured. It feels like I've got cancer and I've got to see what happens over the next five years…” (s8). “But that has been my problem. Even taking the chemotherapy, if you're talking about, well if cancer, prone to cancer. If cancer will come and chemotherapy wouldn't…. if it were there, I agree it would affect it, but if you will come back I don't think chemotherapy can stop it. That's my understanding; maybe I'm wrong” (s9). But some patients did not think their cancer would come back, and this belief seemed to be the basis of their decision to refuse chemotherapy: “I'm sure that if I thought there was a possibility of it coming back, for me, I would…. I would go for treatment. But I'm just hoping that it doesn't happen” (s12). Discussion We developed a DA for patients with stage II colorectal cancer to facilitate an informed decision process for adjuvant chemotherapy. Patient perceptions of the usefulness of the DA were above scores reported in other patient groups,35 but there was evidence of a lack of understanding of key information among some patients, in particular: their baseline risk of recurrence, the net benefit of combination chemotherapy and the rationale for having chemotherapy when cancer had apparently gone. The method of communicating risk of recurrence information based on Adjuvant!online improves patient understanding of the likelihood of disease-free survival.15 Numeric information about risks enhances patient understanding.38 While visual displays can be very helpful, not everyone can extract the relevant information from them.31 Consistent with previous research,15 we found that despite presenting numeric and visual risk information, there was evidence of a lack of understanding of risk information among some patients. Research on members of the general population has shown that understanding of the graphical risk information given in Adjuvant! online is improved if people are presented with fewer options at any one time,39 and that sequential presentation of treatment options leads to improved understanding of key information about risk of recurrence and net benefit of treatment.40 In this study, patients were presented with all three options at once, and patient understanding of combination chemotherapy might have been enhanced by presenting just two options at a time (eg, no treatment vs one drug, no treatment vs two drugs). In addition, any benefit of single agent chemotherapy has only been observed among patients with bad prognostic factors,4 which may be limited to T4.41 42 Significant efforts are being made to develop better prognostic indicators including the search for gene signature sets that predict response to chemotherapy in this patient group.43 44 This suggests that, in future, even more complex information may need to be presented to patients. The values clarification exercise required patients to select arguments for and against chemotherapy. Although the pros and cons method is the most commonly used method for values clarification purposes, a variety of different methods are available, such as ranking, and social matching (how another person values the characteristics of different options and how similar that person is to you), but owing to a lack of research, there are no explicit recommendations about best practice, and no firm evidence about how and whether values clarification exercises actually inform and help the decision-making process.45 Patient feedback showed that some patients found the exercise useful, and for one clarified the bottom line of the decision: prolonging life versus experiencing side effects. However, for others, the process was less helpful, either because understanding of the information contained in the DA had effectively been delegated to a younger member of the family (niece/son) or because the risk of recurrence information was all they felt they needed to make a decision. There is a lack of research about which patients may benefit from values clarification exercises, and the best methods to engage people with different cognitive ability or lower literacy levels and these issues remain to be explored.45 The values clarification exercise used in this study had a higher than recommended readability level, and a simpler method of eliciting patient preferences will be needed for patients with lower levels of literacy. A further suggested modification of the DA is the addition of information that helps people understand the rationale for being given chemotherapy when cancer has apparently gone. The information booklet developed by Jefford et al10 included diagrams showing how colorectal cancer can spread around the body, and similar information should be included in a revised DA. A key limitation of this study is that patients educated to degree-level or above were over-represented, which may have led to an overestimate of perceived usefulness and acceptability of the DA as measured by the PrepDM and other questionnaire items. The American Cancer Society estimated that a third of people diagnosed with colon cancer in 2013 will be diagnosed at stage II,8 resulting in a substantial number of patients potentially faced with making a decision about adjuvant chemotherapy. It is clear that challenges remain in communicating risk of relapse information to patients and the need to check patient understanding of key information should be integrated in future DAs, for example, with the insertion of quizzes and representation of key information that patients appear to have misunderstood. In addition, much more research is needed into the best way to facilitate patients in making judgements about what is most important to them through a greater understanding of values clarification methods in assisting complex decisions. Twitter: Follow Ioannis Chronakis @chronakis Contributors: AMa (guarantor) had the original idea for the study. AMa, IC and AMi designed the trial variables and formed the investigator group which obtained the funding. AMa, IC, JF and AMi developed the decision aid. AMa and AMi were responsible for overseeing study implementation and data collection. They also carried out the analysis, and drafted the manuscript, which was revised by all authors. Funding: This research was supported by the Royal Free Charity. Disclaimer: All researchers were independent of the funders. The study sponsor and funder played no role in study design; the collection, analysis and interpretation of data; the writing of the report; and the decision to submit the article for publication. Competing interests: Over the duration of the project, IC was funded by a grant from the Royal Free Charity and employed by UCL. The Royal Free Charity funded the development of the EPAD software. IC is now employed by Deontics. He is a founder and a shareholder of the company. Deontics is a commercial provider of clinical decision support software. This company was set up after the study reported in the paper was completed. ePAD (the DA tool) was developed using software now owned by Deontics. Over the duration of the project, JF was employed by UCL and Oxford Universities. The Royal Free Charity funded the development of the ePAD software. JF is now employed part-time by Deontics. He is a founder and a shareholder of the company. Ethics approval: Ethical approval was given by the London Multi-Centre Research Ethics Committee (REC 11/LO/0888). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 TSO. UK Department of Health . Equity and excellence: liberating the NHS . 2010 . 2 General Medical Council . Consent: patients and doctors making decisions together . 2016 . 3 Twelves C , Wong A , Nowacki MP Capecitabine as adjuvant treatment for stage III colon cancer . N Engl J Med 2005 ;352 :2696 –704 . 10.1056/NEJMoa043116 15987918 4 André T , Boni C , Navarro M Improved overall survival with oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment in stage II or III colon cancer in the MOSAIC trial . J Clin Oncol 2009 ;27 :3109 –16 . 10.1200/JCO.2008.20.6771 19451431 5 de Gramont A , Larsen AK , Tournigand C Update on targeted agents for adjuvant treatment of colon cancer in 2006 . Gastrointest Cancer Res 2007 ;1 (Suppl 2 ):S47 –9 .19360148 6 Gray R , Barnwell J , McConkey C Adjuvant chemotherapy versus observation in patients with colorectal cancer: a randomised study . Lancet 2007 ;370 :2020 –9 . 10.1016/S0140-6736(07)61866-2 18083404 7 Kuebler JP , Wieand HS , O'Connell MJ Oxaliplatin combined with weekly bolus fluorouracil and leucovorin as surgical adjuvant chemotherapy for stage II and III colon cancer: results from NSABP C-07 . J Clin Oncol 2007 ;25 :2198 –204 . 10.1200/JCO.2006.08.2974 17470851 8 Fang SH , Efron JE , Berho ME Dilemma of stage II colon cancer and decision making for adjuvant chemotherapy . J Am Coll Surg 2014 ;219 :1056 –69 . 10.1016/j.jamcollsurg.2014.09.010 25440029 9 Sanoff HK , Goldberg RM , Pignone MP Assessing the quality of initial consultations regarding adjuvant colon cancer therapy . Clin Colorectal Cancer 2010 ;9 :113 –18 . 10.3816/CCC.2010.n.016 20378506 10 Jefford M , Gibbs A , Reading D Development and evaluation of an information booklet/decision-making guide for patients with colorectal cancer considering therapy in addition to surgery . Eur J Cancer Care (Engl) 2005 ;14 :16 –27 . 10.1111/j.1365-2354.2005.00504.x 15698383 11 Repetto L , Piselli P , Raffaele M Communicating cancer diagnosis and prognosis: when the target is the elderly patient—a GIOGer study . Eur J Cancer 2009 ;45 :374 –83 . 10.1016/j.ejca.2008.08.020 18838268 12 Harrison JD , Solomon MJ , Young JM Patient and physician preferences for surgical and adjuvant treatment options for rectal cancer . Arch Surg 2008 ;143 :389 –94 . 10.1001/archsurg.143.4.389 18427027 13 Ravdin PM , Siminoff IA , Harvey JA Survey of breast cancer patients concerning their knowledge and expectations of adjuvant therapy . J Clin Oncol 1998 ;16 :515 –21 . 10.1200/JCO.1998.16.2.515 9469335 14 Duric VM , Stockler MR , Heritier S Patients’ preferences for adjuvant chemotherapy in early breast cancer: what makes AC and CMF worthwhile now? Ann Oncol 2005 ;16 :1786 –94 . 10.1093/annonc/mdi370 16126738 15 Belkora JK , Hutton DW , Moore DH Does use of the adjuvant! model influence use of adjuvant therapy through better risk communication? J Natl Compr Canc Netw 2011 ;9 :707 –12 .21715722 16 Gigerenzer G , Edwards A Simple tools for understanding risks: from innumeracy to insight . BMJ 2003 ;327 :741 –4 . 10.1136/bmj.327.7417.741 14512488 17 Reyna VF , Nelson WL , Han PK Decision making and cancer . Am Psychol 2015 ;70 :105 –18 . 10.1037/a0036834 25730718 18 Stacey D , Legare F , Col NF Decision aids for people facing health treatment or screening decisions . Cochrane Database Syst Rev 2014 ;(1) :CD001431 10.1002/14651858.CD001431.pub4 24470076 19 Adsul P , Wray R , Spradling K Systematic review of decision aids for newly diagnosed patients with prostate cancer making treatment decisions . J Urol 2015 ;194 :1247 –52 . 10.1016/j.juro.2015.05.093 26055824 20 Jones RA , Steeves R , Ropka ME Capturing treatment decision making among patients with solid tumors and their caregivers . Oncol Nurs Forum 2013 ;40 :E24 –31 . 10.1188/13.ONF.E24-E31 23269778 21 Sawka AM , Straus S , Gafni A A usability study of a computerized decision aid to help patients with, early stage papillary thyroid carcinoma in, decision-making on adjuvant radioactive iodine treatment . Patient Educ Couns 2011 ;84 :e24 –7 . 10.1016/j.pec.2010.07.038 20732775 22 Sawka AM , Straus S , Rotstein L Randomized controlled trial of a computerized decision aid on adjuvant radioactive iodine treatment for patients with early-stage papillary thyroid cancer . J Clin Oncol 2012 ;30 :2906 –11 . 10.1200/JCO.2011.41.2734 22753906 23 Juraskova I , Butow P , Fisher A Development and piloting of a decision aid for women considering participation in the Sentinel Node Biopsy versus Axillary Clearance 2 breast cancer trial . Clin Trials 2015 ;12 :409 –17 . 10.1177/1740774515586404 26033878 24 Fox J , Johns N , Lyons C PROforma: a general technology for clinical decision support systems . Comput Methods Programs Biomed 1997 ;54 :59 –67 . 10.1016/S0169-2607(97)00034-5 9290920 25 Sutton DR , Fox J The syntax and semantics of the PROforma guideline modeling language . J Am Med Inform Assoc 2003 ;10 :433 –43 . 10.1197/jamia.M1264 12807812 26 Bury J , Hurt C , Roy A LISA: a web-based decision-support system for trial management of childhood acute lymphoblastic leukaemia . Br J Haematol 2005 ;129 :746 –54 . 10.1111/j.1365-2141.2005.05541.x 15953000 27 Taylor P , Fox J , Pokropek AT The development and evaluation of CADMIUM: a prototype system to assist in the interpretation of mammograms . Med Image Anal 1999 ;3 :321 –37 . 10.1016/S1361-8415(99)80027-9 10709699 28 Patkar V , Acosta D , Davidson T Using computerised decision support to improve compliance of cancer multidisciplinary meetings with evidence-based guidance . BMJ Open 2012 ;2 :e000439 10.1136/bmjopen-2011-000439 29 Oken MM , Creech RH , Tormey DC Toxicity and response criteria of the Eastern Cooperative Oncology Group . Am J Clin Oncol 1982 ;5 :649 –55 . 10.1097/00000421-198212000-00014 7165009 30 Collins IM , Kelleher F , Stuart C Clinical decision aids in colon cancer: a comparison of two predictive nomograms . Clin Colorectal Cancer 2012 ;11 :138 –42 . 10.1016/j.clcc.2011.07.001 22018885 31 Trevena LJ , Zikmund-Fisher BJ , Edwards A Presenting quantitative information about decision outcomes: a risk communication primer for patient decision aid developers . BMC Med Inform Decis Mak 2013 ;13 (Suppl 2 ):S7 10.1186/1472-6947-13-S2-S7 32 Elwyn G , O'Connor AM , Bennett C Assessing the quality of decision support technologies using the International Patient Decision Aid Standards instrument (IPDASi) . PLoS ONE 2009 ;4 :e4705 10.1371/journal.pone.0004705 19259269 33 Lewis CL , Golin CE , DeLeon C A targeted decision aid for the elderly to decide whether to undergo colorectal cancer screening: development and results of an uncontrolled trial . BMC Med Inform Decis Mak 2010 ;10 :54 10.1186/1472-6947-10-54 20849625 34 Hersch J , Jansen J , Barratt A Overdetection in breast cancer screening: development and preliminary evaluation of a decision aid . BMJ Open 2014 ;4 :e006016 10.1136/bmjopen-2014-006016 35 Bennett C , Graham ID , Kristjansson E Validation of a preparation for decision making scale . Patient Educ Couns 2010 ;78 :130 –3 . 10.1016/j.pec.2009.05.012 19560303 36 Miles A , Wardle J , Atkin W Receiving a screen-detected diagnosis of cancer: the experience of participants in the UK flexible sigmoidoscopy trial . Psychooncology 2003 ;12 :784 –802 . 10.1002/pon.705 14681952 37 Braun V , Clarke V Using thematic analysis in psychology . Qual Res Psychol 2006 ;3 :77 –101 . 10.1191/1478088706qp063oa 38 Trevena LJ , Davey HM , Barratt A A systematic review on communicating with patients about evidence . J Eval Clin Pract 2006 ;12 :13 –23 . 10.1111/j.1365-2753.2005.00596.x 16422776 39 Zikmund-Fisher BJ , Fagerlin A , Ubel PA Improving understanding of adjuvant therapy options by using simpler risk graphics . Cancer 2008 ;113 :3382 –90 . 10.1002/cncr.23959 19012353 40 Zikmund-Fisher BJ , Angott AM , Ubel PA The benefits of discussing adjuvant therapies one at a time instead of all at once . Breast Cancer Res Treat 2011 ;129 :79 –87 . 10.1007/s10549-010-1193-4 20945090 41 Teufel A , Gerken M , Hartl J Benefit of adjuvant chemotherapy in patients with T4 UICC II colon cancer . BMC Cancer 2015 ;15 :419 10.1186/s12885-015-1404-9 25985893 42 Verhoeff SR , van Erning FN , Lemmens VE Adjuvant chemotherapy is not associated with improved survival for all high-risk factors in stage II colon cancer . Int J Cancer 2016 ;139 :187 –93 . 10.1002/ijc.30053 26914273 43 Maak M , Simon I , Nitsche U Independent validation of a prognostic genomic signature (ColoPrint) for patients with stage II colon cancer . Ann Surg 2013 ;257 :1053 –8 . 10.1097/SLA.0b013e31827c1180 23295318 44 Gao S , Tibiche C , Zou J Identification and construction of combinatory cancer hallmark-based gene signature sets to predict recurrence and chemotherapy benefit in stage II colorectal cancer . JAMA Oncol 2016 ;2 :37 –45 . 10.1001/jamaoncol.2015.3413 26502222 45 Fagerlin A , Pignone M , Abhyankar P Clarifying values: an updated review . BMC Med Inform Decis Mak 2013 ;13 (Suppl 2 ):S8 10.1186/1472-6947-13-S2-S8

PMC005xxxxxx/PMC5372113.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01366510.1136/bmjopen-2016-013665Infectious DiseasesResearch150617061692Incidence of bloodstream infections: a nationwide surveillance of acute care hospitals in Switzerland 2008–2014 Buetti Niccolò 1Atkinson Andrew 1Marschall Jonas 1Kronenberg Andreas 12the Swiss Centre for Antibiotic Resistance (ANRESIS) the Swiss Centre for Antibiotic Resistance (ANRESIS) Auckenthaler R Cherkaoui A Gaia V Dubuis O Egli A Koch D Leib S Luyet S Nordmann P Perreten V Piffaretti J-C Prod'hom G Schrenzel J Widmer AF Zanetti G Zbinden R 1 Department of Infectious Diseases, University Hospital Bern, Bern, Switzerland 2 Institute for Infectious Diseases, University of Bern, Bern, SwitzerlandCorrespondence to Dr Niccolò Buetti; niccolo.buetti@gmail.comJM and AK provided equal contribution. 2017 21 3 2017 7 3 e01366530 7 2016 28 11 2016 17 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Background Bloodstream infections are often associated with significant mortality and morbidity. We aimed to investigate changes in the epidemiology of bloodstream infections in Switzerland between 2008 and 2014. Methods Data on bloodstream infections were obtained from the Swiss antibiotic resistance surveillance system (ANRESIS). Results The incidence of bloodstream infections increased throughout the study period, especially among elderly patients and those receiving care in emergency departments and university hospitals. Escherichia coli was the predominant pathogen, with Enterococci exhibiting the most prominent increase over the study period. Conclusions The described trends may impact morbidity, mortality and healthcare costs associated with bloodstream infections. INFECTIOUS DISEASES ==== Body Strengths and limitations of this study We performed a retrospective analysis of bloodstream infections in Switzerland. We analysed data from ANRESIS, a large national bloodstream infection database (Swiss centre for antibiotic resistance surveillance). Twenty-six acute Swiss hospitals were included. No data on mortality was included. Introduction Bloodstream infections (BSIs) are associated with significant morbidity and mortality. The influence of multiple factors, such as the increases in longevity and comorbidities and the use of more invasive technologies,1 could lead to secular changes in the epidemiology of BSIs. Only few nationwide studies have focused on the incidence of BSIs, and their results were often inconsistent. A study from England described an increasing incidence of BSIs until 2006, followed by decline between 2006 and 2008.1 A significant decrease in BSIs was also observed in a Danish population-based study which included data until 2008.2 On the other hand, a report from Finland showed an increase in episodes of BSI until 2007,3 and a European surveillance study from the EARS-Net suggested an increase of the burden of BSI until 2008 based on trends of the five major bacterial pathogens.4 To the best of our knowledge, recent analyses of population-based studies have not been published. The present work aims to describe temporal trends in BSI incidence between 2008 and 2014 in Switzerland, using the national surveillance system. Methods Data on BSIs were obtained from the Swiss centre for antibiotic resistance surveillance (ANRESIS). Since 2004, the ANRESIS programme has collected all routine microbiological data from a representative group of microbiology laboratories located across Switzerland, with blood culture surveillance introduced in 2006. Each participating laboratory sends bacteraemia results on a regular basis to a central database located at the Institute for Infectious Diseases in Bern, Switzerland. We restricted the data set to acute care hospitals that continuously reported ≥5 BSIs per year without major fluctuations from 2008 through 2014. Unfortunately, not all hospitals send negative blood culture results to the central database. The analysed data set represents approximately one-third (26 hospitals) of all acute care hospitals in Switzerland5 with institutions being evenly distributed across the country. Positive blood cultures were grouped as episodes of BSI if they occurred in a 7-day time window for a given patient. Coagulase-negative Staphylococci (CoNS) and other typical skin commensals (see online supplementary appendix) were considered as an infection if identified in >1 blood culture, or in one blood culture and a catheter tip within the same episode; otherwise they were considered as contaminants and excluded from the analysis. A BSI was considered polymicrobial if different microbial species were identified during the same episode. For a more detailed trend analysis, the following microorganisms groups were selected: Escherichia coli, Staphylococcus aureus, non-E. coli Enterobacteriaceae, polymicrobial BSI, CoNS and Enterococcus spp. A subanalysis of the following categories was performed: department (outpatient, emergency or other department vs other), age (<65 vs ≥65 years), gender (male vs female) and type of hospital (community hospital vs university hospital). 10.1136/bmjopen-2016-013665.supp1supplementary appendix Incidence rates were based on the number of reported bacteraemia isolates per 100 000 population. Since our data are from ∼33% of the acute care hospital beds in Switzerland, we attempted to forecast the incidence in the total population by simple linear extrapolation, based on population statistics.5 6 This assumes both that the community is uniformly served by the available beds, and that the 33% is a representative sample of the secondary and tertiary care infrastructures.5 We investigated the latter in the context of a simple sensitivity analysis, whereby the split of 50%:50% primary to tertiary care beds present in our sample data was altered to reflect that in the whole country (80%:20%). Despite being rather crude in nature, the simple change in proportions provides the most straightforward method for investigating plausible differences in the population as a whole. Unadjusted models for the mean increase in bacteraemia per year were fitted using generalised linear models with Poisson distribution and loglink function, using an offset for the sample population estimate in the appropriate year. Adjusted models were investigated by including gender, age (above or below 65 years), microorganism group (6 groups), hospital type (community or university), hospital department (outpatient, emergency or otherwise) and number of contaminants as covariates. Poisson goodness of fit was tested using the χ2 test. Overdispersion was investigated using the likelihood ratio test with subsequent fitting of a negative binomial model, if appropriate. A significance level of 5% was used with p values determined using Wald-type statistics and 95% CIs. All statistical analyses were performed in R. Results A total of 40 378 BSIs were reported between 2008 and 2014, with a mean incidence rate of 220 episodes per 100 000 population. The incidence rate increased by 14%, from 211 in 2008 to 240 per 100 000 population in 2014 (p<0.001, figure 1), while contaminant episodes (not included in the following analysis) remained constant over this period. The sensitivity analysis described above revealed lower incidence rates of BSI (from 174 in 2008 to 197 per 100 000 population in 2014), but the same 14% increase for this period (p<0.001). Figure 1 Incidence in bloodstream infections 2008–2014: trends overall and for subgroups. Note: *significant increase (p<0.001). BSI, bloodstream infection; dept, department. In adjusted models, among the patients with BSI, we observed an increase in number of BSIs in individuals ≥65 years (from 119 to 144 episodes per 100 000 population, mean increase of 3.0% per year, p<0.001). A supplementary analysis eliminating the bias of population growth in the specific at-risk population also revealed an increase in the incidence (from 705 in 2008 to 796 per 100 000, p<0.001) in older people (≥65 years). An additional analysis for lower ages did not indicate any significant trends (refer to the online supplementary appendix table A). We observed a significant increase of BSI diagnosed in emergency departments (from 58 to 79 per 100 000 population, +5.2% per year, p<0.001) and in university hospitals (from 105 to 125 per 100 000 population, +2.8% per year, p<0.001). In 2014, E. coli was the most frequently detected microorganism (73 per 100 000 population, 30.5% of all episodes). Seventy per cent of all BSIs were caused by E. coli, S. aureus, polymicrobial BSI, CoNS, non-E. coli Enterobacteriaceae or Enterococci. All microorganism groups exhibited an increasing trend (figure 2), although this was only marginal for S. aureus (from 28 to 29 per 100 000, p=0.05). Enterococci displayed a steeper relative trend than other pathogens (from 9 to 14 per 100 000 population, +8.4% per year, p<0.001). Figure 2 Incidence of BSIs caused by E. coli, S. aureus, non-E. coli Enterobacteriaceae, polymicrobial episodes, CoNS and Enterococcus spp. Note: all trends were statistically significant (S. aureus p=0.05, rest p<0.001). BSI, bloodstream infection; CoNS, coagulase-negative Staphylococci; E. coli, Escherichia coli; S. aureus, Staphylococcus aureus. Discussion To the best of our knowledge, no national studies on recent BSI trends have been published in the past few years.1 2 4 In this nationwide BSI surveillance study, we observed a 14% increase in BSI incidence in Switzerland between 2008 and 2014. Published European data from before 2008 show variable trends, with increasing bacteraemia rates in Finland from 2004 to 20073 and England from 2004 to 20061 but decreasing rates in England from 2006 to 20081 and Denmark from 2000 to 2008.2 Overall, the increase in BSI was more prominent in emergency departments. This could be due to a decreasing average duration of hospitalisation7 with discharge of sicker patients who are prone to readmission, for example, in case of surgical wound infection. This phenomenon appeared to be more pronounced in elderly people, even after adjusting for changes in the longevity of the underlying population. An alternative explanation could be the increasing number of immigrants6 who tend to seek medical help not in private practices, but in emergency departments, where blood cultures are drawn more liberally. Moreover, as previously observed,8 we cannot rule out an upward trend in healthcare-associated infections overall or subtle changes in blood culture ordering practice leading to improved diagnosis of BSI. The absolute BSI incidence rate observed in our study was higher than that reported in the majority of population-based studies1 3 9 with the exception of a report from Denmark.2 As shown in the sensitivity analysis, our main analysis overestimated the incidence rate due to an over-representation of university hospitals in the data set. It has been described before that BSI incidence rates are higher in university compared with community hospitals,10 and indeed, following correction for hospital type, incidence rates were comparable to the aforementioned population-based studies. As expected,9 E. coli presents as the most important cause of bacteraemia in Switzerland. Our data also revealed increasing incidences of E. coli and non-E. coli Enterobacteriaceae over time, as observed in England between 2004 and 2008.1 This trend could be explained by several factors. BSIs were increasingly detected in emergency departments, a marker for infections acquired in the community, and in older people, who are known to be at higher risk for E. coli BSI.11 12 Moreover, while the incidence of S. aureus bacteraemia remained stable over time, the marked upward trend in BSIs caused by Enterococci is of particular concern, since they may be associated with higher mortality.13 We were unable to identify a predisposing factor for the increase in enterococcal BSIs. However, an association with intensive care unit departments and vulnerable patients with frequent exposure to the healthcare system is conceivable.13 14 Our study has several limitations. First, we extrapolated incidence based on the epidemiology of only 33% of all Swiss acute hospitals and, consequently, the generalisability of our findings may be somewhat limited. A simple sensitivity analysis reflecting the actual percentage split between university and community hospitals did not reveal any significant differences. Even considering this, the study does not include BSIs diagnosed in long-term care facilities and by general practice physicians. The second potential limitation is that we considered an episode as being polymicrobial if multiple microorganisms were isolated within 1 week, which could influence the BSI incidence estimates. The lack of a widely accepted definition for what constitutes polymicrobial BSI complicates the evaluation of its incidence. Finally, no reliable information regarding the total number of drawn blood cultures is available in the ANRESIS database. Unfortunately, not all hospitals send negative blood culture results to the central laboratories and we are therefore unable to present the total number of blood cultures taken. In conclusion, in this extrapolated Swiss population-based study we describe an increase in BSIs between 2008 and 2014, especially in those detected in emergency room patients. The role of E. coli continues to become more prominent, while Enterococci showed the greatest increase over the study period. The described trends may impact morbidity, mortality and healthcare costs associated with BSI. The authors thank all microbiology laboratories participating in the ANRESIS network: Institute for Laboratory Medicine, Cantonal Hospital Aarau; Central Laboratory, Microbiology Section, Cantonal Hospital Baden; Clinical Microbiology, University Hospital Basel; Viollier AG, Basel; Laboratory Medicine EOLAB, Department of Microbiology, Bellinzona; Institute for Infectious Diseases, University Bern; Microbiology Laboratory, Unilabs, Coppet; Central Laboratory, Cantonal Hospital Graubünden; Microbiology Laboratory, Hospital Thurgau; Microbiology Laboratory Hôpital Fribourgeois, Fribourg; Bacteriology Laboratory, Geneva University Hospitals, Geneva; ADMED Microbiology, La Chaux-de-Fonds; Institute for Microbiology, Université de Lausanne; Centre for Laboratory Medicine, Cantonal Hospital Luzern; Centre for Laboratory Medicine, Cantonal Hospital Schaffhausen; Centre for Laboratory Medicine Dr Risch, Schaan; Central Institute, Hôpitaux Valaisans (ICHV), Sitten; Centre of Laboratory Medicine St Gallen; Institute for Medical Microbiology, University Hospital Zürich; Laboratory for Infectious Diseases, University Children's Hospital Zürich. The authors also thank the steering committee of ANRESIS. They thank Paolo Mombelli for his editorial support. Collaborators: R Auckenthaler, Synlab Suisse, Switzerland; A Cherkaoui, Bacteriology Laboratory, Geneva University Hospitals, Switzerland; V Gaia, Department of Microbiology, EOLAB, Bellinzona, Switzerland; O Dubuis, Viollier AG, Basel, Switzerland; A Egli, Clinical Microbiology Laboratory, University Hospital Basel, Switzerland; D Koch, Federal Office of Public Health, Bern, Switzerland; AK, Institute for Infectious Diseases, University of Bern, Switzerland; S Leib, Institute for Infectious Diseases, University of Bern, Switzerland; S Luyet, Swiss Conference of the Cantonal Ministers of Public Health, Switzerland; P Nordmann, Molecular and Medical Microbiology, Department of Medicine, University Fribourg, Switzerland; V Perreten, Institute of Veterinary Bacteriology, University of Bern, Switzerland; J-C Piffaretti, Interlifescience, Massagno, Switzerland; G Prod'hom, Institute of Microbiology, Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland; J Schrenzel, Bacteriology Laboratory, Geneva University Hospitals, Geneva, Switzerland; AF Widmer, Division of Infectious Diseases and Hospital Epidemiology, University of Basel, Switzerland; G Zanetti, Service of Hospital Preventive Medicine, Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland; R Zbinden, Institute of Medical Microbiology, University of Zürich, Switzerland. Contributors: All authors conceived and designed the study. NB and AA analysed the data. NB, JM and AK wrote the manuscript. All authors contributed to the discussion and reviewed the manuscript. All authors commented and approved the final version of the paper. Funding: The ANRESIS database is funded by the Federal office of Public Health, the Conference of cantonal health ministers and the University of Bern, Switzerland. Competing interests: None declared. Patient consent: Obtained. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Wilson J , Elgohari S , Livermore DM Trends among pathogens reported as causing bacteraemia in England, 2004-2008 . Clin Microbiol Infect 2011 ;17 :451 –8 . 10.1111/j.1469-0691.2010.03262.x 20491834 2 Nielsen SL , Pedersen C , Jensen TG Decreasing incidence rates of bacteremia: a 9-year population-based study . J Infect 2014 ;69 :51 –9 . 10.1016/j.jinf.2014.01.014 24576825 3 Skogberg K , Lyytikäinen O , Ollgren J Population-based burden of bloodstream infections in Finland . Clin Microbiol Infect 2012 ;18 :E170 –6 . 10.1111/j.1469-0691.2012.03845.x 22512663 4 de Kraker ME , Jarlier V , Monen JC The changing epidemiology of bacteraemias in Europe: trends from the European Antimicrobial Resistance Surveillance System . Clin Microbiol Infect 2013 ;19 :860 –8 . 10.1111/1469-0691.12028 23039210 5 Bundesamt für Gesundheit (BAG), (Federal Office of Public Health), Spitalstatistiken. (Hospital statistics) 2014 http://www.bag.admin.ch/index.html . 6 Bundesamt für Statistik, (Federal Office of statistics), migration und integration 2015 http://www.bfs.admin.ch/bfs/portal/de/index.html 7 Hospital statistics 2015 http://www.hplus.ch/de/zahlen_fakten/h_spital_und_klinik_monitor/akutsomatik (accessed 11 Jun 2016 ). 8 Lai CC , Chen YH , Lin SH Changing aetiology of healthcare-associated bloodstream infections at three medical centres in Taiwan, 2000-2011 . Epidemiol Infect 2014 ;142 :2180 –5 . 10.1017/S0950268813003166 25116133 9 Laupland KB Incidence of bloodstream infection: a review of population-based studies . Clin Microbiol Infect 2013 ;19 :492 –500 . 10.1111/1469-0691.12144 23398633 10 Rodríguez-Bano J , López-Prieto MD , Portillo MM Epidemiology and clinical features of community-acquired, healthcare-associated and nosocomial bloodstream infections in tertiary-care and community hospitals . Clin Microbiol Infect 2010 ;16 :1408 –13 . 10.1111/j.1469-0691.2009.03089.x 19845694 11 van der Mee-Marquet NL , Blanc DS , Gbaguidi-Haore H Marked increase in incidence for bloodstream infections due to Escherichia coli, a side effect of previous antibiotic therapy in the elderly . Front Microbiol 2015 ;6 :646 10.3389/fmicb.2015.00646 26175721 12 Marchaim D , Zaidenstein R , Lazarovitch T Epidemiology of bacteremia episodes in a single center: increase in Gram-negative isolates, antibiotics resistance, and patient age . Eur J Clin Microbiol Infect Dis 2008 ;27 :1045 –51 . 10.1007/s10096-008-0545-z 18528721 13 Pinholt M , Ostergaard C , Arpi M Incidence, clinical characteristics and 30-day mortality of enterococcal bacteraemia in Denmark 2006-2009: a population-based cohort study . Clin Microbiol Infect 2014 ;20 :145 –51 . 10.1111/1469-0691.12236 23647880 14 Billington EO , Phang SH , Gregson DB Incidence, risk factors, and outcomes for Enterococcus spp. blood stream infections: a population-based study . Int J Infect Dis 2014 ;26 :76 –82 . 10.1016/j.ijid.2014.02.012 24813873

PMC005xxxxxx/PMC5372114.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01336610.1136/bmjopen-2016-013366OphthalmologyResearch150617181719Treatment trends for retinopathy of prematurity in the UK: active surveillance study of infants at risk Adams Gillian G W 1http://orcid.org/0000-0002-0935-3713Bunce Catey 1Xing Wen 1Butler Lucilla 2Long Vernon 3Reddy Aravind 4http://orcid.org/0000-0002-7402-4350Dahlmann-Noor Annegret H 1Joseph Abbott Wagih Aclimandos Gill Adams Ayman Al-Khaier Louise Allen Kayvan Arashvan Jane Ashworth Faye Barampouti Jonathan Barnes Victoria Barrett Sebastian Barry John Adam Bates Tulin Berk Susmito Biswas Andrew Blaikie Rosie Brennan Howard Bunting Jeremy Butcher Tailoi Chan-Ling Jonathan Chan Christopher Child Jessy Choi David Clark David Clark Luke Clifford Ahmad Dabbagh Gervase Dawidek Luna Dhir Karen Drake Richard Edwards Leonard Esakowitz Julia Escardo-Paton Anthony Evans Brian Fleck Vernon Geh Nick George Lawrence Gnanaray Raina Goyal Paul Haigh Joanne Hancox Richard Haynes Dominic Heath Robert Henderson Roxane Hillier Melanie Hingorani Saurabh Jain Sunila Jain David Jones Namir Kafil-Hussain Simon Kelly Nihal Kenawy Tina Kipioti Archana Kulkarni Tim Lavy David Laws Joanna Lawson Jane Leitch Roland Ling Mary Macrae Usman Mahmood Richard Markham Jane Marr Kristina May Eibhlin McLoone Murad Moosa Claire Morton Ali Mount Wisam Muen Alan Mulvihill Vineeta Munshi Mahi Muqit Robert Murray Ranjit Nair William Newman Una O’Colmain Chetan Patel Himanshu Patel Amaya Pedraza Luis Rachel Pilling Narman Puvanachandra Anthony Quinn Dinesh Rathod Ashwin Reddy Alison Rowlands Stephen Scotcher Christopher Scott Rajnish Sekhri Ayad Shafiq Tamsin Sleep Katya Tambe Anamika Tandon Alison Tappin Robert Taylor Maria Theodorou Shery Thomas Graham Thompson Peter Tiffin Aman Ullah Muhammed Patrick Watts Stephanie West Stephanie West Iain Whyte Louisa Wickham Cathy Williams Chien Wong Siobhan Wren Rahila Zakir 1 NIHR Biomedical Research Centre at Moorfields Eye Hospital and UCL Institute of Ophthalmology, London, UK 2 Birmingham and Midland Eye Centre, Birmingham, UK 3 Department of Paediatric Ophthalmology and Strabismus, St James’ University Hospital, Leeds Teaching Hospitals, Leeds, UK 4 Royal Aberdeen Children's Hospital, Aberdeen, UKCorrespondence to Dr Annegret Dahlmann-Noor; annegret.dahlmann-noor@moorfields.nhs.uk2017 21 3 2017 7 3 e0133667 7 2016 9 11 2016 11 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives To estimate the incidence of severe retinopathy of prematurity (ROP) requiring treatment and describe current treatment patterns in the UK. Design Nationwide population-based case ascertainment study via the British Ophthalmic Surveillance Unit and a national collaborative ROP special interest group. Practitioners completed a standardised case report form (CRF). Setting All paediatric ophthalmologists providing screening and/or treatment for retinopathy in the UK were invited to take part. Participants Any baby with ROP treated or referred for treatment between 1 December 2013 and 30 November 2014, treated with laser, cryotherapy, vascular endothelial growth factor (VEGF) inhibitor or vitrectomy/scleral buckling, or a combination. Main outcome measure Incidence of ROP requiring treatment. Results We received 370 CRFs; 327 were included. Denominator from epidemiological data: 8112 infants with birth weight of <1500 g. The incidence of ROP requiring treatment was 4% (327/8112, 95% CI 3.6% to 4.5%). Median gestational age was 25 weeks (IQR 24.3–26.1), and median birth weight 706 g (IQR 620–821). Median age at first treatment was 80 days (IQR 71–96). 204 right eyes (62.39%) had type 1 ROP, and 27 (8.26%) had aggressive posterior ROP. Infants were also treated for milder disease: 9 (2.75%) right eyes were treated for type 2 ROP, and 74 (22.63%) for disease milder than type 1 with plus or preplus, which we defined here as ‘type 2 plus’ disease. First-line treatment was diode laser photoablation of the avascular retina in 90.5% and injection of VEGF inhibitor in 8%. Conclusions ROP treatment incidence in the UK is 2.5 times higher than previously estimated. 8% of treated infants receive intravitreal VEGF inhibitor, currently unlicensed. Research is needed urgently to establish safety and efficacy of this approach. Earlier treatment and increasing numbers of surviving premature infants require an increase in appropriate eye care facilities and staff. Trial registration number NCT02484989. NEONATOLOGY ==== Body Strengths and limitations of this study High case ascertainment using robust epidemiological methodology. High-quality data, based on individual case reports from highly qualified specialists in paediatric ophthalmology. High completion rate of reports. Case ascertainment relying on practitioners notifying cases possibly led to small degree of under-reporting, affecting numerator of incidence estimate. Epidemiological data for denominator (number of low-birthweight infants) not available for full area surveyed and in part extrapolated from reported data. Introduction Retinopathy of prematurity (ROP) is a potentially blinding condition typically affecting preterm neonates of low gestational age and low birth weight.1 ROP is a major cause of preventable blindness in children worldwide: of 15 million children born worldwide in 2010, an estimated 53 000 developed sight-threatening type ROP requiring treatment and 20 000 became blind or severely sight impaired.2 Two-thirds of children suffering sight loss from ROP live in middle-income and moderately developed countries, particularly Latin America.1 2 Incidence of blindness from ROP is lower in highly developed countries (3–13%), where risk factors such as oxygen supplementation and blood oxygen saturation are monitored meticulously, and minimal in poorly developed countries, where premature babies often do not survive.1 In highly and moderately developed countries, the incidence of ROP is increasing, as advances in neonatal management allow more premature infants to survive despite very low gestational age and birth weight.3 4 The current standard treatment for sight-threatening ROP is laser photoablation of the non-vascularised, immature retina. Treatment decisions are based on severity (stage 1–5 or an aggressive posterior form of the condition, aggressive posterior ROP (AP-ROP)) and location (zone 1–3) of the disease and retinal vascular changes (plus disease) as defined by the International Committee for the Classification of ROP.5 Early treatment is now recommended to avoid progression to sight-threatening complications; a clinical algorithm advocates treatment for ‘type 1’ or ‘high-risk prethreshold’ ROP, defined as ‘zone 1 stage 3 ROP with or without plus, or zone 1 stage 1/2 with plus, or zone 2 stage 2/3 with plus disease’.6 However, treatment is occasionally given for earlier forms of ROP, that is, ‘type 2’ or ‘low-risk prethreshold’ ROP (zone 1, stage 1 or 2 without plus, or zone 2, stage 3 without plus disease)6 and ‘mild ROP’ (milder than type 2).7 However, even with timely laser photocoagulation, ROP results in unfavourable structural and visual outcomes in a small, but significant number of children.6 8 9 It usually requires intubation and sedation or general anaesthesia, which may not be safe in systemically unstable infants, and it irreversibly destroys the peripheral retina, reducing the peripheral visual field. Following the successes in the treatment of age-related macular degeneration (AMD) and diabetic retinopathy, recombinant antibodies targeted against vascular endothelial growth factor (VEGF), injected into the vitreous cavity, have recently been used in addition to laser photoablation and as first-line monotherapy to treat the most severe forms of ROP.10 Neither ranibizumab nor bevacizumab is licensed for this indication, though ranibizumab is licensed for intravitreal injection in AMD and few other retinal conditions. Despite being unlicensed,11 bevacizumab is in ophthalmic use, as it incurs only a fraction of the cost of ranibizumab while having equivalent efficacy.12 In ROP, reported benefits of VEGF inhibitors include fast regression of neovascularisation and plus disease, vascularisation of the peripheral retina and the lack of a need for general anaesthesia. However, local and systemic safety is unknown. In infants, VEGF is vital in directing the sequential and orderly development of blood vessels in the retina and systemically.13 14 After injection into the eye, these agents enter the systemic circulation, and there are concerns about dose and timing of administration and potential adverse events, ocular and systemic. Most studies reporting the use of VEGF inhibitors in ROP have used bevacizumab rather than ranibizumab, partly because the larger size of the molecule may make it less likely to induce systemic suppression of VEGF serum levels; there are, however, no pharmacokinetic data to support this concept. The lack of central registries or databases for the treatment of ROP hampers the collection of information about current treatment patterns. To evaluate current practice patterns in the treatment of ROP in the UK, we set up a 1-year national surveillance project. The main objectives of this study are to determine the current incidence of ROP requiring treatment and current treatment preferences. Methods We conducted a prospective epidemiological active surveillance study of ROP treatment in the UK. Study population The UK ROP screening guideline recommends screening of infants <1501 g birth weight and <32 weeks gestational age.15 As denominator, we identified the number of premature births with birth weight of <1500 g in the surveyed area from the Office for National Statistics of England and Wales, http://www.statistics.org.uk. Scotland and Northern Ireland do not report on birth weight, so we estimated the number of low-birthweight (LBW) babies in these areas by assuming that the proportion of live births who were LBW was the same as that observed in England. The latest available birth figures from the Office for National Statistics show that in 2014, there were 661 496 live births in England and 33 544 in Wales; of these, respectively, 6987 and 322 had a birth weight of <1500 g. The number of live births in Northern Ireland and Scotland was 24 394 and 56 725, respectively, so assuming a similar proportion of LBW babies as observed in England, would have resulted in 258 and 545 babies, respectively. The total number of live births with birth weight <1500 g would then be 8112. Inclusion criteria: any baby with ROP treated or referred to another unit for treatment between 1 December 2013 and 30 November 2014, with treatment either in the form of laser therapy, cryotherapy, VEGF inhibitor or vitrectomy/scleral buckling, or a combination of these treatments. Exclusion criteria: any infant not fulfilling the above inclusion criteria. Data collection Incident cases were identified through the existing reporting system set up by the British Ophthalmic Surveillance Unit (BOSU). From December 2013 to December 2014, BOSU mailed cards to all consultant ophthalmologists and associate specialists in the UK once a month, with an invitation to report new cases of treated ROP, defined as above. On receipt of case notifications, we mailed a standardised case report form (CRF) collecting clinical data to the reporting ophthalmologists. We also set up an electronic special interest group (SIG), through which clinicians could inform the research team directly of new cases and send completed CRFs. Definition of ROP severity groups The CRF asked clinicians to specify severity (stage) and location (zone) of ROP based on the International Classification of ROP (ICROP).5 6 We then categorised the data into levels of severity as defined in previous publications6 7 (table 1). However, not all possible scenarios of zone/stage/plus disease status are covered by these classifications. A particular problem is zone 3 disease with plus and zone 2 stage 1 with plus; we categorised these as ‘type 2 plus disease’, which is an addition to existing classifications. A second problem is that preplus disease was not yet defined at the time of the ICROP when type 1 and type 2 disease were described.6 As preplus disease is considered to carry a high risk of progression,16 and as close monitoring is recommended, we also categorised cases of preplus disease as ‘type 2 plus’, with the exception of zone 1 stage 3 disease, which should be treated regardless of plus disease status and is categorised as type 1 disease (table 1). Table 1 Severity of retinopathy classification Zone Stage Plus Severity category Reference AP-ROP 5 1 3 No plus Type 1 6 1 3 Preplus This study 1 3 Plus disease 6 1 2 Plus disease 6 1 1 Plus disease 6 2 3 Plus disease 6 2 2 Plus disease 6 2 1 Plus disease Type 2 plus This study 3 3 Plus disease 3 2 Plus disease 3 1 Plus disease 1 2 Preplus 1 1 Preplus 2 3 Preplus 2 2 Preplus 2 1 Preplus 3 3 Preplus 3 2 Preplus 3 1 Preplus 1 2 No plus Type 2 6 1 1 No plus 2 3 No plus 2 2 No plus Mild 7 2 1 No plus 3 3 No plus 7 3 2 No plus 3 1 No plus 7 4 Partial retinal detachment 5 5 Total retinal detachment 5 AP-ROP, aggressive posterior retinopathy of prematurity. Confounders We reviewed data to exclude duplication of cases arising from children being transferred between neonatal units or consultants, and from reports received via BOSU and via the SIG routes. Statistical analysis Data from the CRFs were entered onto an electronic Red Cap (Research Electronic Data Capture) database. A random sample of forms were inspected to ensure data quality. After data lock, data were transferred into Stata V.13.0 for analysis. Characteristics of infants requiring ROP treatment were summarised using means and SDs for approximately Gaussian continuous variables and medians and IQRs for non-Gaussian continuous variables. Categorical variables are reported as numbers and proportions. Results Participants—numerator: case ascertainment and inclusion During the observation period, BOSU recorded a card return rate of 77.2%. Clinicians notified BOSU of 270 cases (figure 1). We asked clinicians to complete CRF for 268 babies, excluding one who did not meet the inclusion criteria. To reduce bias from under-reporting, we set up a UK ROP-SIG. Members share an electronic mailing list and can report cases of ROP treatment to the research team electronically; this route led to communication of 165 cases. In total, we received 370 completed forms. We excluded one, as treated outside the observation period, and 42 duplicate reports (same child, reported by different clinicians/units). We included 327 cases in the analysis. Figure 1 Case ascertainment, data collection and analysis flow chart (modified from CONSORT, www.consort-statement.org). Treatment incidence Based on the above, the incidence of treatment for ROP during the observation period was 327/8112 or 4% (95% CI 3.6% to 4.5%). Patient characteristics Of the included patients, 57.8% were male (table 2); 69.7% were white, 13.8% Asian, 5.5% black, 5.2% mixed and 5.8% other; and 72.8% were singletons, 24.5% twins and 2.1% triplets. Median (IQR) gestational age at birth was 25 weeks (24.3–26.1), and median (IQR) birth weight was 706 g (620–821). Median (IQR) age at first ROP treatment was 80 days (71–96). Table 2 Demographic, pregnancy and neonatal details of infants requiring ROP treatment Number identified 327 Gender  Female 138 (42.2%)  Male 189 (57.8%) Ethnic group  Asian 45 (13.8%)  Black 18 (5.5%)  Mixed 17 (5.2%)  Other 19 (5.8%)  White 228 (69.7%) Single/multiple birth  Singleton 238 (72.8%)  Twin 80 (24.5%)  Triplet 7 (2.1%)  Other 2 (0.6%) Birth weight in g, median (IQR) 706 (620–821) Gestational age at birth in weeks, median (IQR) 25 (24.3–26.1) Age at first ROP treatment in days, median (IQR) 80 (71–96) ROP, retinopathy of prematurity. Indications for treatment In the following, we report figures for the right eye; figures for the left eye are similar. At first treatment, 204 right eyes (62.39%) had type 1 ROP, and 27 (8.26%) had AP-ROP (table 3). Type 2 plus ROP was present in 74 right eyes (22.63%), and type 2 in 9 (2.75%). Six (1.83%) had mild ROP. One infant had bilateral, and two had unilateral retinal detachments at first treatment (table 3). Table 3 Severity of retinopathy in right and left eyes on the day of first treatment Severity category Number of right eyes treated (% of 327) Number of left eyes treated (% of 327) Same severity in both eyes (% of 327) AP-ROP 27 (8.26) 27 (8.26) 27 (8.26) Type 1 204 (62.39) 202 (61.77) 174 (53.21) Type 2 plus 74 (22.63) 69 (21.10) 43 (13.56) Type 2 9 (2.75) 9 (2.75) 6 (1.83) Mild 6 (1.83) 9 (2.75) 2 (0.61) Unknown 1 (0.31) 2 (0.61) 0 Partial retinal detachment 1 (0.31) 3 (0.92) 1 (0.31) Total retinal detachment 0 0 0 AP-ROP, aggressive posterior retinopathy of prematurity. Primary treatment In 90.5% of right eyes, the first treatment administered was diode laser photoablation of the avascular retina (table 4). One eye received cryotherapy and laser combined (0.3%). Twenty-six infants (8%) received bilateral VEGF inhibitor injections as primary treatment. One child (0.3%) received laser in one and VEGF inhibitor injection in the other eye in the same treatment session, as a vitreous haemorrhage precluded the view of the retina in one eye. Data were missing for three right (0.9%) and six left eyes (1.8%). Table 4 Details of primary treatment First treatment n=327 Right eye n (% of 327) Left eye n (% of 327) Same treatment modality in both eyes n (% of 327) Diode laser 296 (90.5%) 294 (89.9%) 291 (89.0%) VEGF inhibitor injection 26 (8.0%) 26 (8.0%) 26 (8.0%) Cryotherapy and laser 1 (0.3%) 0 VEGF inhibitor injection plus laser* 1 (0.3% 1 (0.3%) 0 Missing data 3 (0.9%) 6 (1.8%) 1 (0.3%) *Infant who received VEGF inhibitor in one eye and laser treatment in the other in the same treatment session. VEGF, vascular endothelial growth factor. Discussion We present the first systematic evaluation of ROP requiring treatment and current treatment preferences in the UK since the introduction of new treatment recommendations and since the first, unlicensed, use of VEGF inhibitors for this condition. The primary objectives of this study were to estimate the current incidence of ROP requiring treatment in the UK and physicians' current preference for treatment modalities, including their use of VEGF inhibitors. Over a 12-month period, we identified 327 premature infants treated for ROP, of whom 90.5% received the current standard treatment, diode laser photoablation, and 8% received VEGF inhibitors. Using epidemiological figures of LBW infants in England and Wales and an estimate of this figure for Scotland and Northern Ireland, we calculated that the incidence of ROP treatment in infants born with birth weight under 1500 g is 4%. The principal limitation of our study is the case ascertainment methodology, which informs the numerator in our calculation of treatment incidence. The BOSU active surveillance system relies on practitioners notifying a central research office of new cases and to complete CRFs. We sought to maximise case ascertainment by setting up a UK ROP-SIG to optimise stakeholder engagement. The BOSU card return rate (response rate) for the observation period was 77.2%. SIG provided notification of 165 cases. Most but not all of the cases notified via the SIG were formally reported by CRFs. Over the same period, the National Neonatal Audit Programme (NNAP), to which most neonatal units in England and Wales contribute, recorded 321 infants receiving treatment for ROP (personal communication, Daniel Grey, Data Analyst, NNAP). The geographic area covered by our study included England and Wales, and Scotland and Northern Ireland. We recorded 292 infants reported by units in England and Wales, 19 from Scotland and 16 from Northern Ireland. There are two possible explanations for the lower number of treatment cases in England and Wales we observed in comparison to NNAP: either our study delivered an underestimate, or NNAP data are an overestimate. During our study period, there was a 77.2% response rate to BOSU. It would seem plausible that cards were more often returned when babies were observed, but we have no evidence to support this. It is possible therefore that cases were omitted. However, our electronic SIG picked up babies who were not reported via the BOSU cards, so we believe that this would mitigate any under-reporting by BOSU. An alternative explanation for the discrepancy is that data entry onto the NNAP database by non-ophthalmic staff may erroneously have recorded ROP screening visits as ROP treatment episodes, leading to an overestimate of treatment numbers. As denominator, we selected infants born with a birth weight of less than 1500 g. ROP screening criteria include a second item to define the infant at risk, that is, birth before 32 weeks gestational age.17 However, figures by gestational age are not routinely captured and so our denominator will have excluded the small number of babies who are born before 32 weeks but weigh more than 1500 g. This approach is consistent with other publications.18 Unfortunately, figures of infants with LBW are not reported throughout the area we surveyed, and we estimated the proportion of LBW infants in Scotland and Ireland from those reported for England. Compared with previous reports, our study provides evidence of an increase in the number of infants treated for ROP in the UK and evidence of a change in treatment pattern. A previous BOSU study detected 223 preterm babies with stage 3 ROP over a 16-month period between 1 December 1997 and 31 March 1999 of whom just 59% were treated—76% of these with laser photocoagulation and 22% with cryotherapy.19 This study used a mixed methodology which included a parental survey, requiring individual consent for study participation. The authors suspected that this led to under-reporting of cases, and they did not present an incidence figure. A report from a single centre in Scotland observed that between 1995 and 2004, 5% of premature infants meeting the UK screening criteria required treatment.20 Two UK-based studies, which used large national databases of routinely collected data, NNAP and Hospital Episode Statistics,21 18 reported significantly lower incidences of ROP requiring treatment (1.5–2%) than our study (4%), probably due to under-reporting, as ophthalmologists were not involved in data collection.22 23 The strengths of our approach are high case ascertainment and data completion. Our treatment incidence figure is also comparable with reports from other countries, which report treatment of 1–5% of infants at risk.24–28 An initially surprising finding was that 27% of eyes received treatment for ROP milder than type 1 and AP-ROP, the current treatment threshold.17 6 8 9 However, the overwhelming majority of these had preplus disease, which we propose to categorise as ‘type 2 plus’ disease: 74 right and 69 left eyes (22.63% and 21.10%, respectively). In fact, previous reports indicate that 70% of those with preplus disease at 33–34 weeks gestational age may progress to requiring laser treatment.16 Treatment for disease earlier than type 1 is also not a phenomenon limited to our setting: a recent report of practice patterns of US-based ROP experts found that a substantial proportion of premature infants, 9.5%, were treated for ROP milder than type 1, as practitioners were concerned about as vascular dragging, tractional membranes, vitreous haemorrhages or persistent ROP, all of which are not captured by the current ICROP classification.7 Our study design did not collect this level of detailed information, but it is clear that in the UK, practitioners are concerned particularly about preplus disease, and provide early treatment. Finally, our study identified 26 infants who received VEGF inhibitor injections as primary and only treatment, and one who received combined VEGF inhibitor and laser treatment. This is the first published national figure for this new treatment modality; it may inform the design of future ROP treatment trials. At present, the medium and longer term safety and efficacy of this approach, that is, risk of recurrence of ROP and effects on cognitive and physical development, are not known, although concerns about neurodevelopmental outcomes and late recurrence are emerging.29 30 Some authors advocate their use for zone 1 ROP or AP-ROP, or in cases of systemically unwell infants, poor visibility of the retina, or after failed laser photocoagulation.10 Research is urgently needed to provide information about safety and efficacy of these treatments, and to integrate them into current treatment algorithms. The treatment incidence we report is likely to be generalisable to other highly developed countries, where facilities and staff are available to provide high-intensity care for infants born prematurely. We wish to thank the British Ophthalmic Surveillance Unit, in particular, Mr Barny Foot, for advice on study design and collection of case notifications. We thank Ms Anneka Tailor for liaising with practitioners across the UK to facilitate data collection and completion of reports, and for maintaining the study database. We thank Mr Zabed Ahmed for setting up a comprehensive electronic database. Collaborators: Members of the UK ROP Special Interest Group: Abbott Joseph, Aclimandos Wagih, Adams Gill, Al-Khaier Ayman, Allen Louise, Arashvan Kayvan, Ashworth Jane, Barampouti Faye, Barnes Jonathan, , Barrett Victoria, , Barry John Sebastian, Bates Adam, Berk Tulin, Biswas Susmito, Blaikie Andrew, Brennan Rosie, Bunting Howard, Butcher Jeremy, LB, Chan-Ling Tailoi, Chan Jonathan, Child Christopher, Choi Jessy, Clark David, Clark David, Clifford Luke, Dabbagh Ahmad, AHD-N, Dawidek Gervase, Dhir Luna, Drake Karen, Edwards Richard, Esakowitz Leonard, Escardo-Paton Julia, Evans Anthony, Fleck Brian, Geh Vernon, George Nick, Gnanaray Lawrence, Goyal Raina, Haigh Paul, Hancox Joanne, Haynes Richard, Heath Dominic, Henderson Robert, Hillier Roxane, Hingorani Melanie, Jain Saurabh, Jain Sunila, Jones David, Kafil-Hussain Namir, Kelly Simon, Kenawy Nihal, Kipioti Tina, Kulkarni Archana, Lavy Tim, Laws David, Lawson Joanna, Leitch Jane, Ling Roland, VL, Macrae Mary, Mahmood Usman, Markham Richard, Marr Jane, May Kristina, McLoone Eibhlin, Moosa Murad, Morton Claire, Mount Ali, Muen Wisam, Mulvihill Alan, Munshi Vineeta, Muqit Mahi, Murray Robert, Nair Ranjit, Newman William, O’Colmain Una, Patel Chetan, Patel Himanshu, Pedraza Luis Amaya, Pilling Rachel, Puvanachandra Narman, Quinn Anthony, Rathod Dinesh, AR, Reddy Ashwin, Rowlands Alison, Scotcher Stephen, Scott Christopher, Sekhri Rajnish, Shafiq Ayad, Sleep Tamsin, Tambe Katya, Tandon Anamika, Tappin Alison, Taylor Robert, Theodorou Maria, Thomas Shery, Thompson Graham, Tiffin Peter, Ullah Muhammed Aman, Watts Patrick, West Stephanie, West Stephanie, Whyte Iain, Wickham Louisa, Williams Cathy, Wong Chien, Wren Siobhan, Zakir Rahila Contributors: All authors meet the ICMJE criteria and have completed the ICJME authorship form. GGWA, CB, LB, AHD-N, VL and AR developed the study protocol. LB, AHD-N, GGWA and CB secured funding. All authors reviewed and discussed and interpreted the data acquired. WX carried out data analysis. AHD-N drafted the manuscript, which was then critically reviewed and modified by all authors. All authors, external and internal, had full access to all of the data (including statistical reports and tables) in the study and can take responsibility for the integrity of the data and the accuracy of the data analysis. The lead author, AHDN, affirms that the manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained. Funding: The study received funding from the Moorfields Special Trustees (grant ST 14 01 D) and the Birmingham Eye Foundation. The research was supported by the National Institute for Health Research (NIHR) Biomedical Research Centre based at Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology. Disclaimer: The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health. Competing interests: All authors have completed the ICMJE uniform disclosure form and declare that this work was funded by Moorfields Eye Charity and the Birmingham Eye Foundation. There have not been any financial relationships with any organisations that might have an interest in the submitted work in the previous 3 years and no other relationships or activities that could appear to have influenced the submitted work. Ethics approval: The study was approved by the Research Ethics Committee North of Scotland, Aberdeen (13/NS/0059). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Gilbert C , Muhit M Twenty years of childhood blindness: what have we learnt? Community Eye Health 2008 ;21 :46 –7 .19030129 2 Blencowe H , Lawn JE , Vazquez T Preterm-associated visual impairment and estimates of retinopathy of prematurity at regional and global levels for 2010 . Pediatr Res 2013 ;74 (Suppl 1 ): 35 –49 . 10.1038/pr.2013.205 24366462 3 Gilbert C , Fielder A , Gordillo L Characteristics of infants with severe retinopathy of prematurity in countries with low, moderate, and high levels of development: implications for screening programs . Pediatrics 2005 ;115 :e518 –25 . 10.1542/peds.2004-1180 15805336 4 Moore T , Hennessy EM , Myles J Neurological and developmental outcome in extremely preterm children born in England in 1995 and 2006: the EPICure studies . BMJ 2012 ;345 :e7961 10.1136/bmj.e7961 23212880 5 International Committee for the Classification of Retinopathy of P . The International Classification of Retinopathy of Prematurity revisited . Arch Ophthalmol 2005 ;123 :991 –9 . 10.1001/archopht.123.7.991 16009843 6 Early Treatment For Retinopathy Of Prematurity Cooperative G . Revised indications for the treatment of retinopathy of prematurity: results of the early treatment for retinopathy of prematurity randomized trial . Arch Ophthalmol 2003 ;121 :1684 –94 . 10.1001/archopht.121.12.1684 14662586 7 Gupta MP , Chan RV , Anzures R Practice Patterns in Retinopathy of Prematurity Treatment for Disease Milder Than Recommended by Guidelines . Am J Ophthalmol 2016 ;163 : 1 –10 . 10.1016/j.ajo.2015.12.005 26705094 8 Good WV , Hardy RJ , Dobson V The incidence and course of retinopathy of prematurity: findings from the early treatment for retinopathy of prematurity study . Pediatrics 2005 ;116 :15 –23 . 10.1542/peds.2004-1413 15995025 9 Good WV , Hardy RJ , Dobson V , Early Treatment for Retinopathy of Prematurity Cooperative G . Final visual acuity results in the early treatment for retinopathy of prematurity study . Arch Ophthalmol 2010 ;128 :663 –71 . 10.1001/archophthalmol.2010.72 20385926 10 Mintz-Hittner HA , Kennedy KA , Chuang AZ Efficacy of intravitreal bevacizumab for stage 3+ retinopathy of prematurity . N Engl J Med 2011 ;364 :603 –15 . 10.1056/NEJMoa1007374 21323540 11 National Institute for Clinical Excellence . Bevacizumab (Avastin) for eye conditions : Report of findings from a workshop held at NICE on 13 July 2010. 12 Wittenberg R , Hu B , Comas-Herrera A Care for older people: projected expenditure to 2022. on social care and continuing health care for England older population. Nuffield Trust in partnership with LSE PSSRU, London, UK. http://eprints.lse.ac.uk/60887/ 13 Thebaud B Angiogenesis in lung development, injury and repair: implications for chronic lung disease of prematurity . Neonatology 2007 ;91 :291 –7 . 10.1159/000101344 17575472 14 Rosenstein JM , Krum JM , Ruhrberg C VEGF in the nervous system . Organogenesis 2010 ;6 :107 –14 . 10.4161/org.6.2.11687 20885857 15 Wilkinson AR , Haines L , Head K UK retinopathy of prematurity guideline . Eye (Lond) 2009 ;23 :2137 –9 . 10.1038/eye.2008.128 18836408 16 Wallace DK , Freedman SF , Hartnett ME Predictive value of pre-plus disease in retinopathy of prematurity . Arch Ophthalmol 2011 ;129 :591 –6 . 10.1001/archophthalmol.2011.63 21555612 17 Wilkinson AR , Haines L , Head K UK retinopathy of prematurity guideline . Early Hum Dev 2008 ;84 :71 –4 . 10.1016/j.earlhumdev.2007.12.004 18280404 18 Painter SL , Wilkinson AR , Desai P Incidence and treatment of retinopathy of prematurity in England between 1990 and 2011: database study . Br J Ophthalmol 2015 ;99 :807 –11 . 10.1136/bjophthalmol-2014-305561 25427778 19 Haines L , Fielder AR , Baker H UK population based study of severe retinopathy of prematurity: screening, treatment, and outcome . Arch Dis Child Fetal Neonatal Ed 2005 ;90 : F240 –4 . 10.1136/adc.2004.057570 15846016 20 Dhaliwal C , Fleck B , Wright E Incidence of retinopathy of prematurity in Lothian, Scotland, from 1990 to 2004 . Arch Dis Child Fetal Neonatal Ed 2008 ;93 :F422 –6 . 10.1136/adc.2007.134791 18463118 21 Wong HS , Santhakumaran S , Statnikov Y Retinopathy of prematurity in English neonatal units: a national population-based analysis using NHS operational data . Arch Dis Child Fetal Neonatal Ed 2014 ;99 :F196 –202 . 10.1136/archdischild-2013-304508 24361602 22 Warden C , George J , Nithyanandarajah A Inaccuracy of ROP screening data in National Neonatal Audit Programme report . Eye (Lond) 2014 ;28 :237 –8 . 10.1038/eye.2013.251 24287898 23 Dahlmann-Noor A , Bunce C , Adams G Data accuracy in estimates of treatment incidence for retinopathy of prematurity . Br J Ophthalmol 2006 ;(3) :CD004863 . 24 Bas AY , Koc E , Dilmen U Incidence and severity of retinopathy of prematurity in Turkey . Br J Ophthalmol 2015 ;99 :1311 –4 . 10.1136/bjophthalmol-2014-306286 25868788 25 Slidsborg C , Olesen HB , Jensen PK Treatment for retinopathy of prematurity in Denmark in a ten-year period (1996 2005): is the incidence increasing? Pediatrics 2008 ;121 :97 –105 . 10.1542/peds.2007-0644 18166562 26 Holmström GE , Hellstrom A , Jakobsson PG Swedish national register for retinopathy of prematurity (SWEDROP) and the evaluation of screening in Sweden . Arch Ophthalmol 2012 ;130 :1418 –24 . 10.1001/archophthalmol.2012.2357 23143441 27 Holmström G , Hellström A , Jakobsson P Evaluation of new guidelines for ROP screening in Sweden using SWEDROP—a national quality register . Acta Ophthalmol 2015 ;93 : 265 –8 . 10.1111/aos.12506 25044161 28 van Sorge AJ , Termote JU , Simonsz HJ Outcome and quality of screening in a nationwide survey on retinopathy of prematurity in The Netherlands . Br J Ophthalmol 2014 ;98 :1056 –60 . 10.1136/bjophthalmol-2013-304493 24682181 29 Morin J , Luu TM , Superstein R Neurodevelopmental Outcomes Following Bevacizumab Injections for Retinopathy of Prematurity . Pediatrics 2016 ;137 :pii: e20153218 10.1542/peds.2015-3218 30 Mintz-Hittner HA , Geloneck MM , Chuang AZ Clinical management of recurrent retinopathy of prematurity after intravitreal bevacizumab monotherapy . Ophthalmology 2016 ;123 :1845 –55 . 10.1016/j.ophtha.2016.04.028 27241619

PMC005xxxxxx/PMC5372115.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01333910.1136/bmjopen-2016-013339Health Services ResearchResearch15061704169116961722Patient experience of different regional models of urgent and emergency care: a cross-sectional survey study Foley Conor 1Droog Elsa 1Boyce Maria 1Healy Orla 2Browne John 1 1 Department of Epidemiology and Public Health, University College Cork, Cork, Ireland 2 Department of Public Health, Health Service Executive, IrelandCorrespondence to Dr Conor Foley; c.foley@ucc.ie2017 20 3 2017 7 3 e0133395 7 2016 18 11 2016 11 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives To compare user experiences of 8 regional urgent and emergency care systems in the Republic of Ireland, and explore potential avenues for improvement. Design A cross-sectional survey. Setting Several distinct models of urgent and emergency care operate in Ireland, as system reconfiguration has been implemented in some regions but not others. The Urgent Care System Questionnaire was used to explore service users' experiences with urgent and emergency care. Linear regression and logistic regression were used to detect regional variation in each of the 3 domains and overall ratings of care. Participants A nationally representative sample (N=8002) of the general population was contacted by telephone, yielding 1205 participants who self-identified as having used urgent and emergency care services in the previous 3 months. Main outcome measures Patient experience was assessed across 3 domains: entry into the system, progress through the system and patient convenience of the system. Participants were also asked to provide an overall rating of the care they received. Results Service users in Dublin North East gave lower ratings on the entry into the system scale than those in Dublin South (adjusted mean difference=−0.18; 95% CI −0.35 to −0.10; p=0.038). For overall ratings of care, service users in the Mid-West were less likely than those in Dublin North East to give an excellent rating (adjusted OR 0.57; 95% CI 0.35 to 0.92; p=0.022). Survey items relating to communication, and consideration of patients' needs were comparatively poorly rated. The use of public emergency departments and out-of-hours general practice care was associated with poorer patient experiences. Conclusions No consistent relationship was found between the type of urgent and emergency care model in different regions and patient experience. Scale-level data may not offer a useful metric for exploring the impact of system-level service change. HEALTH SERVICES ADMINISTRATION & MANAGEMENTPRIMARY CAREACCIDENT & EMERGENCY MEDICINEHealth Research Boardhttp://dx.doi.org/10.13039/501100001590[CARG/2012/28] ==== Body Strengths and limitations of this study This is the first nationally representative study exploring service user experience of urgent and emergency care systems. The study context facilitates the comparison of regions that have undertaken extensive system reconfiguration with those that have not. The analysis identified specific aspects of care that were negatively associated with patient ratings of their experience. Few significant differences were found between regions, suggesting either that patient experience was not linked to system reconfiguration, or the survey measure used was not suitable for making comparisons at system level. Introduction Urgent and emergency care consists of all the services which manage patients seeking immediate attention for a health condition, and the processes for referring and transferring patients between services.1 Studies of patient experience in this field have focused on interactions with individual services rather than whole episodes of care.2 3 Such studies fail to account for the fact that patients often use multiple services before receiving definitive care.4 Policy initiatives designed to improve urgent and emergency care often focus on improving system operation. These include the differentiation of hospital functions, specification of ambulance bypass protocols and the introduction of alternative care models.5 Such initiatives generally aim to ensure that patients receive care in a setting that is appropriate to the severity of their condition as quickly as possible, while also preserving local access and reducing the duplication of services within a critical proximity. In a well-functioning system, patients should choose or be directed to the service that is most appropriate to their condition, progress through the system smoothly and feel as though their care is well co-ordinated.1 Urgent and emergency care services have been reconfigured to varying degrees across the Republic of Ireland since 2006 (outlined in table 1). These initiatives have been planned at a regional level and, where implemented, have attempted to direct patients to settings that are appropriate to the severity of their condition. Common features of these programmes include centralisation of specialist urgent and emergency care at a ‘hub’ hospital, and integrated ambulance and general practice (GP) referral protocols for given conditions. Table 1 Regional characteristics and service reconfiguration summary, 2006 to present Region (constituent counties) Population density (population per km2) Service reconfiguration Clinical governance ED changes GP out-of-hours care North East (Meath, Louth, Cavan, Monaghan) 54.6 Region-specific plan partly implemented in 2006–20107 No unified region-level structure Two EDs reconfigured to local injury units. Some centralisation of trauma, coronary and stroke care One regional cooperative in place Dublin North East (North Dublin City and County) 1093 No reconfiguration of services No unified region-level structure No changes One regional cooperative in place, supplemented by doctor on-call service Dublin South (South Dublin and Wicklow) 259.7 Region-specific plan implemented in 20138 No unified region-level structure One ED reconfigured to local injury unit, reduced operating hours in another. Centralisation of trauma, coronary and stroke care to two hospitals but with limited differentiation and integration Multiple cooperatives in place, supplemented by doctor on-call service Dublin Midlands (South-West Dublin, Offaly, Laois, Kildare, Westmeath, Longford) 89.8 Limited reconfiguration of services No unified region-level structure Some centralisation of trauma, coronary and stroke care Multiple cooperatives in place across the region South East (Waterford, Wexford, Carlow, Kilkenny, South Tipperary) 52.6 Limited reconfiguration of services No unified region-level structure Some centralisation of trauma, coronary and stroke care One regional cooperative in place South (Cork, Kerry) 54.6 Region-specific plan largely implemented in 2012–20139 Region-wide structure established Two EDs reconfigured to local injury units, with another closing. Single hub for acute coronary care, severe stroke and trauma cases, with support services provided at other centres One regional cooperative in place Mid-West (Clare, Limerick, North Tipperary) 46 Region-specific plan largely implemented in 2009–201310 Region-wide structure established Three EDs reconfigured to local injury units. Centralisation of stroke, coronary and trauma at hub One regional cooperative in place West (Galway, Mayo, Roscommon, Leitrim, Sligo, Donegal) 31.1 Some regional reconfiguration of services Region-wide structure established One ED reconfigured to local injury unit in 2011. Single hub for acute coronary care, severe stroke and trauma cases, with support services provided at other centres Multiple cooperatives in place across the region ED, emergency department. In this study, we use a recently developed survey methodology4 to compare patient experience in different regions. We hypothesise that service users in regions which have undertaken reconfiguration of urgent and emergency care will report a better experience than patients in regions which have not. We also seek to identify aspects of patient experience where improvements may be possible. Methods Design and setting A cross-sectional survey of the general public across eight regions covering the Republic of Ireland was conducted from March to June 2015. The characteristics of each region and measures taken to reconfigure services are presented in table 1. Two regions (South and Mid-West) have implemented significant reconfiguration of urgent and emergency care. Four regions (West, North-East, Dublin South, South-East) have introduced some measures designed to reconfigure care but these do not cover all services. Two regions (Dublin Midlands and Dublin North-East) have undertaken no major changes since 2006. Further details on urgent and emergency care provision in Ireland are provided in online supplementary file 1. 10.1136/bmjopen-2016-013339.supp1supplementary file Sampling Data were collected by a market research company using computer-assisted telephone interviewing. Random-digit dialling was used to contact landline and mobile telephone numbers. Quota-controlled sampling was used, whereby the market research company was required to contact a sample in each region that was representative of the age and sex profile of the national population. Approximately 1000 interviews were conducted in each region in order to fill these quotas. A screening question was used to identify recent users of urgent and emergency care. Recent users were defined as adults or parents of children (under the age of 16) who self-identified as having used a healthcare service on an urgent or emergency basis in the previous 3 months. An ‘urgent and emergency basis’ was explained to respondents as a problem where help or advice from a healthcare service was needed on the same day. On the basis of a previous study in England, it was estimated that a survey of 1000 respondents per region would yield ∼150 recent users of urgent and emergency care.2 A sample size of 150 service users would provide 80% power at the 5% significance level for regional patient experience comparisons on the entry into the system scale of the Urgent Care System Questionnaire (UCSQ), assuming an SD of 0.8 and with the intention to detect a difference of half an SD on this scale.4 Measures An adapted version of the UCSQ was used to assess service users' experience of their most recent contact with urgent and emergency care services. Cognitive testing and piloting were conducted to assess and adapt the survey instrument to fit the Irish healthcare context prior to the main data collection phase. Cognitive testing and piloting were conducted to assess and adapt the survey instrument to fit the Irish healthcare context prior to the main data collection phase. Cognitive testing involves testing quantitative questions through a series of qualitative in-depth interviews to see how respondents understand, retrieve information for, decide on and ultimately arrive at responses to those questions. While the cognitive testing interviews indicated that the original questionnaire was working well overall, a number of potential changes were highlighted by the process. A cognitive testing report was created which included minor wording changes, clarification for answer responses and additional or more descriptive interviewer instructions and notes were added to the questionnaire. The full version of the UCSQ instrument is available in online supplementary file 2. 10.1136/bmjopen-2016-013339.supp2supplementary file This instrument measures patient experience using 18 items across three scales: entry into the system, patient convenience of the system and progress through the system. Survey items were derived from qualitative research with users of urgent and emergency care and were designed to capture users' recent experience with the system.6 Participants were asked to indicate their agreement with each of the 18 items along a five-point scale from strongly disagree to strongly agree. Scoring was reversed for negatively worded items. A separate item was used to assess participants' overall rating of the care they had received, with six response options ranging from ‘very poor’ to ‘excellent’. Participants were asked to indicate which of the following services they had used in their episode of care: pharmacist, GP (in-hours and out-of-hours), ambulance service, emergency department (ED), local injury unit, mental health service, public health nurse or other (patient-defined). Finally, participants were asked to indicate the condition that had precipitated their contact with urgent and emergency care service from the following options: illness, injury, adverse reaction, infection, mental health issues, other, not stated. Statistical analysis We compared patient experience ratings across the three UCSQ scales and overall ratings of care across regions. Item-level responses were also described at national and regional levels. Linear regression was used to identify regional differences in the three UCSQ scales, with entry into the system, convenience of the system and progress through the system as the outcome variable, adjusted for sex, age group, services used and condition precipitating care episode. Logistic regression was used to identify regional differences in the proportion of patients who rated their overall care experience as excellent, adjusting for sex, age group, services used and condition precipitating care episode. It has been suggested that evaluations of patient experience should focus on optimal responses (eg, strongly agree, excellent) to provide a guide to the amount of improvement that may be possible and this is the approach we have followed in our analyses.11 12 In all regression models, we used the region with the best performance as the base comparator. All statistical tests were two-sided and p values <0.05 were considered to represent a statistically significant result. SPSS V.22 was used to analyse the data. Results In total, 8002 interviews were conducted, with 1205 interviewees self-identified as recent service users. Since a non-probability quota-based sampling method was used, it is not appropriate to report an overall figure for response rate. There was little variation in the characteristics of respondents across regions (see table 2). There was some variation in the services used, especially out-of-hours GP care which varied from 28.9% of participants in the South region to 12.2% in Dublin South. Participants in the Mid-West used the most services on average (2.48), with participants in Dublin North East using the least (2.05). In addition, there were noteworthy differences in the conditions that precipitated contact with urgent and emergency care services. For example, in the South, illness accounted for 73.2% of contacts, compared with 59.8% in the West. Table 2 Demographic characteristics and services used by participants across regions National South Mid-West South East North East West Dublin North East Dublin Midlands Dublin South Participants 1205 149 155 139 166 127 149 156 164 Sex  Female (%) 50.5% 82 (55) 87 (56.1) 74 (53.2) 98 (59) 79 (62.2) 80 (53.7) 84 (53.8) 93 (56.7) Age  0–15 (%) 21.4% 48 (32.2) 49 (31.6) 39 (28.1) 46 (27.7) 29 (22.8) 42 (28.2) 46 (29.5) 39 (23.8)  16–34 (%) 29.1% 45 (30.2) 42 (27.1) 43 (30.9) 47 (28.3) 40 (31.5) 44 (29.5) 36 (23.1) 53 (32.3)  35–54 (%) 27.8% 29 (19.5) 35 (22.6) 30 (21.6) 37 (22.3) 30 (23.6) 39 (26.2) 47 (30.1) 39 (23.8)  55–74 (%) 16.7% 25 (16.8) 26 (16.8) 22 (15.8) 31 (18.7) 22 (17.3) 21 (14.1) 26 (16.7) 29 (17.7)  75+ (%) 5% 2 (1.3) 3 (1.9) 5 (3.6) 5 (3) 6 (4.7) 3 (2) 1 (0.6) 4 (2.4) Services used  Mean number used and range 2.25 (1–12) 2.27 (1–8) 2.48 (1–9) 2.42 (1–8) 2.16 (1–6) 2.31 (1–6) 2.05 (1–6) 2.23 (1–6) 2.15 (1–12)  Pharmacy 544 (45.1) 63 (42.3) 73 (47.1) 61 (43.9) 76 (45.8) 61 (48) 66 (44.3) 69 (44.2) 75 (45.7)  GP in-hours 771 (64) 92 (61.7) 105 (67.7) 88 (63.3) 101 (60.7) 86 (67.7) 87 (58.4) 108 (69.2) 104 (63.4)  GP out-of-hours 250 (20.7) 43 (28.9) 34 (21.9) 36 (25.9) 44 (26.5) 24 (18.9) 21 (14.1) 28 (17.9) 20 (12.2)  Ambulance 91 (7.6) 10 (6.7) 14 (9) 13 (9.4) 9 (5.4) 8 (6.3) 14 (9.4) 12 (7.7) 11 (6.7)  Public ED 376 (31.2) 39 (26.2) 50 (32.3) 53 (38.1) 45 (27.1) 37 (29.1) 52 (34.9) 57 (36.5) 43 (26.2)  Private ED 44 (3.7) 5 (3.4) 11 (7.1) 5 (3.6) 4 (2.4) 3 (2.4) 5 (3.4) 2 (1.3) 9 (5.5)  Public LIU 68 (5.6) 12 (8.1) 10 (6.5) 11 (7.9) 11 (6.6) 1 (0.8) 4 (2.7) 9 (5.8) 10 (6.1)  Private LIU 67 (5.6) 4 (0.4) 10 (1) 4 (0.4) 5 (0.5) 3 (0.3) 18 (1.8) 7 (4.5) 16 (9.8)  Other 122 (10.1) 11 (7.4) 17 (11) 16 (11.5) 18 (10.8) 17 (13.4) 12 (8.1) 18 (11.5) 13 (7.9) Condition  Illness 798 (66.2) 109 (73.2) 99 (63.9) 100 (71.9) 107 (64.5) 76 (59.8) 101 (67.8) 103 (66) 103 (62.8)  Injury 305 (25.3) 31 (20.8) 44 (28.4) 28 (20.1) 49 (29.5) 33 (26) 38 (25.5) 36 (23.1) 46 (28)  Adverse reaction 9 (0.7) 0 (0) 2 (1.3) 2 (1.4) 2 (1.2) 1 (0.8) 1 (0.7) 1 (0.6) 0 (0)  Infection 23 (1.9) 1 (0.7) 0 (0) 2 (1.4) 2 (1.2) 7 (5.5) 2 (1.3) 4 (2.6) 4 (2.4)  Mental health issue 10 (0.8) 1 (0.7) 1 (0.6) 1 (0.7) 0 (0) 1 (0.8) 0 (0) 5 (3.2) 1 (0.6)  Other 41 (3.4) 5 (3.4) 7 (4.5) 4 (2.9) 3 (1.8) 6 (4.7) 5 (3.4) 4 (2.6) 7 (4.3)  Not stated 20 (1.7) 2 (1.3) 2 (1.3) 2 (1.4) 3 (1.8) 3 (2.4) 2 (1.3) 3 (1.9) 3 (1.8) Values are frequencies with percentages in parentheses, unless otherwise stated. ED, emergency department; GP, general practice; LIU, local injury unit. Table 3 outlines the percentage of respondents who endorsed the optimum response option to each item and the scale scores for each UCSQ domain. At the item level, optimal response proportions varied from as low as 17.6% to as high as 45.8%. The three lowest rated items, when adjusted for negative wording, were: ‘I was told how long I'd have to wait’, ‘Services did not seem to talk to each other’ and ‘Services understood that I had responsibilities, like my need to look after my family’. The three highest rated items were ‘I felt that the first service I tried was the right one to help me’, ‘I did not know which service to go to about this problem’ and ‘I was made to feel like I was wasting everyone's time’. There were no instances where the proportion of patients in a region that endorsed the optimal response differed from the national figures by more than or <10%. Additional detail on item-level responses is available in online supplementary file 3. Table 3 Percentage response to optimal option for survey items (scores reversed for negatively worded items) and mean values for UCSQ domains Description National South* Mid-West* South East North East* West Dublin North East Dublin Midlands Dublin South Overall rating of care (%, excellent) 35.9 37.6 29 33.8 36.7 37 42.6 35.3 35.4 Entry into the system—mean (SD) 4.12 (0.75) 4.14 (0.8) 4.12 (0.73) 4.08 (0.72) 4.11 (0.82) 4.08 (0.7) 4.02 (0.83) 4.20 (0.70) 4.21 (0.69) I did not know which service to go to about this problem. 39.3 44 42 36 39 33.9 35.2 42 40.9 I felt that the first service I tried was the right one to help me. 45.8 43.9 41.6 42.4 50.6 41.7 46.6 49 48.5 I felt sometimes I had ended up in the wrong place. 34.3 36.1 28.4 31.9 36 31.2 32.4 37.4 39.6 Progress through the system—mean (SD) 3.96 (0.75) 4.06 (0.74) 3.88 (0.79) 3.99 (0.67) 4.00 (0.79) 3.87 (0.75) 3.91 (0.83) 3.96 (0.78) 3.97 (0.69) My concerns were taken seriously by everyone. 37.8 42.9 39.4 33.1 38.2 33.9 35.1 43.6 35.6 I was made to feel like I was wasting everyone's time. 39.2 42.6 35.1 33.8 41.2 34.6 39.3 42.6 42.9 I had to push to get the help I needed. 29.9 35.1 28.1 25.9 33.1 21.3 31.1 34.4 28 I moved through the system smoothly. 27.1 30.4 27.6 26.3 27.1 19.7 26.2 29.4 28.4 It took too long to get the care needed. 26.5 27 24.2 23.7 31.5 20.5 27.9 26 29.4 I felt that no one took responsibility and sorted out my problem. 30.4 36.5 26.6 25.2 32.5 26 26.5 35.1 33.3 I saw the right people. 30.0 35.4 28.4 31.2 34.5 23 27.2 29.5 29.2 I felt I was given the wrong advice. 32.2 37.4 23.2 30.9 39.6 26 27.4 35.7 35.8 Services did not seem to talk to each other. 20.1 22.6 17.3 20.3 21.4 19.5 18.7 22.6 18.5 At each stage, I was confident in the advice services gave me. 27.4 29.9 24.8 26.6 29.6 21.3 26.5 30.1 29 Patient convenience of the system—mean (SD) 3.76 (0.76) 3.88 (0.72) 3.69 (0.77) 3.69 (0.74) 3.83 (0.78) 3.70 (0.74) 3.73 (0.83) 3.77 (0.75) 3.81 (0.68) Travelling to services I needed was easy. 34.7 37.7 33.3 27 35 33.3 32.2 39.9 28.3 I was told how long I'd have to wait. 17.6 24.5 15.1 12.1 19.4 14 18.2 16 20.3 Services had the information they needed about me. 32.7 37 31.3 31.2 31.5 32.3 30.3 35.5 32.3 I had to repeat myself too many times. 29.6 27.4 28.9 23.9 32.9 27 27.4 34 34.6 Services understood that I had responsibilities, like my need to look after my family. 25.4 23.8 22.9 15.8 30.3 24.8 29.2 28.9 26.1 *Regions that undertook large-scale reconfiguration programmes. UCSQ, Urgent Care System Questionnaire. 10.1136/bmjopen-2016-013339.supp3supplementary file There was little regional variation in UCSQ scale scores. Dublin North East had significantly lower ratings on the entry into the system scale than Dublin South (adjusted mean difference=−0.18; 95% CI −0.35 to −0.10; p=0.038). Use of GP in-hours (coefficient=−0.13; 95% CI −0.24 to −0.03; p=0.013), GP out-of-hours (coefficient=−0.18; 95% CI −0.30 to −0.62; p=0.003), a public ED (coefficient=−0.31; 95% CI −0.41 to −0.20; p<0.0005), a private ED (coefficient=−0.38; 95% CI −0.61 to −0.15; p=0.001) and other services (coefficient=−0.63; 95% CI −0.99 to −0.27; p=0.001) was independently associated with lower entry into the system scale scores. No significant differences between regions were found for patient convenience. Use of a public ED was the only significant independent predictor of variation in this scale and was associated with lower ratings (coefficient=−0.55; 95% CI −0.67 to −0.41, p<0.0005). For progress through the system, the West region was found to have significantly lower ratings than the South region (−0.18; 95% CI −0.36 to −0.01; p=0.043). No other regional differences were identified. Use of GP in-hours (coefficient=−0.23; 95% CI −0.34 to −0.13; p<0.0005), GP out-of-hours (coefficient=−0.20; 95% CI −0.33 to −0.80; p=0.001), public ED (coefficient=−0.42; 95% −0.50 to −0.31; p<0.0005) and other services (coefficient=−0.43; 95% CI −0.80 to −0.06; p=0.024) were significant and negative independent predictors of variation in this scale. Use of an emergency ambulance (coefficient=0.18; 95% CI 0.02 to 0.34; p=0.032) and an episode precipitated by illness (coefficient=0.21; 95% CI 0.02 to 0.39; p=0.027) were independently associated with higher ratings on this scale. Of the total sample, 35.9% rated their care as ‘excellent’. One statistically significant difference in ratings of excellence across regions was found: service users in the Mid-West were less likely than those in Dublin North East to give an excellent rating (adjusted OR 0.57; 95% CI 0.35 to 0.92; p=0.022). In addition, users of a public ED (adjusted OR 0.60; 95% CI 0.44 to 0.82; p=0.001) and GP out-of-hours (adjusted OR 0.66; 95% CI 0.47 to 0.92; p=0.015) were less likely to report an excellent rating of care than those who had not used these services, as were younger participants (adjusted OR 0.81; 95% CI 1.10 to 1.37; p<0.0005). Discussion Overall, 35.9% of service users in this study rated their care as excellent, slightly lower than figures from two studies conducted in the UK where ratings of 38.5% and 41.5% were reported.4 13 UCSQ mean scale scores were also slightly lower than those reported in the UK.1 4 13 The study has not revealed a strong association between regional care models and patient experience scale scores despite notable regional differences in service users' reasons for contacting urgent and emergency care, and the types of services they used. Indeed, when assessing overall ratings of care, service users in the Mid-West, which has recently undertaken a wide-ranging series of changes to urgent and emergency care services, reported significantly lower ratings than those in Dublin North East, which has undertaken little change. The Mid-West is a predominantly rural region with a relatively low population density, where emergency care is centralised to a single centre, while Dublin North East is a densely populated urban area containing three large public EDs. Participants in the Mid-West used more services than those in Dublin North East, implying a longer pathway to definitive treatment. This has previously been shown to be negatively associated with patient experience.1 The Mid-West reconfiguration programme was designed to ensure that patients received care in a setting that is appropriate to the severity of their condition as quickly as possible:10 our study suggests that this may not have been achieved, but a definitive conclusion is not possible without data from the period before reconfiguration occurred. The other regions where reconfiguration took place, the South and North East, also have relatively low population densities but participants in these regions used comparatively fewer services than those in the Mid-West and gave overall ratings of care that were above the national average. In these regions, some centralisation of emergency care has taken place but at least two EDs remain in operation, compared with just one ED in the Mid-West. In the regions where little or no reconfiguration has taken place, several EDs continue to operate, often with little formal integration of governance structures across sites and services. It is possible that ongoing public negativity around the reconfiguration of acute hospital services in the Mid-West may have influenced participant responses. Changes to healthcare provision often meet with public opposition, characterised by a rejection of the justifications for change advanced by policymakers.14 15 For example, previous studies exploring patient experiences of healthcare reconfiguration have failed to identify consistent positive patient responses to changes in gynaecology services16 and GP out-of-hours provision.17 The respondents in this study were particularly dissatisfied with aspects of their care relating to communication and most satisfied with aspects of care around accessing the appropriate service and feeling cared for. These patterns of item endorsement are very similar to those found in a UK study using the same instrument.4 Previous studies of patient experiences of primary care and EDs have also identified communication around waiting times and physician empathy as important predictors of patient perceptions.18 19 In the Irish context, this may be a signal that while there have been many changes directed at individual urgent and emergency care services, there is still insufficient attention to the delivery of a whole system approach to communication between providers. Our study suggests that there is a particular problem with care episodes that involve contact with public EDs and out-of-hours GP care. Strengths and limitations This is the first national study of the patient experience of different urgent and emergency care models. For patient experience studies to be worthwhile, they must produce recommendations on attributes that are modifiable by policymakers, and in this paper, we present data on the aspects of care that are in greatest need of attention.18 The survey method employed in this study is an innovative way to gain access to people who have had an urgent or emergency care episode. Since recruitment did not require the cooperation of individual service providers, we are able to provide a comprehensive analysis of the experience of service users in all parts of Ireland over the reference period. However, the survey method has several potential sources of bias. First, some participants may have accessed services outside of their region of residence, potentially impacting on the accuracy of regional data. Second, patients who died following contact with urgent and emergency care services are not covered by the study. Finally, only individuals with phones were sampled, potentially excluding some service users. The survey instrument identified few significant differences between regions. It is possible that policy initiatives designed to create streamlined regional care systems have not been successful. It is also possible that it is still too early in the implementation cycle to judge the success of these initiatives, in which case the current study provides a useful baseline for future patient experience studies. Alternatively, it is possible that UCSQ scale scores are not sensitive to the signal of regional differences in patient experience. This may be because the instrument covers a very broad spectrum of severity: it may be more useful to focus only on patients with more serious healthcare episodes where a well-functioning system is a strong determinant of experience. Finally, it is important to note that the generalisability of study results may be somewhat limited outside of the Irish context, which has its own unique configuration of urgent and emergency care provision. However, comparing various models of care situated within the same national healthcare context may be viewed as a strength of the study, as it facilitates comparisons that are not always possible when comparing across national boundaries due to varied system compositions and policy contexts. Conclusion and implications There does not appear to be a consistent relationship between the type of regional care model implemented in different parts of Ireland and patient experience. Scale-level data may not offer a useful metric for comparing different models of care as few differences were detected at this level. Instead, it is recommended that policymakers in Ireland use the item-level data produced in this study as a baseline against which future efforts to improve urgent and emergency care experiences are measured longitudinally. Inability to explore ‘before and after’ comparisons hampered the ability of this study to make meaningful inferences about the impact of changes in healthcare provision in specific regions. It is thus recommended that health system administrators intending to implement changes identify and measure relevant outcome factors prior to implementing changes in order to facilitate examination of their impact on factors of interest. Contributors: JB and OH conceived of the study. ED and MB were involved in piloting, adapting and conducting the survey. CF performed the analysis. CF drafted the initial manuscript and all authors contributed to drafting the final manuscript. Funding: This work was supported by the Health Research Board, Ireland (CARG/2012/28). Competing interests: None declared. Ethics approval: Ethical approval for the study was granted by the Clinical Research Ethics Committee of the Cork Teaching Hospitals. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Knowles E , O'Cathain A , Nicholl J Patients’ experiences and views of an emergency and urgent care system . Health Expect 2012 ;15 :78 –86 . 10.1111/j.1369-7625.2010.00659.x 21281414 2 O'Cathain A , Knowles E , Nicholl J Testing survey methodology to measure patients’ experiences and views of the emergency and urgent care system: telephone versus postal survey . BMC Med Res Methodol 2010 ;10 :52 .20534151 3 Health Service Executive . Emergency departments patient profiles, experiences and perceptions Kildare : Health Service Executive , 2007 . 4 O'Cathain A , Knowles E , Nicholl J Measuring patients’ experiences and views of the emergency and urgent care system: psychometric testing of the urgent care system questionnaire . BMJ Qual Saf 2011 ;20 :134 –40 . 10.1136/bmjqs.2009.036574 5 Spurgeon P , Cooke M , Fulop N Evaluating models of service delivery: reconfiguration principles . Report for the National Institute for Health Research Service Delivery and Organisation programme UK , 2010 . 6 O'Cathain A , Coleman P , Nicholl J Characteristics of the emergency and urgent care system important to patients: a qualitative study J Health Serv Res Policy 2008 ;13 (Suppl 2 ):19 –25 . 10.1258/jhsrp.2007.007097 7 Teamwork Management Services . Improving safety and achieving better standards: an action plan for health services in the North East. UK: Teamwork Management Services Limited, 2006 . 8 Hanly D Report of the National Task Force on Medical Staffing . Dublin : Department of Health & Children , 2003 . 9 HSE South . Reconfiguration of Acute Hospital Services, Cork and Kerry . Ireland: HSE South (Cork and Kerry), 2010 . 10 Horwath Consulting Ireland and Teamwork Management Services . Review of Acute Hospital Services in HSE Mid-West; an action plan for acute and community services 2008 . 11 Collins K , O'Cathain A The continuum of patient satisfaction—from satisfied to very satisfied . Soc Sci Med 2003 ;57 :2465 –70 . 10.1016/S0277-9536(03)00098-4 14572851 12 Otani K , Waterman B , Faulkner KM Patient satisfaction: focusing on “excellent” . J Healthc Manag 2009 ;54 :93 .19413164 13 Turner J , O'Cathain A , Knowles E Evaluation of NHS 111 pilot sites: final report UK : University of Sheffield , 2012 . 14 Fulop N , Walters R , Perri Implementing changes to hospital services: factors influencing the process and ‘results’ of reconfiguration . Health Policy 2012 ;104 :128 –35 . 10.1016/j.healthpol.2011.05.015 21719140 15 Oborn E Legitimacy of hospital reconfiguration: the controversial downsizing of Kidderminster hospital . J Health Serv Res Policy 2008 ;13 (Suppl 2 ):11 –28 . 10.1258/jhsrp.2007.007093 18416924 16 Choo T , Deb S , Wilkins J Evaluating the impact of the reconfiguration of gynaecology services at a University Hospital NHS trust in the United Kingdom . BMC Health Serv Res 2014 ;14 :428 10.1186/1472-6963-14-428 25249035 17 Christensen MB , Olesen F Out of hours service in Denmark: evaluation five years after reform . BMJ 1998 ;316 :1502 10.1136/bmj.316.7143.1502 9582141 18 Otani K , Kurz RS , Harris LE Managing primary care using patient satisfaction measures . J Healthc Manag 2005 ;50 :311 –24 ; discussion 324–5 .16268410 19 Thompson DA , Yarnold PR , Williams DR Effects of actual waiting time, perceived waiting time, information delivery, and expressive quality on patient satisfaction in the emergency department . Ann Emerg Med 1996 ;28 :657 –65 . 10.1016/S0196-0644(96)70090-2 8953956

PMC005xxxxxx/PMC5372116.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01138210.1136/bmjopen-2016-011382Qualitative ResearchResearch150617251845Experience of miscarriage: an interpretative phenomenological analysis Meaney S 12Corcoran P 1Spillane N 2O'Donoghue K 23 1 National Perinatal Epidemiology Centre, University College Cork, Ireland 2 Pregnancy Loss Research Group, Dept. of Obstetrics and Gynaecology, University College Cork, Ireland 3 The Irish Centre for Fetal and Neonatal Translational Research (INFANT), University College Cork, IrelandCorrespondence to Dr S Meaney; s.meaney@ucc.ie2017 27 3 2017 7 3 e0113823 2 2016 16 6 2016 20 7 2016 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective The objective of the study was to explore the experiences of those who have experienced miscarriage, focusing on men's and women's accounts of miscarriage. Design This was a qualitative study using a phenomenological framework. Following in-depth semistructured interviews, analysis was undertaken in order to identify superordinate themes relating to their experience of miscarriage. Setting A large tertiary-level maternity hospital in Ireland. Participants A purposive sample of 16 participants, comprising 10 women and 6 men, was recruited. Results 6 superordinate themes in relation to the participant's experience of miscarriage were identified: (1) acknowledgement of miscarriage as a valid loss; (2) misperceptions of miscarriage; (3) the hospital environment, management of miscarriage; (4) support and coping; (5) reproductive history; and (6) implications for future pregnancies. Conclusions One of the key findings illustrates a need for increased awareness in relation to miscarriage. The study also indicates that the experience of miscarriage has a considerable impact on men and women. This study highlights that a thorough investigation of the underlying causes of miscarriage and continuity of care in subsequent pregnancies are priorities for those who experience miscarriage. Consideration should be given to the manner in which women who have not experienced recurrent miscarriage but have other potential risk factors for miscarriage could be followed up in clinical practice. OBSTETRICSQUALITATIVE RESEARCHpregnancy loss ==== Body Strengths and limitations of this study This study uses interpretative phenomenological analysis in order to interpret the experience of miscarriage. Much of the research in relation to pregnancy loss is focused on women's experience. Purposive sampling was undertaken to ensure that both women's and men's experiences were included in this study. Participants from this study were drawn from a large tertiary maternity hospital with a dedicated pregnancy loss clinic and it may be possible that their experiences may differ from those who attend a hospital where such a clinic is not available to them. Miscarriage is the most common adverse outcome in pregnancy. This study highlights the need for the provision of appropriate clinical information as well as supportive information when counselling individuals who experienced miscarriage. Introduction Improvements in the quality of care provided during pregnancy have led to substantial reductions in perinatal and maternal mortality as well as a reduction in other adverse pregnancy outcomes.1 However, these advances have had little effect on the high rate of miscarriage with between 20% and 30% of pregnancies ending in miscarriage.1 2 Until now, much of the research has aimed to identify potential risk factors as the underlying aetiology of miscarriage is not well understood.2 Studies indicate the need for familial and social support following miscarriage as it can be an extremely painful and upsetting experience,3 4 with some women experiencing medical complications.5 6 Quantitative studies indicate that the experience of miscarriage can negatively impact on the men's and women's psychological well-being.4 7–14 These studies also report that the high levels of stress and anxiety experienced7–9 can endure for 6–12 months following miscarriage.8 In contrast, an interventional study in the USA examined the changes of women's feeling over the course of year following miscarriage. Swanson et al15 found that women's responses recorded at 1 year were not significantly different from those recorded at 6 weeks. Considering the high incidence of miscarriage and the reported impact on the emotional well-being of people, there are comparatively few studies that have qualitatively examined the experience of miscarriage. Of these, most studies focused on the women's experience of miscarriage3 16–18 whereby the male experience has been reported based on the women's perspective.16 19 Our study builds on these findings as it aimed to explore the experiences of people who have experienced miscarriage. The purpose of this study was to focus on men's and women's accounts of miscarriage. Through a qualitative analysis, the objective of the study was to gain detailed insight into their expectations of pregnancy as well as their experience of miscarriage diagnosis and management. Methods An interpretative phenomenological analysis (IPA) was undertaken as this approach has its theoretical foundations in phenomenology.20–22 Phenomenology examines perceptions and engages with the way individuals reflect on the experiences they deem significant in their lives.21 Researchers who engage in IPA acknowledge how experience is subjective and is therefore only accessible through interpretation.20 IPA has an ideographic approach which allows the researcher to rigorously explore how these experiences may affect a person.20 IPA has increasingly been used in healthcare research as its ideographic approach facilitates researchers to rigorously explore how specific phenomena may affect a patient and consequently will impact on patient care.20 The study took place in a large tertiary-level Irish maternity hospital. The sample was initially recruited from a list of women who had previously participated in a prospective cohort study regarding miscarriage23 and agreed to be contacted for future research. It is important to note that there are geographical variations for the definition of miscarriage. For the purposes of this study, miscarriage was defined as any pregnancy loss which occurred before 24 weeks gestation in a fetus weighing <500 g. Participants were eligible for the study if they were aged 18 years and older and had experienced one or more miscarriages. Letters were sent to invite women and their partners to participate in the present study by the primary author. If an opt-out form was not returned, the primary author made contact to provide more detailed information about the study. Over the course of the study, six opt-out forms were returned. Three participants were recruited using snowballing techniques, through contact with the Miscarriage Association of Ireland and/or through the bereavement and loss hospital team. Information on the study was forwarded to them and they made contact with the primary author to become involved in the study. None of the participants were known to the researcher. The primary author recruited until data saturation was met. The final sample consisted of 16 participants (10 female and 6 male), 4 of whom were couples (table 1). All the participants signed an informed consent and were interviewed individually, by the primary author (an experienced female qualitative researcher), using a semistructured interview schedule (table 2). All the interviews were conducted in a room onsite in the maternity hospital or a location convenient to the participant, with the exception of one interview that was undertaken by telephone under participant request. Each interview was digitally recorded and contemporaneous notes were taken immediately after each interview. The average interview was 43 min, ranging from 28 to 69 min in length. Table 1 Overview of the sample Participant Parent Couple Number of miscarriage Time in months since most recent loss Gestation of most recent loss Living children Management Patient status 1 Female 1 2 18 6 0 Expectant Public 2 Male 1 2 18 6 0 Expectant Public 3 Female NA 4 13 10 1 Medical Public 4 Female NA 3 7 9 1 Medical Public 5 Female NA 4 18 5 1 Medical Private 6 Female 2 3 19 15 3 Surgical Private 7 Female NA 2 14 11 2 Medical Private 8 Female 3 7 20 14 1 Expectant Public 9 Male 3 7 20 14 1 Expectant Public 10 Female NA 2 18 12 3 Surgical Public 11 Female NA 3 8 10 3 Medical Public 12 Female 4 2 16 9 3 Expectant Public 13 Male NA 3 27 16 3 Medical Public 14 Male 4 2 16 9 3 Expectant Public 15 Male NA 2 96 7 2 Expectant Public 16 Male 2 3 19 15 3 Surgical Private NA, not available. Table 2 Overview of the semistructured interview schedule Area of interest Example questions/prompts Pregnancy Tell me about your experiences of the pregnancy before you miscarried? What were your expectations? Diagnosis Can you please tell me what happened when you miscarried? Who was with you at the time? Can you remember how you felt and what you thought at that time? Management How were you cared for, by the hospital or GP, when you were miscarrying? Support What supports were offered to you in the hospital following your miscarriage? Did you seek support from family and friends? Miscarriage: knowledge and experience Did you have any knowledge of miscarriage before your experience? Did you seek information about miscarriage? From your medical team? Family and/or friends? Websites? Support groups? Future pregnancies Have you been pregnant or considered another pregnancy since the miscarriage? If they had a pregnancy: can you tell me how you felt during that pregnancy? GP, general practitioner. The IPA involved: first listening and re-reading the interviews a number of times to ensure that a general sense of the participants' accounts were acquired. Second, emergent themes were initially identified which were then refined as similar themes were clustered together and subordinate and superordinate themes were identified. Patterns and connections across each individual transcript were examined. Finally, a master table of themes was created after each transcript was integrated into the final analysis. All analyses were carried out using Nvivo V.10 software (QSR International, Doncaster, Australia) by the primary author, a health sociologist. The analyses were then presented to the co-authors for review. Findings Analysis of the data indicated six superordinate themes in relation to the participant's experience of miscarriage: acknowledgement of miscarriage as a valid loss, misperceptions of miscarriage, the hospital environment, management of miscarriage, support and coping, reproductive history and implications for future pregnancies. Acknowledgement of miscarriage as a valid loss As outlined in box 1, participants gave accounts of the devastation they experienced when they were told that they had miscarried. Participants stated that the miscarriages were all experienced differently but it was important to them that their miscarriages, irrespective of gestation, were acknowledged by healthcare professionals in the first instance but also more broadly throughout society. The men in this study felt that they could not experience the loss in the same manner as their partner. Men, however, did reiterate that although they did not experience the miscarriage physically, they were affected emotionally and, like the women, did go through a grieving process. Box 1 Acknowledgement of miscarriage as a valid loss “But the miscarriage itself, I'd say it was until then…and the whole discussion became a very public thing…it was only at that stage that I started to move on from it and that would have been five years, five years later and it was always something that would of upset me…it is hard to know what you are grieving for in a way because it is fleeting, you know the whole experience of being pregnant and then not being pregnant and thinking if I didn't remember this baby then who would.” (P15, male, two miscarriages) “At this stage I think we had attended a couple of the, of the October, the ahhh annual ahhhh [prompt from interviewer; the annual service of remembrance] yeah. And again they are huge out pouring of grief, and of joy for life, but of grief. The people there and the support, but the fact that there are children and parents and grandparents, it just gives a sense that look it doesn't matter what age you are, doesn't matter how wealthy you are, doesn't matter what colour you are, we have all experienced this in our own way and we are all here today to remember that. And, I think for me, that, that was [pause] I haven't missed one yet and I'll still be going for another while yet. You know, that's a lovely outreach and very important.” (P13 male, two miscarriages) “What I think happens, from my own experience, is I don't think it is recognised enough. Like cancer is recognised, god help us we have all had it and all those things. But a loss, it's a different loss when it's a child. They're still a child, they may not be grown but they're still a child” (P1, female, two miscarriages) The acknowledgement of the loss through miscarriage, both by people and through ritual, was of importance. Participants discussed marking or remembering their loss in a variety of ways such as keeping a diary, writing of poems and songs or through the organisation of a funeral or similar ceremony. Some participants spoke of the importance of rituals particularly around the anniversary of the miscarriage in order to continue to acknowledge their loss. A number of participants remarked about the significance of attending the annual service of remembrance, which is organised by the hospital. Misperceptions of miscarriage All participants spoke about how there is not enough discussion relating to miscarriage in the public domain. It was not until the participants had experienced a miscarriage themselves that they were made aware of a history of miscarriage in their own family or with those in their close network of friends. As the participants recalled the experience of their first miscarriage, they recounted how naive they felt; they said that they had no inclination of what it was that they could and ultimately would experience. All participants asserted that increased discussion and awareness of miscarriage should be promoted in a wide range of contexts beginning with health education in school (box 2). Box 2 Misperceptions of miscarriage “I got spotting and I thought surely it's not going to happen again, cause they [people] always say one spontaneous [miscarriage] but you never (pause) but I think with miscarriage people just don't talk about it and they just don't think that it happens to everybody and they don't think it is as common as it is until you talk to other people about it. So I think the perception I would have had was if you had one you're not really likely to have another, that's what I thought.” (P3, female, four miscarriages) “A friend of mine in work is pregnant and it's her first pregnancy and she's not kind of as worried as I am for her. She is oblivious and naïve and while I'm thinking ‘oh god’ she is saying ‘it's fine’.” (P7, female, two miscarriages) “Well when I did have the miscarriage and I said it to people, everyone says ‘oh you know I had one’ and it all comes out from the woodwork and em everyone knows someone who has had a miscarriage. It's so common how could you not but people generally don't talk about it…you don't have the knowledge…people need to know that this can happen.” (P11, female, three miscarriages) The hospital environment and management of miscarriage When the participants spoke about how they were treated in the hospital, they remarked about how divergent an experience it was. The participants stated that any negative experiences in the hospital were related to the administration and/or physical design of the hospital specifically relating to the emergency department and the general clinics. When the women were miscarrying, they first attended the emergency department and found it difficult to be sitting in the waiting area surrounded by women attending with varying symptoms. This was considered one of the hardest aspects of the miscarriage experience as they felt they could not express any emotion (eg, anger or upset) relating to their loss, as they did not wish to distress the other pregnant women. Once admitted for care in the emergency department, they felt that the physical space heightened their distress. With only a curtain between them, the participants recalled hearing other fetal monitors recording an audible heartbeat or conversations among staff, as they received confirmation that they had miscarried. It was felt that having to be in this environment while miscarrying exacerbated the distress experienced by participants, who as a result believed that the hospital administration should be more sensitive to the situation. The women recalled how the early pregnancy clinic provided them with a better environment as there was more privacy which allowed them to more openly express their worry, anxiety or upset (box 3). Box 3 The hospital environment and management of miscarriage “We came straight up here [the maternity hospital] and we went into the emergency place downstairs and we were seen straight away. But there were other patients and staff behind curtains, we were behind ours waiting on the doctor to come round. And there were nurses in there chatting and they were laughing and chatting and jokes and stuff, which they are entitled to have…but I was there with [husband] and we were worried sick that we were losing our baby and the doctor came in and she went through all the things and said ‘No, I can't find a fetal heartbeat, it's gone’. Well, I started roaring crying, I was so upset but all the life was happening all around us, carrying on you know happily in behind the curtains…it was absolutely horrendous. But they organised for me to come back to the early pregnancy clinic, you know I didn't have to speak to anybody we just left the hospital [pause] that was hard.” (P8, female, seven miscarriages) “That was hugely traumatic, cause em, I didn't miscarry the same as the last time it just went on and on and on. I was in and out of here [maternity hospital] every second day for blood tests. The first day they went up a bit and then they went down a bit and then it was kind of, like, and it was just two weeks really of turmoil.” (P4, female, three miscarriages) “I woke up an hour later and I just completely haemorrhaged and I passed out a couple of times. Then I got in the bath and em, I was saying god people should warn people or prepare people if they are going to have miscarriages, cause I didn't know what was happening to me. And em what I excreted was unbelievable cause I was 12 weeks. And I started vomiting and I passed out again and then he rang the hospital. I tried talking to the hospital but I couldn't get the words out I was so weak at this point, you know, and they told me to come straight in. So I did and they killed me [slang: were annoyed with me] when I got in cause they said I should've called the ambulance.” (P3, female, three miscarriages) “The first and the last were spontaneous and the last two I had to take medication but it would of happened inevitably but I, I just wanted to speed up [the miscarriage]…” (P5, female, four miscarriages) Participants experienced anxiety about attending the hospital to get tests over a number of days to confirm the loss of their baby. This was relatively impractical for some with work commitments, but was also difficult as they did not want to reattend the hospital to face the inevitable diagnosis. Many of the women expressed how they had suspected that something was wrong but had no knowledge of what to expect or what is considered normal while miscarrying. Those who miscarried at a later gestation discussed how they were wholly unprepared for the extent of the bleeding when they miscarried (box 3). When women had a choice, most chose to have some form of medical intervention. A number of factors influenced the decision to choose to intervene with women citing other commitments such as having to take care of other children in the family. Support and coping Keeping busy helped participants cope with their loss; this was particularly evident in the participants who already had children. Participants were hesitant to receive formal support by way of counselling and most opted for support from family, friends and/or support groups instead. Men felt that their primary role was to support their partners through the loss and, at times reluctantly, while planning subsequent pregnancies. During subsequent pregnancies, the participants disclosed that high levels of anxiety were experienced. They spoke of how they navigated through the pregnancy focusing on specific gestational weeks as goals, including exceeding the gestation they had experienced their miscarriage(s) at, as well as those coinciding with clinic appointments at the maternity hospital. Many of the participants detailed how these actions meant they could not fully enjoy the experience of being pregnant. Throughout the subsequent pregnancy(ies), participants indicated their satisfaction with the service offered to them in the early pregnancy clinic. It was felt that the staff in the clinic were knowledgeable and cared for the women in a sensitive and understanding manner. The early pregnancy clinic provided reassurance to participants by facilitating appointments whereby the women could be scanned at earlier gestations and more frequently. Of those who were under the care of the specialist bereavement team, both men and women commented on the emotional support provided to them by the specialist midwife. The relationship between the women and midwife, in particular, was considered vital as they felt that these midwives, and the dedicated team, truly cared about their welfare and well-being (box 4). Box 4 Support and coping “I was upset for a good while after but I had the other three [children] to keep me going [slang: busy] with school and everything…I had the D&C the same week as my daughter's communion, so I had to just go ahead and get on with things you know, I had to be happy for her.” (P10, female, two miscarriages) “I'd say we were slightly different in that if we had called it a day at the end of number seven, we both would have been extremely disappointed but you know I think, em, it's more about protection I suppose, I didn't want to have to go through it again. The decision was extremely difficult, now I mean [wife] was very much in favour of going forward and trying again, em, I would have been a bit more reticent I suppose, em a bit more, you know, a bit more nervous about it. Obviously she had major concerns but you I think, I think it was a case of a tough decision but we just went for it.” (P9, male, seven miscarriages) “I love babies and if someone was to say on Friday that you are pregnant and you are going to have to have the baby tomorrow, I would say yeah that's great but I just can't do the, the nine months of worrying.” (P7, female, two miscarriages) “I went up to the [early pregnancy clinic] and they said ‘the next time you get pregnant call us here and come in and we will do a scan, we will do an early scan, we will give you that reassurance’. That made a huge difference, it made a huge difference because it felt like ok someone is not saying ‘yeah, yeah, yeah move it along, move it along, next person’ someone is actually saying ‘we care about you, we know this is hard and the next time you get pregnant we know it's going to be distressful for the first few weeks so come in and we will give you scans’. And they were so good about it and when I did get pregnant it was one of the first calls I did make.” (P12, female, two miscarriages) Reproductive history and implications for future pregnancies Whether there were children in the family before the miscarriage made a difference to how the experience impacted on each individual. Those who already had children were better able to reassure themselves that they could successfully get pregnant and give birth. Participants who did not have children before experiencing a miscarriage recalled their concerns about their health, behaviour and/or fertility (box 5). Box 5 Reproductive history and implications for future pregnancies “We already had a loss, I know they were two, two different losses but I was thinking not again, what is going on, is there something wrong with me, am I ever going to have children.” (P1, female, two miscarriages) “One of the things that I asked for was an appointment with [the specialist in pregnancy loss]to have tests done to see why I was having the miscarriages but I was told I would have to have 3 miscarriages before they would see me and I was kind of thinking, do they not take age into account? You know, you have to have three and I think two is an adequate level at my age. If I was in my twenties maybe you'd manage the three but not at my age.” (P4, female, three miscarriages) Medical investigations, such as karyotyping, are not offered to women unless they have experienced recurrent miscarriage (three consecutive miscarriages).24 Participants expressed frustration that these tests were not offered to them following a second miscarriage. This dissatisfaction was heightened in women who felt that other risk factors, such as advancing maternal age, should be considered (box 5). Discussion The findings of this qualitative study indicate that the experience of miscarriage has a considerable impact on men and women. Findings from this study support what has been reported by others, that there is a need for increased awareness in relation to the frequent occurrence of miscarriage. Miscarriage is a common occurrence, yet as revealed by these participants it is not until a miscarriage was experienced that the participants were made aware of these high rates. A study from the USA also indicated that people believe that miscarriage is a rare complication of pregnancy.25 The participants from this study believed that improvement of information provision would be beneficial in allowing individuals to better prepare for the possibility that their pregnancy could end in miscarriage and, if it does occur, that support is available. Second, given that a cause cannot be determined in as many as 50% of miscarriages, it was felt that having this information in advance may alleviate some of the guilt experienced. Participants emphasised that such information provision should also focus on the physical aspects of miscarrying. These findings mirror those of Moohan et al,26 whereby women felt unprepared when miscarrying spontaneously and were questioning of whether what they had experienced was normal. Wong et al27 support this finding by detailing how miscarriage may be a physically traumatic event as women may experience considerable and sudden pain, loss of blood and may need to be hospitalised. Similar to the longitudinal study by Côté-Arsenault,28 the participants in this study indicated how pregnancy following miscarriage was stressful. There is a need for improved communication between healthcare professionals and patients to better counsel patients through the miscarriage and provide reassurance in subsequent pregnancies. One coping strategy adopted by men and women was focusing on commitments, particularly taking care of other children in the family. In a review of the literature on grief following miscarriage, Brier states that having living children has also been used as an indicator for the importance attached to the pregnancy. This belief is based on the assumption that the absence of living children is associated with a relatively greater desire for children.29 Wong et al27 also highlight how, given this belief, it is also assumed that women with children will be less emotionally distressed and are less likely to receive emotional support from nursing staff. In contrast, the findings from this study illustrated that these participants were affected emotionally and did go through a grieving process irrespective of gestation of the pregnancy loss or whether they had living children or not. The findings also indicated the importance that healthcare professionals acknowledge miscarriage and how appreciative participants were of the support given to them. It has been documented that men and women grieve differently following miscarriage in the literature,30 31 and these findings are also reflected in the accounts of the participants in this study. Similar to Johnson and Puddifoot,31 the men in this study indicated that they were less likely to openly discuss the miscarriage unless prompted by another person with a similar experience. This was also the case with discussing the impact of the miscarriage on them with their partners with the men identifying their primary role as that of a support to their partner. However, as outlined by Brier,29 this could suggest differences in the general expression of emotion and grief rather than affective reactions to miscarriage. Although the men in this study did not actively seek out support, they did reiterate that certain experiences and rituals were helpful for their grieving process as they allowed them to mark and remember their loss. Participants in this study were reasonably satisfied with the care provided to them by the hospital. However, a number of shortcomings with the system were identified. When miscarrying, the first contact with the maternity hospital was with the emergency room. It was felt that waiting for extended periods of time in an area with other pregnant women was particularly difficult and a situation which hospital management should be more sensitive to. Wong et al27 outlined that in previous studies women believed that medical staff do not consider miscarriage as either important or an emergency and considered medical staff insensitive and unsympathetic about accommodation. Our findings build on these results whereby participants identified this insensitivity to be as a result of the hospital setting rather than medical staff. Participants were appreciative of staff, especially those whom they considered to be knowledgeable and those who displayed understanding and compassion. The dedicated early pregnancy clinic was an environment they believed could be further developed to enhance the care currently provided to women when they are miscarrying. Consistent with a number of other studies,1 4 27 all the participants expressed a desire to determine the cause of the miscarriage. Participants expressed dissatisfaction that they were ineligible to have tests to fully investigate the cause of their miscarriage as they had not experienced the requisite three consecutive miscarriages. In our study, this perceived inadequacy in service provision was amplified in women of advancing maternal age. As Brier29 outlines, maternal age can potentially influence an individual's goals with regard to childbearing. Advancing maternal age in combination with a number of losses experienced by a woman may impact on the duration and intensity of grief experienced. The women in this study expressed dissatisfaction with their ineligibility for investigations, maintaining that staff should appreciate that although they had not experienced recurrent miscarriage, there were other risk factors, such as their age, to be considered. As part of the analysis, it is important to consider any factors which may influence the results. The participants in the study all made reference to the dedicated early pregnancy loss clinic. This clinic is staffed by a dedicated pregnancy loss team. Such a dedicated clinic is not available in all hospitals. Thus, the presence of such a team in the hospital may have raised awareness about miscarriage among other medical staff and influenced how they cared for the participants sampled here. It is important to note, that although a qualitative methodology was deemed appropriate for this study, the findings of such studies are context-specific. The experiences of the women and men in this study may or may not reflect the experiences of those who attend other units with differing resources and practices. Notwithstanding these limitations, given the level of agreement with other studies, we feel that these results add additional insight into the experiences of miscarriage. Conclusions This study highlights that a thorough investigation of the underlying causes of miscarriage and continuity of care in subsequent pregnancies are priorities for those who experience miscarriage. The provision of appropriate clinical information as well as supportive information when counselling individuals who are experiencing a miscarriage is important. Consideration should be given to the manner in which women who have not experienced recurrent miscarriage but have other potential risk factors for miscarriage could be followed up in clinical practice. The authors are grateful to the participants for participating in the study and giving of their time freely. Twitter: Follow Sarah Meaney @sarahmeaney5 Contributors: SM and KOD contributed to and were responsible for the conception and design of the study. SM and NS were responsible for data collection. SM was responsible for transcription, data analysis and the initial drafting of the article. SM, PC, NS and KOD contributed to revising the manuscript critically for important intellectual content; final approval of the version to be published and the decision to submit the article for publication. Funding: This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Competing interests: None declared. Ethics approval: Ethical approval for the study was provided by the Clinical Research Ethics Committee of the Cork Teaching Hospitals (CREC; Reference: ECM 4 (iii) 10/01/12). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Simmons RK , Singh G , Maconochie N Experience of miscarriage in the UK: qualitative findings from the National Women's Health Study . Soc Sci Med 2006 ;63 :1934 –46 . 10.1016/j.socscimed.2006.04.024 16781809 2 Maconochie N , Doyle P , Prior S Risk factors for first trimester miscarriage—results from a UK-population-based case-control study . BJOG 2007 ;114 :170 –86 . 10.1111/j.1471-0528.2006.01195.x 17305901 3 Adolfsson A Applying Heidegger's interpretive phenomenology to women's miscarriage . Psychol Res Behav Manag 2010 ;3 :75 –9 . 10.2147/PRBM.S4821 22110331 4 Murphy F The experience of early miscarriage from a male perspective . J Clin Nurs 1998 ;7 :325 –32 . 10.1046/j.1365-2702.1998.00153.x 9830973 5 Nanda K , Peloggia A , Grimes D Expectant care versus surgical treatment for miscarriage . Cochrane Database Syst Rev 2006 ;(2 ):CD003518 . 6 Saraiya M , Berg CJ , Shulman H Estimates of the annual number of clinically recognized pregnancies in the United States, 1981–1991 . Am J Epidemiol 1999 ;149 :1025 –9 . 10.1093/oxfordjournals.aje.a009747 10355378 7 Lee C , Slade P Miscarriage is a traumatic event: a review of the literature and new implications for intervention . J Psychosom Res 1996 ;40 :235 –44 . 10.1016/0022-3999(95)00579-X 8861119 8 Lok IH , Neugebauer R Psychological morbidity following miscarriage . Best Pract Res Clin Obstet Gynaecol 2007 ;21 : 229 –47 . 10.1016/j.bpobgyn.2006.11.007 17317322 9 Nikcevic AV , Tunkel SA , Kuczmierczyk AR Investigation of the cause of miscarriage and its influence on women's psychological distress . Br J Obstet Gynaecol 1999 ;106 :808 –13 . 10.1111/j.1471-0528.1999.tb08402.x 10453831 10 Geller PA , Kerns D , Klier CM Anxiety following miscarriage and the subsequent pregnancy: a review of the literature and future directions . J Psychosom Res 2004 ;56 :35 –45 . 10.1016/S0022-3999(03)00042-4 14987962 11 Craig M , Tata P , Regan L Psychiatric morbidity among patients with recurrent miscarriage . J Psychosom Obstet Gynaecol 2002 ;23 :157 –64 . 10.3109/01674820209074668 12436801 12 Kagami M , Maruyama T , Koizumi T Psychological adjustment and psychosocial stress among Japanese couples with a history of recurrent pregnancy loss . Hum Reprod 2012 ;27 :787 –94 . 10.1093/humrep/der441 22215625 13 Klock SC , Chang G , Hiley A Psychological distress among women with recurrent spontaneous abortion . Psychosomatics 1997 ;38 :503 –7 . 10.1016/S0033-3182(97)71428-2 9314720 14 Serrano F , Lima ML Recurrent miscarriage: psychological and relational consequences for couples . Psychol Psychother 2006 ;79 :585 –94 . 10.1348/147608306X96992 17312873 15 Swanson KM , Connor S , Jolley SN Contexts and evolution of women's responses to miscarriage during the first year after loss . Res Nurs Health 2007 ;30 :2 –16 . 10.1002/nur.20175 17243104 16 Gerber-Epstein P , Leichtentritt RD , Benyamini Y The experience of miscarriage in first pregnancy: the women's voices . Death Stud 2008 ;33 :1 –29 . 10.1080/07481180802494032 17 Limbo R , Glasser JK , Sundaram ME “Being sure”: women's experience with inevitable miscarriage . MCN 2014 ;39 :165 –74 . 10.1097/NMC.0000000000000027 18 Wojnar DM , Swanson KM , Adolfsson AS Confronting the inevitable: a conceptual model of miscarriage for use in clinical practice and research . Death Stud 2011 ;35 :536 –58 . 10.1080/07481187.2010.536886 24501829 19 Swanson KM , Karmali ZA , Powell SH Miscarriage effects on couples’ interpersonal and sexual relationships during the first year after loss: women's perceptions . Psychosom Med 2003 ;65 :902 –10 . 10.1097/01.PSY.0000079381.58810.84 14508039 20 Biggerstaff D , Thompson AR Interpretive phenomenological analysis (IPA): a qualitative methodology of choice in health care research . Qual Res Psychol 2008 ;5 :214 –24 . 10.1080/14780880802314304 21 Smith JA , Flowers P , Larkin M Interpretive phenomenological analysis: theory method and research . London : SAGE Publications , 2009 . 22 Willig C Introducing qualitative research in psychology . 2nd edn New York : Open University Press , 2008 . 23 Meaney S , Lutomski JE , Corcoran P PPO.50 Evaluation of non-response in the Women's Health Cohort Study . Arch Dis Child Fetal Neonatal Ed 2014 ;99 :A166 –168 . 10.1136/archdischild-2013-303999 24 RCOG Green-top Guideline No. 17 . The investigation and treatment of couples with recurrent first-trimester and second-trimester miscarriage . Royal College of Obstetricians and Gynaecologists , April 2011 . 25 Bardos J , Hercz D , Friedenthal J A national survey on public perceptions of miscarriage . Am J Obstet Gynaecol 2015 ;125 :1313 –20 . 10.1097/AOG.0000000000000859 26 Moohan J , Ashe RG , Cecil R The management of miscarriage results from a survey at one hospital . J Reprod Health Infant Psychol 1994 ;12 :17 –19 . 10.1080/02646839408408863 27 Wong MKY , Crawford TJ , Gask L A qualitative investigation into women's experiences after a miscarriage: implications for the primary healthcare team . Br J Gen Pract 2003 ;53 :697 –702 .15103877 28 Côté-Arsenault D Threat appraisal, coping, and emotions across pregnancy subsequent to perinatal loss . Nurs Res 2007 ;56 :108 –16 . 10.1097/01.NNR.0000263970.08878.87 17356441 29 Brier N Grief following miscarriage: a comprehensive review of the literature . J Womens Health 2008 ;17 :451 –64 . 10.1089/jwh.2007.0505 30 Beutel M , Willner H , Deckardt R Similarities and differences in couples’ grief reactions following a miscarriage: results from a longitudinal study . J Psychosom Res 1996 ;40 :245 –53 . 10.1016/0022-3999(95)00520-X 8861120 31 Johnson MP , Puddifoot JE The grief response in the partners of women who miscarry . Br J Med Psychol 1996 ;69 (Pt 4 ):313 –27 . 10.1111/j.2044-8341.1996.tb01875.x 8989980

PMC005xxxxxx/PMC5372117.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01490210.1136/bmjopen-2016-014902Emergency MedicineProtocol15061691169216911683Public health surveillance of automated external defibrillators in the USA: protocol for the dynamic automated external defibrillator registry study Elrod JoAnn Broeckel 1Merchant Raina 23Daya Mohamud 4Youngquist Scott 5Salcido David 67Valenzuela Terence 89Nichol Graham 1 1 Department of Medicine, University of Washington-Harborview Center for Prehospital Emergency Care, University of Washington, Seattle, Washington, USA 2 Salt Lake City Fire Department, University of Utah School of Medicine, Salt Lake City, Utah, USA 3 Department of Surgery, University of Utah School of Medicine, Salt Lake City, Utah, USA 4 Department of Emergency Medicine, Oregon Health and Sciences University, Portland, Oregon, USA 5 Department of Emergency Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania, USA 6 Tucson Fire Department, University of Arizona, Tucson, Arizona, USA 7 Department of Emergency Medicine, University of Arizona, Tucson, Arizona, USA 8 Department of Emergency Medicine, University of Pennsylvania, Pennsylvania, USA 9 Penn Medicine Social Media and Health Innovation Laboratory, University of Pennsylvania, Philadelphia, Pennsylvania, USACorrespondence to Professor Graham Nichol; nichol@uw.edu2017 29 3 2017 7 3 e01490225 10 2016 10 2 2017 27 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction Lay use of automated external defibrillators (AEDs) before the arrival of emergency medical services (EMS) providers on scene increases survival after out-of-hospital cardiac arrest (OHCA). AEDs have been placed in public locations may be not ready for use when needed. We describe a protocol for AED surveillance that tracks these devices through time and space to improve public health, and survival as well as facilitate research. Methods and analysis Included AEDs are installed in public locations for use by laypersons to treat patients with OHCA before the arrival of EMS providers on scene. Included cases of OHCA are patients evaluated by organised EMS personnel and treated for OHCA. Enrolment of 10 000 AEDs annually will yield precision of 0.4% in the estimate of readiness for use. Enrolment of 2500 patients annually will yield precision of 1.9% in the estimate of survival to hospital discharge. Recruitment began on 21 Mar 2014 and is ongoing. AEDs are found by using multiple methods. Each AED is then tagged with a label which is a unique two-dimensional (2D) matrix code; the 2D matrix code is recorded and the location and status of the AED tracked using a smartphone; these elements are automatically passed via the internet to a secure and confidential database in real time. Whenever the 2D matrix code is rescanned for any non-clinical or clinical use of an AED, the user is queried to answer a finite set of questions about the device status. The primary outcome of any clinical use of an AED is survival to hospital discharge. Results are summarised descriptively. Ethics and dissemination These activities are conducted under a grant of authority for public health surveillance from the Food and Drug Administration. Results are provided periodically to participating sites and sponsors to improve public health and quality of care. STATISTICS & RESEARCH METHODSPUBLIC HEALTH ==== Body Strengths and limitations of this study Study tracks individual automated external defibrillators over time. Outcome is ascertained by public health surveillance. Study has potential for selection bias due to incomplete case finding. Introduction Out-of-hospital cardiac arrest (OHCA) is defined as a sudden and unexpected pulseless condition attributable to cessation of cardiac mechanical activity.1 Cardiac arrest has multiple aetiologies, and the aetiology of arrest influences treatment decisions. Underlying mechanisms for non-traumatic cardiac arrest are crudely categorised as (1) conductive abnormalities of the myocardium leading to arrhythmias, (2) chronically weakened myocardium leading to end-stage pump failure and (3) acute occlusion of a coronary artery leading to myocardial infarction. Of these mechanisms of OHCA, resuscitation is generally most successful for isolated conductive abnormalities or for acute coronary thrombosis that is treated rapidly. Acute occlusion is common among patients with a first recorded rhythm of ventricular fibrillation (VF), which hereafter includes pulseless ventricular tachycardia as well as rhythms interpreted as shockable by an automated external defibrillator (AED). Recognition2 and successful treatment of VF is highly time-dependent.3 AEDs are medical devices (MDs) intended to simplify and improve treatment of OHCA. The timely use of AEDs by laypersons in conjunction with cardiopulmonary resuscitation (CPR including manual chest compressions with or without ventilation) is the only field intervention that has been shown to significantly increase the number of individuals who survive to discharge after OHCA.4 Since the effectiveness of use of AEDs by laypersons was demonstrated more than a decade ago, more than 2.4 million AEDs have been sold for use by laypersons in the USA.5 As more of these AEDs are placed in the community, the Food and Drug Administration (FDA) has received reports of potential adverse events associated with them. Since the overall number of AEDs installed is not public information and it is difficult to track individual AEDs through time and space, agency staff have experienced challenges in trying to interpret these reports.6 As well, each major manufacturer of AEDs in the USA has recalled their devices to address one or more potential safety issues over the past 5 years. A need for a more comprehensive surveillance system has been identified.7 Specifically, FDA's advisory panel recommended “Registry information should be established to collect more data. An active reporting registry would help with recalls and would help to maintain the quality and confidence of the data collected. Sold devices could be tracked prospectively, and long term device performance would be captured.” In this paper, we describe the rationale for and design of an ongoing dynamic registry for surveillance of AED location and use. This registry has been created through a partnership of academics, FDA, key stakeholders and manufacturers of AEDs. The long-term goals of the registry are to provide reliable, valid and sustainable postmarket surveillance of AEDs; to provide important and timely information to patients, providers and the public to improve MD device development including the quality of AEDs and the quality of care for patients with OHCA and to provide a framework for prospective embedded studies of the effectiveness of next-generation AEDs. The Dynamic AED Registry represents a novel approach to AED surveillance, leverages a clinical registry infrastructure and can be applied to other MDs that are mobile in time and space. Methods and analysis Study populations Included AEDs are those: Pre-existing in community settings and Identified by crowdsourcing techniques, or Identified by manufacturer or distributor. Newly installed in community setting: Identified by owner during at time of installation, or Identified by manufacturer at time of sale. Included patients are those treated for OHCA: Evaluated by organised emergency medical services (EMS) personnel and receive attempts at external defibrillation (by lay, police or EMS), or Receive chest compressions by organised EMS personnel. Although a minority of patients with OHCA have an AED applied by a layperson before the arrival of EMS providers on scene (ref 8, p. 339), we monitor all cases of EMS-treated cardiac arrest in participating communities to obtain complete case finding. Monitoring the outcome of every AED use is necessary to identify those uses that may have caused or contributed to survival or death (21CFR803), since the majority of patients treated for OHCA die. To date, studies reported from smaller selective registries were difficult to interpret because of potential selection bias and lack of knowledge of the overall population. This potential selection bias should be mitigated as the Dynamic AED Registry evolves over time to include a high percentage of treated patients in the USA. Several strategies are used to achieve a high rate of enrolment in the Dynamic AED Registry. Initially, AEDs were enrolled by their identification in community settings by using crowdsourcing techniques.9 According to these methods, individuals or teams of volunteers register to participate in a scavenger hunt contest. Those who identify the most AEDs during the contest period receive a prize. As well, the first individual or team to report one of a finite number of preselected AED receive a prize. Now, enrolment has been extended to include prospective registry of AEDs sold by a manufacturer for use by lay persons in public locations. We also collate information about individual AEDs reported by laypersons who have registered online and then used their smartphone (https://heartmap.uwctc.org/, accessed on 25 May 2016). Individual device identifiers are provided by the investigators to owners or responsible individuals of these AEDs, with information about the purpose and methods of AED surveillance, as well as instructions on where to apply the identifier to the AED and how to register the AED in the dynamic registry by scanning the identifier. A series of individual identifiers are provided to each sponsoring manufacturer for inclusion with the product registration materials provided with AEDs newly sold for use in community settings. Organisations that provide physician oversight and AED readiness services encourage client participation by placing an individual identifier on each AED under their supervision then scanning the identifier to record information about the location of the device in our secure database. With permission of the client, this information is shared with the relevant EMS dispatch centre so as to improve the community response to OHCA for the client's staff and customers. The lightweight and portability of AEDs makes them easily movable. If the AED is moved later and then the identifier is rescanned, we provide information about the updated location of the device to the oversight organisation so that the location of the AED is known at all times. Participating EMS agencies are recruited as interest and resources allow. Initially, this included a mix of EMS agencies that participated in the Resuscitation Outcomes Consortium (funded by the National Institutes of Health, American Heart Association and other agencies), are participating in the Mission:Lifeline Cardiac Resuscitation program (funded by American Heart Association) or associated with a dispatch centre using software to enable faster lifesaving AED response within the community (PulsePoint Responder, PulsePoint Foundation, San Ramon, CA). These vanguard sites included Philadelphia, PA, Pittsburgh, PA, Portland, OR, Salt Lake City, UT, Seattle, WA and Tucson, AZ. We expect that participation will expand over time to include as many EMS agencies as possible throughout the USA. Furthermore, EMS agencies or receiving hospitals may enrol patients to provide an analysis of their programme as well as to benchmark the process and outcome of care compared to other participant's experience. Data acquisition EMS providers and personnel responsible for maintenance of AEDs are encouraged to scan the individual identifier after each AED use. On scanning of the 2D matrix code, the date, time and location of use are automatically identified, recorded and transmitted to the dynamic registry. As well, users are asked to respond to a finite set of queries confirming the location of the AED as well as describing the reason for its scan. In the event of a clinical use, EMS providers who scan the individual identifier are queried about the patient's intended hospital destination to facilitate subsequent tracking of patient vital status at hospital discharge. After the event, EMS providers are queried about a finite set of patient and EMS factors to facilitate adjustment of outcomes for case-mix as well as assessment of rates of potential adverse events in a manner consistent with the Utstein template for reporting of outcomes after OHCA.1 As well, investigators ascertain the participant's vital status at discharge from the receiving hospital. Anonymised quarterly reports are provided to participating EMS systems, which describe the process and outcome of care for patients with OHCA as an incentive to facilitate continued surveillance of AEDs, their use and factors that contribute to positive clinical outcomes. The quality of the data included in the registry is monitored and improved by using regular checks and audits. Data elements Baseline characteristics, covariates and outcomes included in the registry were developed and implemented by consensus of the executive committee. In doing so, emphasis was placed on developing a finite set of key variables so as to ensure that the registry was simple, sustainable and scalable. These can be revised periodically as required. Data quality control The Dynamic AED Registry uses multiple mechanisms to ensure data completeness and accuracy (table 1). Monitoring and auditing use standardised data quality checks. Events are adjudicated as required for prespecified outcomes. Table 1 Techniques for optimising data quality control Applied to all dynamic AED Registry Data Applied to postapproval studies Site training and support from Registry staff for queries Yes Yes Data cleaning: data integrity checks using range validation and other measures Yes Yes Audit portion of data Yes Yes Collection of source documents and verification of prespecified key events No Yes Adjudication of selected outcomes Yes Yes Potential adverse events No shock advised by AED but first EMS rhythm shockable Yes Use of data to improve care The Dynamic AED Registry provides feedback to participating sites including periodic quality benchmarking. Participants have access to a repository of their own data and tools to evaluate their local practice and conduct user-specified local data queries. This is especially valuable for communities that alert lay responders to the need for bystander CPR and AED need through dispatch-identified OHCA. Data are also provided to FDA to address specific surveillance questions. Use of data to facilitate research The data will be used to facilitate monitoring of enrolment in prospective embedded evaluations of interventions in patients with cardiac arrest. For example, the majority of vanguard sites are planning to participate in large randomised trials of resuscitation interventions as soon as feasible in patients treated for OHCA. As sites implement this trial, trial investigators can monitor whether all eligible patients were enrolled by comparing the volume of patients enrolled in the AED registry to that enrolled in the trial. As well, the characteristics of the patients who accrue to the trial can be compared to data obtained from the AED registry. In order to adhere to patient privacy requirements, data from the registry will not be linked to individual patients who might also be enrolled in the clinical trial. Instead, aggregate statistics for the population of all patients with OHCA treated within the study communities will be compared to aggregate statistics for the patients enrolled in this trial. From such a comparison, the generalisability of the trial results can be investigated. Innovative methods The Dynamic AED Registry uses multiple innovative methods to enhance surveillance of AEDs. Crowdsourcing We have adapted crowdsourcing methods to find AEDs installed in public locations for use by laypersons in major metropolitan areas that are participating in the Dynamic AED Registry. Merchant pioneered use of crowdsourcing to identify AEDs installed in community settings.9 According to these methods, help is solicited for tasks usually performed by particular individuals from an undefined large group (ie, a crowd). Since the task is offered to a large group with diverse backgrounds, it has the potential to attract individuals who are interested in the problem, likely to finish the task, and likely to contribute innovative ideas. On the basis of methods developed by the Defense Advanced Research Projects Agency (DARPA),10 adults were invited to identify AEDs installed in community settings in Philadelphia County, Pennsylvania. According to state law, AEDs installed in community settings in Philadelphia County, Pennsylvania, should be registered with the local EMS authority. Before the contest began, 57 devices were registered with Philadelphia EMS. Monetary prizes were awarded to those individuals or teams who reported the most AEDs, as well as to those who were first to report any of 200 preselected AEDs. During the contest, 313 individuals and teams identified 852 unique AEDs in 528 locations. These locations were in 94% of eligible census tracts and included a mix of public (59%) and private (41%) locations. 50% of AEDs were identified as either non-functional or having unknown functional status. The investigators concluded that crowdsourcing is a feasible approach for identifying AEDs installed in community settings in a large urban city. Able to track individual medical devices We have pioneered the use of an individual identifier to track AEDs or other MDs through time and space in community settings. According to this method, each AED is tagged with a label which consists of a unique two-dimensional (2D) matrix code (QR code, Denso Wave, Chita-gun, Japan); then the QR code or 2D matrix code is recorded as well as the location and status of the AED by using open-source software native to any contemporary smartphone (eg, Google Goggles, Google, Mountain View, CA) and finally these elements are automatically passed via the internet to a secure and confidential database in real time (Provisional Patent # 61/498 424, figure 1). The user is prompted to rescan the label after any use (figure 2). Note that by design each label has a unique alphanumeric code that provides built in redundancy with the 2D matrix code. The user is queried to answer a finite set of questions when the smartphone is connected to the database (table 2). If the device has a clinical use, these queries include verification of the intended receiving hospital for transported cases to facilitate longitudinal follow-up of patient outcome to hospital discharge (table 3). Using the limited data collected in the field with the 2D matrix code, additional variables are abstracted later from the participant's record to further characterise the process and outcome of care. Table 2 Questions for EMS on scanning after clinical use Question Definition Device not used Medic, bystander or other person tried to apply AED to patient but was unable to use device; reason device not used Date of cardiac arrest Date cardiac arrest occurred Time of cardiac arrest Time cardiac arrest occurred Age Patient's age in years Gender Patient's gender Destination hospital Hospital to which patient was transported First responder agency Name and state-assigned code number from first responder vehicle Table 3 Patient outcome variables Variable Definition Emergency department disposition Final disposition of patient from emergency department: patient documented to be discharged from ED alive, patient documented to die in ED, transferred to a different hospital, admitted to same hospital. Vital status at discharge Patient's vital status at discharge from hospital: died in hospital, discharged alive, not yet determined. Neurologic status at discharge Patient's neurologic status based on CPC at discharge from hospital: good cerebral performance, moderate cerebral disability, severe cerebral disability, coma/vegetative state. CPC, cerebral performance category. Figure 1 Use of unique device identifier for AED surveillance. AED, automated external defibrillator. Figure 2 Two-dimensional matrix code for AED tagging. AED, automated external defibrillator. The method for scanning the 2D matrix code is deliberately platform independent (ie, agnostic). We opted to design and implement a website that can be accessed by any web-capable smartphone rather than take on the burden of designing, disseminated and maintaining multiple different smartphone applications. Most EMS providers carry their own smartphone. Since no specific smartphone application is required to access the website, our web developer is able to update this centrally as needed. This device identifier complement's the FDA's Global Use Device Identification Database (GUDID) initiative.11 This establishes that a unique device identifier number is assigned by the device manufacturer to each version or model of a device. This identifier is in human-readable format and in machine-readable format. It is intended to contain information about the type of device as well as information about its manufacture. The primary difference is that GUDID does not require or enable tracking of the location of a device through time and space. Sample size and analysis Recruitment began on 21 Mar 2014. Enrolment of AEDs and patients with OHCA in this cohort study is ongoing. There is no planned end. As such, there is no maximum expected enrolment necessary to detect a significant difference between groups, as there would be in a clinical trial. The primary analysis will be descriptive. If 95% of AEDs are ready for use, enrolment of 10 000 AEDs annually will yield precision (measured as half of the 95% CI) of 0.4% in the estimate of readiness for use. If 10% of patients treated for OHCA survive to discharge, enrolment of 2500 patients annually will yield precision of 1.9% in the estimate of survival to discharge. Dissemination The Dynamic AED Registry is the first registry to be conducted under a grant of authority for public health surveillance from FDA. As such, these activities have been determined by multiple institutional review boards (IRBs) to not be subject to human subjects' regulations and to exercise the public health exception within the Health Insurance Portability and Accountability Act (HIPAA, 45 CFR 164.512(b)). These methods are adaptable to public health surveillance of other MDs. Discussion Lay use of AEDs before the arrival of EMS providers on scene increases survival after OHCA.4 Large numbers of AEDs have been placed in public locations to improve the community response to cardiac arrest.5 It is sometimes difficult to determine whether these devices are maintained in a state of readiness for use or whether they functioned as intended when applied to someone in presumed cardiac arrest. Importantly, there is no widely deployed method of tracking their location and use in community settings. We describe the implementation and maintenance of a dynamic registry for surveillance of AEDs through time and space. This registry serves multiple purposes, including public health surveillance of AEDs, quality improvement of AED use as well as overall EMS care, and serving as a framework for embedded research studies. We note that local and multicentre registries of patients with OHCA are often used for multiple purposes. For example, the ROC cardiac arrest registry12 measured the public health impact of cardiac arrest in participating communities,13 provided a mechanism for local quality improvement14 as well as a mechanism for embedded randomised trials.15–17 Unfortunately funding for ROC has ceased, and the ROC registry collated too many variables to be sustainable over the long term without dedicated funding. The Dynamic AED Registry has several strengths. By design, it is simple, sustainable and scalable. Since the registry functions under a grant of authority for public health surveillance, the base surveillance and quality improvement functionality is not subject to human subjects' research regulations and is exempted from HIPAA. Data collected for public health surveillance can, subject to appropriate permissions, can be used to improve the process and outcome of care for patients with OHCA in participating communities. Like other registries, it can also be used as a framework for embedded prospective research evaluations of existing or next-generation defibrillators, as well as other interventions intended to improve outcomes after OHCA. What does the Dynamic AED Registry offer over other registries? It tracks the location and use of individual AEDs in diverse communities. This identifies the specific AED used, allowing verification of the diagnostic algorithm in no-shock advised cases. It supports passive reporting of potential AED failure that is supposed to occur. AED manufacturers have recognised the value of and subsequently invested in the Dynamic AED Registry to provide ongoing safety and effectiveness data, which can be useful as real-world evidence to support regulatory decisions for AEDs. Recognising that data collated in different resuscitation registries may not be comparable, the academic leadership of the Dynamic AED Registry is working with leaders of other resuscitation-related registries to encourage harmonisation of data elements, quality control techniques, outcomes and adverse event classification. This registry has some limitations. The robustness of the data collated by it depends in part on the completeness of case finding by EMS providers and participating sites may have variable success in enrolling consecutive AEDs or patients. On the basis of a comparison of patients not enrolled versus enrolled in a registry intended to enrol consecutive patients with acute coronary syndrome, patients who are not included in a registry may have higher risk, receive poorer quality of care and have a higher mortality than those who are included.18 Such selection bias limits the ability of registries to reliably assess the effectiveness of study interventions.19 At present, the Dynamic AED registry has a small footprint but this is expected to grow over time. Although crowdsourcing methods are used to identify AEDs that were placed in public locations, it is not feasible to apply crowdsourcing in every urban and rural community. Instead, the registry is evolving to include prospective tagging of AEDs through manufacturers as they are placed in public locations. The distribution of AEDs by geography may be non-random (ie, one manufacturer may have more devices placed in a given community than another). Since geographic region is a strong predictor of outcome after OHCA,20 comparison of outcomes by different manufacturers needs to account for differences in geography. The small footprint of the registry impacts on its representativeness and may be associated with selection bias. Communities choose to participate or not participate in the registry. To the extent that participating communities have greater (or lesser) survival than other communities, the registry may over (under) estimate the effectiveness of AEDs. It is plausible that incomplete identification and scanning of QR codes may impact on the registry's representativeness and may be associated with selection bias. To try to mitigate this, participating communities are encouraged to track and report all patients with OHCA treated by EMS providers, regardless of whether they are initially identified as having been treated with an AED by a layperson. The Center for Devices and Radiological Health (CDRH) of the US FDA of the USA is responsible for regulating firms who manufacture, distribute or import MDs sold in the USA. MDs are instruments or related articles which are intended for use in the diagnosis or treatment of disease, are applied externally or internally to patients to affect the structure or function of their body and do not achieve their purpose through chemical action or metabolism.21 FDA relies on registries for postmarket surveillance of MDs to ensure their safety and efficacy. The US congress has directed the agency to implement an active system capable of identifying safety issues and communicating them in a timely manner to patients, providers and the public.22 Policymakers have suggested that increased surveillance of MDs is necessary to enhance and maintain patient safety.23 The Institute of Medicine (IOM) made recommendations about how to improve the approval process and postmarket surveillance of MDs.24 As well, the IOM made recommendations about how to improve outcomes after OHCA that included the establishment and maintenance of a sustainable method of national surveillance of the process and outcome of care of all patients with OHCA.25 Experts have recommended postmarket surveillance of any AED introduced for in-hospital, emergency medical services or public access defibrillation.26 Our registry can be placed in the context of FDA's evolving medical device regulatory process (table 4) as well as in the context of other resuscitation-related registries. It is a next-generation registry implemented with the multiple purposes of surveillance of MDs as well as measurement and improvement of the process and outcome of care for patients in real-world settings, as well as serving as a framework for embedded studies of the effectiveness of defibrillator as well as other interventions in patients with OHCA. The Dynamic AED Registry is designed to evolve into a comprehensive but scalable and sustainable national registry in the USA. Multiple other resuscitation-related registries exist in the USA, and the rest of the world that seek to characterise the process and outcome of care related to OHCA.27 However, to the best of our knowledge, none are capable of tracking clinical and non-clinical use of individual AEDs including vital status at hospital discharge. As well, some have experienced challenges related to sustainability. Table 4 Dynamic AED Registry in context of FDA approval process for medical devices Premarket Hybrid Postmarket Dynamic AED Registry Phase I Phase II Phase III Postapproval Postmarket Clinical registry Aim Safety Efficacy Safety, efficacy and effectiveness Effectiveness Inclusion criteria Restrictive Either Broad Intervention Tight protocol Either Implemented in usual clinical practice Cointervention Based on protocols for many aspects of care Either Based on local practice; monitored but minimal control Adherence to protocol Required Either Expected and considered in sample size and analysis plan Events Related to biological effect (eg, conversion out of shockable rhythm) Either Related to patient outcome (eg, survival) Analysis Treatment received Both Intention to treat Sample size Usually <1000 Either Usually > 1000 Data burden Large Core supplemented by study-specific Minimal and simple core Study management Significant interventions and support from research staff Minimal support and interventions from research team Multiple other registries exist that describe the process and outcome of care for patients with OHCA. In the USA, the Resuscitation Outcomes Consortium (ROC) Cardiac Arrest Registry enrolled such patients in geographic sites that were selected by a competitive process to participate in a clinical research network.12 This registry included all patients with non-traumatic OHCA assessed or treated by participating EMS providers. Data reported by participating EMS agencies were routinely screened for missing cases by comparing the observed number of cases in any month to the expected number for that agency as estimated by averaging the observed cases over a 12-month period. As a consequence of this, the incidence of EMS-treated OHCA enrolled in this registry increased by 25% over time.13 28 The registry used multiple methods to monitor and improve data quality and completeness. It was intended to facilitate case identification and data collection for ROC intervention trials so it used more than 200 variables to characterise each enrolled episode. The registry informed the design and implementation of multiple pragmatic trials in patients with OHCA,15 17 29–31 and secondary analyses about the association between patient and treatment characteristics.13 32–40 Although funding to support data collection in US ROC sites has ceased, most sites are continuing to enrol cases in this registry during ROC's current no-cost extension phase. We adapted and extended the ROC registry as a simple, sustainable, registry focused on tracking the process and outcome of care associated with use of AEDs. The Cardiac Arrest Registry to Enhance Survival (CARES) Registry initially focused on enrolment of cardiac arrest of presumed cardiac aetiology, but has broadened its inclusion criteria to include cardiac arrest of any aetiology.41 Since there is at least a twofold variation in the reported proportion of cardiac arrests that are of non-cardiac aetiology,42 43 assessments of the effectiveness of interventions for OHCA that include only cases with presumed cardiac aetiology may be susceptible to selection bias. Reports from CARES emphasise survival among patients with witnessed VF rather than among all treated patients. Importantly, the likelihood that an initial rhythm will be VF depends in part on the time of rhythm assessment.2 A consequence of emphasising survival in this subpopulation rather than all patients treated for OHCA is that assessments of the effect of interventions intended to improve survival (eg, AEDs) may be susceptible to bias or confounding. CARES reports that 32 states contribute data to it, but the catchment population within each ranges from 50% to 100%.44 As well, multiple EMS agencies that participate in CARES report that 100% of their cases received bystander CPR and 100% survived to discharge, which suggests that they may be missing cases that did not receive bystander CPR (and did not have a good outcome) (ref. 44, p. 16 and 17). Collectively, these observations suggest that the external validity of CARES data may require further assessment. CARES was initially funded by the Centers for Disease Control. It is currently funded by the American Heart Association and Medtronic Foundation, as well as a subscription-based funding model. The National Emergency Medical Services Information System (NEMSIS) is an electronic documentation system that is intended for use in every local EMS system.45 Local data are collated by each state or territory and then contributed to a national data warehouse. Over 90% of states and territories have a NEMSIS-compliant system.46 Challenges to use of NEMSIS for public health surveillance of AEDs include that many states are working to revise data elements, improve data capture and ensure compliance with the latest NEMSIS data standard. Another challenge is that historically NEMSIS has had limited hospital outcome data. NEMSIS is supported by state public health departments. The PulsePoint program incorporates information about AED locations (as opposed to individual AEDs) into dispatch software used by Public Safety Answering Points to improve the community response to OHCA. The Dynamic AED Registry provides location information to PulsePoint to supplement and periodically update their location information. When a dispatcher identifies that someone has had OHCA in a community that is participating in the PulsePoint program, citizens who have previously downloaded a location-aware application to their smartphone are alerted to the event if they are close to the patient. They can then respond to provide CPR and use a nearby AED before the arrival of EMS providers on scene. PulsePoint does not routinely collate outcome on patients with OHCA for whom citizens are activated. As of 25 May 2016, the PulsePoint program has been adopted by more than 1600 EMS agencies that are primarily based in the USA. (Price R. Personal communication, 25 May 2016). PulsePoint is supported by a subscription-based funding model. Lay use of AEDs before the arrival of EMS providers on scene increases survival after OHCA. Large numbers of AEDs have been placed in public locations to improve the community response to cardiac arrest. These devices may be moved or may not be ready for use when needed. Our dynamic AED registry surveys the location of AEDs dynamically through time and space so as to improve the community response to and outcome of OHCA. Twitter: Follow Graham Nichol @grahamnichol Collaborators: Ben Eloff, PhD. Contributors: GN and RM contributed to the study conception and design. MD, SY, DS, TV and RM contributed to acquisition of data. GN, JBE, MD, SY, DS, TV and RM participated in drafting the article, revising it for critically important intellectual content and gave final approval of the submitted version. Funding: This work is supported by 1U01FD004587-04 from the Food and Drug Administration, Silver Spring, Maryland and ZOLL Medical Corporation, Chelmsford, MA. Competing interests: The Dynamic AED Registry is funded by the Food and Drug Administration, Silver Spring, Maryland and ZOLL Medical Corporation, Chelmsford, Massachusetts. Provenance and peer review: Not commissioned; externally peer reviewed. ==== Refs References 1 Jacobs I , Nadkarni V , Bahr J Cardiac Arrest and Cardiopulmonary Resuscitation Outcome Reports: Update and Simplification of the Utstein Templates for Resuscitation Registries. A Statement for Healthcare Professionals from a Task Force of the International Liaison Committee on Resuscitation (American Heart Association, European Resuscitation Council, Australian Resuscitation Council, New Zealand Resuscitation Council, Heart and Stroke Foundation of Canada, Interamerican Heart Foundation, Resuscitation Council of Southern Africa) . Resuscitation 2004 ;63 :233 –49 .15582757 2 Holmberg M , Holmberg S , Herlitz J An alternative estimate of the disappearance rate of ventricular fibrillation in our-of-hospital cardiac arrest in Sweden . Resuscitation 2001 ;49 :219 –20 . 10.1016/S0300-9572(00)00376-2 11428360 3 Valenzuela TD , Roe DJ , Cretin S Estimating effectiveness of cardiac arrest interventions: a logistic regression survival model . Circulation 1997 ;96 :3308 –13 . 10.1161/01.CIR.96.10.3308 9396421 4 Hallstrom AP , Ornato JP , Weisfeldt M Public-access defibrillation and survival after out-of-hospital cardiac arrest . N Engl J Med 2004 ;351 :637 –46 . 10.1056/NEJMoa040566 15306665 5 Nichol G , Elrod JA , Becker LB Treatment for out-of-hospital cardiac arrest: is the glass half empty or half full? Circulation 2014 ;130 :1844 –6 . 10.1161/CIRCULATIONAHA.114.013263 25399398 6 https://www.Federalregister.Gov/Documents/2015/01/29/2015-01619/Effective-Date-of-Requirement-for-Premarket-Approval-for-Automated-External-Defibrillator-Systems (27 Sep 2016). 7 http://www.Fda.gov/Downloads/Advisorycommitees/Committeesmeetingmaterials/Medi caldevices/Medicaldevicesadvisorycommittee/Circulatorysystemdevicespanel/Ucm241780.pdf 8 Mozaffarian D , Benjamin EJ , Go AS Heart disease and stroke statistics—2016 update: a report from the American Heart Association . Circulation 2016 ;133 :e38 –e360 . 10.1161/CIR.0000000000000350 26673558 9 Merchant RM , Asch DA , Hershey JC A crowdsourcing innovation challenge to locate and map automated external defibrillators . Circ Cardiovasc Qual Outcomes 2013 ;6 :229 –36 . 10.1161/CIRCOUTCOMES.113.000140 23481522 10 Tang JC , Cebrian M , Giacobe NA Reflecting on the Darpa Red Balloon Challenge . Commun ACM 2011 ;54 :78 –85 . 10.1145/1924421.1924441 11 http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/UniqueDeviceIdentification/default.htm (25 May 2016 ). 12 Morrison LJ , Nichol G , Rea TD Rationale, development and implementation of the Resuscitation Outcomes Consortium Epistry-Cardiac Arrest . Resuscitation 2008 ;78 :161 –9 . 10.1016/j.resuscitation.2008.02.020 18479802 13 Nichol G , Thomas E , Callaway CW Regional variation in out-of-hospital cardiac arrest incidence and outcome . JAMA 2008 ;300 :1423 –31 . 10.1001/jama.300.12.1423 18812533 14 Daya MR , Schmicker RH , Zive DM Out-of-hospital cardiac arrest survival improving over time: results from the Resuscitation Outcomes Consortium (ROC) . Resuscitation 2015 ;91 :108 –15 . 10.1016/j.resuscitation.2015.02.003 25676321 15 Hostler D , Everson-Stewart S , Rea TD Effect of real-time feedback during cardiopulmonary resuscitation outside hospital: prospective, cluster-randomised trial . BMJ 2011 ;342 :d512 10.1136/bmj.d512 21296838 16 Stiell IG , Nichol G , Leroux BG Early versus later rhythm analysis in patients with out-of-hospital cardiac arrest . N Engl J Med 2011 ;365 :787 –97 . 10.1056/NEJMoa1010076 21879896 17 Nichol G , Leroux B , Wang H Trial of continuous or interrupted chest compressions during CPR . N Engl J Med 2015 ;373 :2203 –14 . 10.1056/NEJMoa1509139 26550795 18 Ferreira-Gonzalez I , Marsal JR , Mitjavila F Patient registries of acute coronary syndrome: assessing or biasing the clinical real world data? Circ Cardiovasc Qual Outcomes 2009 ;2 :540 –7 . 10.1161/CIRCOUTCOMES.108.844399 20031891 19 Krumholz HM Registries and selection bias: the need for accountability . Circ Cardiovasc Qual Outcomes 2009 ;2 :517 –8 . 10.1161/CIRCOUTCOMES.109.916601 20031886 20 Rea TD , Cook AJ , Stiell IG Predicting survival after out-of-hospital cardiac arrest: role of the Utstein data elements . Ann Emerg Med 2010 ;55 :249 –57 . 10.1016/j.annemergmed.2009.09.018 19944488 21 http://www.FDA.gov/Medicaldevices/Deviceregulationandguidance/Overview/Classifyyourdevice/Ucm051512.htm (27 Sep 2016 ). 22 Platt R , Wilson M , Chan KA The New Sentinel Network—the evidence of medical-product safety . N Engl J Med 2009 ;361 :645 –7 . 10.1056/NEJMp0905338 19635947 23 http://www.Grassley.Senate.Gov/News/News-Releases/Qa-Medical-Device-Safety (25 May 2016 ). 24 Institute of Medicine (US) . Committee on the Public Health Effectiveness of the FDA 510(k) Clearance Process. Medical devices and the public's health: the FDA 510(K) clearance process at 35 years . Washington (DC ): National Academies Press , 2011 :298 . 25 IOM (Institute of Medicine). Strategies to improve cardiac arrest survival: A time to act. Washington, DC: The National Academies Press, 2015:458. 26 Kerber RE , Becker LB , Bourland JD Automatic external defibrillators for public access defibrillation: recommendations for specifying and reporting arrhythmia analysis algorithm performance, incorporating new waveforms, and enhancing safety. A Statement for Health Professionals from the American Heart Association Task Force on Automatic External Defibrillation, Subcommittee on AED Safety and Efficacy . Circulation 1997 ;95 :1677 –82 .9118556 27 Nishiyama C , Brown SP , May S Apples to apples or apples to oranges? International variation in reporting of process and outcome of care for out-of-hospital cardiac arrest . Resuscitation 2014 ;85 :1599 –609 . 10.1016/j.resuscitation.2014.06.031 25010784 28 Roger VL , Go AS , Lloyd-Jones DM Heart disease and stroke statistics—2012 update: a report from the American Heart Association . Circulation 2012 ;125 :e2 –e220 . 10.1161/CIR.0b013e31823ac046 22179539 29 Aufderheide TP , Kudenchuk PJ , Hedges JR Resuscitation Outcomes Consortium (ROC) PRIMED cardiac arrest trial methods part 1: rationale and methodology for the impedance threshold device (ITD) protocol . Resuscitation 2008 ;78 :179 –85 . 10.1016/j.resuscitation.2008.01.028 18487005 30 Stiell IG , Callaway C , Davis D Resuscitation Outcomes Consortium (ROC) primed cardiac arrest trial methods. Part 2: rationale and methodology for “analyze later vs. analyze early” protocol . Resuscitation 2008 ;78 :186 –95 .18487004 31 Kudenchuk PJ , Brown SP , Daya M Amiodarone, lidocaine, or placebo in out-of-hospital cardiac arrest . N Engl J Med 2016 ;374 :1711 –22 . 10.1056/NEJMoa1514204 27043165 32 Cheskes S , Schmicker RH , Verbeek PR The impact of peri-shock pause on survival from out-of-hospital shockable cardiac arrest during the Resuscitation Outcomes Consortium Primed Trial . Resuscitation 2014 ;85 :336 –42 . 10.1016/j.resuscitation.2013.10.014 24513129 33 Cheskes S , Schmicker RH , Christenson J Perishock pause: an independent predictor of survival from out-of-hospital shockable cardiac arrest . Circulation 2011 ;124 :58 –66 . 10.1161/CIRCULATIONAHA.110.010736 21690495 34 Stiell IG , Brown SP , Christenson J What is the role of chest compression depth during out-of-hospital cardiac arrest resuscitation? Crit Care Med 2012 ;40 :1192 –8 . 10.1097/CCM.0b013e31823bc8bb 22202708 35 Stiell IG , Brown SP , Nichol G What is the optimal chest compression depth during out-of-hospital cardiac arrest resuscitation of adult patients? Circulation 2014 ;130 :1962 –70 . 10.1161/CIRCULATIONAHA.114.008671 25252721 36 Idris AH , Guffey D , Aufderheide TP Relationship between chest compression rates and outcomes from cardiac arrest . Circulation 2012 ;125 :3004 –12 . 10.1161/CIRCULATIONAHA.111.059535 22623717 37 Idris AH , Guffey D , Pepe PE Chest compression rates and survival following out-of-hospital cardiac arrest . Crit Care Med 2015 ;43 :840 –8 . 10.1097/CCM.0000000000000824 25565457 38 Christenson J , Andrusiek D , Everson-Stewart S Chest compression fraction determines survival in patients with out-of-hospital ventricular fibrillation . Circulation 2009 ;120 :1241 –7 . 10.1161/CIRCULATIONAHA.109.852202 19752324 39 Vaillancourt C , Everson-Stewart S , Christenson J The impact of increased chest compression fraction on return of spontaneous circulation for out-of-hospital cardiac arrest patients not in ventricular fibrillation . Resuscitation 2011 ;82 :1501 –7 . 10.1016/j.resuscitation.2011.07.011 21763252 40 Glover BM , Brown SP , Morrison L Wide variability in drug use in out-of-hospital cardiac arrest: a report from the Resuscitation Outcomes Consortium . Resuscitation 2012 ;83 :1324 –30 . 10.1016/j.resuscitation.2012.07.008 22858552 41 McNally B , Stokes A , Crouch A Cares: cardiac arrest registry to enhance survival . Ann Emerg Med 2009 ;54 :674 –83 e2 . 10.1016/j.annemergmed.2009.03.018 19394110 42 Drory Y , Turetz Y , Hiss Y Sudden unexpected death in persons less than 40 years of age . Am J Cardiol 1991 ;68 :1388 –92 . 10.1016/0002-9149(91)90251-F 1951130 43 Kuisma M , Alaspaa A Out-of-hospital cardiac arrests of non-cardiac origin. Epidemiology and outcome . Eur Heart J 1997 ;18 :1122 –8 .9243146 44 Vellano K , Crouch A , Rajdev M Cardiac arrest registry to enhance survival (cares) report on the public health burden of out-of-hospital cardiac arrest. Prepared for Institute of Medicine. 15 May 2015 [internet] , 2016 . 45 Dawson DE National Emergency Medical Services Information System (NEMSIS) . Prehosp Emerg Care 2006 ;10 :314 –16 . 10.1080/10903120600724200 16801268 46 http://Nemsis.Org/Theproject/Whatisnemsis/Goalsandobjectives.html (17 June 2016 ).

PMC005xxxxxx/PMC5372118.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01378510.1136/bmjopen-2016-013785Cardiovascular MedicineProtocol15061683170217041724Protocol for validating cardiovascular and cerebrovascular ICD-9-CM codes in healthcare administrative databases: the Umbria Data Value Project Cozzolino Francesco 1http://orcid.org/0000-0002-5440-775XAbraha Iosief 1Orso Massimiliano 1Mengoni Anna 2Cerasa Maria Francesca 2Eusebi Paolo 1Ambrosio Giuseppe 2Montedori Alessandro 1 1 Health Planning Service, Regional Health Authority of Umbria, Perugia, Italy 2 Division of Cardiology, Santa Maria della Misericordia Hospital, University of Perugia School of Medicine, Perugia, ItalyCorrespondence to Dr Iosief Abraha; iosief_a@yahoo.it2017 29 3 2017 7 3 e0137856 8 2016 19 1 2017 23 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction Administrative healthcare databases can provide a comprehensive assessment of the burden of diseases in terms of major outcomes, such as mortality, hospital readmissions and use of healthcare resources, thus providing answers to a wide spectrum of research questions. However, a crucial issue is the reliability of information gathered. Aim of this protocol is to validate International Classification of Diseases, 9th Revision—Clinical Modification (ICD-9-CM) codes for major cardiovascular diseases, including acute myocardial infarction (AMI), heart failure (HF), atrial fibrillation (AF) and stroke. Methods and analysis Data from the centralised administrative database of the entire Umbria Region (910 000 residents, located in Central Italy) will be considered. Patients with a first hospital discharge for AMI, HF, AF or stroke, between 2012 and 2014, will be identified in the administrative database using the following groups of ICD-9-CM codes located in primary position: (1) 410.x for AMI; (2) 427.31 for AF; (3) 428 for HF; (4) 433.x1, 434 (excluding 434.x0), 436 for ischaemic stroke, 430 and 431 for haemorrhagic stroke (subarachnoid haemorrhage and intracerebral haemorrhage). A random sample of cases, and of non-cases, will be selected, and the corresponding medical charts retrieved and reviewed for validation by pairs of trained, independent reviewers. For each condition considered, case adjudication of disease will be based on symptoms, laboratory and diagnostic tests, as available in medical charts. Divergences will be resolved by consensus. Sensitivity and specificity with 95% CIs will be calculated. Ethics and dissemination Research protocol has been granted approval by the Regional Ethics Committee. Study results will be disseminated widely through peer-reviewed publications and presentations at national and international conferences. Atrial fibrillationValidationadministrative database ==== Body Strengths and limitations of this study The study will evaluate the validity of the International Classification of Diseases, 9th Revision—Clinical Modification (ICD-9-CM) codes for cerebrovascular and cardiovascular diseases (acute myocardial infarction, heart failure, atrial fibrillation and stroke), using the administrative database of Umbria Region. The strength of this study is that it will review original source documentation available in medical charts to adjudicate case of cerebrovascular and cardiovascular diseases. Once the administrative database will be validated for the considered diseases, it can be used for outcome research including pharmacoepidemiology, health service research and quality of care research. Validation studies of administrative data are related to their context, and are not immediately generalisable to other settings. Introduction As computer technology continues to advance, use of digital administrative databases is increasingly growing in healthcare settings worldwide. These databases anonymously store patients' data about healthcare assistance they received, including birth, death or disease treatment. Usually, diagnosis of the disease is associated with a specific code from the International Classification of Diseases (ICD-9) 9th or 10th revision. The ICD is designed to map health conditions to corresponding generic categories along with specific variations. Merging of individual patient data from administrative databases with other sources (eg, prescription and laboratory data) allows investigating a wide range of relevant and often unique public health questions1 monitoring population health status over time, and performing population-based pharmacoepidemiological research.1 These databases, therefore, have the potential to address important issues in postmarketing surveillance2 3 epidemiology4 quality performance, and health services research.5 However, there is a concern that the power of administrative databases as a source of healthcare information cannot be fully exploited unless they are thoroughly validated.6–9 For instance, a systematic review10 of ICD-9 code validation in Italian administrative databases reported that only a few regional databases have been validated, and just for a limited number of ICD-9 codes of diseases.11–18 While non-clinical information in healthcare databases, such as demographic and prescription data, are highly accurate,19 20 reliability of registered diagnoses and procedures is variable.20 21 Determining the accuracy of the latter two categories of clinical information is important to all potential users, and involves confirming the consistency of information within the databases with the corresponding clinical records of patients.19 Hence, it is imperative that health authorities systematically validate their databases for major diseases to productively use the information they contain. In Western countries, cardiovascular and cerebrovascular diseases are the leading cause of mortality and morbidity, and represent a major social and economic problem. In this context, most of the burden is due to occurrence of acute myocardial infarction (AMI),22 heart failure (HF),23 24 atrial fibrillation (AF)25 and stroke.17 Administrative healthcare databases can provide a comprehensive understanding of the burden of these diseases in terms of important outcomes such as mortality, hospital readmissions and use of other healthcare resources. In addition, such databases can aid in monitoring adherence to essential drug therapies, including the use of evidence-based therapies. The Regional Health Authority of Umbria has started a research regarding case definitions of diseases26–28 as well as validating ICD-9 codes for diseases.29 The objective of the present protocol is to evaluate the accuracy of the ICD-9-CM codes related to AMI, HF, AF, ischaemic and haemorrhagic stroke (intracerebral and subarachnoidal haemorrhage) in the administrative database of the Regional Health Authority of Umbria. Methods Setting and data source Administrative database The Regional Administrative Database of Umbria gathers information regarding all hospital admission medical records on all 910 000 residents, including personal demographics, hospital admission and discharge dates, vital status, hospital department, primary and secondary diagnoses and surgical or diagnostic procedures. In addition, the database records all drug prescriptions listed in the National Drug Formulary, and it allows identification of the prescriber. In Italy, all hospital medical doctors, including those who fill discharge diagnoses, are salaried. The Regional Administrative Database has been used for pharmacoepidemiology and drug-related outcome researches.2 30–32 Each resident has a unique national identification code, with which it is possible to link the various types of information, corresponding to each person, within the database. In Italy, healthcare assistance is covered almost entirely by the Italian National Health System (NHS), therefore most residents' significant healthcare information can be found within the healthcare databases. Source population Source population will be represented by permanent residents aged 18 or above of Umbria Region. Any resident that has been discharged alive (with exclusion of voluntary discharge and interhospital transfer of patients) from a hospital with a diagnosis of AMI, AF, HF and stroke (ischaemic stroke, intracerebral haemorrhage, subarachnoidal haemorrhage) will be considered. Residents that have been hospitalised outside the regional territory of competence will be excluded from analysis. Case selection and sampling method Patients with the first occurrence of below diagnosis located in primary positioning of the administrative database between 2012 and 2014 will be identified using the following groups of ICD-9-CM codes: (1) 410 for AMI; (2) 427.31 for AF and 427.32 for atrial flutter; (3) 428 for HF; (4) 433.×1, 434 (excluding 434.x0), 436 for ischaemic stroke, 430 and 431 for haemorrhagic stroke (subarachnoid haemorrhage and intracerebral haemorrhage). It is worth mentioning that according to the Italian legislation, the principal diagnosis coding needs to be based on the condition identified at the end of hospitalisation, which constitutes the main cause of the need of treatment and/or diagnostic tests, and is mainly responsible for the use of resources.33 34 Table 1 displays the description of the ICD-9-CM codes for each disease of interest. Table 1 Description of ICD-9-CM codes ICD-9-CM diagnosis code Diagnosis Description Acute myocardial infarction 410.0 Anterolateral wall 410.1 Other anterior wall 410.2 Inferolateral wall 410.3 Inferoposterior wall 410.4 Other inferior wall 410.5 Other lateral wall 410.6 True posterior wall infarction 410.7 Subendocardial infarction 410.8 Other specified sites 410.9 Unspecified site Atrial fibrillation and flutter 427.31 Atrial fibrillation 427.32 Atrial flutter Heart failure 428.0 Congestive heart failure, unspecified 428.1 Left heart failure 428.2 Systolic heart failure  428.20 Unspecified  428.21 Acute  428.22 Chronic  428.23 Acute on chronic 428.3 Diastolic heart failure  428.30 Unspecified  428.31 Acute  428.32 Chronic  428.33 Acute on chronic 428.4 Combined systolic and diastolic heart failure  428.40 Unspecified  428.41 Acute  428.42 Chronic  428.43 Acute on chronic  428.9 Heart failure, unspecified Haemorrhagic stroke   430 Subarachnoid haemorrhage 431 Intracerebral haemorrhage Ischaemic stroke   433 Occlusion and stenosis of precerebral arteries with cerebral infarction  433.01 Occlusion and stenosis of basilar artery  433.11 Occlusion and stenosis of carotid artery  433.21 Occlusion and stenosis of vertebral artery  433.31 Occlusion and stenosis of multiple and bilateral  433.81 Other specified occlusion and stenosis of precerebral artery  433.91 Unspecified occlusion and stenosis of precerebral artery 434 Occlusion of cerebral arteries with cerebral infarction  434.01 Cerebral thrombosis with cerebral infarction  434.11 Cerebral embolism with cerebral infarction  434.91 Cerebral artery occlusion, unspecified with cerebral infarction Validation criteria Criteria used for validation of the considered ICD-9-CM codes will be those derived from international guidelines if available, or systematic reviews published on these topics. Acute myocardial infarction For validation of AMI, we will consider the Guidelines of the European Society of Cardiology35 using the relative set of guidelines available at the time of each patient's discharge. To adjudicate the event as an AMI, in addition to troponin release (with at least one peak level greater than the upper limit of reference range), at least one of the following criteria of the reference standard needs to be met: Cardiac symptoms at rest (eg, such as angina, chest pain, chest discomfort, chest heaviness, or substernal chest pain), consistent with coronary artery disease/AMI. Evidence for a new AMI on the presenting ECG, as demonstrated by any one of these: ST elevation (≥1 mm) in two contiguous leads; new Q waves; new left bundle branch block. Presence of myocardial infarction/injury (exclude old myocardial infarction) on any ECG during index hospital stay. Imaging evidence of new loss of viable myocardium, or new regional wall motion abnormality. Identification of an intracoronary thrombus by angiography. Atrial fibrillation To validate ICD-9 code 427.31 (index test) related to AF, we will require at least one ECG tracing documenting the presence of AF when present in the medical record. However, in the medical literature,36 alternative criteria have been used to identify cases of AF: (1) any mention of current AF, without a reference ECG; (2) two ECG measurements documenting the presence of AF in the same hospital admission; (3) the presence of at least two ECG documenting the presence of AF in more than one hospital admissions for AF in the same period. Accordingly, we will also record and analyse charts according to these different criteria for subsequent comparison, in order to understand which algorithm gives the best diagnostic yield. Heart failure For the validation of the ICD-9-CM 428.x relating to HF, we will consider the European Society of Cardiology Heart Failure Guidelines published at the time of patient's discharge.37 Those guidelines consider the algorithm for the diagnosis of HF in the non-acute setting and in the acute setting. Since the ICD-9 codes do not distinguish between the acute and non-acute setting, we will combine the clinical presentation of both settings. Diagnosis of HF will be adjudicated when, in addition to the presence of symptoms (such as dyspnoea, orthopnoea), or presence of signs at physical examination (rales, bilateral ankle oedema, increased jugular venous pressure, displaced apical beat), at least one of the following conditions are found in the medical chart: any abnormality in resting ECG (ie, sinoatrial disease, atrioventricular block, or abnormal intraventricular conduction); plasma natriuretic peptides (NPs) concentration (elevated levels of NPs are considered brain natriuretic peptide (BNP) ≥35 pg/mL and/or N-terminal pro-brain natriuretic peptide (NT-proBNP) ≥125 pg/mL); echocardiography abnormalities (ventricular and atrial volumes and function) attributable to heart failure. Stroke For the validation of stroke, we will consider the ICD-9 codes related to stroke valid when both of the following conditions will be present:6 38 39 (1) detection of focal lesions by neurological examination; (2) imaging test (CT or MRI). Neuroimaging will be the main discriminator among the two types of stroke: negative imaging for haemorrhage will classify the case as ischaemic stroke (ICD-9-CM codes: 433.x1, 434.xx (excluding 434.x0), 436, 437.1); neuroimaging that shows the presence of haemorrhagic lesion will classify the case as haemorrhagic stroke (codes 431, 430). Chart abstraction and case ascertainment The corresponding medical charts of the randomly selected sample cases for each ICD-9 code will be obtained from hospitals for validation purposes. From each medical chart, the following information will be retrieved: unique identification patient code, date of birth, gender, dates of hospital admission and discharge, any diagnostic procedure and treatment that contributed to the diagnosis of the disease. For each target disease, we will abstract data regarding clinical, laboratory and instrumental data, including the date or dates of performance. Specifically, for AMI, we will record any cardiac signs and symptoms, cardiac enzymes (eg, troponin levels), ECG data, other diagnostic instrumental examinations (ie, angiography); for AF, we will record any cardiac signs and symptoms, ECG data or other instrumental examinations (eg, echocardiography); for HF, we will record any cardiac symptoms and signs, plasma NPs concentration, ECG data and echocardiography measurements or other instrumental examinations of interest; for stroke, we will record any neurological signs and symptoms, imaging test (CT or MRI), any other laboratory or instrumental measurement. Two medical doctors acting as chart reviewers will receive specific training on data abstraction. An initial chart review will be performed, with each reviewer independently examining the same medical charts (n=20). The interreviewer agreement between the two reviewers regarding presence or absence of the diseases considered among the pairs of reviewers will be calculated using κ statistics. This process will be reiterated until the strength of agreement among the pairs of reviewers will be optimal (κ statistics between 0.81 and 1.00). Any discrepancies between the two reviewers will be resolved by consensus and where necessary, a third expert will be involved. Case adjudication of disease within medical charts will be based on symptoms, laboratory and diagnostic tests for each cerebrovascular and cardiovascular disease considered. Statistical analysis For each condition, we anticipate that a sample of 121 charts of cases will be necessary to obtain an expected sensitivity of 80% with a precision of 8% and a power of 80% according to binomial exact calculation. For an expected specificity of 90% (precision of 8% and a power of 80%), we will randomly select from the same administrative database 73 non-cases, that is, medical records belonging to diseases of the circulatory system (ICD-9-CM codes 390–459) but without the ICD-9-codes of interest. Expected accuracy figures were based on the published literature.6–9 Sensitivity and specificity will be analysed separately for each ICD-9-CM code by constructing 2×2 tables. Sensitivity expresses the proportion of ‘true positives’ (ie, AMI classified as positive by the administrative database and medical record review) and all cases deemed positive by medical chart review. Specificity expresses the proportion of ‘true negatives’ (ie, AMI identified as negative by the administrative database and medical record review), and with all cases deemed negative by medical chart review. For both sensitivity and specificity, 95% CIs will be calculated. We account for a 10% of missing charts for sensitivity and specificity. Positive and negative predicting values will also be reported along with their 95% CI. Reporting Complete, transparent and accurate reporting is essential in research,40 because it allows readers to assess internal validity as well as to evaluate the generalisability and applicability of results.41 To ensure quality and thoroughness of reporting, any dissemination or publication of the results from the present study will follow recommended guidelines based on the criteria published by the Standards for Reporting of Diagnostic accuracy (STARD) initiative for the accurate reporting of investigations of diagnostic studies.41–43 The project will be completed within 2 years. Discussion Administrative databases constitute a valid instrument to get insights into epidemiology and health management, and to get ‘real-world’ knowledge, of major disease under situations in which randomised trials are not able to provide the required evidence, for practical or economic reasons, or because of selective inclusion criteria. Epidemiological studies are frequently based on administrative claims databases to identify cases of specific diseases, such as AMI, and often contain ICD-9 codes. These codes have the advantage of being widely available and require lower effort and cost than consulting medical charts.20 However, their effective exploitation hinges on the prerequisite that coding does reliably reflect the clinical condition of a given patient. In this protocol, we present the approach to be used to analyse the validity of ICD-9-CM codes for important cardiovascular diseases, such as AMI, HF, AF, and stroke using the Regional Health database of Umbria. We seek to verify whether an acceptable level (at least 80% for sensitivity and 90% for specificity) of operating characteristics is in fact achieved for each identified ICD-9-CM codes. In case of inadequate level of validation, algorithms will be considered to explore potential elements to enhance the level of validation. Elements of an algorithm may include the type of the data source, number of years of administrative data, other diagnostic codes, medication use or laboratory data. The algorithms selected for evaluation will be based on the literature review and consultations with clinicians, health services researchers. If the desired accuracy is not reached despite the use of algorithms, results of our research will be used to identify potential areas for improvement for future validation. Dissemination Study results will be disseminated widely through peer-reviewed publications and presentations at national and international conferences. Twitter: Follow Paolo Eusebi @paoloeusebi Contributors: IA, FC, MO, AMo and GA conceived the study. IA, FC, MO, GA, AMe, MFC, PE and AMo were responsible for designing the protocol. IA, FC, MO, GA and AMo drafted the protocol manuscript. IA, FC and MO developed the search strategy. IA, FC, MO, GA, AMe, MFC, PE and AMo critically revised the successive versions of the manuscript and approved the final version. Funding: This review protocol was funded by the Regional Health Authority of Umbria. [Progetto Data-Value: valorizzazione del dato sanitario regionale per la Ricerca dei Servizi Sanitari (Health Services Research)]. Disclaimer: The study funder was not involved in study design or writing of the protocol. Competing interests: None declared. Ethics approval: Ethics approval has been obtained from the Regional Ethics Committee of Umbria (CEAS). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Study results will be disseminated widely through peer-reviewed publications and presentations at national and international conferences. ==== Refs References 1 Jutte DP , Roos LL , Brownell MD Administrative record linkage as a tool for public health research . Annu Rev Public Health 2011 ;32 :91 –108 . 10.1146/annurev-publhealth-031210-100700 21219160 2 Traversa G , Bianchi C , Da Cas R Cohort study of hepatotoxicity associated with nimesulide and other non-steroidal anti-inflammatory drugs . BMJ 2003 ;327 :18 –22 . 10.1136/bmj.327.7405.18 12842950 3 Trifiro G , Patadia V , Schuemie MJ EU-ADR healthcare database network vs. spontaneous reporting system database: preliminary comparison of signal detection . Stud Health Technol Inform 2011 ;166 :25 –30 .21685607 4 Gini R , Francesconi P , Mazzaglia G Chronic disease prevalence from Italian administrative databases in the VALORE project: a validation through comparison of population estimates with general practice databases and national survey . BMC Public Health 2013 ;13 :15 10.1186/1471-2458-13-15 23297821 5 Colais P , Pinnarelli L , Fusco D The impact of a pay-for-performance system on timing to hip fracture surgery: experience from the Lazio Region (Italy) . BMC Health Serv Res 2013 ;13 :393 10.1186/1472-6963-13-393 24099264 6 McCormick N , Bhole V , Lacaille D Validity of diagnostic codes for acute stroke in administrative databases: a systematic review . PLoS ONE 2015 ;10 :e0135834 10.1371/journal.pone.0135834 26292280 7 McCormick N , Lacaille D , Bhole V Validity of heart failure diagnoses in administrative databases: a systematic review and meta-analysis . PLoS ONE 2014 ;9 :e104519 10.1371/journal.pone.0104519 25126761 8 McCormick N , Lacaille D , Bhole V Validity of myocardial infarction diagnoses in administrative databases: a systematic review . PLoS ONE 2014 ;9 :e92286 10.1371/journal.pone.0092286 24682186 9 Jensen PN , Johnson K , Floyd J A systematic review of validated methods for identifying atrial fibrillation using administrative data . Pharmacoepidemiol Drug Saf 2012 ;21 (Suppl 1) :141 –7 . 10.1002/pds.2317 22262600 10 Abraha I , Orso M , Grilli P The current state of validation of administrative healthcare databases in Italy: a systematic review . Int J Stat Med Res 2014 ;3 :309 –20 . 11 Galdarossa M , Vianello F , Tezza F Epidemiology of primary and secondary thrombocytopenia: first analysis of an administrative database in a major Italian institution . Blood Coagul Fibrinolysis 2012 ;23 :271 –7 . 10.1097/MBC.0b013e328351882d 22343688 12 Franchi C , Giussani G , Messina P Validation of healthcare administrative data for the diagnosis of epilepsy . J Epidemiol Community Health 2013 ;67 :1019 –24 . 10.1136/jech-2013-202528 24022813 13 Tinelli M , Mannino S , Lucchi S Healthcare-acquired infections in rehabilitation units of the Lombardy Region, Italy . Infection 2002 ;39 :353 –8 . 14 Fano V , D'Ovidio M , del Zio K [The role of the quality of hospital discharge records on the comparative evaluation of outcomes: the example of chronic obstructive pulmonary disease (COPD)] . Epidemiol Prev 2011 ;36 :172 –9 . 15 Faustini A , Canova C , Cascini S The reliability of hospital and pharmaceutical data to assess prevalent cases of chronic obstructive pulmonary disease . COPD 2012 ;9 :184 –96 . 10.3109/15412555.2011.654014 22409483 16 Bogliun G , Beghi E Validity of hospital discharge diagnoses for public health surveillance of the Guillain-Barré syndrome . Neurol Sci 2002 ;23 :113 –17 . 10.1007/s100720200036 12391495 17 Baldi I , Vicari P , Di Cuonzo D A high positive predictive value algorithm using hospital administrative data identified incident cancer cases . J Clin Epidemiol 2008 ;61 :373 –9 . 10.1016/j.jclinepi.2007.05.017 18313562 18 Schifano P , Papini P , Agabiti N Indicators of breast cancer severity and appropriateness of surgery based on hospital administrative data in the Lazio Region, Italy . BMC public health 2006 ;6 :25 10.1186/1471-2458-6-25 16464258 19 Rawson NSB , Shatin D Assessing the validity of diagnostic data in large administrative healthcare utilization databases . In: Hartzema A , Tilson H , Chan K , eds. Pharmacoepidemiology and therapeutic risk management . Harvey Whitney Books , 2008 :495 –518 . 20 West SL , Ritchey ME , Poole C Validity of pharmacoepidemiologic drug and diagnosis data . In: Strom BL , Kimmel SE , Hennessy S , eds. Pharmacoepidemiology . Wiley-Blackwell , 2012 :757 –94 . 21 Campbell SE , Campbell MK , Grimshaw JM A systematic review of discharge coding accuracy . J Public Health Med 2001 ;23 :205 –11 .11585193 22 World Health Organisation . Global Health Estimates . 2014 Summary tables: Deaths by cause, age and sex, by WHO region, 2000–2012, 2014 2014. http://www.who.int/healthinfo/global_burden_disease/estimates/en/index1.html (accessed 23 Oct 2015 ). 23 McMurray JJ , Pfeffer MA Heart failure . Lancet 2005 ;365 :1877 –89 . 10.1016/S0140-6736(05)66621-4 15924986 24 Carluccio E , Dini FL , Biagioli P The ‘Echo Heart Failure Score’: an echocardiographic risk prediction score of mortality in systolic heart failure . Eur J Heart Fail 2013 ;15 :868 –76 . 10.1093/eurjhf/hft038 23512095 25 Angeli F , Reboldi G , Garofoli M Atrial fibrillation and mortality in patients with acute myocardial infarction: a systematic overview and meta-analysis . Curr Cardiol Rep 2012 ;14 :601 –10 . 10.1007/s11886-012-0289-3 22821004 26 Abraha I , Giovannini G , Serraino D Validity of breast, lung and colorectal cancer diagnoses in administrative databases: a systematic review protocol . BMJ Open 2016 ;6 :e010409 10.1136/bmjopen-2015-010409 27 Montedori A , Abraha I , Chiatti C Validity of peptic ulcer disease and upper gastrointestinal bleeding diagnoses in administrative databases: a systematic review protocol . BMJ Open 2016 ;6 :e011776 10.1136/bmjopen-2016-011776 28 Rimland JM , Abraha I , Luchetta ML Validation of chronic obstructive pulmonary disease (COPD) diagnoses in healthcare databases: a systematic review protocol . BMJ Open 2016 ;6 :e011777 10.1136/bmjopen-2016-011777 29 Abraha I , Serraino D , Giovannini G Validity of ICD-9-CM codes for breast, lung and colorectal cancers in three Italian administrative healthcare databases: a diagnostic accuracy study protocol . BMJ Open 2016 ;6 :e010547 10.1136/bmjopen-2015-010547 30 Menniti-Ippolito F , Maggini M , Raschetti R Ketorolac use in outpatients and gastrointestinal hospitalization: a comparison with other non-steroidal anti-inflammatory drugs in Italy . Eur J Clin Pharmacol 1998 ;54 :393 –7 .9754982 31 Abraha I , Montedori A , Stracci F Statin compliance in the Umbrian population . Eur J Clin Pharmacol 2003 ;59 :659 –61 . 10.1007/s00228-003-0675-2 14508622 32 Raschetti R , Maggini M , Da Cas R Time trends in the coprescribing of cisapride and contraindicated drugs in Umbria, Italy . JAMA 2001 ;285 :1840 –1 .11308395 33 Italian Ministerial Decree no. 380 of 27 October 2000. http://wwwprivacyit/dm2000380html 34 Guidelines for national hospital discharge data coding, approved by the Italian State-Regions Conference www.statoregioni.it/DettaglioDoc.asp?IDDoc=26555&IdProv=8266&tipodoc=2&CONF=CSR (rep. Acts 64 / CSR of 29 April 2010) . 35 Thygesen K , Alpert JS , Jaffe AS Third universal definition of myocardial infarction . Eur Heart J 2012 ;33 :2551 –67 . 10.1093/eurheartj/ehs184 22922414 36 Camm AJ , Lip GY , De Caterina R 2012 focused update of the ESC Guidelines for the management of atrial fibrillation: an update of the 2010 ESC Guidelines for the management of atrial fibrillation. Developed with the special contribution of the European Heart Rhythm Association . Eur Heart J 2012 ;33 :2719 –47 . 10.1093/eurheartj/ehs253 22922413 37 McMurray JJ , Adamopoulos S , Anker SD ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure 2012: The Task Force for the Diagnosis and Treatment of Acute and Chronic Heart Failure 2012 of the European Society of Cardiology. Developed in collaboration with the Heart Failure Association (HFA) of the ESC . Eur Heart J 2012 ;33 :1787 –847 . 10.1093/eurheartj/ehs104 22611136 38 Stroke . National clinical guideline for diagnosis and initial management of acute stroke and Transient Ischaemic Attack (TIA) . Royal College of Physicians, UK London , 2008 . 39 Sacco RL , Kasner SE , Broderick JP An updated definition of stroke for the 21st century: a statement for healthcare professionals from the American Heart Association/American Stroke Association . Stroke 2013 ;44 :2064 –89 . 10.1161/STR.0b013e318296aeca 23652265 40 Macura A , Abraha I , Kirkham J Selective outcome reporting: telling and detecting true lies. The state of the science . Intern Emerg Med 2010 ;5 :151 –5 . 10.1007/s11739-010-0371-z 20300879 41 Bossuyt PM , Reitsma JB , Bruns DE Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD initiative . BMJ 2003 ;326 :41 –4 . 10.1136/bmj.326.7379.41 12511463 42 Benchimol EI , Manuel DG , To T Development and use of reporting guidelines for assessing the quality of validation studies of health administrative data . J Clin Epidemiol 2011 ;64 :821 –9 . 10.1016/j.jclinepi.2010.10.006 21194889 43 De Coster C , Quan H , Finlayson A Identifying priorities in methodological research using ICD-9-CM and ICD-10 administrative data: report from an international consortium . BMC Health Serv Res 2006 ;6 :77 10.1186/1472-6963-6-77 16776836

PMC005xxxxxx/PMC5372119.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01415410.1136/bmjopen-2016-014154Qualitative ResearchResearch1506172516961715168417341694Understanding Dutch practice nurses' adherence to evidence-based smoking cessation guidelines and their needs for web-based adherence support: results from semistructured interviews de Ruijter D 1Smit E S 2de Vries H 1Goossens L 1Hoving C 1 1 Department of Health Promotion, Care and Public Health Research Institute (CAPHRI), Maastricht University, Maastricht, The Netherlands 2 Department of Communication Science, Amsterdam School of Communication Research (ASCoR), University of Amsterdam, Amsterdam, The NetherlandsCorrespondence to Dr C Hoving; c.hoving@maastrichtuniversity.nl2017 22 3 2017 7 3 e0141546 9 2016 29 12 2016 16 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives Practice nurses in general practices suboptimally adhere to smoking cessation guidelines. Since the effectiveness of their smoking cessation support is greatest when full adherence to these guidelines is achieved, interventions need to be developed to improve practice nurses' guideline adherence, for example, by tailoring their content to adherence determinants. However, the sociocognitive determinants explaining adherence have not yet been investigated. Therefore, this qualitative needs assessment aimed to explore practice nurses' current counselling practices, as well as their sociocognitive beliefs related to their smoking cessation guideline adherence and their needs regarding web-based adherence support. Setting Primary care; general practices in the Netherlands. Participants 19 practice nurses, actively involved in smoking cessation counselling. Methods Semistructured individual interviews, based on the I-Change Model and the Diffusion of Innovations Theory, were conducted from May to September 2014. Data were systematically analysed using the Framework Method and considered reliable (κ 0.77; % agreement 99%). Results Respondents felt able to be empathic and collaborative during smoking cessation consultations. They also reported psychological (eg, low self-efficacy to increase patient motivation and arranging adequate follow-up consultations) and practical barriers (eg, outdated information on quit support compensation and a perceived lack of high-quality trainings for practice nurses) to smoking cessation guideline adherence. Most respondents were interested in web-based adherence support to overcome these barriers. Conclusions Sociocognitive determinants influence practice nurses' smoking cessation guideline adherence. To improve their adherence, web-based tailored adherence support can provide practice nurses with personally relevant feedback tailored to individually perceived barriers to smoking cessation guideline adherence. More specifically, low self-efficacy levels can be increased by peer modelling (eg, presenting narratives of colleagues) and up-to-date information can be presented online, enabling practice nurses to use it during patient consultations, resulting in more effective communication with their smoking patients. Trial registration number NTR4436; Pre-results. QUALITATIVE RESEARCHPRIMARY CAREPREVENTIVE MEDICINEsmoking cessationtailored supportKWF Kankerbestrijdinghttp://dx.doi.org/10.13039/501100004622UM2013-6107 ==== Body Strengths and limitations of this study Sociocognitive determinants of practice nurses' smoking cessation guideline adherence were not investigated yet. Interviews explored determinants of smoking cessation guideline adherence, both from a user's and from an innovation's perspective. Qualitative data were systematically coded and analysed using the Framework Method. Despite purposeful sampling, participants might differ from the ‘average’ practice nurse as they were possibly more involved in smoking cessation. Introduction Globally, smoking continues to be the leading cause of preventable disease and premature death.1 2 Since 23% of the Dutch adult population still smokes,3 successful smoking cessation strategies are needed. Advice and counselling by a general practitioner (GP) is a cost-effective strategy to increase patients' quit rates.4 However, a shift in cessation counselling can be observed, where it is increasingly provided by trained practice nurses (PNs) in Dutch general practices.5 6 PNs are predominantly responsible for chronic patient care and lifestyle counselling, applying evidence-based guidelines. PNs are highly educated (ie, college degree) and employed in 80% of Dutch general practices.7 Regarding smoking cessation, PNs are trained to use one of several national smoking cessation guidelines that include similar counselling steps (figure 1) and of which the STIMEDIC guideline is the most recent.8 In the Netherlands, the ‘Stop Smoking Partnership’, an alliance of several parties involved in smoking cessation in the healthcare sector, is responsible for developing and regularly updating national smoking cessation guidelines. These guidelines can be used by healthcare professionals to structure consultations with smoking patients and full adherence to these evidence-based guidelines leads to more effective health communication and hence positively contributes to quality of smoking cessation care.9 10 Adequate guideline adherence is also known to have stronger effects on patients' quit rates than a brief quit advice.11 Figure 1 Flow chart Dutch smoking cessation guidelines (adapted from guideline treatment tobacco addiction revision 20098). However, PNs regularly report suboptimal adherence to smoking cessation guidelines,12–14 due to practical factors such as time constraints or the perception that most smoking patients are unwilling to quit.12 15 16 Moreover, research has especially shown that sociocognitive determinants consistently influence guideline adherence among healthcare professionals. A study among cardiac nurses working in a hospital17 found significant positive correlations between self-efficacy expectations, the perceived simplicity and advantages of the guideline, and intention towards continued use of a smoking cessation guideline. Relatedly, Segaar et al13 concluded that a positive attitude, positive social influences and higher self-efficacy best explained adoption of the same guideline by cardiac nurses. Though these studies illustrate the importance of sociocognitive determinants in explaining smoking cessation guideline adoption and adherence, such data are not available for PNs working in general practices. It might be that different determinants influence the behaviour of PNs as they have different types of training and expertise, and work in a different setting with other types of patients compared with cardiac nurses. Therefore, it is important to investigate PNs' current counselling practices and the role of sociocognitive determinants in explaining their smoking cessation guideline adherence in a general practice setting. Web-based computer tailored (CT) programmes have been shown to be able to effectively change various health behaviours and sociocognitive determinants among the general population and patient groups.18–21 CT programmes provide content adapted to characteristics (eg, demographic factors and sociocognitive determinants) of the individual user.22 A CT programme can, for instance, provide an individual with personal feedback about their behaviour, which is based on that individual's answers to a questionnaire. Consequently, individually tailored content is personally relevant and more likely to be read and remembered compared with more static tools such as leaflets or generic websites.22 Moreover, through web-based support, the intervention elements can be provided when and where PNs require them at a relatively low cost.23 24 Offering a minimally time-consuming approach enables PNs to integrate adherence support in their busy schedules, which was found to be important when asking nurses regarding their preferences for smoking cessation training.25 26 However, CT support has only been sparsely applied to increase smoking cessation guideline adherence among healthcare professionals and none have been offered in a web-based format. One study assessed the use of printed CT advice to increase smoking cessation counselling,27 demonstrating improved guideline implementation by physicians and increased patients' quit rates 6 months postintervention. Despite these promising results of printed CT advice among physicians, until now neither printed nor web-based CT support exists to stimulate smoking cessation guideline adherence among PNs. To develop a web-based CT adherence support program for PNs, it is essential to assess PNs' current counselling practices and how determinants influence their adherence to smoking cessation guidelines, and to what extent. Furthermore, PNs' needs regarding the content and design of web-based support need to be assessed to ensure compatibility with their daily practice, as this contributes to higher exposure to an intervention's content.28 Therefore, we describe a qualitative needs assessment aimed at exploring PNs' current counselling practices, the sociocognitive determinants of their smoking cessation guideline adherence and their needs regarding web-based adherence support. Methods Study design Data were gathered by means of semistructured interviews with individual PNs (N=19), because individual interviews yield more in-depth information than group interviews.29 PNs provided oral informed consent. Interview recordings and verbatim transcriptions were anonymised and stored in a password-protected database. Evaluation of this study by the Medical Ethics Committee Atrium-Orbis-Zuyd (14-N-17) revealed that no medical ethical clearance for this study was needed according to the rules of the Medical Research Involving Human Subjects Act (WMO). The study is registered with the Dutch Trial Register (NTR4436). Participants and procedure Dutch PNs actively engaged in smoking cessation counselling in a general practice were recruited via telephone between May and September 2014. We invited a heterogenic sample of PNs to obtain diverse responses to the interview questions. Through purposeful sampling, we included PNs from (1) different geographical regions in the Netherlands; (2) different types of practices (urban and rural); and (3) practices with different sizes (in terms of patients and employees). Furthermore, recruited PNs differed in their (4) use of smoking cessation guidelines; (5) amount of working experience and (6) educational level. Recruitment continued until data saturation was reached; PNs did not introduce any new issues during the interviews at this point. Successfully contacted PNs who declined to participate (n=7) mentioned a lack of time and interest in the study as main reasons. All participating PNs were rewarded with a €10 gift card. Interviews were conducted at the PNs' workplace (n=17) or via telephone (n=2) and only the PN and one or two researchers (ie, DdR and LG) were present. No prior relationship between PNs and researchers was established before conducting the interviews. Both researchers had prior experience with conducting individual (ie, DdR) and group interviews (ie, LG). Before starting the interviews, PNs were informed about the aims of the interview. Interviews were audio-recorded and notes were taken to keep track of the issues and themes that were covered during each interview. Interviews lasted for 25–62 min (m=43 min). The interview guide A semistructured interview guide was developed based on the I-Change Model30 and the Diffusion of Innovations Theory,31 a combination which has been shown to be successful in explaining the implementation of smoking cessation interventions by primary healthcare professionals.13 32 In the interview guide, follow-up questions relating to PNs' current counselling practices and prompt questions regarding the STIMEDIC guideline were especially informed by the concepts attitude, social influence, self-efficacy, perceived barriers and skills from the I-Change Model.30 Questions relating to PNs' needs for web-based support (ie, the innovation) were guided by Rogers'31 five innovation characteristics: the relative advantage, compatibility, complexity, trialability and observability of the innovation. Two primary care professionals engaged in smoking cessation counselling pretested the interview guide, which was adapted according to their feedback. The final interview guide (appendix 1) consisted of open-ended questions and prompts to encourage PNs to discuss their current smoking cessation practices, guideline adherence and desired programme characteristics. 10.1136/bmjopen-2016-014154.supp1supplementary appendix Interviews were started with an assessment of PNs' personal (age, gender, personal smoking history and professional education) and work-related characteristics (years of experience as PN, the number of practices employed in, and the size, location and type of the(se) general practice(s)). Subsequently, PNs were asked to describe their smoking cessation counselling routine and discuss barriers encountered during counselling. Third, the STIMEDIC smoking cessation guideline was introduced and PNs were asked to identify guideline steps that they would find difficult to perform and to describe situations they might find difficult to handle when applying these steps. Finally, PNs' needs were explored, including interest in and preferences regarding the content and design of web-based adherence support aiming to improve their smoking cessation counselling. Data analysis Data analysis using the Framework Method33 was conducted using Nvivo (V.9). First, interviews were transcribed verbatim and two coders (ie, DdR and LG) familiarised themselves with the interview content by listening to audio records of the interviews and reading transcripts. Next, both coders independently developed a coding tree by analysing the same single transcript and the coding trees were compared for consistency;34 any inconsistencies were discussed and resolved. Then three similar rounds of independent coding of three additional transcripts resulted in a final coding tree consisting of multiple major and minor themes for every main interview question. This coding tree was perceived to cover all relevant information and agreed on by both coders. Subsequently, both coders independently applied the coding tree to the same five randomly selected interview transcripts and the intercoder reliability was assessed, resulting in a value for per cent agreement and Cohen's κ. Per cent agreement reflects the degree of similarity between coders in assigning the same code to the same piece of text34 and a coefficient of ≥90% is considered acceptable.35 36 Cohen's κ takes into account that agreement between coders might occur due to chance and is therefore a more conservative coefficient;35 a κ of ≥0.70 is considered acceptable.35 It was determined that the intercoder agreement was sufficiently high (% agreement 99%; Cohen's κ 0.72). Nonetheless, coding inconsistencies were again discussed and resolved. Thereafter, the first coder (ie, DdR) completed coding the remaining transcripts (n=14), whereas the second coder (ie, LG) only coded the final 10% of the transcripts (n=2). This revealed an improved intercoder reliability of 99% (% agreement) and 0.77 (Cohen's κ). Once all transcripts were coded, themes were grouped together and data were clustered based on their importance. Finally, data clusters were interpreted by looking for patterns and identifying answers to the research questions. Results Participating PNs were on average 46.1 years old and mostly female; the majority had never smoked (table 1). All PNs had obtained a Bachelor's degree in nursing, while some were additionally trained in pulmonary or diabetic patient care. PNs' mean work experience was 8.3 years; they were mostly employed at urban practices and reported a mean number of 8.4 employees (range 4–20) and a patient population of nearly 4500 patients (range 500–12 000) per practice. Table 1 Descriptive characteristics of participating PNs (N=19) Personal characteristics Mean age in years (SD) 46.1 (9.2) Female (%) 17 (89.5%) Smoking status  Never smoked (%) 11 (57.9%)  Ex-smoker (%) 8 (42.1%) Professional education  Postbachelor PN training (%) 19 (100%)  Specialised diabetes nurse (%) 3 (15.8%)  Specialised pulmonary nurse (%) 4 (21.1%) Mean working experience as PN in years (SD) 8.3 (4.0) Practice characteristics Urban practice location (%) 12 (63.2%) Mean number of employees in practice (SD) 8.4 (4.4) Mean patient visits per year* (SD) 4478.4 (3016.4) Counselling elements Adherence by PNs (n)† Provide a quit advice 10 (52.6%) Assess patients' smoking profile 18 (94.7%) Assess patients’ motivation 18 (94.7%) Increase patients’ motivation 7 (36.8%) Assess patients’ barriers 15 (78.9%) Remove patients’ barriers 7 (36.8%) Discuss use of cessation aids 16 (84.2%) Develop a quit plan 16 (84.2%) Follow-up after the quit date 18 (94.7%) *Standardised practice capacity: 2168 patients per full-time general practitioner. †The data represent the number (and percentage) of adherent PNs. PN, practice nurse. Current smoking cessation counselling practices All PNs reported using a national guideline to structure their counselling, but described varying levels of guideline adherence. Smoking cessation counselling would usually be initiated by GPs' patient referral to a PN (ie, GP provides a short quit advice, then refers to PN) or by PNs themselves giving patients a brief quit advice, especially ‘people who visit with lung problems or diabetes’. After a quit advice, additional counselling was predominantly conducted by PNs, mostly by first assessing their patients' smoking profile (eg, amount smoked, smoking pattern, addiction level) during an intake interview. Most PNs also reported assessing patients' motivation to quit, but more than half indicated discontinuing counselling if “they [patients] are not internally motivated…Becoming internally motivated often takes longer than quitting itself.” Those PNs attempting to actively increase their patients' motivation used techniques like reflecting on patients' intrinsic reasons for quitting and providing patients with homework assignments.When they go home I ask [the patient] to write down their personal pros and cons of smoking… When they come back we especially discuss alternatives for the pros of smoking. The majority of PNs reported assessing patients' barriers for quitting, but less than half also attempted to remove present barriers, which was usually done by discussing distraction strategies to deal with craving.Craving for cigarettes only lasts a minute…I especially discuss distraction strategies to deal with this minute. Most PNs discussed the use of cessation aids (eg, nicotine replacement therapy, medication, group counselling), especially focusing on their working mechanisms and possible side effects. Patients' preferences often influenced the decision to prescribe a particular cessation aid.Most patients know about Champix (varenicline) and specifically request it because they have heard about it from others. Subsequently, the majority of PNs reported developing a quit plan with their patients, including a quit date, the use of cessation aids and follow-up appointments. They described that the duration, frequency and format of follow-up appointments after the quit date strongly depend on patients' preferences and their level of motivation, ranging from multiple face-to-face appointments to a single telephone appointment.First, patients often did not return…By planning follow-up appointments more frequently…I have more success in [helping them] quitting smoking. Concerning their role as a smoking cessation counsellor, PNs perceived themselves able to ‘adapt counselling to a (patient's) personal situation’, to stimulate patients to make their own decisions by weighing personal pros and cons, and to provide patients with relevant information ‘which is evidence-based’ about the disadvantages of smoking. Experienced barriers during counselling PNs found it difficult to handle unmotivated patients, because they struggle to decide how much they should try to increase their patients' motivation. Since they believed that the success of their counselling mainly depends on their patients' level of knowledge of and awareness about possibilities for smoking cessation and not so much on their own efforts to enhance patients' motivation, PNs sometimes decided to only provide patients with awareness-enhancing information.If a patient reports not to need counseling, I experience that it [counseling] has no effect. I am glad when they [patients] agree to take home some information. Several PNs described that this often happens after GPs directly referred patients without first assessing patients' motivation or providing a brief quit advice. They believed GPs' advice could be motivation-enhancing for patients, but felt ‘it (GPs’ quit advice) does not happen enough’. If the counselling process was continued, PNs did not attempt to remove patients' barriers when their patients' motivation was low, as they perceived that ‘unmotivated patients should first work on their motivation themselves’. PNs generally believed that a sufficient level of motivation is first needed (before dealing with the presence of barriers), but that achieving this is mainly the patient's own responsibility. Regarding follow-up after the quit date, PNs felt that smoking cessation guidelines lack detailed instructions on which topics to discuss during follow-up appointments.You call them [patients] or they visit the practice and of course you talk about their quit attempt, but there is no specific plan. Moreover, they described that patients do not always perceive the need for follow-up and regularly skip or cancel scheduled appointments, making it difficult for PNs to ‘get them (patients) to return to practice’ for follow-up and to determine their success rate, since ‘it is their (the patients’) own responsibility’. Several PNs reported difficulties staying up to date regarding smoking cessation counselling as they perceived a low supply of high-quality training classes. Furthermore, associated monetary and time investments were considered barriers for attending classes that do exist. Moreover, most PNs reported troubles with staying informed about compensation of counselling by different health insurances, as they all have their own regulations (eg, a maximum number of consultations), which change regularly. Several PNs reported that ‘online medication orders are not delivered on time’, which means that ‘a well-prepared quit attempt falls to pieces’. Obtaining reliable information about both health insurances and ordering online prescriptions was described as ‘very time-consuming and sometimes leading to confusion’. Barriers for providing patients with educational materials included a lack of time and high-quality brochures, containing educational and visually attractive content. Moreover, patients were perceived as often not using provided materials and consequently lacking awareness of and knowledge about smoking cessation aids and counselling.It is not the case that you take a single pill of Champix and suddenly do not like cigarettes anymore…Sometimes it is necessary to adjust this expectation. Needs regarding adherence support Most PNs stated interest in web-based support to improve their smoking cessation counselling. Several PNs were already familiar with other web-based programmes for primary care professionals and positively evaluated these programmes. They expressed interest in being able to “browse through a program and choose only those things that I can personally apply during counseling.” Regarding compatibility, PNs expected the content to be up to date, individually relevant and applicable in practice. Moreover, a programme should be easy to use (low complexity) and require little time investment.I want to know the latest developments about smoking cessation…and easily retrieve them from a web-based program, without having to search through databases. Long questionnaires and extensive texts were perceived as potential disadvantages and availability after working hours was perceived as a potential advantage of web-based support. Furthermore, PNs wanted a programme free to visit and revisit at their convenience without obligations (high trialability). Visually attractive content (ie, images or videos) and experiences of other PNs were preferred over a text-only form of support (relative advantage).Experiences of others would be interesting…[]…so you know what was successful and which strategies did not work. Most PNs also expressed a need for printable patient-centred educational materials to use during counselling, such as an overview of their patients' counselling trajectory or a timeline depicting physical advantages of quitting smoking.The information should be suitable for patients…simple and clear, without complicated terminology. Preferably a combination of text and figures. Discussion and conclusion Our results add a novel perspective to the scientific literature on health communication by describing PNs' smoking cessation counselling practices, determinants of their adherence to evidence-based guidelines as well as their needs regarding a web-based adherence support programme. Results indicate that while all PNs reported the use of an evidence-based guideline, their guideline adherence was suboptimal, which was influenced by psychological and practical barriers. Most PNs showed interest in web-based adherence support that could help them to overcome these barriers and reported specific needs that could be addressed to ensure a programme's optimal compatibility to their work situation. Discussion Despite specific difficulties that PNs perceive regarding full adherence to smoking cessation guidelines, they also reported being confident that they could adapt their counselling to the situation of smoking patients, to stay close to a patient's preferences and beliefs with respect to a quit attempt. In other words, PNs felt able to provide empathic counselling in a collaborative way, which are core elements of motivational interviewing,37 and hence an important competence to possess and apply during smoking cessation consultations. Nevertheless, PNs also reported psychological and practical barriers regarding their adherence to evidence-based guidelines. A psychological barrier to guideline adherence was PNs' perceived inability to motivate patients to quit smoking; they reported to rather provide awareness-enhancing information than to actively try to increase patients' motivation, a finding also described in studies investigating nurses' adherence to motivational interviewing techniques.38 39 PNs' perceived inability to motivate smokers could indicate a lack of self-efficacy, which was also found in other studies investigating determinants of nurses' use of guidelines or protocols.17 40 Low self-efficacy could possibly originate from PNs' perceptions that they are of minor influence on patients' intrinsic motivation (ie, low confidence that counselling will make a difference). Relatedly, PNs indicated that they often did not actively try to remove barriers when dealing with unmotivated patients. However, patients' lack of intrinsic motivation could be the result of seeing many barriers for quitting and hence not discussing them could be a missed opportunity to increase patients' motivation to quit.41 Furthermore, PNs reported difficulties in ensuring their patients' presence at follow-up appointments. This potentially originates from their perception that it is their patients' responsibility to show up. Consequently, they do not repeatedly contact their patients about missed appointments, as was also found in a previous study among primary care professionals' smoking cessation counselling practices.16 Concerning practical barriers for guideline adherence, PNs perceived a lack of high-quality smoking cessation training classes and reported a lack of uniformity about health insurances' compensation policies. Accordingly, PNs had difficulties providing patients with personally relevant and up-to-date information. Additionally, retrieving information about such topics was considered to be very time-consuming, though necessary to enable patients to make an informed choice about their quit attempt. Such time-consuming tasks may cause time constraints regarding actual cessation counselling activities, a commonly mentioned barrier among primary care professionals,15 42 and consequently interfere with PNs' adherence to smoking cessation guidelines. When considering solutions for the identified barriers, most PNs were interested in web-based adherence support. To overcome these barriers, PNs expressed the need for tailored support (ie, instead of offering generic information), compatible with their situation and with low complexity and high trialability. PNs' perceived accessibility after office hours as a potential advantage of the web-based nature of such a programme. These findings are in line with previous studies that related innovation characteristics to increased adoption and implementation of novel interventions by nurses.43 44 Furthermore, PNs would like to be supported with practical tools (eg, printable patient materials) to apply during counselling and hence improve effective communication about smoking cessation. Practice implications The development of web-based adherence support seems suitable to help PNs to overcome psychological and practical barriers to improve their smoking cessation guideline adherence. That is, a web-based CT programme could provide PNs with personal feedback about performing specific counselling elements they find difficult to apply, based on evidence-based methodologies to change sociocognitive determinants as described elsewhere.45 For instance, role modelling could be used to increase self-efficacy, which can be operationalised by presenting PNs with narratives of peers (ie, fellow PNs) via a web-based programme.46 Targeting sociocognitive determinants through a tailored web-based support programme is a well-established method to successfully influence these determinants and ultimately behaviour change.22 47 Furthermore, tailored feedback can be supplemented with up-to-date information about smoking cessation counselling in primary care, like an overview of available training classes or health insurances' compensation policies. This could reduce PNs' time needed to find this information and consequently increase their time spent on counselling. Finally, PNs can be supported with practical tools, like visually attractive educational materials for lower educated patients and patients with different cultural backgrounds. Such strategies potentially enable PNs to optimally adhere to evidence-based guidelines and subsequently promote the quality of their smoking cessation care. Strengths and limitations A strength of our study was that the interview guide was based on the I-Change Model30 and the Diffusion of Innovations Theory.31 This enabled us to explore determinants of smoking cessation guideline adherence from both the users' (ie, PNs) and the innovation's (ie, web-based adherence support) perspective, a combination which has previously been associated with nurses' adoption and implementation behaviour of intervention programmes.13 43 Second, owing to the semistructured nature of the interview guide, PNs had ample opportunity to elaborate on their experiences during smoking cessation counselling, providing a vast amount of data per interview. Moreover, the interviews enabled us to assess PNs' individual preferences for web-based adherence support to optimise such a programme's compatibility.48 Third, qualitative data were systematically analysed using the Framework Method, ensuring optimal coverage of relevant content through the development of an extensive coding tree.33 Moreover, intercoder reliability was high, indicating that the coding was sufficiently objective and valid. A limitation of the study was that PNs interested in smoking cessation research may have been more likely to participate in the interviews than non-interested PNs. This may have threatened the generalisability of the results to the total population of PNs. Nonetheless, by purposeful sampling, we recruited a heterogenic group of PNs who provided a great variability of responses. Moreover, data saturation occurred within the 19 interviews conducted. Conclusion Dutch PNs experienced several barriers to completing smoking cessation guideline adherence. They reported low self-efficacy to motivate patients and to remove patients' barriers to quit, and struggled with successfully organising patient follow-up. Moreover, a perceived lack of smoking cessation training classes and uniform insurance policies was identified as practical barriers for optimal guideline adherence. PNs were interested in web-based adherence support to overcome these barriers, especially if a web-based programme is easy to use, free to (re)visit and compatible with their current smoking cessation practices. Taking PNs' needs into account when developing a programme is likely to stimulate the programme's usage and maximise exposure to its content, making it more likely to be effective in improving PNs' smoking cessation guideline adherence. The authors wish to thank all participating practice nurses for their time. They would also like to thank the research assistants for assisting with transcribing the interview audiotapes. Contributors: All authors significantly contributed to the drafting of this manuscript. ESS, HdV andCH developed the proposal for the present research. DdR was responsible for conducting and reporting on the research. LG assisted in data collection and analysis and ESS, HdV and CH supervised the process and provided feedback. Funding: This work was supported by the Dutch Cancer Society, grant number UM2013-6107. Competing interests: None declared. Ethics approval: Medical Ethics Committee Atrium-Orbis-Zuyd. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Requests to share anonymised data will be considered. Please contact the first author at d.deruijter@maastrichtuniversity.nl ==== Refs References 1 WHO . WHO report on the global tobacco epidemic 2013: enforcing bans on tobacco advertising, promotion and sponsorship . Geneva : World Health Organization , 2013 . 2 USDHHS . The health consequences of smoking—50 years of progress. A report of the surgeon general . Atlanta, GA: USA Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health , 2014 . 3 RIVM . Volksgezondheid Toekomst Verkenning (VTV) ‘Een gezonder Nederland’ . Bilthoven : 2014 . 4 Stead LF , Buitrago D , Preciado N Physician advice for smoking cessation . Cochrane Database Syst Rev 2013 ;(5) :CD000165 10.1002/14651858.cd000165.pub4 23728631 5 Nederlands Huisartsen Genootschap & Landelijke Huisarsten Vereniging . NHG/LHV Standpunt: Het (ondersteunend) Team in de huisartsenvoorziening . Utrecht : Nederlands Huisartsen Genootschap and Landelijke Huisartsen Vereniging , 2011 . 6 Freund T , Everett C , Griffiths P Skill mix, roles and remuneration in the primary care workforce: who are the healthcare professionals in the primary care teams across the world? Int J Nurs Stud 2015 ;52 :727 –43 . 10.1016/j.ijnurstu.2014.11.014 25577306 7 Heiligers PJM , Noordman J , Korevaar JC Kennisvraag: praktijkondersteuners in de huisartspraktijk (POH's), klaar voor de toekomst? Utrecht : NIVEL, 2012 . 8 Stop Smoking Partnership . Guideline treatment tobacco addiction [Richtlijn behandeling van tabaksverslaving] . Alphen aan den Rijn, The Netherlands : 2009 . 9 Miller M , Kearney N Guidelines for clinical practice: development, dissemination and implementation . Int J Nurs Stud 2004 ;41 :813 –21 . 10.1016/j.ijnurstu.2003.09.005 15288803 10 Woolf SH , Grol R , Hutchinson A Clinical guidelines: potential benefits, limitations, and harms of clinical guidelines . BMJ 1999 ;318 :527 –30 . 10.1136/bmj.318.7182.527 10024268 11 Rice VH , Hartmann-Boyce J , Stead LF Nursing interventions for smoking cessation . Cochrane Database Syst Rev 2013 ;(8) :CD001188 10.1002/14651858.cd001188.pub4 23939719 12 Whitehead D , Zucker SB , Stone J Tobacco cessation education for advanced practice nurses . Nurse Educ 2014 ;39 :252 –5 . 10.1097/NNE.0000000000000056 24867074 13 Segaar D , Willemsen MC , Bolman C Nurse adherence to a minimal-contact smoking cessation intervention on cardiac wards . Res Nurs Health 2007 ;30 :429 –44 . 10.1002/nur.20204 17654478 14 Glasgow RE , Lichtenstein E , Marcus AC Why don't we see more translation of health promotion research to practice? Rethinking the efficacy-to-effectiveness transition . Am J Public Health 2003 ;93 :1261 –7 . 10.2105/AJPH.93.8.1261 12893608 15 Studts JL , Flynn SM , Dill TC Nurse practitioners’ knowledge, attitudes, and clinical practices regarding treatment of tobacco use and dependence . J Nurse Pract 2010 ;6 :212 –19 . 10.1016/j.nurpra.2009.06.003 16 van Rossem C , Spigt MG , Kleijsen JR Smoking cessation in primary care: exploration of barriers and solutions in current daily practice from the perspective of smokers and healthcare professionals . Eur J Gen Pract 2015 ;21 :1 –7 . 10.3109/13814788.2014.990881 17 Bolman C , de Vries H , Mesters I Factors determining cardiac nurses’ intentions to continue using a smoking cessation protocol . Heart Lung 2002 ;31 :15 –24 . 10.1067/mhl.2002.119834 11805745 18 Lustria MLA , Noar SM , Cortese J A meta-analysis of web-delivered tailored health behavior change interventions . J Health Commun 2013 ;18 :1039 –69 . 10.1080/10810730.2013.768727 23750972 19 Krebs P , Prochaska JO , Rossi JS A meta-analysis of computer-tailored interventions for health behavior change . Prev Med 2010 ;51 :214 –21 . 10.1016/j.ypmed.2010.06.004 20558196 20 Kohl LFM , Crutzen R , de Vries NK Online prevention aimed at lifestyle behaviors: a systematic review of reviews . J Med Internet Res 2013 ;15 :e146 10.2196/jmir.2665 23859884 21 Civljak M , Stead LF , Hartmann-Boyce J Internet-based interventions for smoking cessation . Cochrane Database Syst Rev 2013 ;(7) :CD007078 10.1002/14651858.cd007078.pub4 23839868 22 Kreuter MW , Wray RJ Tailored and targeted health communication: strategies for enhancing information relevance . Am J Health Behav 2003 ;27 (Suppl 3) :S227 –32 . 10.5993/AJHB.27.1.s3.6 14672383 23 Griffiths F , Lindenmeyer A , Powell J Why are health care interventions delivered over the internet? A systematic review of the published literature . J Med Internet Res 2006 ;8 :e10 10.2196/jmir.8.2.e10 16867965 24 Tate DF , Finkelstein EA , Khavjou O Cost Effectiveness of Internet Interventions: review and recommendations . Ann Behav Med 2009 ;38 :40 –5 . 10.1007/s12160-009-9131-6 19834778 25 Smith PM , Sellick SM , Spadoni MM Tobacco cessation Clinical Practice Guideline use by rural and urban hospital nurses: a pre-implementation needs assessment . BMC Nurs 2012 ;11 :6 10.1186/1472-6955-11-6 22545579 26 Katz DA , Paez MW , Reisinger HS Implementation of smoking cessation guidelines in the emergency department: a qualitative study of staff perceptions . Addict Sci Clin Pract 2014 ;9 :1 10.1186/1940-0640-9-1 24460974 27 Unrod M , Smith M , Spring B Randomized controlled trial of a computer-based, tailored intervention to increase smoking cessation counseling by primary care physicians . J Gen Intern Med 2007 ;22 :478 –84 . 10.1007/s11606-006-0069-0 17372796 28 Schneider F , van Osch L , de Vries H Identifying factors for optimal development of health-related websites: a Delphi study among experts and potential future users . J Med Internet Res 2012 ;14 :e18 10.2196/jmir.1863 22357411 29 Bourgeault I , Dingwall R , Vries RD The SAGE handbook of qualitative methods in health research . London: SAGE Publications Ltd , 2010 :186 . 30 de Vries H , Mudde A , Leijs I The European Smoking Prevention Framework Approach (EFSA): an example of integral prevention . Health Educ Res 2003 ;18 :611 –26 . 10.1093/her/cyg031 14572020 31 Rogers EM Diffusion of innovations . 5th edn. New York : Free Press , 2003 . 32 Smit ES , de Vries H , Hoving C The PAS study: a randomized controlled trial evaluating the effectiveness of a web-based multiple tailored smoking cessation programme and tailored counselling by practice nurses . Contemp Clin Trials 2010 ;31 :251 –8 . 10.1016/j.cct.2010.03.001 20215047 33 Gale NK , Heath G , Cameron E Using the framework method for the analysis of qualitative data in multi-disciplinary health research . BMC Med Res Methodol 2013 ;13 :117 10.1186/1471-2288-13-117 24047204 34 Campbell JL , Quincy C , Osserman J Coding in-depth semistructured interviews: problems of unitization and intercoder reliability and agreement . Sociol Methods Res 2013 ;42 :294 –320 . 10.1177/0049124113500475 35 Lombard M , Snyder-Duch J , Bracken CC Content analysis in mass communication: assessment and reporting of intercoder reliability . Hum Commun Res 2002 ;28 :587 –604 . 10.1111/j.1468-2958.2002.tb00826.x 36 Krippendorff K Reliability in content analysis . Hum Commun Res 2004 ;30 :411 –33 . 10.1111/j.1468-2958.2004.tb00738.x 37 Copeland L , McNamara R , Kelson M Mechanisms of change within motivational interviewing in relation to health behaviors outcomes: a systematic review . Patient Educ Couns 2015 ;98 :401 –11 . 10.1016/j.pec.2014.11.022 25535015 38 Efraimsson EO , Fossum B , Ehrenberg A Use of motivational interviewing in smoking cessation at nurse-led chronic obstructive pulmonary disease clinics . J Adv Nurs 2012 ;68 :767 –82 . 10.1111/j.1365-2648.2011.05766.x 21771043 39 Noordman J , van Lee I , Nielen M Do trained practice nurses apply motivational interviewing techniques in primary care consultations? J Clin Med Res 2012 ;4 :393 –401 . 10.4021/jocmr1120w 23226172 40 Cassista J , Payne-Gagnon J , Martel B Applying theory to understand and modify nurse intention to adhere to recommendations regarding the use of filter needles: an intervention mapping approach . Nurs Res Pract 2014 ;2014 :356153 10.1155/2014/356153 25120927 41 Bundy C Changing behaviour: using motivational interviewing techniques . J R Soc Med 2004 ;97 (Suppl 44) :43 –7 .15239293 42 Berndt NC , Bolman C , de Vries H Smoking cessation treatment practices: recommendations for improved adoption on cardiology wards . J Cardiovasc Nurs 2013 ;28 :35 –47 . 10.1097/JCN.0b013e318231f424 22048618 43 Smit ES , de Vries H , Hoving C Determinants of practice nurses’ intention to implement a new smoking cessation intervention: the importance of attitude and innovation characteristics . J Adv Nurs 2013 ;69 :2665 –74 . 10.1111/jan.12153 23600904 44 van der Weide M , Smits J Adoption of innovations by specialised nurses: personal, work and organisational characteristics . Health Policy 2004 ;68 :81 –92 . 10.1016/j.healthpol.2003.09.007 15033555 45 de Ruijter D , Smit ES , de Vries H Web-based computer-tailoring for practice nurses aimed to improve smoking cessation guideline adherence: a study protocol for a randomized controlled effectiveness trial . Contemp Clin Trials 2016 ;48 :125 –32 . 10.1016/j.cct.2016.04.007 27103232 46 Kok G , Schaalma H , Ruiter RA Intervention mapping: protocol for applying health psychology theory to prevention programmes . J Health Psychol 2004 ;9 :85 –98 . 10.1177/1359105304038379 14683571 47 Lustria ML , Cortese J , Noar SM Computer-tailored health interventions delivered over the Web: review and analysis of key components . Patient Educ Couns 2009 ;74 :156 –73 . 10.1016/j.pec.2008.08.023 18947966 48 Avis JL , van Mierlo T , Fournier R Lessons learned from using focus groups to refine digital interventions . JMIR Res Protoc 2015 ;4 :e95 10.2196/resprot.4404 26232313

PMC005xxxxxx/PMC5372121.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01361110.1136/bmjopen-2016-013611NeurologyResearch15061713173017121698Instruments measuring the disease-specific quality of life of family carers of people with neurodegenerative diseases: a systematic review Page Thomas E 1Farina Nicolas 2Brown Anna 1Daley Stephanie 2Bowling Ann 3Basset Thurstine 4Livingston Gill 5Knapp Martin 6Murray Joanna 7Banerjee Sube 2 1 School of Psychology, University of Kent, Canterbury, UK 2 Centre for Dementia Studies, Brighton and Sussex Medical School, Brighton, UK 3 Department of Health Sciences, University of Southampton, Southampton, UK 4 Department of Lived Experience Advisory Panel, Sussex Partnership NHS Foundation Trust, Hove, UK 5 Division of Psychiatry, University College London, London, UK 6 Department of Social Policy, London School of Economics, London, UK 7 Institute of Psychiatry, Psychology & Neuroscience, King's College London, London, UKCorrespondence to Professor Sube Banerjee; s.banerjee@bsms.ac.uk2017 29 3 2017 7 3 e01361124 7 2016 5 2 2017 8 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective Neurodegenerative diseases, such as dementia, have a profound impact on those with the conditions and their family carers. Consequently, the accurate measurement of family carers' quality of life (QOL) is important. Generic measures may miss key elements of the impact of these conditions, so using disease-specific instruments has been advocated. This systematic review aimed to identify and examine the psychometric properties of disease-specific outcome measures of QOL of family carers of people with neurodegenerative diseases (Alzheimer's disease and other dementias; Huntington's disease; Parkinson's disease; multiple sclerosis; and motor neuron disease). Design Systematic review. Methods Instruments were identified using 5 electronic databases (PubMed, PsycINFO, Web of Science, Scopus and the International Bibliography of the Social Sciences (IBSS)) and lateral search techniques. Only studies which reported the development and/or validation of a disease-specific measure for adult family carers, and which were written in English, were eligible for inclusion. The methodological quality of the included studies was evaluated using the COnsensus based Standards for the selection of health Measurement Instruments (COSMIN) checklist. The psychometric properties of each instrument were examined. Results 676 articles were identified. Following screening and lateral searches, a total of 8 articles were included; these reported 7 disease-specific carer QOL measures. Limited evidence was available for the psychometric properties of the 7 instruments. Psychometric analyses were mainly focused on internal consistency, reliability and construct validity. None of the measures assessed either criterion validity or responsiveness to change. Conclusions There are very few measures of carer QOL that are specific to particular neurodegenerative diseases. The findings of this review emphasise the importance of developing and validating psychometrically robust disease-specific measures of carer QOL. Neurodegenerative diseasesquality of lifepsychometricsmeasurementcarersAlzheimer's Societyhttp://dx.doi.org/10.13039/501100000320234ASPG14017 ==== Body Strengths and limitations of this study This study provides the first comprehensive review of disease-specific instruments measuring quality of life of family carers of people with neurodegenerative diseases. The findings of the review offer guidance to researchers and clinicians in the selection of appropriate and psychometrically strong disease-specific instruments. Only instruments specific to neurodegenerative diseases were examined. We did not review the performance of generic measures of quality of life. The instruments identified in this review were developed specifically for family carers rather than professional (ie, paid) carers of people with neurodegenerative diseases. Introduction Neurodegenerative diseases have a profound impact as they lead to a prolonged and irreversible decline in global intellectual, social and physical functioning. People living with neurodegenerative diseases tend to progressively lose their independence and require increased levels of care and support. Dementia is the most common type of neurodegenerative disease, affecting 800 000 people in the UK alone,1 and costing £26.3 billion to society each year.2 Family members, friends and neighbours who act as carers are a vital determinant of positive outcomes for people with neurodegenerative diseases such as dementia. They provide a wide range of practical and emotional support, social care, and assistance with activities of daily living.3 These caregiving tasks include, for example, help with personal care, managing finances and legal affairs, social activities, mobility, and administering and coordinating medication.3 Caring for a person with a neurodegenerative disease may result in multiple negative outcomes, including increased anxiety and depression, stress, exhaustion, social isolation, and concerns about the future.1 Nevertheless, caring can also have positive effects such as greater closeness to the person with the condition, reciprocity and spiritual growth.4–6 In recent years, there has been an increasing research focus on assessing quality of life (QOL) as an informative and important patient-reported outcome measure in the person with the disease and their carers. Although QOL is subjective in regard to how it is perceived by the individual, there is growing consensus that it represents a multidimensional construct encompassing various domains such as physical health, socioeconomic status; psychological, emotional and social well-being, and that it is a useful way of capturing the broad impacts of complex disorders such as neurodegenerative disease.7 In this paper, following the recommendations of our carer consultation, the term ‘family carer’ is used to refer to all ‘informal’ carers (ie, not paid carers), including family members, neighbours and friends of a person with a neurodegenerative disease. This also includes people who are not, in fact, family members, since the term ‘informal’ was seen as minimising and trivialising the nature of the care provided, and the term ‘carer’ by itself was seen as too imprecise. Given the important role of family carers, it is essential for the person with the illness as well as for the carer, that carers of people with neurodegenerative diseases maintain a good QOL. Carer QOL is likely to be associated with patient QOL, and when carer QOL deteriorates, there is likely to be a higher risk that the person with dementia will need admission into a care home, so driving lower life quality and higher societal cost. It is also important to measure the broad QOL carer impacts of interventions for the person with dementia and for carers. To ensure this can be monitored effectively, it is necessary to measure carer QOL accurately using psychometrically robust measures. To date, QOL of affected individuals and their carers has been predominantly assessed using generic health status instruments such as the SF-368 and the EQ-5D.9 Disease-specific instruments have been used much less frequently. This is problematic as generic instruments miss issues that are pertinent to specific conditions and, consequently, are less responsive to detecting changes in carer QOL over time.10 11 For example, an increased level of support received as a carer has been associated with better carer QOL in dementia,4 but neither the EQ-5D nor the SF-36 capture such a construct. An earlier review has examined the effects (eg, physical, social, emotional, financial) of caring for an elderly family member with dementia on carer QOL.11 Disease-specific and generic instruments were identified that assess QOL of either patients with dementia or their carers. However, the psychometric properties of these measures were not reported or discussed, and the review focused exclusively on dementia and Alzheimer's disease. Building on this, we therefore completed a systematic review to identify and examine the disease-specific instruments that measure QOL of family carers of people with a neurodegenerative disease. Here, we examine the psychometric properties of these instruments. Methods The systematic review was conducted and reported in accordance with the recommendations from the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.12 Inclusion criteria Instruments were included in the review if they satisfied the following criteria: (1) the study described the development (inauguration paper) and/or the evaluation of the psychometric properties and relationships with other relevant constructs (validation paper) of a measure of carer QOL; (2) the study population were adult (aged 18 years or above) family (defined as family members, neighbours or friends) carers of people with a neurodegenerative disease (dementia, Alzheimer's disease, Parkinson's disease (PD), Huntington's disease (HD), multiple sclerosis (MS) and motor neuron disease); (3) the instrument was designed to be disease-specific and not a generic measure of QOL; (4) studies only documented instruments that were self-report in their design. Any studies that reported an eligible measure (eg, as an outcome in a clinical trial) without any explicit validation were excluded; and (5) studies were written in English. Search strategy Articles were identified from initial searches of five electronic databases: PubMed, PsycINFO, Scopus, Web of Science and the International Bibliography of the Social Sciences (IBSS). The searches were conducted from inception until 8 February 2016. All articles identified up until this time were screened regardless of their publication date. We used the following five combined sets of search terms: (1) ‘quality of life’; AND (2) caregiver* OR carer*; AND (3) informal OR unpaid OR spous* OR family; AND (4) dementia OR alzheimer* OR parkinson* OR huntington* OR multiple sclerosis OR motor neurone disease; AND (5) valid* OR reliab* OR development OR psychometric. The search terms were intentionally broad and sensitive enough to ensure that all potentially relevant articles were identified (please see online supplementary material file for full details of search strategy). Six additional articles were obtained using lateral searching techniques.13 These searches involved the manual checking of reference lists of included studies (snowballing), citation searches using the ‘Cited by’ option on Web of Science, Google Scholar, and Scopus, and the ‘Related articles’ option on PubMed and Web of Science (lateral searching). The grey literature (ie, unpublished studies) was also searched using the following databases: OpenGrey and Patient-Reported Outcome and Quality of Life Instruments Database (PROQOLID). We attempted to make direct contact (via email) with the authors of a manuscript when no published psychometric data were available for the instrument being reported. Two independent reviewers (TEP and NF) screened article titles and abstracts against the predefined inclusion criteria. Full-text articles were sought for all potentially relevant studies. Any disagreements concerning inclusion were resolved through discussion and advice from a third reviewer (SD). 10.1136/bmjopen-2016-013611.supp1supplementary file Data extraction Two reviewers (TEP and NF) extracted, independently, the following data for studies that met the inclusion criteria: name of the instrument, country, sample characteristics (gender, age), study design/setting, measurement domains, number of items, response format, evidence of reliability and validity. Quality assessment The methodological quality of the studies was assessed using the COnsensus based Standards for the selection of health Measurement Instruments (COSMIN) checklist.14–16 This is a standardised tool which assesses the measurement properties of health-related instruments across nine domains (internal consistency, reliability, measurement error, content validity (including face validity), construct validity (subdivided into structural validity, hypotheses testing and cross-cultural validity), criterion validity and responsiveness) with each domain rated using 5–18 items. Each item is rated as ‘excellent’, ‘good’, ‘fair’ or ‘poor’ quality. A methodological quality score for each measurement property is obtained by taking the lowest rating of any item in that box (worst score counts). Two independent reviewers (TEP and NF) assessed the methodological quality of the included studies using the checklist. Any disagreements in scoring were resolved through discussion and advice from a third reviewer (SD). Results Search results The initial database searches identified 676 articles, of which 232 were deleted because of duplicates. After title and abstract screening, we assessed seven articles as potentially relevant and obtained full texts. From reviewing the full texts of these remaining articles, we found two studies that met inclusion criteria (figure 1). Six additional articles were identified using lateral search techniques. Figure 1 PRISMA flow diagram of study selection. PRISMA,Preferred Reporting Items for Systematic Reviews and Meta-Analyses; QOL, quality of life. In total, we included eight studies in the systematic review and these reported seven carer QOL instruments across the neurodegenerative diseases. All of them were self-report measures and consisted of the HD quality-of-life battery for carers (HDQoL-C);17 HD quality-of-life battery for carers-short form (HDQoL-C-SF);18 CAREQOL-MS for MS;19 Alzheimer's Carers Quality of Life Instrument (ACQLI);20 Parkinson Disease Questionnaire for Carers (PDQ-Carer);21 Parkinsonism Carers QoL (PQoL Carers);22 and the Caregiver Quality Of Life for dementia (CGQOL).6 We did not identify any instruments designed to measure the QOL of carers of people with motor neuron disease. Study characteristics The sample characteristics for the eight studies assessing the carer QOL instruments are presented in table 1. Table 1 Characteristics of the included studies Carer QOL Instrument Author(s) Country Sample Total N (% female) Mean age years (SD) Study design/setting Measurement domains Number of items/response format HDQoL-C Aubeeluck and Buchanan17 UK 87 (62.1) 58.2 (14.2) Longitudinal/cohort: spousal carers recruited through the Huntington's Disease Association UK Demographic and objective information; practical aspects of caregiving; satisfaction with life; feelings about living with Huntington's disease 34 items 11-point Likert scale HDQoL-C-SF Aubeeluck et al18 France Italy 301 (60.5) 57.1 (13.1) Cross-sectional survey: family or friend carers of people with Huntington's disease Satisfaction with life; feelings about living with Huntington's disease 20 items 11-point Likert scale CAREQOL-MS Benito-Leon et al19 Spain 276 (56.5) 50.2 (12.6) Longitudinal/cohort: family carers of people with MS recruited from 19 Spanish outpatient clinics Physical burden and global health; social impact; emotional impact; need of support; emotional reactions to patient's psychic status 24 items 5-point Likert scale ACQLI Doward20 UK France Germany Italy Spain 192 (72.3) 60.0 (12.2) Longitudinal/cohort: family carers of people with dementia Single domain of carer QOL 30 items Dichotomous (true/not true) PDQ-Carer Jenkinson et al21 UK 236 (63.5) 68.2 (9.5) Cross-sectional survey: family carers of people with Parkinson's disease who were members of Parkinson's UK Social and personal activities; anxiety and depression; self-care; stress 29 items 5-point Likert scale PDQ-Carer-SI Morley et al23 UK 236 (63.5) 68.2 (9.5) Cross-sectional survey: family carers of people with Parkinson's disease who were members of Parkinson's UK Single summary index score computed using the four subscales of the PDQ-Carer 29 items 5-point Likert scale PQoL Carers Pillas et al22 UK 430 (62.4) 66.2 (8.5) Cross-sectional survey: family carers of people with MSA and PSP Single domain of carer QOL 26 items 5-point Likert scale CGQOL Vickrey et al6 USA 200 (79.0) 61.5 (13.5) Longitudinal/cohort: family carers of people with dementia Assistance with instrumental activities of daily living; assistance with activities of daily living; role limitations due to caregiving; personal time; family interaction; demands of caregiving; worry; spirituality and faith; benefits of caregiving; caregiver feelings 80 items 3-point and 5-point Likert scales specific to groups of items ACQLI, Alzheimer's Carers Quality of Life Instrument; CGQOL, Caregiver Quality Of Life; HDQoL-C, Huntington's disease quality-of-life battery for carers; HDQoL-C-SF, Huntington's disease quality-of-life battery for carers short form; MS, multiple sclerosis; MSA,multiple system atrophy; PDQ-Carer, Parkinson Disease Questionnaire for Carers; PDQ-Carer-SI, Parkinson Disease Questionnaire for Carers Summary Index; PQoL Carers, Parkinsonism Carers QoL; PSP,progressive supranuclear palsy; QOL, quality of life. The total sample sizes of these studies ranged from 8717 to 430.22 The studies recruited participants from the UK,17 21–23 Spain19 and the USA.6 The remaining two studies recruited participants across multiple countries which included France, Italy and Germany.18 20 The mean age of participants across the eight studies was 61.2 years. The number of measurement domains of the instruments ranged from 1 (ACQLI, PQoL Carers) to 10 (CGQOL). The number of items of these instruments ranged from 20 (HDQoL-C-SF) to 80 (CGQOL). All measures used Likert-type rating scales (ranging from 3-point to 11-point), except for the ACQLI which used a dichotomous (true/not true) response format. Methodological quality of studies Table 2 provides a summary of the scores from the COSMIN checklist. Six of the eight studies (75%) included in the review had at least one methodological domain rated as ‘poor’ quality.6 17–21 Across all studies, the measurement property that received the highest number of ‘poor’ ratings was content validity (4/8 studies). Five of the eight studies (63%) had at least one methodological domain rated as of ‘fair’ quality.6 18 21–23 The measurement property that received the highest number of ‘fair’ ratings was internal consistency (5/8 studies). Of the 8 studies, 2 (26%) had at least one area of methodological quality rated as ‘good’ or ‘excellent’.19 22 The measurement properties that received ‘good’ ratings were reliability and measurement error.19 The measurement properties that received ‘excellent’ ratings were internal consistency, structural validity19 and content validity.22 Table 2 Results of COSMIN checklist Carer QOL instrument Study reference Internal consistency Reliability Measurement error Content validity Structural validity Hypotheses testing Cross-cultural validity Criterion validity Responsiveness HDQoL-C Aubeeluck and Buchanan17 1 1 0 1 1 0 0 0 0 HDQoL-C-SF Aubeeluck et al18 2 0 0 0 2 2 1 0 0 CAREQOL-MS Benito-Leon et al19 4 3 3 1 4 0 0 0 0 ACQLI Doward20 1 1 0 1 0 0 1 0 0 PDQ-Carer Jenkinson et al21 2 0 0 1 2 0 0 0 0 PDQ-Carer-SI Morley et al23 2 0 0 0 2 0 0 0 0 PQoL Carers Pillas et al22 2 0 0 4 2 0 0 0 0 CGQOL Vickrey et al6 2 1 0 0 1 1 1 0 0 Key: 4: excellent, 3: good, 2: fair, 1: poor, 0: no information available. ACQLI, Alzheimer's Carers Quality of Life Instrument; CGQOL, Caregiver Quality Of Life; COSMIN,COnsensus-based Standards for the selection of health Measurement Instruments;HDQoL-C, Huntington's disease quality-of-life battery for carers; HDQoL-C-SF, Huntington's disease quality-of-life battery for carers short form; MS, multiple sclerosis; PDQ-Carer-SI, Parkinson Disease Questionnaire for Carers Summary Index; PQoL Carers, Parkinsonism Carers QoL; QOL, quality of life. Psychometric properties of the carer QOL instruments The instruments that had allowed the most comprehensive evaluation of psychometric properties across the 9 domains of the COSMIN checklist were the CAREQOL-MS (5/9 domains assessed) and the CGQOL (5/9 domains assessed). The HDQoL-C, HDQoL-C-SF and ACQLI each had four of the nine domains evaluated. The measures that had the least evidence available for their psychometric properties were the PQoL Carers (3/9 domains assessed), the PDQ-Carer (3/9 domains assessed) and the summary index of the PDQ-Carer (2/9 domains assessed). Internal consistency was the most widely assessed measurement property, and was examined in all seven instruments. Cronbach's α coefficients for the total scale and instrument subscales ranged from 0.75 to 0.94 (see table 3 for an overview of the psychometric properties of each carer QOL instrument). Table 3 Evidence of the reliability and validity of the carer QOL instruments Reliability Validity Carer QOL instrument Author(s) Internal consistency Test–retest Measurement error Content validity Construct validity HDQoL-C Aubeeluck and Buchanan17 Not provided for total scale. Subscale α: 0.80, 0.84, 0.89 Test–retest (N=10) over 2 weeks. Pearson correlation: 0.78, 0.86, 0.90 for subscales Not assessed Two experts in the field of QOL and two experts in the field of HD commented on item content Convergent validity: positive correlations of HDQoL-C subscales with WHOQOL-BREF, Satisfaction with Life Scale, and Perceived Health Visual Analogue Scale HDQoL-C-SF Aubeeluck et al18 Not provided for total scale. Subscale α: 0.88, 0.80 Not assessed Not assessed Not assessed Known groups analyses showed that HDQoL-C-SF scores were higher (better carer QOL) for carers of patients in the low disease severity group compared with the moderate and high severity groups CAREQOL-MS Benito-Leon et al19 Not provided for total scale. Subscale α: 0.90, 0.85, 0.81, 0.78, 0.75 Test–retest (N=276) over 2 weeks. ICC for total scale score=0.96 and ranged from 0.75 to 0.95 for subscales. Cohen's κ ranged from 0.46 to 0.93. One item had κ<0.60 and 15 items had κ≥0.80. SE of measurement ranged from 0.91 to 2.43 across the five subscales over 2 weeks time interval. Item content analysed by MS experts. Focus groups of MS carers and patients discussed item pool. Items rated for clarity and meaning by MS experts and a separate carer sample Convergent validity: positive correlations of CAREQOL-MS subscales with Zarit Burden Interview. Moderate-to-high negative correlations of CAREQOL-MS subscales with physical and mental components of SF-36 ACQLI Doward20 α ranged between 0.87 and 0.95 across the UK, France, Germany, Italy and Spain for times 1 and 2 administrations of the ACQLI. Test–retest over 2 weeks. Spearman correlations were 0.93, 0.92, 0.95, 0.94, 0.90 for UK, France, Germany, Italy and Spain, respectively. Based on very small N per country Not assessed Field test of items with samples of carers. States that carers found the ACQLI to be understandable, acceptable and relevant across all countries Convergent validity: positive spearman correlations of ACQLI with General Well-being Index in UK and Italy. Known groups analysis demonstrated that the ACQLI can distinguish between carers based on their current health status. PDQ-Carer Jenkinson et al21 Not provided for total scale. Subscale α: 0.92, 0.87, 0.86, 0.83 Not assessed Not assessed Very limited assessment of content validity. Items were evaluated by focus groups of researchers and a pilot sample of carers. Convergent validity: moderate-to-high negative correlations of PDQ-Carer subscales with physical and mental components of SF-36 PDQ-Carer-SI Morley et al23 α=0.94 Not assessed Not assessed Not assessed Convergent validity: moderate-to-high negative correlations of PDQ-Carer-SI score with physical and mental components of SF-36 Parkinsonism Carers QoL (PQoL Carers) Pillas et al22 α=0.96 Not assessed Not assessed Very limited assessment of content validity. The questionnaire was pilot tested in a small group of carers Convergent validity: Subscales of PQoL Carers correlated with Caregiver Burden Inventory, Hospital Anxiety and Depression Scale and the EQ-5D. ANOVA revealed that PQoL Carers scores differentiate between carers based on their current health status. Caregiver Quality Of Life (CGQOL) Vickrey et al6 Not provided for total scale. Subscale α: 0.88, 0.93, 0.78, 0.83, 0.86, 0.86, 0.82, 0.94, 0.92, 0.89 Test-retest (N=71) between 11 and 63 days following first administration (75% within 21 days). ICC ranged from 0.53 to 0.89, exceeding 0.70 for 6 of the 10 subscales Not assessed Not assessed Convergent validity: Regression and correlation analyses of CGQOL subscales with a range of patient and carer characteristics (e.g. number of hours spent caregiving, duration of being a carer) ACQLI, Alzheimer's Carers Quality of Life Instrument; CGQOL, Caregiver Quality Of Life; HD, Huntington's disease; HDQoL-C, Huntington's disease quality-of-life battery for carers; HDQoL-C-SF, Huntington's disease quality-of-life battery for carers short form; ICC, intraclass correlation coefficient; MS, multiple sclerosis; PDQ-Carer-SI, Parkinson Disease Questionnaire for Carers Summary Index; PQoL Carers, Parkinsonism Carers QoL; QOL, quality of life. Reliability of the measures was reported in 4/8 studies.6 17 19 20 These studies computed test–retest reliability over 2–3 weeks time interval through calculation of Pearson or Spearman correlations. Intraclass correlation coefficients (ICCs) were also used to assess the test–retest reliability of individual subscales in two instruments: the CAREQOL-MS and CGQOL. ICCs ranged between 0.53 and 0.95 for all subscales across these two measures. The CAREQOL-MS was the only instrument to calculate measurement error. Content validity was evaluated in 5/8 studies.17 19 20–22 Structural validity was assessed in 7/8 studies.6 17–19 21–23 Principal components analysis (PCA) was the most frequently employed method of statistical analysis for examining structural validity. No studies included the use of confirmatory factor analysis (CFA) with either continuous or categorical variables (also known as item response theory (IRT)) to assess factorial structure. Correlational and known group analyses were most commonly used to assess convergent and discriminant validity. However, hypotheses testing was carried out for only two instruments; the HDQoL-C-SF18 and the CGQOL.6 Cross-cultural validity was evaluated for three measures: the HDQoL-C-SF,18 ACQLI20 and CGQOL.6 None of the instruments assessed criterion validity or responsiveness. Constructs assessed by the carer quality of life instruments Huntington's disease As displayed in table 3, the HDQoL-C consists of four measurement domains (demographic and objective information; practical aspects of caregiving; satisfaction with life; feelings about living with HD) which assess the QOL of spousal carers of people with HD. It is a 34-item instrument which employs an 11-point Likert-type response scale. PCA using varimax rotation was performed separately on each set of items for three of the four domains (not the component of demographic and objective information). This identified three subcomponents for the first component of ‘practical aspects of care giving’ (levels of support and access to professionals; long-term and genetic issues; daily hassles). PCA on the third domain of ‘satisfaction with life’ extracted two subcomponents (overall QOL issues; personal issues). Finally, PCA on the fourth domain of ‘feelings about living with HD’ resulted in the identification of two subcomponents (negative feelings about life; positive feelings about life). The short form of the HDQoL-C measures two of the four domains of the original instrument (satisfaction with life; feelings about living with HD). It is a 20-item measure using an 11-point Likert-type response scale. Exploratory factor analysis (EFA) using varimax rotation was performed on the 17 items of the ‘feelings about living with HD’ subscale. This resulted in the extraction of the two subcomponents of ‘negative feelings about life’ and ‘positive feelings about life’ which accounted for 84% of the total variance. Factor analysis was not performed on the ‘satisfaction with life’ subscale as only three items were retained from the full-length measure. In addition to the English language version of the HDQoL-C, the measure was translated into French and Italian. Multiple sclerosis The CAREQOL-MS measures QOL in carers of people with MS. It consists of 24 items and uses a five-point Likert-type response format. PCA using orthogonal rotation (unspecified method of rotation) extracted four factors identified as ‘physical burden and global health’, ‘emotional impact’, ‘need of support’ and ‘emotional reactions to patient's psychic status’ explaining 60% of the total variance. The first factor was later subdivided to include a separate subscale of ‘social impact’. Parkinson's disease The PDQ-Carer measures QOL in carers of people with PD. This instrument contains 29 items and uses a five-point Likert-type response format. EFA using varimax rotation identified four factors: social and personal activities, anxiety and depression, self-care, and stress. These factors explained 60% of the total variance. The PQoL Carers is an alternative measure of QOL in family carers of people with PD. This instrument contains 26 items and uses a five-point Likert-type response format. Parallel analysis was performed to assess the dimensionality of the scale which identified a single factor structure (representing overall QOL), explaining 54% of the total variance. Dementia and Alzheimer's disease The ACQLI measures QOL in carers of people with Alzheimer's disease and dementia. This instrument consists of 30 items which use a dichotomous (true/not true) response format. Structural validity of the ACQLI has not been evaluated. The items of the measure assess QOL as a unidimensional construct, but no factor analysis was reported in the development of this scale. Five language versions of the ACQLI are available: English, French, German, Italian and Spanish. Finally, the CGQOL is the most recently developed measure of QOL in carers of people with dementia and Alzheimer's disease. The instrument contains 80 items across 10 domains (assistance with instrumental activities of daily living; assistance with activities of daily living; role limitations due to caregiving; personal time; family interaction; demands of caregiving; worry; spirituality and faith; benefits of caregiving; caregiver feelings) using either a three-point or a five-point Likert-type response format, with categories specific to groups of items. These 10 subscales were categorised under three superordinate factors labelled ‘tangible assistance’ (comprised of assistance (to the person with dementia) with instrumental activities of daily living; assistance (to the person with dementia) with activities of daily living; personal time; role limitations due to caregiving), ‘psychosocial’ (comprised of role limitations due to caregiving (cross-loaded with tangible assistance); family involvement; demands of caregiving; worry; caregiver feelings) and ‘benefits/faith’ (comprised of spirituality and faith, and benefits of caregiving) using EFA with promax rotation. The CGQOL was originally developed in English and was later translated into Spanish. It is to be noted that items were translated into Spanish, reviewed by a second translator and interviews were then conducted with eight Spanish-speaking carers in which to assess and refine item wording. Discussion We identified eight studies which report the development of seven carer QOL measures in neurodegenerative diseases (ie, dementia and Alzheimer's disease; HD; PD; MS). No instruments were identified that measure carer QOL in motor neuron disease. The findings of the studies, taken together, are potentially of value in guiding researchers and health professionals in the selection of an appropriate and psychometrically robust disease-specific instrument. The studies demonstrate some key methodological problems in the current instruments available to measure carer QOL in these neurodegenerative diseases and highlight avenues for future research. Overall, limited information was available concerning the psychometric properties of the instruments that were identified. Collectively, the studies have shown that many important elements of psychometric evaluation were either absent or not sufficiently tested during the development phases of these disease-specific measures. This review found that the CAREQOL-MS19 and the CGQOL6 received the most comprehensive psychometric evaluation, whereas the PQoL Carers22 and the PDQ-Carer21 (including the validation of the summary index23) had the least information available on their psychometric properties. Internal consistency was the most widely reported psychometric property and was available for all seven instruments. This was quite strong across the studies with Cronbach's α coefficients ranging from 0.75 to 0.94. Less information was provided for the temporal stability of the measures. Only four of the instruments reported test–retest reliability (HDQoL-C, CAREQOL-MS, ACQLI, CGQOL), and this ranged from being adequate to excellent over 2–3 weeks period. The CAREQOL-MS was the only instrument to report measurement error, which is the more useful figure than reliability since it controls for the population variance and is therefore more readily generalisable to other populations.24 The assessment of factorial structure was generally limited across all studies. EFA and PCA were the methods predominantly used by the test developers to assess dimensionality. However, no factor analysis was reported for the development of the ACQLI, despite the underlying assumption that it is a unidimensional measure of carer QOL. Moreover, none of the studies had performed CFA as a hypothesis-driven test of structural validity. All identified studies disregarded the categorical nature of responses and treated them as continuous for the purposes of factor analysis. While this approximation may be acceptable for items using five or more categories,25 it is generally untenable for dichotomous items24 such as used in the ACQLI. Future research using these instruments should investigate whether the factor structures that were identified in EFA will replicate in CFA when the fit of alternative measurement models are tested. It would also be advantageous for researchers to explore factorial structures using methods for categorical variables (IRT), such as Rasch analysis.26 A strong advantage of Rasch analysis is that different measures of the same attribute can be calibrated using the same scale and items can be used for computerised adaptive testing.27 IRT methods can thus control for item properties that are difficult in classical measurement, such as the item difficulty or ability to discriminate between varying levels of the attribute. Interestingly, the factor analyses reported in the studies reviewed suggest that similar constructs are being measured by all carer QOL instruments across the various diseases. These measurement domains include, for example, an appraisal of caring demands, evaluation of support received, positive and negative feelings towards caring, the social impact of caring and negative health effects due to caring, such as anxiety, depression, stress and fatigue. In contrast to the other measures, the CGQOL includes a wider array of QOL domains believed to be affected by the caring role such as role limitations and family interaction. Correlations with external measures were examined in all seven instruments. As evidence of convergent validity, the disease-specific measures were often found to correlate with generic measures of QOL such as the SF-36 and the EQ-5D.19 21–23 Known group analysis was frequently used to evaluate construct validity. For example, the HDQoL-C-SF differentiated between carers on the basis of the patients' disease severity. As reported previously, it was shown that carers' scores on the HDQoL-C-SF were higher (ie, representing better QOL) when caring for family members whose HD was less severe.18 A limitation of the studies reviewed is that hypotheses were not formulated in the assessment of construct validity. The HDQoL-C-SF and CGQOL were the only two instruments in which hypotheses had been postulated a priori about the expected relationships among measures. As such, it was unclear how researchers had anticipated their measures to be associated with, or distinct from, existing scales that measure similar constructs. Another limitation of the studies included in the review is that criterion-related validity had not been examined or even considered as an important issue. More importantly, none of the instruments that we examined had assessed responsiveness to change. Thus, it is not known whether these instruments are sensitive to detecting changes in carers' QOL over time. Researchers and clinicians need to measure whether interventions and services are effective in improving carer QOL. The lack of known responsiveness to change in these existing measures could, therefore, be problematic for those seeking to administer these measures within intervention programmes and social care settings. The growing use of QOL as a primary outcome measure in interventional studies highlights the importance of conducting responsiveness analysis on these measures and including such analyses in the development of future QOL instruments. Limitations There are several limitations of the present review that warrant consideration. First, only articles written in English were sourced, and this may have led to the exclusion of carer QOL measures that were developed and/or validated in other languages. Second, this review only focused on neurodegenerative diseases and did not examine the psychometric properties of carer QOL measures developed for other medical conditions. For example, the Caregiver Quality of Life Index-Cancer (CQOLC);28 scale is a self-report measure of QOL in family carers of patients with cancer. In future research, it would be interesting to compare the psychometric properties of neurodegenerative disease-specific measures with carer QOL instruments across a broader range of disorders. Third, we only examined the measurement of QOL of family carers (defined for our purposes as family members, neighbours or friends) and did not consider professional (paid) carers. It should be noted, however, that no articles in this review were excluded during the screening process on the basis of the carer samples that were reported. It appears that there are no instruments developed to measure the specific QOL impacts on paid carers. Our systematic review appears to have been successful in detecting carer QOL measures that currently exist across the various neurodegenerative diseases. Conclusions In this systematic review, we provide a comprehensive overview of disease-specific instruments to measure QOL of family carers of people with a neurodegenerative disease. The included studies had key methodological limitations associated with the measurement of QOL using these disease-specific instruments. These findings indicate that there is a need to further develop and refine these measures and potentially to develop new measures, in order to improve the psychometric quality of the measures available. Given the considerable overlap in the constructs measured by the QOL instruments that we reviewed, it might be useful for researchers to explore whether a single measure of carer QOL for all neurodegenerative diseases would be feasible and valid. However, due to the heterogeneity in symptoms and disease course, a rigorous development process is needed before assuming a single measure would be sufficient. Moreover, it would be advantageous to create shorter and more concise measures of carer QOL. This would offer greater brevity and flexibility to researchers and clinicians who need to administer these instruments in tandem with a number of other scales. Long instruments such as the 80-item CGQOL6 arguably impose a greater burden on respondents. Thus, shorter measures that contain fewer items would benefit respondents by reducing completion time and reporting burden. Overall, the findings of this review should be helpful in guiding researchers and health professionals in the selection of an appropriate, and psychometrically robust, disease-specific instrument. The accurate assessment of carer QOL is a growing research priority, and the findings of this review are a useful foundation for researchers seeking to develop and validate new measures of carer QOL in neurodegenerative disorders. The development of psychometrically strong, disease-specific measures of QOL is important for the generation of better treatments, services, care and support for people with neurodegenerative disorders and their carers. As demonstrated in this review, there are few instruments that measure carer QOL in neurodegeneration, and the psychometric properties of the few available measures are limited. Further psychometric testing is needed on existing measures; future validation studies should include the use of IRT in conjunction with traditional methods of assessment. There is room for new instruments with stronger psychometric development, evaluation and properties in this important area. Contributors: All authors were involved in the design of the study. TEP designed the literature search and screened the titles and abstracts of the identified studies in collaboration with NF. TEP and NF performed the data extraction and quality assessment. Preparation of the manuscript was completed by TEP. AB and SB edited and reviewed the manuscript. All authors approved the manuscript. The views expressed are the authors' own. Funding: This work was supported by an Alzheimer's Society Project Grant (234ASPG14017). Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Department of Health . Living well with dementia: a national dementia strategy . London : TSO , 2009 . 2 Prince M , Knapp M , Guerchet M Dementia UK . 2nd edn London : Alzheimer's Society , 2014 . 3 Manthorpe J , Bowling A Quality of life measures for carers for people with dementia: measurement issues, gaps in research and promising paths . Res Policy Plan 2016 ;31 :163 –78 . 4 Vellone E , Piras G , Talucci C Quality of life for caregivers of people with Alzheimer's disease . J Adv Nurs 2008 ;61 :222 –31 . 10.1111/j.1365-2648.2007.04494.x 18186913 5 Vellone E , Piras G , Venturini G Quality of life for caregivers of people with Alzheimer's disease living in Sardinia, Italy . J Transcult Nurs 2012 ;23 :46 –55 . 10.1177/1043659611414199 21807960 6 Vickrey BG , Hays RD , Maines ML Development and preliminary evaluation of a quality of life measure targeted at dementia caregivers . Health Qual Life Outcomes 2009 ;7 :1 –12 . 10.1186/1477-7525-7-1 19134191 7 Lim JW , Zebrack B Caring for family members with chronic physical illness: a critical review of caregiver literature . Health Qual Life Outcomes 2004 ;2 :50 .15377384 8 Ware JE Jr, Gandek B The SF-36 Health Survey: development and use in mental health research and the IQOLA Project . Int J Ment Health 1994 ;23 :49 –73 . 10.1080/00207411.1994.11449283 9 Brooks R EuroQol: the current state of play . Health Policy 1996 ;37 :53 –72 . 10.1016/0168-8510(96)00822-6 10158943 10 Ettema TP , Droes RM , de Lange J A review of quality of life instruments used in dementia . Qual Life Res 2005 ;14 :675 –86 . 10.1007/s11136-004-1258-0 16022061 11 Zacharopoulou G , Zacharopoulou V , Lazakidou A Quality of life for caregivers of elderly patients with dementia and measurement tools: a review . Int J Health Res Innov 2015 ;3 :49 –64 . 12 Moher D , Liberati A , Tetzlaff J Preferred reporting items for systematic reviews and meta-analyses: the PRISMA Statement . Ann Intern Med 2009 ;151 :264 –9 . 10.7326/0003-4819-151-4-200908180-00135 19622511 13 Greenhalgh T , Peacock R Effectiveness and efficiency of search methods in systematic reviews of complex evidence: audit of primary sources . BMJ 2005 ;331 :1064 –5 . 10.1136/bmj.38636.593461.68 16230312 14 Mokkink LB , Terwee CB , Patrick DL The COSMIN study reached international consensus on taxonomy, terminology, and definitions of measurement properties for health-related patient-reported outcomes . J Clin Epidemiol 2010a ;63 :737 –45 . 10.1016/j.jclinepi.2010.02.006 20494804 15 Mokkink LB , Terwee CB , Patrick DL The COSMIN checklist for assessing the methodological quality of studies on measurement properties of health status measurement instruments: an international Delphi study . Qual Life Res 2010b ;19 :539 –49 . 10.1007/s11136-010-9606-8 20169472 16 Terwee CB , Mokkink LB , Knol DL Rating the methodological quality in systematic reviews of studies on measurement properties: a scoring system for the COSMIN checklist . Qual Life Res 2012 ;21 :651 –7 . 10.1007/s11136-011-9960-1 21732199 17 Aubeeluck A , Buchanan H The Huntington's disease quality of life battery for carers: reliability and validity . Clin Genet 2007 ;71 :434 –45 . 10.1111/j.1399-0004.2007.00784.x 17489849 18 Aubeeluck A , Dorey J , Squitieri F Further evidence of reliability and validity of the Huntington's disease quality of life battery for carers: Italian and French translations . Qual Life Res 2012 ;22 :1093 –8 . 10.1007/s11136-012-0227-2 22821667 19 Benito-Leon J , Rivera-Navarro J , Guerrero AL The CAREQOL-MS was a useful instrument to measure caregiver quality of life in multiple sclerosis . J Clin Epidemiol 2011 ;64 :675 –86 . 10.1016/j.jclinepi.2010.08.003 21071173 20 Doward LC The development of the Alzheimer's carers quality of life instrument . Qual Life Res 1997 ;6 :639 . 21 Jenkinson C , Dummett S , Kelly L The development and validation of a quality of life measure for the carers of people with Parkinson's disease (the PDQ-Carer) . Parkinsonism Relat Disord 2012 ;18 :483 –7 . 10.1016/j.parkreldis.2012.01.007 22321866 22 Pillas M , Selai C , Quinn NP Development and validation of a carers quality-of-life questionnaire for parkinsonism (PQoL Carers) . Qual Life Res 2016 ;25 :81 –8 . 10.1007/s11136-015-1071-y 26264781 23 Morley D , Dummett S , Kelly L The PDQ-Carer: development and validation of a summary index score . Parkinsonism Relat Disord 2013 ;19 :448 –9 . 10.1016/j.parkreldis.2012.11.018 23238067 24 McDonald RP Test theory. A unified approach . Mahwah, NJ : Lawrence Erlbaum , 1999 . 25 Rhemtulla M , Brosseau-Liard PÉ , Savalei V When can categorical variables be treated as continuous? A comparison of robust continuous and categorical SEM estimation methods under suboptimal conditions . Psychol Methods 2012 ;17 :354 –73 . 10.1037/a0029315 22799625 26 Fisher GH , Molenaar IW Rasch models: foundations, recent developments, and applications . New York : Springer , 1995 . 27 de Boer MR , Moll AC , de Vet HCW Psychometric properties of vision-related quality of life questionnaires: a systematic review . Ophthalmic Physiol Opt 2004 ;24 :257 –73 . 10.1111/j.1475-1313.2004.00187.x 15228503 28 Weitzner MA , Jacobsen PB , Wagner H Jr , et al The Caregiver Quality of Life Index-Cancer (CQOLC) scale: development and validation of an instrument to measure quality of life of the family caregiver of patients with cancer . Qual Life Res 1999 ;8 :53 –63 . 10.1023/A:1026407010614

PMC005xxxxxx/PMC5372122.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01221210.1136/bmjopen-2016-012212Medical Publishing and Peer ReviewResearch150617111730Time to publication for publicly funded clinical trials in Australia: an observational study Strand Linn Beate 12Clarke Philip 3Graves Nicholas 1http://orcid.org/0000-0001-6339-0374Barnett Adrian G 1 1 Institute of Health and Biomedical Innovation, School of Public Health and Social Work, Queensland University of Technology, Brisbane, Queensland, Australia 2 Department of Public Health and General Practice, Norwegian University of Science and Technology, Trondheim, Norway 3 Centre for Health Policy, School of Population and Global Health, The University of Melbourne, Melbourne, Queensland, AustraliaCorrespondence to Dr Linn Beate Strand; linn.b.strand@ntnu.no2017 22 3 2017 7 3 e01221211 4 2016 30 9 2016 3 10 2016 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective To examine the length of time between receiving funding and publishing the protocol and main paper for randomised controlled trials. Design An observational study using survival analysis. Setting Publicly funded health and medical research in Australia. Participants Randomised controlled trials funded by the National Health and Medical Research Council of Australia between 2008 and 2010. Main outcome measures Time from funding to the protocol paper and main results paper. Multiple variable survival models examining whether study characteristics predicted publication times. Results We found 77 studies with a total funding of $A59 million. The median time to publication of the protocol paper was 6.4 years after funding (95% CI 4.1 to 8.1). The proportion with a published protocol paper 8 years after funding was 0.61 (95% CI 0.48 to 0.74). The median time to publication of the main results paper was 7.1 years after funding (95% CI 6.3 to 7.6). The proportion with a published main results paper 8 years after funding was 0.72 (95% CI 0.56 to 0.87). The HRs for how study characteristics might influence timing were generally close to one with narrow CIs, the notable exception was that a longer study length lengthened the time to the main paper (HR=0.62 per extra study year, 95% CI 0.43 to 0.89). Conclusions Despite the widespread registration of clinical trials, there remain serious concerns of trial results not being published or being published with a long delay. We have found that these same concerns apply to protocol papers, which should be publishable soon after funding. Funding agencies could set a target of publishing the protocol paper within 18 months of funding. ==== Body Strengths and limitations of this study We had a relatively small sample size of 77 studies. The protocol paper was examined as a time-dependent intermediate event when examining the time taken to publish the main paper. We did not contact authors to confirm publication dates. Introduction Randomised controlled trials (RCTs) represent the gold standard when evaluating healthcare interventions, mainly because they minimise selection bias.1 Policymakers often depend on the published results from RCTs and systematic reviews based on RCTs to learn about the efficacy of interventions and to apply the findings in health policy.2 Previous research has found 30–50% of clinical trial results do not get published, even years after completion of the project.3–7 This missing research is a large part of the overall ‘waste in research’, with the estimate that 85% of current research investment is wasted.8 National Health and Medical Research Council (NHMRC) Project Grants are the biggest source of funding in Australia for clinical trials. Project Grants aim to support individual researchers as individuals or small teams for a defined project from 1 to 5 years. Applications are extensive, including sections on past funding, budget and progress reports, but the key sections are the researchers' track records and the nine-page detailed background and research plan. Applications for funding large clinical trials are generally peer reviewed by a separate panel of experts who only consider clinical trials. The NHMRC have supported policies to improve clinical trial reporting. In 2005, an NHMRC grant helped establish the Australian and New Zealand Clinical Trial Registry (ANZCTR). In 2007, the NHMRC published a code for the responsible conduct of research, which stated that researchers must register clinical trials and have a responsibility to their colleagues and the wider community to disseminate a full account of their research as broadly as possible.9 In 2014, the NHMRC supported the Alltrials campaign, which calls for all past and present clinical trials to be registered and their results reported. To date, there is no NHMRC policy that mentions publishing research protocols. The aim of this paper was to examine the time delay between funding and publication for government-funded RCTs in Australia. As well as the time to the publication of the main paper, we also examined the time to a protocol paper. The main purpose of the protocol is to predetermine the methods and primary outcomes, so that issues such as outcome switching are reduced or at least made explicit.10 Publishing a protocol paper also allows others to see what research is ongoing and so helps avoid duplicated research being funded. Ideally, protocols would be published quickly, preferably soon after the research is funded. We were also interested in whether publication of a protocol paper was associated with time to publication of the main paper and whether there was an association between the study characteristics (ie, funding amount, number of investigators and number of participants) and time to publication. Methods We examined studies funded by the NHMRC of Australia. We combined data on the funding amount, estimated number of research participants, number of investigators and publication times in peer-reviewed journals of the protocol paper and main results paper. We examined the time to publication of the protocol paper and the main paper, and whether characteristics of the study predicted these times. Data collection The NHMRC is Australia's main funding body for medical research. The NHMRC administers funding for health and medical research on behalf of the Australian Government, and provides funding for all areas of research relevant to human health and medicine. This study examined Project Grants which are the NHMRC's main scheme for individuals and teams of researchers undertaking RCTs in any field relevant to human health. Applicants can apply for grants of up to 5 years in duration.11 Every RCT should be registered. The three most relevant trial registries for this study are the: ANZCTR, International Standard Randomised Controlled Trial Number (ISRCTN) and ClinicalTrials.gov. They represent a comprehensive source for information about ongoing and completed publicly and privately funded trials within and outside Australia. All three registries use web-based systems to facilitate registration of clinical trials.12–14 The registries include compulsory and optional data items, and trials will not be registered without completion of all compulsory data items and approval by a human participant ethics review board. Additional information about the registries is available from their websites.12–14 Search strategy We aimed to find all RCTs funded by the NHMRC between 2008 and 2010. We used these years to strike a balance between examining recent data and allowing a reasonable time for trial completion and publication. We identified 77 RCTs by searching for the words ‘RCT’, ‘randomised’ and ‘trial’ in NHMRC documents available on the NHMRC website (http://www.nhmrc.gov.au/grants-funding/outcomes-funding-rounds). From the same data we also extracted the funding amount ($A), number of investigators and planned study duration. To find the number of participants and trial registration number, we first searched for the project name and principal investigator within ANZCTR. When a study could not be located in ANZCTR, we searched for it within ISRCTN and ClinicalTrials.gov. Some studies may have been registered in multiple registries, but we did not record this. The data were entered into an Excel spreadsheet. Two investigators (LBS and AGB) independently determined the publication status of each trial in January 2016. To locate a protocol paper or main results paper, we searched for the unique NHMRC project number and the unique trial registration number in PubMed and in Google Scholar. If we were unable to locate the papers by this method, we searched for the name of the principal investigator and the name of the project. We matched the articles that were identified to the corresponding trial using information about location, enrolment, study start and completion dates, and outcome measures described in the clinical trial registers. We also searched all available NHMRC grant summaries from 2004 to 2012 (the last year available from the NHMRC). If we found a protocol paper, we searched all the papers that cited the protocol paper to look for the main paper. Similarly, if we found the main paper, we searched the text for mention of the protocol paper. We only included protocols published in peer-reviewed journals. The data collection process is shown in figure 1. Figure 1 Flow chart showing the data collection process. ANZCTR, Australian and New Zealand Clinical Trial Registry NHMRC, National Health and Medical Research Council; RCT, randomised controlled trial. Publication status and time to publication For all published papers, we calculated time to publication from the date of the funding announcement. For papers published online ahead of print, we used the date the publication was available electronically. For trials where we could not find a protocol paper or a main results paper, we used a censored follow-up time as the last day we performed our search (31 January 2016). We were interested in whether some trials were more likely to be completed and wanted to examine whether time to publication of the protocol paper and main results paper was associated with funding amount, number of investigators or participant numbers. We were also interested in whether publishing a protocol paper influenced the time it took for the main results paper to be published. Statistical analysis Kaplan-Meier survival plots were used to show the time to publication of protocol paper and main results paper. The y-axes in the plots were reversed to show the increasing proportion of published papers over time. We calculated the median time to publication and 95% CIs, and the mean survival and 95% CI at the maximum follow-up time of 8 years. We used Cox proportional hazard models with time since funding date as the time scale, and presented results as HRs with 95% CIs. The predictor variables for the publication of a protocol paper were funding amount (per $A100K), number of investigators, project duration and participant numbers (per 100). These variables were chosen as indicators of complexity that may make a trial more difficult to complete. When examining the time to publication of the main results paper, we added the protocol paper as a time-dependent intermediate variable (figure 2).15 The alternative approach of using a binary variable for protocol published (yes/no) ignores time and assumes the protocol was published on the funding date. This lengthens the time between the protocol and main paper and would bias the HR downwards.16 All analyses were performed using R V.3.1.0. Figure 2 The three key events over time. Studies may go straight from funding to main paper publication or first publish a protocol paper, which is therefore a time-dependent variable. The illustration shows how ignoring the timing of the protocol lengthens the time between the protocol paper and the main paper which would bias the HR downwards. The thicker part of the arrow from protocol to main paper in the right panel is the arrow length in the left panel. Missing data We were unable to find four studies in any of the trial registers, and so we did not have information on the number of participants. We therefore randomly imputed these missing numbers from the available participant numbers. We created 10 imputed data sets and combined the results using the ‘mitools’ package in R.17 Results We found 77 studies with a total funding of ∼$A59 million (see online supplementary appendix for list of studies). Forty-three studies published a protocol paper and 39 studies published a main results paper (figure 3). Thirty million dollars and 50% of the funding went to studies that have not yet published the main results paper during the mean follow-up time of 6.2 years. Figure 3 Flow chart showing publication status (protocol and main paper) of the 77 funded studies. Characteristics of the RCTs by publication status are in table 1. Table 1 Characteristics of the randomised controlled trials funded by the National Health and Medical Research Council (NHMRC) Project Grants 2008–2010 by publication status of main results paper   Main results paper published Yes No Characteristic Mean SD Mean SD Funding amount ($1000) 756 539 788 509 Participant numbers 562 680 545 1049 Median IQR Median IQR Number of investigators 5 4–6 5 4–6 Project duration (years) 4 4–5 5 4–6 Protocol paper n Per cent n Per cent Yes 20 51 23 61 No 19 49 15 39 Total 39 51 38 49 10.1136/bmjopen-2016-012212.supp1supplementary appendix Time to publication From the funding date, the median time to publication of the protocol paper was 6.4 years (figure 4), with a 95% CI from 4.1 to 8.1 years. The estimated proportion with a published protocol paper 8 years after funding was 0.61 (95% CI 0.48 to 0.74). One protocol was published before funding. Figure 4 Kaplan-Meier survival curve (and 95% CIs as dashed lines) for time from funding to protocol paper published (n=77). Horizontal lines on the survival curve indicate censoring. The median time to publication of the main results paper was 7.1 years (figure 5), with a 95% CI 6.3 to 7.6. The estimated proportion with a published main results paper 8 years after funding is 0.72 (95% CI 0.56 to 0.87). As expected, shorter studies were generally published faster. Figure 5 Kaplan-Meier survival curves (and 95% CIs as dashed lines) for time from funding to main paper published for all papers combined (left) and split into relatively long and short studies by median study length (right; n=77). Horizontal lines on the survival curves indicate censoring. Multivariable analysis We found no evidence for an association between funding amount, number of investigators or participant numbers with time to publication of protocol paper or main results (table 2). Publication of a protocol paper was not associated with time to publication of the main results paper. Project duration was not associated with time to protocol paper, but not surprisingly longer projects took longer to publish their main results paper, with a HR of 0.62 per extra year (95% CI 0.43 to 0.89). Table 2 HRs and 95% CIs for time to publication of protocol paper and main results paper dependent on study characteristics Study characteristic Protocol paper Main paper Funding amount (per $100K) 1.01 (0.95 to 1.07) 1.02 (0.95 to 1.09) Number of investigators 1.03 (0.84 to 1.27) 1.03 (0.81 to 1.30) Project duration (years) 1.00 (0.86 to 1.16) 0.62 (0.43 to 0.89) Participant numbers (per 100) 1.01 (0.98 to 1.04) 1.01 (0.97 to 1.04) Protocol paper as intermediate event 0.69 (0.34 to 1.39) Discussion The median time to publishing a protocol paper from NHMRC-funded research was 6.4 years after funding. The median time to publication of the main results paper was 7.2 years, well beyond the duration of funding for NHMRC grant schemes. It is hard to understand the reasons for these delays, because the study design should largely be finalised at the time the researchers receive funding, and a key criteria the NHMRC uses for funding research is the feasibility of the project and the investigators' track record.18 Some delays may be due to protracted peer review, which is no fault of the investigators.19 Other delays outside the investigators' control are sponsor termination and contractual issues.20 Publishing a protocol Publishing a protocol early increases transparency and gives the research community and the public a better idea of ongoing studies. Protocols are often included in systematic reviews or assist in development of prospective meta-analyses, and can be particularly important for research questions with little current evidence. While some journals (eg, BMJ Open) will not accept protocol papers if the data collection has concluded, there are few guidelines on the timing of the publication of protocol paper in relation to funding or initiation of the study. While there is an established guideline for the reporting of a protocol,21 these only cover the content and not the timing of protocol papers. The guidelines could be extended to include a recommended target date for publication after funding, or after ethics approval for unfunded studies. A period of 18 months after funding seems a sensible target. This time allows researchers to transfer the information from the grant application into the protocol format and allows time for peer review. Delays due to peer review could be avoided by adding the protocol to a clinical trial registry. Funding agencies such as the NHMRC could mandate a target and withhold further funding if the protocol is not published or at least submitted to a journal to provide a clear incentive to publish the protocol. There is a precedent for this, as the UK National Institute for Health Research include in their funding contract that a publicly available final report must be submitted, and this agency has a 98% publication rate.22 Current editorial processes at many journals mean it is now much harder to get an RCT published without a protocol, and so the problem we identified may be a historic concern. However, the median time to protocol paper was 7.3 years for studies funded in 2008, 3.9 years for studies funded in 2009 and 6.0 years for studies funded in 2010. So our data do not show clear signs of improvement between 2008 and 2010. Predictors of publication timing Our sample size was relatively small, but the HRs for how study characteristics might influence timing were generally close to one with narrow CIs (table 2), which suggests that the issues with timely publication are common to large and small studies. The two exceptions were study length and protocol paper. The mean HR for the protocol paper is <1, suggesting that publishing a protocol slows the time to the main paper. This could be because not having a protocol paper allows authors to focus on statistically significant results that may not have been the primary outcome, and there is a bias towards journals accepting papers with statistically significant results,23 and it generally takes longer to publish negative results.24 However, the CIs around the HR for the protocol were wide, and a larger study is needed to examine this potential association. Previous studies Our results are similar to related studies. A large study of US RCTs found that only 36% had published their results within 2 years of study completion7 and another US study of clinical trials found that only 46% were published within 30 months of trial completion.4 These results and ours highlight the serious problem of missing research and the timeliness of completing studies. Registering trials has been an important step in avoiding the waste of missing research, but it has not solved the problem as there are many registered randomised trials that remain unpublished. Ours is the first study to use Australian data and the first to additionally examine the protocol paper. Limitations There are a number of limitations. Our sample size was relatively small compared with related studies. We did not email authors to confirm publication status, but a related study found that emailing authors did not provide much additional information.4 Our search to identify RCTs would have missed studies that did not include the three phrases we searched for in their title. Conclusions In a time when there is very high competition for scarce research funds, we found that both the protocol paper and the results of clinical trials are not being published in a timely fashion. Funding bodies need to focus more on the outputs from research and timely publication of a protocol paper should be the first milestone to assess whether trials will produce results which can be added to the body of evidence used for medical decision-making. Twitter: Follow Adrian Barnett @aidybarnett Contributors: PC, NG and AGB came up with the idea for the study while all authors contributed to the overall conception and design. AGB and LBS independently determined the publication status of each trial and carried out the statistical analysis. LBS and PC wrote the first draft and all authors contributed to the interpretation of results and helped with the manuscript writing. All authors had full access to all of the data (including statistical reports and tables) in the study and can take responsibility for the integrity of the data and the accuracy of the data analysis. Funding: This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: The data are available on figshare: https://figshare.com/articles/Time_to_publication_data/4054878. ==== Refs References 1 Odgaard-Jensen J , Vist GE , Timmer A Randomisation to protect against selection bias in healthcare trials . Cochrane Database Syst Rev 2011 ;(4) :MR000012 10.1002/14651858.MR000012.pub3 21491415 2 Whitty CMJ What makes an academic paper useful for health policy? BMC Med 2015 ;13 :1 –5 . 10.1186/s12916-015-0544-8 25563062 3 Ross JS , Mulvey GK , Hines EM Trial publication after registration in ClinicalTrials.Gov: a cross-sectional analysis . PLoS Med 2009 ;6 :e1000144 10.1371/journal.pmed.1000144 19901971 4 Ross JS , Tse T , Zarin DA Publication of NIH funded trials registered in ClinicalTrials.gov: cross sectional analysis . BMJ 2012 ;344 :d7292 10.1136/bmj.d7292 22214755 5 Chinnery F , Young A , Goodman J Time to publication for NIHR HTA programme-funded research: a cohort study . BMJ Open 2013 ;3 :e004121 10.1136/bmjopen-2013-004121 6 Doshi P Data too important to share: do those who control the data control the message? BMJ 2016 ;352 :i1027 10.1136/bmj.i1027 26936297 7 Chen R , Desai NR , Ross JS Publication and reporting of clinical trial results: cross sectional analysis across academic medical centers . BMJ 2016 ;352 :i637 10.1136/bmj.i637 26888209 8 Chalmers I , Glasziou P Avoidable waste in the production and reporting of research evidence . Lancet 2009 ;374 :86 –9 . 10.1016/S0140-6736(09)60329-9 19525005 9 National Health and Medical Research Council, Australian Research Council, Universities Australia . Australian Code for the Responsible Conduct of Research 2007 https://www.nhmrc.gov.au/guidelines-publications/r39 10 Goldacre B How to get all trials reported: audit, better data, and individual accountability . PLoS Med 2015 ;12 :e1001821 10.1371/journal.pmed.1001821 25874719 11 National Health and Medical Research Council . Project grants 2015 https://www.nhmrc.gov.au/grants-funding/apply-funding/project-grants 12 Australia and New Zealand Clinical Trials Registry. About us 2016 http://www.anzctr.org.au/Support/AboutUs.aspx 13 ClinicalTrials.gov . ClinicalTrials.gov: A service of the U.S. National Institutes for Health 2016 https://clinicaltrials.gov 14 ISRCTN . ISRCTN registry 2016 http://www.controlled-trials.com/page/about 15 Beyersmann J , Allignol A , Schumacher M Competing risks and multistate models with R . New York : Springer , 2012 . 16 Beyersmann J , Gastmeier P , Wolkewitz M An easy mathematical proof showed that time-dependent bias inevitably leads to biased effect estimation . J Clin Epidemiol 2008 ;61 : 1216 –21 . 10.1016/j.jclinepi.2008.02.008 18619803 17 Lumley T mitools: Tools for multiple imputation of missing data. R package version 2.0.1. 2010 https://cran.r-project.org/web/packages/mitools/index.html 18 National Health and Medical Research Council . NHMRC Project Grants peer review guidelines for funding commencing in 2014 2013 http://www.nhmrc.gov.au/grants-funding/apply-funding/project-grants 19 Mooney LA , Fay L Cross-sectional study of Pfizer-sponsored clinical trials: assessment of time to publication and publication history . BMJ Open 2016 ;6 :e012362 10.1136/bmjopen-2016-012362 20 Tice DG , Carroll KA , Bhatt KH Characteristics and causes for non-accrued clinical research (NACR) at an academic medical institution . J Clin Med Res 2013 ;5 :185 –93 . 10.4021/jocmr1320w 23671544 21 Chan AW , Tetzlaff JM , Altman DG SPIRIT 2013 statement: defining standard protocol items for clinical trials . Ann Intern Med 2013 ;158 :200 –7 . 10.7326/0003-4819-158-3-201302050-00583 23295957 22 Turner S , Wright D , Maeso R Publication rate for funded studies from a major UK health research funder: a cohort study . BMJ Open 2013 ;3 :pii: e002521 10.1136/bmjopen-2012-002521 23 Emerson GB , Warme WJ , Wolf FM Testing for the presence of positive-outcome bias in peer review: a randomized controlled trial . Arch Intern Med 2010 ;170 :1934 –9 . 10.1001/archinternmed.2010.406 21098355 24 Ioannidis JA Effect of the statistical significance of results on the time to completion and publication of randomized efficacy trials . JAMA 1998 ;279 :281 –6 .9450711

PMC005xxxxxx/PMC5372123.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01527610.1136/bmjopen-2016-015276Research MethodsResearch1506173017041724Recruitment and retention of participants in randomised controlled trials: a review of trials funded and published by the United Kingdom Health Technology Assessment Programme http://orcid.org/0000-0001-9000-8126Walters Stephen J Bonacho dos Anjos Henriques-Cadby Inês Bortolami Oscar Flight Laura Hind Daniel http://orcid.org/0000-0001-6710-5403Jacques Richard M Knox Christopher Nadin Ben Rothwell Joanne Surtees Michael Julious Steven A School of Health and Related Research, University of Sheffield, Sheffield, UKCorrespondence to Professor Stephen J Walters; s.j.walters@sheffield.ac.uk2017 20 3 2017 7 3 e01527622 11 2016 23 1 2017 2 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Background Substantial amounts of public funds are invested in health research worldwide. Publicly funded randomised controlled trials (RCTs) often recruit participants at a slower than anticipated rate. Many trials fail to reach their planned sample size within the envisaged trial timescale and trial funding envelope. Objectives To review the consent, recruitment and retention rates for single and multicentre randomised control trials funded and published by the UK's National Institute for Health Research (NIHR) Health Technology Assessment (HTA) Programme. Data sources and study selection HTA reports of individually randomised single or multicentre RCTs published from the start of 2004 to the end of April 2016 were reviewed. Data extraction Information was extracted, relating to the trial characteristics, sample size, recruitment and retention by two independent reviewers. Main outcome measures Target sample size and whether it was achieved; recruitment rates (number of participants recruited per centre per month) and retention rates (randomised participants retained and assessed with valid primary outcome data). Results This review identified 151 individually RCTs from 787 NIHR HTA reports. The final recruitment target sample size was achieved in 56% (85/151) of the RCTs and more than 80% of the final target sample size was achieved for 79% of the RCTs (119/151). The median recruitment rate (participants per centre per month) was found to be 0.92 (IQR 0.43–2.79) and the median retention rate (proportion of participants with valid primary outcome data at follow-up) was estimated at 89% (IQR 79–97%). Conclusions There is considerable variation in the consent, recruitment and retention rates in publicly funded RCTs. Investigators should bear this in mind at the planning stage of their study and not be overly optimistic about their recruitment projections. recruitmentRetentionRandomised Controlled TrialsHealth Technology AssessmentPublicly fundedreview ==== Body Strengths and limitations of this study Substantial amounts of public funds are spent on healthcare research and randomised controlled trials (RCTs) and this is potentially wasted if a trial fails to recruit to time and target sample size. Trialists and funders have highlighted recruitment and retention as a key issue for the conduct of RCTs. This study reports the recruitment and retention rates for 151 single and multicentre randomised control trials funded by the UK's National Institute for Health Research (NIHR) Health Technology Assessment (HTA) Programme and published in the HTA Journal between 2004 and 2016. There is considerable variation in the consent, recruitment and retention rates in publicly funded RCTs. Crude recruitment rates, assuming all centres were recruiting for the same time period were calculated; as such the recruitment rate estimates may be an underestimation of the true recruitment rate. The study was restricted to publicly funded RCTs published as reports in the HTA Journal and not commercially sponsored trials. Recruitment to trials is complex and the complete picture cannot be untangled in a simple review. Introduction Substantial amounts of public funds are invested in medical research worldwide with an estimate of US$100 billion in 2012.1 In 2014/2015, the National Institute for Health Research (NIHR) in England spent £237.6 million across a broad range of research programmes and initiatives to ensure that patients and the public benefit from the most cost-effective, up-to-date health interventions and treatments.2 A substantial proportion of this research expenditure was invested in Randomised Controlled Trials (RCTs) to assess the clinical and cost-effectiveness of new health technologies. RCTs are widely regarded as the most powerful research design for evaluating new health technologies and decision makers, such as the UK's National Institute for Health and Care Excellence (NICE), are increasingly using the results of RCTs to guide practice and policy. A frequently reported problem with publicly funded RCTs is that the recruitment of participants is often slower or more difficult than expected. Many trials fail to reach their planned sample size within the originally envisaged trial timescale and trial funding envelope. A review of a cohort of 122 trials funded by the UK Medical Research Council (MRC) and the NIHR Health Technology Assessment Programme (HTA), between 1994 and 2002 found that less than a third (31%) of the trials achieved their original patient recruitment target; 55/122 (45.1%) achieved < 80% of their original target and half (53%) were awarded an extension.3 This situation has improved marginally over time, with a recent review of 73 HTA/MRC funded studies recruiting between 2002 and 2008,4 finding that 55% (40/73) of the trials achieved their original patient recruitment target; 16/73 (22%) achieved <80% of their original target and 45% (33/73) were awarded an extension.4 A HTA commissioned review recommended further research is required, particularly in relation to: problems being experienced and solutions employed in current RCTs; the optimum structure, staffing and organisation for the conduct of large and small trials; and the factors which influence the participation5 in RCTs. Several Cochrane systematic reviews have suggested strategies to improve the recruitment6 7 and retention8 of participants to RCTs. These recruitment strategies include: telephone reminders; requiring potential participants to opt-out of being contacted by the trial team regarding participation; and open (unblinded) designs. Other HTA commissioned reviews have shown that participant and clinician preferences, for the intervention, can affect trial recruitment9 and that payment to healthcare professionals for patient recruitment to trials acts as a limited incentive.10 However despite the growing literature summarising the barriers and facilitators to recruitment to RCTs only 55% of trials are recruiting to within 80% of the target. A recent survey among the directors of the clinical trials units registered with the UK NIHR Clinical Research Network identified priorities for research into the methodology of trials. The top three priorities were improving recruitment, choice of outcomes and improving retention.11 The Consolidated Standards of Reporting Trials (CONSORT) Statement, first published in 1996,12 13 and revised in 200114 and 2010,15 is a set of standards for publication of results of RCTS in medical journals. They are for the article itself and the article abstract.15 The CONSORT statement includes details of the number of eligible patients; number of patients randomised; number of recruiting centres and recruitment time period (start and finish time of recruitment). A review of publicly funded RCTs was carried out to evaluate how well recruitment and retention figures are reported; how successful RCTs are in reaching their target sample size and retaining participants and to assess recruitment rates. Methods Trial identification Reports of individually RCTs published in the NIHR HTA Journal from January 2004 to April 2016 were reviewed. The HTA Journal publishes research on the effectiveness, costs and broader impact of health technologies for those who use, manage and provide care in the NHS. Reports are published in the HTA Journal if (1) they have resulted from work for the NIHR HTA Programme and (2) they are of a sufficiently high scientific quality as assessed by the external reviewers and journal editors (http://www.journalslibrary.nihr.ac.uk/hta/about-the-journal). Trial reports published in the HTA Journal were chosen as they are of high quality and provide detailed trial and recruitment information including the number of centres and recruitment period. A pilot review of 30 trials reported in five major journals: British Medical Journal (BMJ), The Lancet, New England Journal of Medicine, The Journal of the American Medical Association (JAMA) and Annals of Internal Medicine as well as six trials reported in the HTA Journal, found that there was sufficient information to calculate the recruitment rate for only 23 out of the 30 trials reported in the journals. This information was available for all six trials reported in the HTA Journal. The pilot identified six trials published in the HTA journal over a 7 months period and with a basic extrapolation over a 12 year publication period it was deemed this would provide a large and manageable number of trials for review. Limiting the review to publicly funded trials published in the HTA Journal identified trials from medicine, surgery and therapy as well as from a range of disease areas. The HTA Journal reports were obtained from the NIHR Journals Library website (http://www.journalslibrary.nihr.ac.uk/hta—date last accessed 10 August 2016) along with any previously published trial paper, protocol paper or trial protocol, where available. For trials that had a published International Standardised Randomised Controlled Trial Number (ISRCTN) number this was used to check the ISRCTN register of clinical trials for any additional information, a trial website or any previously unobtainable trial reports (cf. http://www.isrctn.com/). The trial report published in the HTA Journal was used as the main resource where there were discrepancies in reporting. The titles and abstracts of all reports published in the HTA journal from January 2004 to April 2016 were checked for relevance. January 2004 was chosen as a start date for the review because there were relatively few reports of RCTs in the first 7 years of the HTA Journal from 1997 to 2003 (13 RCTs out of 208 reports). Inclusion/exclusion criteria Trials included in the review were single and multicentre RCTs that were either fully or partially randomised and recruitment to the trial had finished. Where trials had reported early termination, either prior to completion of recruitment or following recruitment but prior to completion of follow-up, the trials were retained in the review and the reason for the termination was noted. Trials that were nested parallel trials as part of another RCT were included as were trial reports of two or more parallel RCTs. Trials that were excluded include: cluster randomised trials as these have separate specific recruitment issues;16 adaptive designs and pilot trials. HTA reports of a pilot trial that went on to a full trial were retained and the results from the full trial were extracted and included in the review. Trials of influenza vaccination were excluded as these recruit patients over a short period of time, usually 1–3 months, and so have an exceptionally high recruitment rate. Data extraction Once HTA reports had been selected for inclusion, information was extracted, using a standardised data extraction form. For each trial the following general trial information was extracted: the trial design, the clinical area, use of blinding, intervention type, type of control, number of arms, single or multicentre and number of centres, recruitment setting and the number and timing of follow-up visits. Data relating to the sample size and recruitment rate was extracted including: the target and actual sample size, the overall and centre-specific recruitment period and CONSORT information on the numbers screened, consented, randomised and analysed with the primary outcome.17 Where available more detailed trial information was recorded including: use of a pilot, whether there was support from a trials unit, geographical region, recruitment strategy and country where the trial took place. Data extraction was carried out by a team of reviewers. Each article was independently reviewed by a second member of the review team. Any uncertainties were resolved by discussion. The standard of reporting of trial information was good but for some variables the details were not always available. There was limited information about whether trials had any form of pilot phase or had involvement from a clinical trials unit. These features were recorded as absent where they were not mentioned. The primary outcome for the review was considered to be the recruitment rate for each trial. To calculate this accurately the centre-specific recruitment periods within the trials were extracted. However, this was generally poorly reported. Analysis The recruitment rate was defined as the number of participants recruited and randomised per centre per month. This was summarised and compared using median rates and the IQR due to the skewness of the distribution of the data.18 For secondary outcomes, the percentage of eligible patients randomised and randomised patients assessed and analysed with the primary outcome (retention), were expressed in terms of the median and IQR. Comparisons were made between categories of different characteristics using appropriate non-parametric tests; Wilcoxon rank sum (for characteristics with two levels), Kruskal-Wallis (three or more nominal levels) and non-parametric test of trend (three or more ordered levels)19). Additionally, associations between trial characteristics and recruitment rates were investigated individually using Wilcoxon rank sum tests to compare trials on the presence and absence of certain characteristics. Analysis was carried out on a complete case basis so where the characteristic information, the recruitment or retention data were missing these were excluded. Data were collected in excel and transferred to the R statistical software for analysis.20 The recruitment period was calculated as the time between dates of recruitment start and recruitment completion. If only months were reported the recruitment period was estimated as the time between the 1st of the first month and the end of the final month unless explicitly stated otherwise. If the date of the first recruit was reported instead of the date of start of recruitment then the 1st of the month of the first recruit was taken as the start of recruitment. Start of screening was used to calculate the recruitment period where the start of recruitment was not reported. In cases where information on the start and end of the recruitment period was not explicitly reported this was estimated from subtracting the length of the follow-up period from the length of study period where this was suitably reported. The recruitment rate was calculated in two distinct ways. First, to calculate the overall recruitment rate, the total number of patients recruited was divided by the maximum number of sites recruiting, then divided by total number of months that the trial recruited for. In reality the opening of trial sites is likely to be staggered. For the majority of trials most sites do not recruit for the entire recruitment period. For this reason this estimate of the overall recruitment rate for multicentre trials is likely to be an underestimate. To account for the differences in start-up times for sites and the corresponding site-specific recruitment periods, where available, the centre-specific recruitment periods were extracted. These were averaged over the number of sites to give an average centre-specific recruitment period. An average recruitment rate was calculated as the total number of patients recruited, divided by the maximum number of sites and then divided by the average number of months recruiting. Results In total 778 reports were published between January 2004 and April 2016 in the HTA Journal and 596 of these were excluded following screening of all titles and abstracts. The search produced 191 trial reports of randomised trials of which a further 40 were excluded for various reasons (18 cluster RCTs; 15 pilot/feasibility studies; 3 influenza vaccination trials and four excluded for other reasons). The 15 pilot/feasibility studies were standalone/external trials at the outset and were not definitive trials that were changed to a pilot study as a result of poor recruitment. In total, 151 individually RCTs were included in the review and analysed as shown in figure 1. Figure 1 Flow diagram of search process for a review of trial reports published in the Health Technology Assessment Journal between 2004 and end of April 2016. RCT, randomised controlled trials. Trial characteristics The characteristics of the 151 trials included in the review are summarised in table 1. The majority of trials were two armed, parallel group, multicentre trials. Trials were identified from a variety of different clinical areas with 18% (27/151) of trials in mental health, including neurosciences, psychiatry and psychology, and 13% (19/151) of trials of musculoskeletal conditions, including orthopaedics, rheumatology and back pain. Trials were most commonly set in hospitals (54% (82/151)) and 91% (137/151) took place solely in the UK. Drug trials were as common as therapy trials, both occurred more frequently than surgery trials. 30% (46/151) of trials used an intervention that was not easily categorised and there was a far greater number of trials that used an active control (80% (121/151)) compared with a placebo (20% (30/151)). Table 1 Characteristics of the trials included in the review Characteristic n (%) Trial design (n=151)  Parallel 129 (85)  Factorial 10 (7)  Crossover 1 (1)  Other (patient preference/Zelen's) 11 (7) Arms (n=151)  2 101 (67)  3 30 (21)  4 13 (9)  >4 7 (5) Clinical area (n=151)  Cancer/oncology 8 (5)  Mental health 27 (18)  Musculoskeletal 19 (13)  Obstetrics and gynaecology 9 (6)  Primary care 7 (5)  Cardiovascular 12 (8)  Gastrointestinal 6 (4)  Respiratory 14 (9)  Stroke 4 (3)  Diabetes 4 (3)  Dermatology (including ulcers) 10 (7)  Other* 31 (21) Setting (n=151)  Hospital 82 (54)  General practice 20 (13)  Mixed 25 (17)  Community 16 (11)  Other† 8 (5) Intervention type (n=151)  Drug intervention 37 (25)  Therapy 36 (24)  Surgery 19 (13)  Complex intervention 13 (9)  Other‡ 46 (30) Control type (n=151)  Placebo 30 (20)  Active 121 (80) Patient blinded? (n=147)  No 29 (19)  Yes 118 (79) Centres outside the UK? (n=151)  No 138 (91)  Yes 13 (9) Geographical spread (n=148)  Multiple regions 119 (80)  Regional 29 (20) Some form of pilot§?  Yes 59 (41)  No 87 (60)  Not stated 5 (3) *Alcohol abuse, chronic fatigue, nutrition, infectious diseases, paediatric (general, dermatology, anaesthesiology), gerontology, hepatology (hepatitis C), intensive care, multiple sclerosis, minor surgery, neurology (Bell's palsy, cerebral palsy epilepsy), neurosurgery, ophthalmology, otorhinolaryngology, physical exercise, rehabilitation, resuscitation, sleep disorders, speech therapy, urology (general, urinary tract infections, incontinence, prostate disorders), vascular. †HIV Clinical Centres, University Clinics, Sexual Health Clinics, Primary and Secondary Strike Care service, Intellectual Disability Services, Public Schools, Leisure Centres, Physical Therapy Classes and Specialist care centres. ‡Technique, equipment, diagnostic intervention, advice and information, consultation, patient pathway, drug versus surgery, health professional. §Any mention of pilot work or feasibility study recorded. The trial characteristics relating to recruitment and sample size are summarised in table 2. The final recruitment/sample size target ranged from 44 participants to 28 000 and final total actual/achieved recruitment ranged from 19 participants to 24 510. Two trials targeted and achieved recruitment of 20 000 or more participants; one of these was a trial of trauma patients21 and the other was a cervical screening trial.22 Overall 56% (85/151) of trials recruited to their final recruitment target and 79% (119/151) managed to recruit to within 80% of the recruitment target. For 34% (52/151) of trials the original sample size target was revised (downward in 79% (41/52)). Eight single-centre trials were identified. Five trials recruited in more than 100 centres; the maximum number of centres was 274. The majority of trials had a final follow-up visit at 18 months or less postrandomisation and the longest reported final follow-up was 10 years postrandomisation. Table 2 Recruitment and sample size characteristics of the trials included in the review Characteristic (N=151) n (%) Mean (SD) Median Range No. of centres  1 8 (5) 29 (35) 15 1–274  2–5 23 (15)  6–10 21 (14)  11–20 33 (22)  21–50 35 (23)  51–100 22 (15)  >100 5 (3)  Missing 4 (3) Original target recruitment  ≤200 14 (9) 1231 (2946) 545 90–28 000  201–400 40 (26)  401–600 31 (21)  601–800 15 (10)  >800 50 (33)  Missing 1 (1) Final target recruitment  ≤200 17 (11) 1132 (2926) 480 44–28 000  201–400 49 (32)  401–600 27 (18)  601–800 13 (9)  >800 45 (30) Final total recruitment  ≤200 24 (16) 1014 (2673) 424 19–24 510  201–400 48 (32)  401–600 28 (19)  601–800 12 (8)  >800 39 (26) Final recruitment target achieved  Yes 85 (56)  No, but with 80% of target 119 (79)  No, <80% of target 32 (21) Timing of primary outcome follow-up (months postrandomisation)  <1 month 27 (18) 9 (10) 6 0–48  1–6 months 54 (36)  6–18 months 36 (24)  >18 months 21 (14)  Missing 13 (9) Timing of final follow-up (months postrandomisation)  <1 month 9 (6) 15 (18) 12 0.066–120  1–6 months 20 (13)  6–18 months 84 (56)  >18 months 33 (22)  Missing 5 (3) CONSORT and recruitment data The data completeness in relation to CONSORT and recruitment information is summarised in table 3. Out of the 151 trials identified 95 (63%) demonstrated complete compliance with the CONSORT statement and reported each of the number: screened, eligible, declined consent, recruited and assessed in their primary outcome. The number of participants recruited, randomised and assessed for the primary outcome, used to measure retention, was available for all 151 trials. To calculate the recruitment rate 144 out of 151 trials reported the maximum length of the recruitment period, from first centre opening to completion of recruitment, and 106 reported the total number of centres that recruited at least one participant. Centre-specific recruitment information, used to calculate an average recruitment period per centre, could only be extracted from 34 of the 111 trials (25%). The overall recruitment rate, based on the maximum recruitment length, was calculated for 142 out of 151 RCTs. Table 3 Data completeness in relation to CONSORT guidelines and recruitment information Trial characteristic (N=151) n (%) Number screened 127 (84) Number eligible 109 (72) Number refused/declined consent 106 (70) Total recruitment 151 (100) Number included in primary analysis (retention) 151 (100) Number of centres 106 (70) Maximum recruitment length 144 (95) Centre-specific recruitment length 34 (23) Recruitment rate can be calculated 142 (94) CONSORT, Consolidated Standards of Reporting Trials. Recruitment and retention rates From the 142/151 (94%) trials with sufficient information the median recruitment rate was found to be 0.92 patients recruited per centre per month. This ranged from 0.04 to 57.75 patients per centre per month, with 80th and 90th percentiles of 4.4 and 10.1 patients recruited per centre per month, respectively. The two studies found to have the largest recruitment rates were single-centre studies,23 24 recruited from obstetrics and gynaecological populations (figure 2). The eight single-centre studies produced five of the nine23–31 largest recruitment rates ranging from 16 to 58 patients per centre per month. Taking the multicentre studies on their own the median recruitment was 0.86 patients recruited per centre per month with a range from 0.04 to 30.11 patients recruited per centre per month. A median of 70% (IQR 51–87%) of eligible patients were consented and randomised and a median 89% (IQR 79–97%) of randomised patients had valid primary outcome data for analysis (table 4). Table 4 Overall recruitment and retention rates Median IQR Range Eligible patients consented and randomised (N=109) 70% 51–87% 14–100% Recruited per centre per month (N=142) 0.92 0.43–2.79 0.04–57.75 Randomised patients retained and assessed in primary outcome (N=151) 89% 79–97% 23–100% Figure 2 Recruitment Rates by clinical area for the 151 Health Technology Assessment trials considered. Tables 5 and 6 summarise the trial recruitment and retention rates by various trial characteristics (setting, number of arms, control type, original and final recruitment targets, total number recruited and time of follow-up). There is some statistical evidence of an association between trial setting, final recruitment target and the total number recruited although there is no clear pattern to these associations. Table 5 Association between recruitment rate (number of patients/centre/month) and trial characteristics Characteristic n Median IQR p Value Setting  Hospital 82 1.22 0.58–2.61  General practice 20 0.52 0.23–0.85  Mixed 25 0.98 0.46–3.57  Community 16 1.62 0.38–4 0.043*†  Other 8 3.62 0.53–11.48 Arms  2 101 0.98 0.44–3.01  3 30 0.89 0.39–5.86  4 13 1.04 0.76–2.45  >4 7 0.61 0.39–2.43 0.889† Control type  Placebo 30 1.29 0.54–4.01  Active 121 0.88 0.42–2.6 0.427‡ Original target recruitment  ≤200 14 0.49 0.21–2.23  201–400 40 1.30 0.51–2.26  401–600 31 0.87 0.42–2.33 0.033§  601–800 15 0.87 0.39–2.61  >800 50 1.34 0.58–5.73 Final target recruitment  ≤200 17 0.87 0.59–3.5  201–400 49 1.96 0.72–5.68  401–600 27 0.72 0.42–1.67  601–800 13 0.41 0.07–1.14 <0.001§  >800 45 0.89 0.39–4.42 Total recruitment  ≤200 24 0.60 0.34–1.72  201–400 48 1.40 0.42–4.28 <0.001§  401–600 28 0.84 0.39–1.61  601–800 12 1.51 0.28–2.17  >800 39 1.38 0.43–5.48 Timing of final follow-up  <1 month 9 1.77 0.39–7.48  1–6 months 41 1.11 0.73–5.43 0.352§  6–18 months 63 0.62 0.31–1.98  >18 months 33 0.87 0.42–3.85 *The category ‘other’ was not included in Kruskal-Wallis test. †p Values are reported from a Kruskal-Wallis test. ‡p Values are reported from a Wilcoxon rank sum test. §p Values are reported from a nonparametric test of trend (Cuzick). Table 6 Association between the trial retention rate (% of randomised participants with valid primary outcome data for analysis) and trial characteristics Characteristic n Median IQR p Value Setting  Hospital 81 92.4 81.7–99.1    General practice 20 85.6 77.4–91.0    Mixed 25 90.0 84.8–97.4    Community 16 85.4 79.1–96.1 0.019*†  Other 8 99.4 98.8–100.0   Arms  2 100 0.90 0.82–0.98    3 30 0.89 0.79–0.97    4+ 20 0.92 0.83–0.97 0.747† Control type  Placebo 30 90.0 88.7–99.4    Active 120 89.9 81.0–97.4 0.166‡ Final target recruitment  ≤200 17 93.7 87.6–98.3    201–400 49 89.2 79.8–96.8 <0.001§  401–600 27 86.7 72.2–100.0    601–800 13 86.3 83.2–89.9    >800 44 94.0 76.4, 99.4   Total recruitment  ≤200 23 94.7 86.4–100.0    201–400 48 89.1 79.3–96.4 <0.001§  401–600 28 85.7 81.7–92.5    601–800 12 89.9 88.5–94.6    >800 39 94.0 77.8–99.3   Timing of final follow-up  <1 month 9 99.3 77.4–100.0    1–6 months 41 94.6 84.8–100.0 0.693§  6–18 months 62 86.2 75.1–96.8    >18 months 33 89.2 85.6–95.4   *The category ‘other’ was not included in Kruskal-Wallis test. †p Values are reported from a Kruskal-Wallis test. ‡p Values are reported from a Wilcoxon rank sum test. §p Values are reported from a nonparametric test of trend (Cuzick). Table 7 compares the results of the current review, in terms of successful recruitment to target sample size, with two previous reviews.3 4 It should be noted that there is some overlap in the trials included in our review and Sully et al; so we have included a column with the non-overlapping time interval for the 2009–2016 data. Table 7 shows that reaching 100% of the original sample size target is lower in 2009–2016 than previous periods/reviews; with only 50% (45/90) achieving the original sample size target. The original sample size target was revised in 39% (35/90) of trials; and this revision was downwards for the majority of trials, 71% (25/35). Table 7 Comparison of current review with results of two previous reviews in terms of successful recruitment to target sample size and extensions to recruitment Review McDonald et al3 Sully et al4 This study This study Recruitment period 1994–2002 2002–2008 2009–2016 2004–2016 Number of trials in the study N=122 N=73 N=90 N=151 Recruited 100% of original target 38 of 122 (31%) 40 of 73 (55%) 45 of 90 (50%) 61 of 151 (40%) Original target was revised 42 of 122 (34%) 14 of 73 (19%) 35 of 90 (39%) 52 of 151 (34%) Original target revised upward 6 of 42 (14%) 5 of 14 (36%) 10 of 35 (29%) 11 of 52 (21%) Original target revised downward 36 of 42 (86%) 9 of 14 (64%) 25 of 35 (71%) 41 of 52 (79%) Recruited 80% of original target 67 of 122 (55%) 57 of 73 (78%) 65 of 90 (72%) 95 of 151 (63%) Recruited 100% of revised target 19 of 42 (45%) 10 of 14 (71%) 26 of 35 (74%) 28 of 52 (54%) Recruited 80% of revised target 34 of 42 (80%) 13 of 14 (93%) 31 of 35 (89%) 48 of 52 (92%) Extended their recruitment 65 of 122 (54%) 33 of 73 (45%) 28 of 90 (31%) 49 of 151 (32%) Discussion and conclusions This study provides a comprehensive review of the recruitment and retention data of a cohort of 151 RCTs funded and published by the UK NIHR HTA Programme from 2004 to 2016. This review found that the final recruitment target sample size was achieved in 56% (85/151) of the RCTs; the median recruitment rate (participants per centre per month) was 0.92 (IQR 0.43–2.79) and the median retention rate 89% (IQR 79–97%). This review found that 56% of publicly funded RCTs achieve their target sample size, a similar figure to that found in the most recent review of 55%, by Sully et al which covered the period of 2002 to 2008.4 However, there is still a suggestion that recruitment success is improving slightly compared with the previous review covering the period of 1994 to 2002.3 Even though the recruitment picture is improving there is certainly still room for improvement with more than half of the 151 publicly funded RCTs not recruiting to target which in some cases was revised down during the course of the trial. These findings are congruent with the concerns of clinical trials unit directors.11 There is a possible relationship between planned sample size and recruitment rate with recruitment rate increasing as the target sample size increases. Sample sizes are inflated for expected attrition or non-response and this is commonly set at 10–20%.18 32 The estimate of average retention was 89% suggesting that the current inflation of sample sizes for attrition is reasonable. Overall retention is not as big an issue as recruitment in terms of fulfilling a sample size for a primary outcome. These findings slightly contrast with the concerns of clinical trials unit directors'.11 The retention figure, however, will be affected by the number of trials with short-term outcomes and the use of survival analysis methods with time to event outcomes, where missing outcomes are typically censored at the time of any loss to follow-up and but included in the analysis. This study has several limitations. Data extraction was carried out by two independent reviewers. Reviewers conferred to try to ensure consistency in the interpretation of data extraction items; but it is possible that errors have occurred. Crude recruitment rates, assuming all centres were recruiting for the same time period were calculated; as such the recruitment rate estimates may be an underestimation of the true recruitment rate. The study was restricted to publicly funded RCTs published as reports in the HTA Journal and not commercially sponsored trials. There is a possibility of publication bias as this study is restricted to trials that have had their results published in the NIHR HTA Journal; as not all funded trials are actually published. However the possibility of publication bias is small as a review of projects funded by the NIHR HTA Programme, between 2002 and 2011, found that 98% (274/280) published in the programme's journal.33 The HTA's expectation (in line with their contract) is that all HTA Programme funded studies publish in the NIHR Journals Library, even when they have had to close early because of, for example, poor recruitment. The HTA will ask investigators to include a section/chapter on the challenges faced and lessons learnt that will then inform other researchers who might be considering similar research. There are limitations in drawing conclusions from this data, not least in the accuracy of the recruitment rates calculated, potential confounders not accounted for and in some cases underlying factors that cannot be measured in the data. Recruitment to trials is complex and the complete picture cannot be untangled in a simple review. This review does, however, provide some pointers to factors that might need to be considered when estimating recruitment periods for RCTs and could be used in models of recruitment projection. Recent qualitative research has also highlighted that realistic estimation of recruitment rates is complex and that early planning and pilot and feasibility work to help project trial recruitment is important.34 In practice, recruitment rates will vary, depending on whether the target population is acute, where opportunistic recruitment will target incident cases, or chronic, where database recruitment can effectively target prevalent cases.34 35 It will also vary according to whether the intervention is therapeutic or preventive36 and the base incidence and prevalence rate of the condition. Based on this review for most publicly funded trials the recruitment rate is likely to be between 1 and 2 participants per centre per week (4–10 a month). There is considerable variation in the consent, recruitment and retention rates in publicly funded RCTs. Investigators should bear this in mind at the planning stage of their study and not be overly optimistic about their recruitment projections. Contributors: SJW is the guarantor of the study, had full access to all the data in the study, and is responsible for the integrity of the data and the accuracy of the data analysis. SJW contributed to the study conception and design, acquisition of data, analysis and interpretation of data, and writing of the report. JR contributed towards the selection of relevant data and extraction of data, as well as the drafting of the paper. CK contributed to the study design and the selection and extraction of the data. IBdAH-C contributed towards the selection, extraction and analysis of the data, as well as the drafting of the paper and graphics within it. OB contributed to the selection and extraction of the data, as well as the drafting of the paper. RMJ contributed to the selection and extraction of the data, as well as the drafting of the paper. MS contributed to the selection and extraction of the data, as well as the drafting of the paper. DH contributed to the study conception and design, the acquisition of data, the interpretation of data, as well as the drafting of the paper. SAJ contributed to the selection and extraction of the data, as well as the drafting of the paper. BN contributed to the selection and extraction of the data, as well as the drafting of the paper. LF contributed to the selection and extraction of the data, as well as the drafting of the paper. Funding: This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Competing interests: All authors have completed the Unified Competing Interest form at http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare (1) no financial support for the submitted work from anyone other than their employer; (2) no financial relationships with commercial entities that might have an interest in the submitted work; (3) no spouses, partners or children with relationships with commercial entities that might have an interest in the submitted work; (4) No non-financial interests that may be relevant to the submitted work. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Data set available on request from the corresponding author at s.j.walters@sheffield.ac.uk ==== Refs References 1 Chakma J , Gordon H , Sun MD Asia's ascent—global trends in biomedical R&D expenditures . N Engl J Med 2014 ;370 :3 –6 . 10.1056/NEJMp1311068 24382062 2 National Institute for Health Research Annual Report 2014/15 (online). http://www.nihr.ac.uk/news/nihr-annual-report-201415/3111 (accessed 28 Feb 2017) . 3 McDonald AL , Knight RC , Campbell MK What influences recruitment to randomised controlled trials? A review of trials funded by two UK funding agencies . Trials 2006 ;7 :9 10.1186/1745-6215-7-9 16603070 4 Sully BG , Julious SA , Nicholl J A reinvestigation of recruitment to randomised, controlled, multicenter trials: a review of trials funded by two UK funding agencies . Trials 2013 ;14 :166 10.1186/1745-6215-14-166 23758961 5 Prescott RJ , Counsell CE , Gillespie WJ Factors that limit the quality, number and progress of randomised controlled trials . Health Technol Assess (Winchester, England) . 1999 ;3 :1 –143 . 6 Treweek S , Mitchell E , Pitkethly M Strategies to improve recruitment to randomised controlled trials . Cochrane Database Syst Rev 2010 ;(4) :MR000013 10.1002/14651858.MR000013.pub5 7 Rendell JM , Merritt RD , Geddes JR Incentives and disincentives to participation by clinicians in randomised controlled trials . Cochrane Database Syst Rev 2007 ;(2 ) :MR000021 http://dx.doi.org/10.1002/14651858.MR000021.pub3 17443636 8 Brueton VC , Tierney J , Stenning S Strategies to improve retention in randomised trials . Cochrane Database Syst Rev 2013 ;(12) :MR000032 10.1002/14651858.MR000032.pub2 24297482 9 King M , Nazareth I , Lampe F Conceptual framework and systematic review of the effects of participants’ and professionals’ preferences in randomised controlled trials . Health Technol Assess 2005 ;9 :1 –186 . 10.3310/hta9350 10 Raftery J , Bryant J , Powell J Payment to healthcare professionals for patient recruitment to trials: systematic review and qualitative study . Health Technol Assess 2008 ;12 :1 –128 . 11 Smith TC , Hickey H , Clarke M The trials methodological research agenda: results from a priority setting exercise . Trials 2014 ;15 :32 .24456928 12 Begg C , Cho M , Eastwood S Improving the quality of reporting of randomized controlled trials. The CONSORT statement . JAMA 1996 ;276 :637 –9 . 10.1001/jama.1996.03540080059030 8773637 13 Altman DG Better reporting of randomised controlled trials: the CONSORT statement . BMJ 1996 ;313 :570 –1 . 10.1136/bmj.313.7057.570 8806240 14 Moher D , Schulz KF , Altman DG The CONSORT statement: revised recommendations for improving the quality of reports of parallel-group randomized trials . Ann Intern Med 2001 ;134 :657 –62 . 10.7326/0003-4819-134-8-200104170-00011 11304106 15 Schulz KF , Altman DG , Moher D , for the CONSORT Group . CONSORT 2010 Statement: updated guidelines for reporting parallel group randomised trials . BMJ 2010 ;340 :c332 10.1136/bmj.c332 20332509 16 Edwards SJL , Braunholtz DA , Lilford RJ Ethical issues in the design and conduct of cluster randomised controlled trials . BMJ 1999 ;318 :1407 –9 . 10.1136/bmj.318.7195.1407 10334756 17 Schulz KF , Altman DG , Moher D Factors that limit the quality, number and progress of randomised controlled trials . BMC Med 2010 ;8 :18 10.1186/1741-7015-8-18 20334633 18 Campbell MJ , Machin D , Walters SJ Medical statistics: a text book for the health sciences , 4th edn . Chichester : Wiley , 2007 . 19 Cuzick J A Wilcoxon-type test for trend . Stat Med 1985 ;4 :87 –90 . 10.1002/sim.4780040112 3992076 20 R Core Team . R: a language and environment for statistical computing . Vienna, Austria : R Foundation for Statistical Computing , 2016 https://www.r-project.org/ . 21 Roberts I , Shakur H , Coats T The CRASH-2 trial: a randomised controlled trial and economic evaluation of the effects of tranexamic acid on death, vascular occlusive events and transfusion requirement in bleeding trauma patients . Health Technol Assess 2013 ;17 :1 –79 . 10.3310/hta17100 22 Kitchener HC , Almonte M , Gilham C ARTISTIC: a randomised trial of human papillomavirus (HPV) testing in primary cervical screening . Health Technol Assess 2009 ;13 :1 –150 , iii–iv 10.3310/hta13510 23 Robson SC , Kelly T , Howel D Randomised preference trial of medical versus surgical termination of pregnancy less than 14 weeks’ gestation (TOPS) . Health Technol Assess 2009 ;13 :1 –124 . 10.3310/hta13530 24 Townsend J , Wolke D , Hayes J Routine examination of the newborn: the EMREN study. Evaluation of an extension of the midwife role including a randomised controlled trial of appropriately trained midwives and paediatric senior house officers . Health Technol Assess 2004 ;8 :iii –iv , ix–xi, 1–100 . 25 Wiggins M , Oakley A , Roberts I The social support and family health study: a randomised controlled trial and economic evaluation of two alternative forms of postnatal support for mothers living in disadvantaged inner-city areas . Health Technol Assess 2004 ;8 :iii , ix–x, 1–120 10.3310/hta8320 26 Kalra L , Evans A , Perez I A randomised controlled comparison of alternative strategies in stroke care . Health Technol Assess 2005 ;9 : iii –iv , 1–79 10.3310/hta9180 27 Sharples L , Hughes V , Crean A Cost-effectiveness of functional cardiac testing in the diagnosis and management of coronary artery disease: a randomised controlled trial. The CECaT trial . Health Technol Assess 2007 ;11 :iii –iv , ix-115 . 28 Goodacre S , Bradburn M , Fitzgerald P The RATPAC (Randomised Assessment of Treatment using Panel Assay of Cardiac markers) trial: a randomised controlled trial of point-of-care cardiac markers in the emergency department . Health Technol Assess 2011 ;15 :iii –xi , 1–102 10.3310/hta15230 29 Pickard R , Lam T , Maclennan G Types of urethral catheter for reducing symptomatic urinary tract infections in hospitalised adults requiring short-term catheterisation: multicentre randomised controlled trial and economic evaluation of antimicrobial- and antiseptic-impregnated urethral catheters (the CATHETER trial) . Health Technol Assess 2012 ;16 :1 –197 . 10.3310/hta16470 30 Allen SJ , Wareham K , Wang D A high-dose preparation of lactobacilli and bifidobacteria in the prevention of antibiotic-associated and Clostridium difficile diarrhoea in older people admitted to hospital: a multicentre, randomised, double-blind, placebo-controlled, parallel arm trial (PLACIDE) . Health Technol Assess 2013 ;17 :1 –140 . 10.3310/hta17570 31 Nicholson KG , Abrams KR , Batham S Randomised controlled trial and health economic evaluation of the impact of diagnostic testing for influenza, respiratory syncytial virus and Streptococcus pneumoniae infection on the management of acute admissions in the elderly and high-risk 18- to 64-year-olds . Health Technol Assess 2014 ;18 :1 –274 , vii–viii 10.3310/hta18360 32 Walters SJ Quality of life outcomes in clinical trials and health care evaluation: a practical guide to analysis and interpretation . Chichester : Wiley , 2009 . 33 Turner S , Wright D , Maeso R Publication rate for funded studies from a major UK health research funder: a cohort study . BMJ Open 2013 ;3 :e002521 10.1136/bmjopen-2012-002521 34 White D , Hind D Projection of participant recruitment to primary care research: a qualitative study . Trials 2015 ;16 :473 10.1186/s13063-015-1002-9 26482231 35 Stuardi T , Cox H , Torgerson DJ Database recruitment: a solution to poor recruitment in randomized trials? Fam Pract 2011 ;28 :329 –33 . 10.1093/fampra/cmq108 21156752 36 Cooper CL , Hind D , Duncan R A rapid review indicated higher recruitment rates in treatment trials than in prevention trials . J Clin Epidemiol 2015 ;68 :3 , 347–354 10.1016/j.jclinepi.2014.09.010 25441702

PMC005xxxxxx/PMC5372131.txt

==== Front RMD OpenRMD OpenrmdopenrmdopenRMD Open2056-5933BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR 28405474rmdopen-2016-00041210.1136/rmdopen-2016-000412Autoimmunity1506Clinical caseRheumatic immune-related adverse events of checkpoint therapy for cancer: case series of a new nosological entity Calabrese C 12Kirchner E 12Kontzias K 12Velcheti V 13Calabrese L H 12 1 Cleveland Clinic Foundation, Cleveland, Ohio, USA 2 Department of Rheumatology and Immunology, Cleveland Clinic, Cleveland, Ohio, USA 3 Department of Hematology and Oncology, Cleveland Clinic, Cleveland, Ohio, USACorrespondence to Dr Leonard H Calabrese; calabrl@ccf.org2017 20 3 2017 3 1 e00041221 11 2016 19 1 2017 5 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Immunotherapy of cancer with checkpoint inhibitors has been associated with a spectrum of autoimmune and systemic inflammatory reactions known as immune-related adverse events (irAEs). Rheumatic irAEs are infrequently reported and extensively described. Here, we report our experience over an 18-month period with 15 patients evaluated in the rheumatology department for rheumatic irAEs. We identified 13 patients without pre-existing autoimmune disease (AID) who subsequently developed rheumatic irAEs, and two with established AID referred pre-emptively. irAEs encountered included: inflammatory arthritis, sicca syndrome, polymyalgia rheumatica-like symptoms and myositis. All cases required glucocorticoids, and three required a biological agent. Rheumatic irAEs led to temporary or permanent cessation of immunotherapy in all but five patients. One patient with pre-existing AID experienced a flare after starting immunotherapy. Our findings underscore that rheumatic irAEs are complex, at times require additional immunosuppressive therapy, and may influence ongoing immunotherapy regimens for the primary disease. Similar irAEs will be increasingly seen as checkpoint inhibitors adopted as standard of care in the community. Autoimmune DiseasesInflammationMultidisciplinary team-care ==== Body Introduction The introduction of biological agents targeting immunological checkpoints represents a major advance in the field of oncology. At the present time, there are four Food and Drug Administration (FDA)-approved drugs: ipilimumab, targeting cytotoxic T-lymphocyte-associated protein (CTLA-4), nivolumab and pembrolizumab, targeting programmed cell death protein 1 (PD-1), and atezolizumab which targets programmed cell death ligand 1 (PD-L1). These medications have produced significant survival benefits in patients with metastatic melanoma, non-small cell lung cancer, renal cell carcinoma, Hodgkin lymphoma and urothelial carcinoma and are in investigation for many others. Many other targets for checkpoint therapy are now in clinical trials.1 Checkpoint inhibitors exploit suppressor and regulatory pathways, thereby boosting integrated immunity against tumours. Unfortunately, these new therapies are attended by a unique spectrum of immune-related adverse events (irAEs) related to overactivation of the immune system with resultant autoimmune disease (AID). The most commonly affected systems are the dermatological, gastrointestinal and endocrine. Reports of rheumatic irAEs have been sparse, not systematically reported, and have only been described in case reports or small series. These adverse events have recently been described in what was the largest case series to date.2 Most clinical trials for immunotherapy agents do not report on the rheumatic manifestations and have excluded patients with pre-existing AID.3 At our institution, we created a multidisciplinary referral process to evaluate and manage irAEs. In this article, we report a series of patients evaluated at the Cleveland Clinic Foundation from 2015 to 2016 with rheumatic irAEs as a result of immunotherapy, as well as patients with pre-existing rheumatic AID who were evaluated pre-emptively. Methods In February 2016, an interdisciplinary group was created at the Cleveland Clinic Foundation to manage irAEs occurring in patients on approved and experimental immune-based therapies for cancer. Patients were identified by the treating oncologist and then triaged by a designated advanced practitioner and seen in a facilitated fashion. Two designated rheumatologists saw all patients referred to the rheumatology arm of the multidisciplinary clinic. Two types of referrals were made: (1) patients without pre-existing AID who developed a rheumatic irAE after start of immunotherapy and (2) patients with pre-existing AID referred for pre-emptive evaluation. Patients were determined to have no pre-existing AID based on no prior diagnosis in their medical record, as well as through history taking during clinic visits with the treating rheumatologist. All patients were over the age of 18 and receiving or scheduled to receive ipilimumab, nivolumab, tremelimumab (anti-CTLA-4), durvalumab (anti-PD-L1) or atezolizumab either as monotherapy or in combinations. Patients were classified as having sicca syndrome, polymyalgia rheumatica (PMR)-like symptoms, inflammatory arthritis or myositis based on history, examination, imaging and laboratory findings as determined by the treating rheumatologist. All patients were included in a database of information culled from the electronic medical record including: gender, date of birth, age at diagnosis of malignancy, type and stage of malignancy, prior treatment (chemotherapy, radiation, surgery), checkpoint inhibitor (drug(s), date started, date of last dose), pre-existing autoimmune history, nosology of irAE (type, date of onset, diagnostic testing), irAE treatment and global response to treatment, and prior autoimmune serology. Response was clinically defined as significant on near-complete resolution of rheumatic irAE symptoms, moderate on improvement to the point that symptoms were tolerable but still present and minimal if symptoms remained severe despite treatment. Results Demographics Rheumatic irAEs were evaluated in 15 patients between February 2015 and September 2016. Thirteen patients without pre-existing rheumatic AID were referred to our rheumatology department for evaluation after onset of irAE. Two patients with established AID (one rheumatoid arthritis, one psoriatic arthritis) were evaluated pre-emptively prior to starting immunotherapy. In the entire group, the median age was 63 years and 67% were male. The most common malignancy was melanoma (seven), followed by non-small cell lung cancer (four) and renal cell carcinoma (four) (table 1). All had been previously treated with surgery, chemotherapy, radiation or a combination of two or all three treatments. Seven patients received combination therapy with ipilimumab and nivolumab, one patient received tremelimumab followed by durvalumab, one received ipilimumab followed by pembrolizumab, and the remaining six received monotherapy with either nivolumab (five) or atezolizumab (one). Patients were receiving immunotherapy as standard of care and as participants in clinical trials. Table 1 Demographic features, cancer types, immunotherapy and rheumatic immune-related adverse events (irAEs) Patient Age Sex Malignancy Immunotherapy irAE Serology Time to onset (weeks) Treatment Improvement Immunotherapy held for irAE 1 74 F NSCL Nivolumab Arthritis ANA 1:160 Anti-dsDNA 77 7.3 Prednisone 40 mg Significant Y 2 49 F Melanoma Ipilimumab Pembrolizumab Arthritis 52.7 Prednisone 20 mg HCQ Moderate Y 3 42 F RCC Ipilimumab/nivolumab Arthritis 3 Prednisone Infliximab, MTX Etanercept Moderate N 4 57 M RCC Ipilimumab/nivolumab Arthritis RF 214 48.4 Prednisone MTX Etanercept Adalimumab Significant N 5 59 F Melanoma Ipilimumab/nivolumab Arthritis 21.7 Prednisone 60 mg Minimal N 6 81 M Melanoma Ipilimumab/nivolumab Arthritis ANA 1.5 13.1 Prednisone 15 mg Moderate Y 7 57 F Melanoma Ipilimumab/nivolumab Arthritis Sicca ANA 1:320 6.7 Prednisone 30 mg Significant Y 8 61 M Melanoma Ipilimumab/nivolumab Sicca 5.3 Prednisone 60 mg* Significant Y† 9 63 M RCC Atezolizumab Sicca 21.9 Prednisone 60 mg* Significant Y† 10 68 M Melanoma Ipilimumab/nivolumab Sicca PMR ANA 1:1280 SSA 8.1 Prednisone 30 mg Significant Y 11 79 M Melanoma Nivolumab PMR Sicca 2 Prednisone 20 mg Moderate Y 12 63 M RCC Nivolumab PMR 213 Prednisone 40 mg Infliximab Minimal Y 13 68 M NSCL Tremelimumab Durvalumab Myositis 4.6 IV methylpred Prednisone 60 mg Moderate Y *Prednisone given for hypophysitis. †Immunotherapy held for hypophysitis. Atezolizumab, anti-PD-L1; durvalumab, anti-PD-L1; HCQ, hydroxychloroquine; MTX, methotrexate; NSCL, non-small cell lung cancer; PMR, polymyalgia rheumatica; RCC, renal cell carcinoma; RF, rheumatoid factor; tremelimumab, anti-CTLA-4. No pre-existing autoimmune disease In the group without AID (n=13) the median age was 63 years with median age at diagnosis of malignancy of 58 years. Rheumatic irAEs included seven patients with inflammatory arthritis, three with PMR-like syndrome, five with sicca syndrome and one with myositis. The majority of patients had more than one irAE (table 2) with patient 10 experiencing irAEs involving five different systems. Patients with arthritis exhibited different clinical phenotypes with the majority having a combination of small and large joint involvement (table 3). Regarding the patients with a PMR-like phenotype, all three had clinical features consistent with PMR including pain and stiffness involving the shoulders, hips/lower extremities and neck with associated severe morning stiffness. Two of the three had elevated inflammatory markers, and the third had normal inflammatory markers but levels had not been checked prior to initiation of prednisone to treat these symptoms. None had symptoms concerning for giant cell arteritis. Four of the five patients with sicca syndrome were xerostomic without keratoconjunctivitis. ANA was positive in two of the five, and SSA was positive in one (table 1). One sicca patient had a Schirmer's test performed, which was negative, and none underwent minor salivary gland biopsy. The myositis patient presented with proximal muscle weakness with diaphragmatic dysfunction, diplopia and dysphagia all attributable to myositis. Testing confirmed respiratory neuromuscular dysfunction and absence of primary oesophageal peristalsis. Electromyogram was consistent with severe inflammatory/necrotising myopathy. Imaging evaluation in our cohort was limited, but patient 11 did have a shoulder MRI to evaluate PMR symptoms which revealed extensive rotator cuff tendinosis and bursitis. Table 2 Non-rheumatic immune-related adverse events (irAEs) Patient Non-rheumatic irAE 2 Hypothyroid 4 Colitis 5 Rash Hypothyroid Colitis 8 Hypophysitis Thyroiditis 9 Hypophysitis 10 Hypophysitis Pneumonitis Neuropathy 11 Hypophysitis 12 Colitis Table 3 Clinical phenotypes of inflammatory arthritis Patient Joint pattern Symmetrical Tenosynovitis 1 PIPs, MCPs, wrists, elbows, knees Yes 2 Generalised involvement of small hand joints Yes Yes 3 PIPs, MCPs, PIPs, elbows, knees, ankles, feet, toes Yes 4 PIPs, MCPs, ankles, knees Yes 5 PIPs, MCPs, wrists, knees Yes 6 Generalised involvement of small hand joints, wrists Yes 7 Generalised involvement of small hand joints, left knee No Yes MCP, metacarpal phalangeal joints; PIP, proximal interphalangeal joint. With the exception of two patients who experienced irAEs over 1 year after starting immunotherapy, the median time to onset of irAE was 7.3 weeks (range 2–48.4). The longest time between start of immunotherapy and development of irAE was 213 weeks in a patient who developed a PMR-like syndrome after over 4 years on nivolumab for renal cell carcinoma. Rheumatic irAEs led to holding of immunotherapy in eight patients and immune-related hypophysitis led to cessation of therapy in two of the remaining patients. Autoimmune testing was performed in the majority of patients (table 1). Four patients had a positive ANA with one also having anti-double stranded DNA antibodies and another anti-SSA antibodies. One patient with inflammatory arthritis was positive for rheumatoid factor (RF) but negative for anti-cyclic citrullinated protein antibodies. Eleven patients were treated with glucocorticoids for their rheumatic irAE; five required additional therapy with either anti-tumour necrosis factor (TNF) α medications, intravenous immunoglobulin or hydroxychloroquine. Treatment of irAEs led to significant improvement in six patients, moderate improvement in five patients and only minimal improvement in two. Non-rheumatic irAEs (table 2) were addressed per guidelines on an individualised basis. Patients with pre-existing autoimmune disease Of the two patients with pre-established AID, one experienced a disease flare after starting immunotherapy: a patient with psoriatic arthritis previously treated with apremilast had experienced remission of his psoriatic arthritis (PsA) while on chemotherapy. He experienced a psoriasis flare 2.8 weeks after starting nivolumab and apremilast was resumed. He experienced mild psoriasis flares on his face during therapy but inflammatory arthritis remained quiet. The patient with rheumatoid arthritis had seropositive, non-erosive disease; he remained without disease activity on hydroxychloroquine throughout his course of immunotherapy. Discussion irAEs of any type are common in patients receiving immunotherapy, occurring in up to 90% of patients receiving anti-CTLA-4 agents4 and 70% of those receiving anti-PD-1/PD-L1 agents.5 6 Somewhat lower rates of irAEs have been seen with anti-PD-1 agents, but incidence increases when they are used in combination with inhibitors of CTLA-4.7 The most common irAEs are gastrointestinal, hepatic, endocrine and dermatological events.7 8 Rheumatic irAEs, however, have been infrequently reported in clinical trials9 and generally have been the subject of isolated case reports.9 One single-centre case series of rheumatic irAEs was recently published, in which Cappelli et al2 described inflammatory arthritis and sicca syndrome in 13 patients receiving nivolumab and/or ipilimumab.2 In this report, they described nine patients who developed inflammatory arthritis; synovitis was confirmed by imaging in four. Overall, a variety of rheumatic manifestations have been described in these reports including arthralgia, inflammatory arthritis, sicca complex as well as rare reports of vasculitis, myositis and lupus. Our series of 15 patients now expands the description of rheumatic irAEs and describes a wider variety of rheumatic irAEs than seen in previous reports, including inflammatory arthritis, sicca syndrome, a PMR-like syndrome and myositis in patients receiving monotherapy and combination therapy with checkpoint inhibitors. In our study, in 13 of 15 subjects the median time to onset of rheumatic irAE was 7.3 weeks after initiation of immunotherapy, which is consistent with previous reports. In clinical trials for ipilimumab, the majority of all irAEs have been reported to occur within 12 weeks of initial dosing, and this seems to be consistent across numerous studies.4 7 Interestingly there were two outliers in our cohort with one patient developing a PMR-like syndrome after over 4 years of being stable on nivolumab. Reasons for the delay in irAE presentation are unknown. Several patients were noted to have persistent symptoms for months after immunotherapy was stopped. The longest persistence of symptoms in one patient was over 2 years after the last dose of immunotherapy. This persistence of rheumatic symptoms long after discontinuation of immunotherapy has also been previously described.2 The pathophysiology of each respective irAE remains to be fully elucidated albeit it seems that they are at least partially T cell-mediated as expected by the mechanism of action of these medications.10 To date, there have been no detailed pathophysiological studies investigating the mechanism of any rheumatic irAE. Management of rheumatic irAEs remains an area of uncertainty. In general, irAEs have been reported to be steroid sensitive and in most cases resolve within 6–12 weeks.8 While some patients in our series were responsive to glucocorticoids, three out of 13 required more aggressive immunosuppression with TNF inhibitors. We also observed that these patients required higher doses of glucocorticoids and largely more intense therapy than our traditional experience with inflammatory arthritis seen in rheumatoid arthritis or associated with connective tissue disease. An important clinical question is whether patients with pre-existing rheumatic diseases undergoing cancer immunotherapy with checkpoint inhibitors are at increased risk for flares of their underlying diseases. The largest series to date describes 30 patients with pre-existing AIDs of a variety of sorts; eight experienced flares 3 days to 7 months after starting ipilimumab.11 In our study, one of two patients experienced a flare of underlying AID. The patient had psoriatic arthritis and the flare was managed with apremilast without interruption of immunotherapy. In our entire cohort, treatment discontinuation rate owing to irAEs was significantly higher (10 of 13 patients) compared with the literature. Immunotherapy was discontinued for rheumatic irAEs in eight of the 10, and for endocrine irAE in two patients. This needs to be validated in larger prospective cohort trials. A comprehensive grading system for rheumatic irAEs is lacking as opposed to other more common irAEs. Additionally, despite a standardised grading system, toxicity evaluation and treatment for all irAEs is highly subjective based on the treating oncologist.9 Incorporating rheumatic irAEs in the grading system of adverse effects of immunotherapy would facilitate early recognition, referral and expeditious management which may translate to decreased need for immunosuppressive medications and/or cessation of immunotherapy. It would also allow a more accurate assessment of the incidence and prevalence of these complications. New-onset inflammatory symptoms pertaining to the broad spectrum of rheumatological conditions postadministration of immune checkpoint inhibitors should be further investigated. There are several limitations of our study. While our series is the largest to date, it is a single-centre experience and retrospective in nature. Accordingly diagnostic testing, including laboratory tests, imaging and organ-specific testing, was not standardised and thus subject to diagnostic bias. For those patients who did have positive tests (ie, ANA, RF, etc), there was no testing prior to immunotherapy for comparison. There was limited use of imaging and no synovial fluid analysis to confirm inflammatory arthritis. All rheumatic irAE diagnoses were based on expert opinion of the rheumatologist. We also were unable to determine the true incidence of rheumatic irAEs given the observational basis of this cohort. At present, it is clear that we are in the early stages of diagnosing and treating rheumatic irAEs secondary to immunotherapy of cancers with checkpoint inhibitors. Numerous questions are unanswered regarding these complications. These questions include: what is the true incidence of these disorders with individual and combined therapies? What are the risk factors and what is the underlying pathophysiology of these disorders? What are the risks of disease flare with these types of immunotherapy in patients with pre-existing autoimmune conditions? In terms of therapy, what is the optimal treatment of these types of complications and finally what are the ramifications of concomitant immunosuppressive therapy for irAEs on antitumorous responses? Given the proliferation of checkpoint inhibitor therapy into the general oncology armamentarium and the rise of new therapeutics with similar mechanisms of action, it is assured that rheumatic adverse events will be seen by general rheumatologists in academic and community settings. Rheumatologists must be alerted to these complications and acquire knowledge to accurately diagnose and manage these disorders in collaboration with treating oncologists. Twitter: Follow cassandra calabrese @CCalabreseDO Contributors: All authors contributed to data acquisition, analysis and manuscript preparation. Competing interests: LHC reports personal fees from Bristol-Myers Squibb, outside the submitted work. VV reports grants and personal fees from Bristol-Myers Squibb, grants and personal fees from Genentech, grants and personal fees from Merck, grants and personal fees from Astra Zeneca, personal fees from Celgene, grants and personal fees from Genoptix, personal fees from Foundation medicine, outside the submitted work. CC, KK and EK have nothing to disclose. Ethics approval: Cleveland Clinic Foundation IRB. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Topalian SL , Drake CG , Pardoll DM Immune checkpoint blockade: a common denominator approach to cancer therapy . Cancer Cell 2015 ;27 :450 –61 . 10.1016/j.ccell.2015.03.001 25858804 2 Cappelli LC , Gutierrez AK , Baer AN Inflammatory arthritis and sicca syndrome induced by nivolumab and ipilimumab . Ann Rheum Dis 2017 ;76 :43 –50 . 10.1136/annrheumdis-2016-209595 27307501 3 Donia M , Pedersen M , Svane IM Cancer immunotherapy in patients with preexisting autoimmune disorders . Semin Immunopathol 2016 Published Online First 11 October 2016 . 10.1007/s00281-016-0595-8 4 Hodi FS , O'Day SJ , McDermott DF Improved survival with ipilimumab in patients with metastatic melanoma . N Engl J Med 2010 ;363 :711 –23 . 10.1056/NEJMoa1003466 20525992 5 Topalian SL , Hodi FS , Brahmer JR Safety, activity, and immune correlates of anti-PD-1 antibody in cancer . N Engl J Med 2012 ;366 :2443 –54 . 10.1056/NEJMoa1200690 22658127 6 Brahmer JR , Tykodi SS , Chow LQM Safety and activity of anti-PD-L1 antibody in patients with advanced cancer . N Engl J Med 2012 ;366 :2455 –65 . 10.1056/NEJMoa1200694 22658128 7 Boutros C , Tarhini A , Routier E Safety profiles of anti-CTLA-4 and anti-PD-1 antibodies alone and in combination . Nat Rev Clin Oncol 2016 ;13 :473 –86 . 10.1038/nrclinonc.2016.58 27141885 8 Michot JM , Bigenwald C , Champiat S Immune-related adverse e vents with immune checkpoint blockade: a comprehensive review . Eur J Cancer 2016 ;54 :139 –48 . 10.1016/j.ejca.2015.11.016 26765102 9 Horvat TZ , Adel NG , Dang TO Immune-related adverse events, needs for systemic immunosuppression, and effects on survival and time to treatment failure in patients with melanoma treated with ipilimumab at Memorial Sloan Kettering Cancer Center . J Clin Onc 2015 ;33 :3193 –8 . 10 Johnson DB , Balko JM , Compton ML Fulminant myocarditis with combination immune checkpoint blockade . N Engl J Med 2016 ;375 :1749 –55 . 10.1056/NEJMoa1609214 27806233 11 Johnson DB , Sullivan RJ , Ott PA Ipilimumab therapy in patients with advanced melanoma and preexisting autoimmune disorders . JAMA Oncol 2016 ;2 :234 –40 . 10.1001/jamaoncol.2015.4368 26633184

PMC005xxxxxx/PMC5372132.txt

==== Front BMJ Open Diabetes Res CareBMJ Open Diabetes Res CarebmjdrcbmjdrcBMJ Open Diabetes Research & Care2052-4897BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjdrc-2016-00031710.1136/bmjdrc-2016-000317Perspectives in Diabetes15061612Diabetes incidence and glucose intolerance prevalence increase with higher outdoor temperature Blauw Lisanne L 12Aziz N Ahmad 3Tannemaat Martijn R 3Blauw C Alexander 4de Craen Anton J 5Pijl Hanno 1Rensen Patrick C N 16 1 Division of Endocrinology, Department of Medicine, Leiden University Medical Center, Leiden, The Netherlands 2 Department of Epidemiology, Leiden University Medical Center, Leiden, The Netherlands 3 Department of Neurology, Leiden University Medical Center, Leiden, The Netherlands 4 Delft University of Technology, Delft, The Netherlands 5 Department of Gerontology and Geriatrics, Leiden University Medical Center, Leiden, The Netherlands 6 Einthoven Laboratory for Experimental Vascular Medicine, Leiden University Medical Center, Leiden, The NetherlandsCorrespondence to Dr Lisanne L Blauw; l.l.blauw@lumc.nlLLB, NAA, HP and PCNR contributed equally. In memoriam (17 January 2016) Dr Anton J de Craen, Department of Gerontology and Geriatrics, Leiden University Medical Center, Leiden, The Netherlands 2017 20 2 2017 5 1 e00031726 8 2016 16 12 2016 3 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective Rising global temperatures might contribute to the current worldwide diabetes epidemic, as higher ambient temperature can negatively impact glucose metabolism via a reduction in brown adipose tissue activity. Therefore, we examined the association between outdoor temperature and diabetes incidence in the USA as well as the prevalence of glucose intolerance worldwide. Research design and methods Using meta-regression, we determined the association between mean annual temperature and diabetes incidence during 1996–2009 for each US state separately. Subsequently, results were pooled in a meta-analysis. On a global scale, we performed a meta-regression analysis to assess the association between mean annual temperature and the prevalence of glucose intolerance. Results We demonstrated that, on average, per 1°C increase in temperature, age-adjusted diabetes incidence increased with 0.314 (95% CI 0.194 to 0.434) per 1000. Similarly, the worldwide prevalence of glucose intolerance increased by 0.170% (95% CI 0.107% to 0.234%) per 1°C rise in temperature. These associations persisted after adjustment for obesity. Conclusions Our findings indicate that the diabetes incidence rate in the USA and prevalence of glucose intolerance worldwide increase with higher outdoor temperature. Type 2 DiabetesEpidemiologyEnvironmentBrown Adipose Tissuespecial-featurepress-release ==== Body Significance of this study What is already known about this subject? The prevalence of type 2 diabetes is increasing rapidly worldwide. Interestingly, it was recently shown that acclimatization of patients with type 2 diabetes to moderate cold for only 10 days already improves insulin sensitivity. Physiologically, cold exposure activates brown adipose tissue (BAT) that has been identified to combust large amounts of lipids to generate heat. Previously, it has been shown that BAT activity is negatively associated with outdoor temperature. It is thus conceivable that an increased flux of fatty acids toward BAT may result in a compensatory increased flux of glucose to other metabolically active tissues, explaining improved insulin sensitivity at lower temperature. What are the new findings? On the basis of the putative role of BAT in the control of insulin action and the effect of ambient temperature on BAT activity, we hypothesized that the global increase in temperature contributes to the current type 2 diabetes epidemic. Therefore, in this study, we aimed to assess the association of outdoor temperature with diabetes incidence and prevalence of glucose intolerance, on a countrywide as well as a global scale. Here we show that the diabetes incidence rate in the USA and the prevalence of glucose intolerance worldwide increase with higher outdoor temperature. How might these results change the focus of research or clinical practice? On the basis of our results, a 1°C rise in environmental temperature would account for over 100 000 new diabetes cases per year in the USA alone, given a population of nearly 322 million people in 2015. This emphasizes the importance of future research into the effects of environmental temperature on glucose metabolism and the onset of diabetes, especially in view of the global rise in temperatures with a new record set for the warmest winter in the USA last year. Introduction The prevalence of type 2 diabetes is increasing rapidly worldwide. In 2015, 415 million adults globally were suffering from diabetes, and expectations are that the prevalence will rise by almost 55%, up to 642 million cases by 2040.1 In high-income countries, 91% of adults affected by diabetes have type 2 diabetes.1 The type 2 diabetes epidemic accompanies the increasing prevalence of obesity.2 With increasing body mass index (BMI), glucose and lipids are initially stored in expanding (subcutaneous) adipose tissue compartments. However, according to the lipid overflow hypothesis, when the storage capacity of adipose tissue is exceeded, lipids can accumulate in organs (steatosis) including the pancreas, liver, heart and skeletal muscle, resulting in insulin resistance of those organs.3 The variability in the degree of steatosis and the heterogeneity of body fat distribution over subcutaneous and visceral fat depots probably explains the only modest association between measures of overall body fat and insulin resistance.3–5 For example, South Asians are prone to develop type 2 diabetes at a relatively low BMI, presumably because of the limited lipid storage capacity of their adipose tissue depots.6 Interestingly, a very low-calorie diet can rapidly diminish steatosis and insulin resistance independent of weight loss7 indicating dissociation between insulin resistance and obesity perse. Recently, brown adipose tissue (BAT) has emerged as an organ that is capable of combusting large amounts of lipids to generate heat.8 Physiologically, BAT is activated by cold exposure.9 10 Indeed, prolonged cold acclimatization recruits BAT activity10 and is able to induce modest weight loss.11 A recent landmark paper showed that acclimatization of patients with type 2 diabetes to moderate cold for only 10 days already improved insulin sensitivity as determined by a markedly higher glucose infusion rate during a hyperinsulinemic–euglycemic clamp, while body weight was unaltered.12 It is conceivable that an increased flux of fatty acids toward BAT will result in a compensatory increased flux of glucose to other metabolically active tissues, thereby improving systemic insulin sensitivity.13 It has previously been shown that BAT activity is negatively associated with outdoor temperature and is highest in winter.14–16 Considering the putative role of BAT in the control of insulin action, combined with the effect of ambient temperature on BAT activity, we hypothesized that the global increase in temperature contributes to the current type 2 diabetes epidemic. Recently, a positive association was found between outdoor temperature and glycated hemoglobin (HbA1c),17 indicating that systemic glucose homeostasis is influenced by environmental temperature. However, the importance of this association, especially in relation to the increasing diabetes burden, has never been studied on the population level. Therefore, in the present study, we aimed to assess the association between outdoor temperature and glucose metabolism on a countrywide as well as a global scale. We specifically hypothesized that diabetes incidence and prevalence of glucose intolerance increase with rising outdoor temperatures. Research design and methods Diabetes incidence rate in the USA Data on diabetes incidence in 50 states of the USA and 3 territories (ie, Guam, Puerto Rico and Virgin Islands) were available for the years 1996–2013. Via the National Diabetes Surveillance System of the Centers for Disease Control and Prevention (CDC), age-adjusted diabetes incidence rates per 1000, along with the corresponding 95% CI, were available for all states and territories per year.18 The diabetes incidence rate was self-reported via the Behavioral Risk Factor Surveillance System (BRFSS) and defined as the rate of adults (18–76 years) who reported to be told by a health professional in the last year they had diabetes (type 1 or 2). Data were age-adjusted to the 2000 US standard population using age strata of 18–44, 45–64 and 65–76.7 Detailed information on survey methods can be found elsewhere.19 We restricted our analysis to the 1996–2009 period as major changes in the survey methods were implemented between 2010 and 2011. Data on the mean annual temperature per state were collected through the National Centers for Environmental Information.20 Obesity prevalence (ie, BMI≥30) per state or territory per year, expressed as a percentage with a corresponding 95% CI, was also obtained via the BRFSS of the CDC.19 21 Prevalence of raised fasting blood glucose worldwide Data on country-wise prevalence rates of raised fasting blood glucose and obesity (ie, BMI≥30) were available and obtained through the WHO's Global Health Observatory online data repository system, which provides various health-related statistics for its member states.22 In this database, the prevalence of raised fasting blood glucose is defined as the percentage of the population with fasting blood glucose ≥7.0 mmol/L or on medication for raised blood glucose.23 We used the most recent age-adjusted and sex-adjusted estimates available which were based on data from 2014. Country-wise mean annual temperature data for 2014 were obtained via the Climatic Research Unit, University of East Anglia, using the 3.23 release of its data set.24 25 Statistical analyses First, for each state or territory of the USA separately, we performed a weighted meta-regression analysis to estimate the association between mean annual temperatures (°C) and age-adjusted diabetes incidence rates, both measured yearly during the period 1996–2009. Age-adjusted diabetes incidence rate was included as the dependent variable and mean annual temperature was included as the independent variable. Moreover, in order to adjust for the underlying secular trends in temperature and diabetes incidence, all meta-regression models were adjusted for the effect of time passage by including the variable year as an additional independent variable in the models. For Guam, Illinois and the Virgin Islands, there were insufficient data to perform a meta-regression analysis, and therefore these were excluded from further analyses. Second, a meta-analysis was performed to integrate the results of the meta-regression analyses into an overall effect estimate, representing the mean strength of the association between mean annual temperature and diabetes incidence rate in the USA during the period 1996–2009. For the worldwide data, we performed a meta-regression analysis with age-adjusted and sex-adjusted country-wise prevalence of raised fasting blood glucose as the dependent variable and mean annual temperature as well as the World Bank income group (ie, low, lower middle, upper middle and high income)22 as the independent variables. For meta-regression analysis and meta-analysis, the DerSimonian and Laird random-effects model was used.26 27 Weights were calculated as with being the SE of the age-adjusted incidence rate in year i and the estimated between-study variance.28 SEs of the age-adjusted incidence rates were unknown and therefore estimated as with UL the upper limit and LL the lower limit of the 95% CI of the age-adjusted incidence rate. To establish whether the association between mean annual temperature and diabetes incidence rate was mediated through effects on body weight, we first calculated an overall effect estimate for the association between mean annual temperature and obesity prevalence in the USA using a similar approach as described above. Since this overall effect size significantly differed from zero, we added obesity prevalence as an independent variable in the meta-regression models relating mean annual temperature and diabetes incidence rate. Subsequently, we again performed a meta-analysis to integrate the estimates of the obesity prevalence corrected meta-regression analyses into an overall effect estimate for the association between mean annual temperature and diabetes incidence rate adjusted for obesity prevalence in the USA during the period 1996–2009. For the global data, we performed a similar analysis by considering obesity prevalence as a mediator variable in the meta-regression models. We also composed a map of the USA showing the mean annual temperature over the period 1996–2009 and β coefficients from the obesity-adjusted meta-regression analyses per state. This map provides insight into potential geographically determined confounding factors (eg, sea influences, biomes or altitudes), which could be apparent from a non-random distribution of β coefficients overlapping with geographical patterns. Programming was performed in STATA Statistical Software V.12.0 (Statacorp, College Station, Texas, USA) and R V.3.1.0 using the ‘metafor’ package.29 Maps were computed using MATLAB V.R2015a (The MathWorks, Natick, Massachusetts, USA). Results National level: USA Table 1 shows data on annual temperature, diabetes incidence rate and obesity prevalence in the period 1996–2009 for each included US state or territory. During these 14 years, the age-adjusted diabetes incidence rate was highest in Puerto Rico, West Virginia and South Carolina and lowest in Minnesota, Massachusetts and Colorado. The mean change (Δ) in annual temperature between two consecutive years ranged from −0.11°C (Minnesota) to +0.09°C (Hawaii). Table 1 Temperature characteristics, mean diabetes incidence, and mean obesity prevalence in the USA* over the period 1996–2009 Mean annual temperature (°C) Δ Temperature per year (°C)† Age-adjusted diabetes incidence rate per 1000 Obesity prevalence (%) Alabama 17.46 −0.03 9.7 26.3 Alaska −2.48 −0.07 6.5 23.5 Arizona 16.13 0.03 7.7 19.3 Arkansas 16.11 −0.02 8.6 24.6 California 14.85 0.03 8.6 20.9 Colorado 7.86 0.03 5.2 15.9 Connecticut 9.78 −0.02 6.5 18.3 Delaware 13.27 −0.06 8.2 22.1 Florida 21.72 −0.05 8.4 20.8 Georgia 17.74 −0.03 9.2 23.0 Hawaii 23.32 0.09 6.4 17.9 Idaho 6.54 0.01 7.3 21.1 Indiana 11.26 −0.05 8.3 24.4 Iowa 9.12 −0.07 6.6 23.4 Kansas 12.81 −0.02 7.2 22.2 Kentucky 13.46 −0.04 8.1 25.3 Louisiana 19.48 −0.03 9.5 25.7 Maine 5.37 0.00 7.9 21.1 Maryland 12.87 −0.06 8.1 22.1 Massachusetts 9.16 −0.02 5.1 17.8 Michigan 7.37 −0.05 8.5 24.9 Minnesota 5.48 −0.11 4.6 20.9 Mississippi 17.82 −0.03 9.1 27.7 Missouri 12.88 −0.04 7.5 24.3 Montana 5.91 −0.08 5.6 18.9 Nebraska 9.76 −0.04 6.1 22.8 Nevada 10.50 0.05 7.7 20.3 New Hampshire 6.59 0.00 6.4 19.9 New Jersey 11.84 −0.05 7.0 19.8 New Mexico 12.42 0.02 7.6 20.2 New York 7.73 −0.02 8.2 20.4 North Carolina 15.17 −0.05 8.8 23.9 North Dakota 5.04 −0.09 5.9 23.1 Ohio 10.84 −0.04 8.4 23.9 Oklahoma 15.71 −0.02 8.9 23.9 Oregon 8.75 0.02 6.7 21.7 Pennsylvania 9.61 −0.04 7.5 23.2 Puerto Rico 26.94 −0.07 10.9 22.7 Rhode Island 10.20 −0.03 7.6 18.8 South Carolina 17.28 −0.04 10.2 24.6 South Dakota 7.71 −0.08 5.9 22.3 Tennessee 14.63 −0.04 9.6 24.9 Texas 18.74 −0.01 9.0 24.4 Utah 9.55 0.00 6.4 19.4 Vermont 6.07 −0.01 5.9 19.0 Virginia 13.20 −0.05 7.7 21.8 Washington 8.36 −0.03 8.1 21.2 West Virginia 11.34 −0.04 10.7 26.8 Wisconsin 6.68 −0.08 7.0 22.3 Wyoming 5.70 0.00 6.9 20.2 Results are presented as mean values over the period 1996–2009. *Forty-nine states and one territory (Puerto Rico): insufficient data were available for Guam, Illinois and the Virgin Islands. †Mean change (Δ) over the period 1996–2009 in mean annual temperature between two consecutive years. Both diabetes incidence rate (figure 1) and obesity prevalence (see online supplementary figure S1) increased with higher mean annual temperature. The overall effect estimate for the association between mean annual temperature and diabetes incidence rate was 0.314 (95% CI 0.194 to 0.434), indicating a 0.314 increase in the diabetes incidence rate per 1000 for each degree Celsius increase in mean annual temperature. Similarly, obesity prevalence increased by 0.173% (95% CI 0.050% to 0.296%) for each degree Celsius increase in mean annual temperature. Figure 1 The association between mean annual temperature and diabetes incidence in the USA over the period 1996–2009. The forest plot represents the differencea,b in diabetes incidence rate per 1°C increase in temperature.a The β coefficient from meta-regression analysis. Error bars represent a 95% CI.b Adjusted for the effect of time passage. Diabetes incidence rate is the age-adjusted diabetes incidence rate per 1000. 10.1136/bmjdrc-2016-000317.supp1supplementary figures After adjustment for obesity prevalence, the positive association between mean annual temperature and diabetes incidence rate attenuated only slightly (see online supplementary figure S2). Per degree Celsius increase in mean annual temperature, the obesity-adjusted diabetes incidence rate increased by 0.290 (95% CI 0.164 to 0.416) per 1000. Figure 2 shows a map of the USA, illustrating the relation between the mean annual temperature per state over the period 1996–2009 and the magnitude of the corresponding change in obesity-adjusted diabetes incidence rate. As can be appreciated from this figure, the diabetes incidence rate increases in most states with higher mean annual temperature, while there is no apparent geographical pattern which could explain this association. Figure 2 Map of the USA (including Alaska, Hawaii and Puerto Rico), showing the mean annual temperature and the magnitude of the β coefficientsa from the obesity-adjusted meta-regression analysis,b per state or territory over the period 1996–2009.a The β coefficient from meta-regression analysis, representing the difference in diabetes incidence rate per 1°C increase in temperature; red circles indicate a positive β coefficient, while blue circles indicate a negative β coefficient.b Adjusted for the effect of time passage. Diabetes incidence rate is the age-adjusted diabetes incidence rate per 1000. Global level Prevalence data on raised fasting blood glucose and obesity were available for 190 countries. The worldwide aggregated age-adjusted and sex-adjusted prevalences of raised fasting blood glucose levels and obesity were 9.65% (95% CI 9.11% to 10.13%) and 19.51% (95% CI 17.94% to 21.07%), respectively. There was a positive association between mean annual temperature and country-wise age-adjusted, sex-adjusted and income-adjusted prevalence of raised fasting blood glucose. The prevalence increased by 0.170% for each degree Celsius rise in temperature (95% CI 0.107% to 0.234%). There was also a positive association between mean annual temperature and country-wise age-adjusted, sex-adjusted and income-adjusted obesity prevalence. For each degree Celsius rise in temperature, the obesity prevalence increased by 0.295% (95% CI 0.137% to 0.454%). Interestingly, the increase in obesity prevalence for each degree Celsius rise in average temperature was larger in women (0.503%) as compared with men (0.111%; mean difference 0.391%, 95% CI 0.160% to 0.622%). After additional adjustment for obesity prevalence, the association between mean annual temperature and country-wise age-adjusted, sex-adjusted and income-adjusted prevalence of raised fasting glucose slightly attenuated (0.106% increase for each degree Celsius rise in temperature, 95% CI 0.057% to 0.155%; figure 3). Figure 3 The worldwide association between mean annual temperature and age-adjusted, sex-adjusted, income-adjusted and obesity-adjusted prevalence of raised fasting blood glucose for 190 countries in 2014. Colors indicate the six WHO regions.40 The size of each circle is inversely proportional to the SE of the estimate of the prevalence of raised FBG. FBG, fasting blood glucose. Given the extremely high prevalence of raised fasting glucose in some relatively small countries (figure 3), we performed a sensitivity analysis excluding all countries (n=29) with a prevalence of raised fasting blood glucose higher than 15%. Under this restrictive scenario, the association between mean annual temperature and country-wise age-adjusted, sex-adjusted, income-adjusted and obesity-adjusted prevalence of raised fasting blood glucose remained present: 0.051% increase in prevalence for each degree Celsius rise in temperature (95% CI 0.003% to 0.099%; see online supplementary figure S3). Conclusions The association between outdoor temperature and fasting blood glucose concentration has rarely been examined in the past, and studies showed discordant results.30 31 These studies were performed in selected study populations from specific cities and the conflicting results indicate that the results are not generalizable to other places. To the best of our knowledge, this study is the first to assess the association of outdoor temperature with diabetes incidence and the prevalence of raised fasting blood glucose on a national and global level. Using 14-year longitudinal state-level data from the USA, we show that the overall diabetes incidence rate is higher in warmer years. Per 1°C increase in temperature, we found an overall increase in age-adjusted diabetes incidence of 0.314 per 1000. Importantly, we expect confounding in this association to be minimal, as temperature itself is hardly influenced by any extraneous factors, apart from the passage of time and geographical location. Hence, we adjusted the models for the underlying secular trends in temperature and diabetes incidence by adjusting for time passage. Also, by first determining the association for each state separately before pooling for obtaining the overall effect estimate, we minimized the effect of potential geographically determined confounding factors, and in this way we also precluded all other differences between states that could influence or bias the association between temperature and diabetes incidence. Furthermore, the distribution of the magnitudes and directions of the β coefficients over the USA (figure 2) appeared to be random as we could not distinguish a pattern overlapping with geographic (eg, sea influences, biomes or altitudes) or demographic characteristics, indicating that these factors do not explain the association. Our results are therefore unlikely to be merely due to interstate demographic, socioeconomic or other differences. On the global scale, detailed longitudinal international data on diabetes incidence rates were unfortunately unavailable for a large number of countries. Therefore, we used country-wise estimates of the prevalence of raised fasting blood glucose instead. In accordance with the results of the USA, we found that on a global level the prevalence of glucose intolerance is higher in warmer countries. This association between temperature and raised fasting blood glucose cannot be merely due to international differences in age, sex, income or obesity prevalence, as our analyses were adjusted for these variables. Although we cannot exclude residual confounding by population stratification, the fact that this association closely parallels our findings in the USA supports the notion that the ambient temperature affects the occurrence of glucose intolerance worldwide. The associative design of our study does not allow us to draw conclusions on causality. However, it is tempting to speculate that the mechanism underlying our present findings is related to an interplay between BAT activity and glucose clearance from the circulation by metabolically active tissues. Recently, peripheral insulin sensitivity in patients with type 2 diabetes was shown to increase markedly after 10 days of mild cold exposure.12 These patients had an increased glucose uptake by skeletal muscle, probably explained by an increased overall flux of fatty acids toward BAT, reducing the fatty acid flux to other tissues.13 Interestingly, in this short-term cold acclimation study, body weight was unaffected,12 indicating that the effects of temperature on glucose metabolism can occur independently of body weight. This is further supported by another study showing that higher BAT activity after cold exposure was associated with lower blood glucose and HbA1c levels independent of body fat.32 In line with these findings, we found that adjustment for obesity prevalence only slightly attenuated the association of outdoor temperature with both diabetes incidence in the USA and prevalence of glucose intolerance worldwide. Therefore, our data are consistent with the hypothesis that a decrease in BAT activity with increasing environmental temperature may deteriorate glucose metabolism and increase the incidence of diabetes. The role of BAT activity as an underlying pathway in the association between outdoor temperature and diabetes incidence may even be genetic in origin. According to the cold climate genes hypotheses, DNA areas involved in thermogenesis may have encountered selection pressure during evolution in areas with cold climates.33 The biological effects of the gene products may also decrease susceptibility to diabetes. Interestingly, DNA areas that regulate uncoupling proteins, including uncoupling protein 1 that drives heat production in BAT, have been proposed as candidates for these cold climate genes.33 Besides BAT activity, physical activity could be a potential mediator in the association between temperature and diabetes incidence, as physical activity varies with environmental temperature.34 Although a decrease in physical activity with increasing mean annual temperature may theoretically explain our findings, physical activity has been shown to be highest between daily temperatures of 15°C and 20°C and decreases with higher but also lower temperatures.35 36 Therefore, we do not expect that physical activity completely explains the positive association between outdoor temperature and diabetes incidence. Unfortunately, data on physical activity for the USA were not available for the period 1996–2009, and therefore we were not able to further elucidate the mediating effects. This topic evidently requires further investigation. There are several potential limitations to our study. First, we used state-level and country-level aggregated data on incidence and prevalence rates. Although individual-level data would have yielded more accurate results, acquiring individual-level data on such a large scale was not feasible. Moreover, we used statistical techniques, that is, meta-regression and meta-analysis, which can appropriately account for the uncertainty in the aggregated effect estimates. Second, available data on BMI were categorically organized (ie, obesity prevalence) and BMI could therefore not be applied as a continuous variable in the models. Although continuous data would have been more accurate, the prevalence of obesity is by definition closely related to the average BMI and its distribution in the general population, implying that obesity prevalence can be used as a valid surrogate measure for average BMI in a given population. It should be emphasized that we were not able to evaluate the linearity of the association between BMI and diabetes incidence or prevalence in our models, as continuous data on BMI were not available. Although adjustment for obesity prevalence only slightly attenuated the association between outdoor temperature and diabetes, mediation via BMI should be further evaluated in future studies to assess whether a non-linear association between BMI and diabetes might account for our results. Third, owing to the observational nature of our analyses, we cannot exclude residual confounding. For the USA, however, we largely circumvented this issue by analyzing longitudinal state-level data for each state separately before pooling the results. Furthermore, adjustment for important intercountry differences, that is, age, sex, income and obesity, did not materially change the results on the global level. Importantly, adjusting for income also takes along other intercountry differences, as many surrogate measures which are often applied for adjusting for global inequalities in the degree of social development are based on the gross national income (GNI) per capita. In fact, the WHO uses a four-group classification scheme, based on the Work Bank Atlas methodology to estimate GNI per capita,37 according to which each country is categorized as low, lower middle, upper middle or high income.22 In conclusion, diabetes incidence rate in the USA and prevalence of glucose intolerance worldwide increase with higher outdoor temperature. On the basis of our results, a 1°C rise in environmental temperature would account for over 100 000 new diabetes cases per year in the USA alone, given a population of nearly 322 million people in 2015.38 These findings emphasize the importance of future research into the effects of environmental temperature on glucose metabolism and the onset of diabetes, especially in view of the global rise in temperatures with a new record set for the warmest winter in the USA last year.39 Contributors: LLB and NAA were involved in study concept and design, acquisition of data, analysis and interpretation of data, statistical analysis, and drafting of the manuscript. MRT was involved in study concept and design, analysis and interpretation of data, and critical revision of the manuscript. CAB was involved in MATLAB programming, analysis and interpretation of data, and critical revision of the manuscript. AJdC was involved in study concept and design, analysis and interpretation of data, and statistical analysis. HP and PCNR were involved in study concept and design, analysis and interpretation of data, critical revision of the manuscript, and study supervision. Funding: LLB is supported by an excellence grant from the Board of Directors of the Leiden University Medical Center. NAA is supported by a VENI grant (#91615080) from the Netherlands Organization of Scientific Research. PCNR is an established investigator of the Dutch Heart Foundation (2009T038). Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 International Diabetes Federation . IDF Diabetes Atlas . 7th edn Brussels, Belgium : International Diabetes Federation , 2015 http://www.diabetesatlas.org (accessed May 2016 ). 2 American College of Cardiology/American Heart Association Task Force on Practice Guidelines, Obesity Expert Panel . Executive summary: guidelines (2013) for the management of overweight and obesity in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the Obesity Society published by the Obesity Society and American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Based on a systematic review from the the Obesity Expert Panel, 2013 . Obesity (Silver Spring) 2014 ;22 (Suppl 2) :S5 –39 . 10.1002/oby.20821 24961825 3 Tchernof A , Despres JP Pathophysiology of human visceral obesity: an update . Physiol Rev 2013 ;93 :359 –404 . 10.1152/physrev.00033.2011 23303913 4 Ferrannini E , Natali A , Bell P Insulin resistance and hypersecretion in obesity. European Group for the Study of Insulin Resistance (EGIR) . J Clin Invest 1997 ;100 :1166 –73 . 10.1172/JCI119628 9303923 5 Weiss R , Taksali SE , Dufour S The “obese insulin-sensitive” adolescent: importance of adiponectin and lipid partitioning . J Clin Endocrinol Metab 2005 ;90 :3731 –7 . 10.1210/jc.2004-2305 15797955 6 Bakker LE , Sleddering MA , Schoones JW Pathogenesis of type 2 diabetes in South Asians . Eur J Endocrinol 2013 ;169 :R99 –r114 .23939919 7 Lara-Castro C , Newcomer BR , Rowell J Effects of short-term very low-calorie diet on intramyocellular lipid and insulin sensitivity in nondiabetic and type 2 diabetic subjects . Metab Clin Exp 2008 ;57 :1 –8 . 10.1016/j.metabol.2007.05.008 18078853 8 Hoeke G , Kooijman S , Boon MR Role of brown fat in lipoprotein metabolism and atherosclerosis . Circ Res 2016 ;118 :173 –82 . 10.1161/CIRCRESAHA.115.306647 26837747 9 Cypess AM , Lehman S , Williams G Identification and importance of brown adipose tissue in adult humans . N Engl J Med 2009 ;360 :1509 –17 . 10.1056/NEJMoa0810780 19357406 10 van Marken Lichtenbelt WD , Vanhommerig JW , Smulders NM Cold-activated brown adipose tissue in healthy men . N Engl J Med 2009 ;360 :1500 –8 . 10.1056/NEJMoa0808718 19357405 11 Yoneshiro T , Aita S , Matsushita M Recruited brown adipose tissue as an antiobesity agent in humans . J Clin Invest 2013 ;123 :3404 –8 . 10.1172/JCI67803 23867622 12 Hanssen MJ , Hoeks J , Brans B Short-term cold acclimation improves insulin sensitivity in patients with type 2 diabetes mellitus . Nat Med 2015 ;21 :863 –5 . 10.1038/nm.3891 26147760 13 Schilperoort M , Hoeke G , Kooijman S Relevance of lipid metabolism for brown fat visualization and quantification . Curr Opin Lipidol 2016 ;27 :242 –8 . 10.1097/MOL.0000000000000296 27023630 14 Ouellet V , Routhier-Labadie A , Bellemare W Outdoor temperature, age, sex, body mass index, and diabetic status determine the prevalence, mass, and glucose-uptake activity of 18F-FDG-detected BAT in humans . J Clin Endocrinol Metab 2011 ;96 :192 –9 . 10.1210/jc.2010-0989 20943785 15 Saito M , Okamatsu-Ogura Y , Matsushita M High incidence of metabolically active brown adipose tissue in healthy adult humans: effects of cold exposure and adiposity . Diabetes 2009 ;58 :1526 –31 . 10.2337/db09-0530 19401428 16 Shao X , Shao X , Wang X Characterization of brown adipose tissue (1)(8)F-FDG uptake in PET/CT imaging and its influencing factors in the Chinese population . Nucl Med Biol 2016 ;43 :7 –11 . 10.1016/j.nucmedbio.2015.09.002 26702781 17 Lee P , Bova R , Schofield L Brown adipose tissue exhibits a glucose-responsive thermogenic biorhythm in humans . Cell Metab 2016 ;23 :602 –9 . 10.1016/j.cmet.2016.02.007 26972823 18 Centers for Disease Control and Prevention . National Diabetes Surveillance System. http://www.cdc.gov/diabetes/data/index.html (accessed Sep 2015 ). 19 Vyth EL , Steenhuis IH , Mallant SF A front-of-pack nutrition logo: a quantitative and qualitative process evaluation in the Netherlands . J Health Commun 2009 ;14 :631 –45 . 10.1080/10810730903204247 19851915 20 NOAA National Centers for Environmental information, Climate at a Glance: U.S. Time Series, Average Temperature. http://www.ncdc.noaa.gov/cag/ (accessed Sep 2015 ). 21 Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Division of Population Health . BRFSS Prevalence & Trends Data. http://www.cdc.gov/brfss/brfssprevalence/ (accessed: Sep 2015 ). 22 Lumley T Analysis of complex survey samples. http://www.jstatsoft.org/v09/i08/paper. Secondary Analysis of complex survey samples 2004 http://www.jstatsoft.org/v09/i08/paper (accessed 02.05.13). http://www.jstatsoft.org/v09/i08/paper 23 Ministerie van VWS. NdMG . Hoeveel mensen hebben overgewicht? http://www.rivm.nl/dsresource?objectid=rivmp:76024&type=org&disposition=inline&ns_nc=1. Secondary Hoeveel mensen hebben overgewicht? 2013 http://www.rivm.nl/dsresource?objectid=rivmp:76024&type=org&disposition=inline&ns_nc=1 (accessed 02.05.13). http://www.rivm.nl/dsresource?objectid=rivmp:76024&type=org&disposition=inline&ns_nc=1 24 Harris I , Jones PD , Osborn TJ Updated high-resolution grids of monthly climatic observations—the CRU TS3.10 Dataset . Int J Climatol 2014 ;34 :623 –42 . 25 Wouters K , van Gorp PJ , Bieghs V Dietary cholesterol, rather than liver steatosis, leads to hepatic inflammation in hyperlipidemic mouse models of nonalcoholic steatohepatitis . Hepatology 2008 ;48 :474 –86 . 10.1002/hep.22363 18666236 26 DerSimonian R , Laird N Meta-analysis in clinical trials . Control Clin Trials 1986 ;7 :177 –88 .3802833 27 Kontopantelis E , Reeves D metaan: Random-effects meta-analysis . Stat J 2010 ;10 :395 –407 . 28 Harbord RM , Higgins JPT Meta-regression in Stata . Stat J 2008 ;8 :493 –519 . 29 Viechtbauer W Conducting meta-analyses in R with the metafor Package . J Stat Software 2010 ;36 :48 . 30 Li S , Zhou Y , Williams G Seasonality and temperature effects on fasting plasma glucose: a population-based longitudinal study in China . Diabetes Metab 2016 ;42 :267 –75 . 10.1016/j.diabet.2016.01.002 26851820 31 Suarez L , Barrett-Connor E Seasonal variation in fasting plasma glucose levels in man . Diabetologia 1982 ;22 :250 –3 .6980156 32 Matsushita M , Yoneshiro T , Aita S Impact of brown adipose tissue on body fatness and glucose metabolism in healthy humans . Int J Obes (Lond) 2014 ;38 :812 –7 . 10.1038/ijo.2013.206 24213309 33 Fridlyand LE , Philipson LH Cold climate genes and the prevalence of type 2 diabetes mellitus . Med Hypotheses 2006 ;67 :1034 –41 . 10.1016/j.mehy.2006.04.057 16797871 34 Chan CB , Ryan DA Assessing the effects of weather conditions on physical activity participation using objective measures . Int J Environ Res Public Health 2009 ;6 :2639 –54 . 10.3390/ijerph6102639 20054460 35 Togo F , Watanabe E , Park H Meteorology and the physical activity of the elderly: the Nakanojo Study . Int J Biometeorol 2005 ;50 :83 –9 . 10.1007/s00484-005-0277-z 16044348 36 Tucker P , Gilliland J The effect of season and weather on physical activity: a systematic review . Public Health 2007 ;121 :909 –22 . 10.1016/j.puhe.2007.04.009 17920646 37 Feunekes IJ , Van Staveren WA , Graveland F Reproducibility of a semiquantitative food frequency questionnaire to assess the intake of fats and cholesterol in the Netherlands . Int J Food Sci Nutr 1995 ;46 :117 –23 .7621083 38 United Nations Population Division . World Population Prospects. http://esa.un.org/unpd/wpp/Download/Standard/Population/ (accessed Mar 2016 ). 39 National Centers for Environmental Information . Climate monitoring: national temperature and precipitation maps . http://www.ncdc.noaa.gov/temp-and-precip/us-maps/3/201602—us-maps-select (accessed Mar 2016 ). 40 Li Z , Wang Y , van der Sluis RJ Niacin reduces plasma CETP levels by diminishing liver macrophage content in CETP transgenic mice . Biochem Pharmacol 2012 ;84 :821 –9 . 10.1016/j.bcp.2012.06.020 22750059

PMC005xxxxxx/PMC5372136.txt

==== Front BMJ Open Diabetes Res CareBMJ Open Diabetes Res CarebmjdrcbmjdrcBMJ Open Diabetes Research & Care2052-4897BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjdrc-2016-00035110.1136/bmjdrc-2016-000351Clinical Care/Education/Nutrition/Psychosocial Research15061866Lower glycemic load meals reduce diurnal glycemic oscillations in women with risk factors for gestational diabetes Kizirian Nathalie V 12Goletzke Janina 123Brodie Shannon 124Atkinson Fiona S 12Markovic Tania P 45Ross Glynis P 56Buyken Anette 3Brand-Miller Jennie C 124 1 Charles Perkins Centre, The University of Sydney, Sydney, New South Wales, Australia 2 School of Life and Environmental Sciences, The University of Sydney, Sydney, New South Wales, Australia 3 IEL-Nutritional Epidemiology, University of Bonn, DONALD Study, Dortmund, Germany 4 Boden Institute of Obesity, Nutrition, Exercise & Eating Disorders, The University of Sydney, Sydney, New South Wales, Australia 5 Department of Endocrinology, Royal Prince Alfred Hospital, Camperdown, New South Wales, Australia 6 Sydney Medical School, The University of Sydney, Sydney, New South Wales, AustraliaCorrespondence to Professor Jennie C Brand-Miller; jennie.brandmiller@sydney.edu.au2017 29 3 2017 5 1 e00035127 10 2016 15 2 2017 9 3 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective Maternal glycemia plays a key role in fetal growth. We hypothesized that lower glycemic load (GL) meals (lower glycemic index, modestly lower carbohydrate) would substantially reduce day-long glucose variability in women at risk of gestational diabetes mellitus (GDM). Research design and methods A crossover study of 17 women (mean±SD age 34.8±4 years; gestational weeks 29.3±1.3; body mass index 23.8±4.7 kg/m2) who consumed a low GL or a high GL diet in random order, 1-day each, over 2 consecutive days. Diets were energy-matched and fiber-matched with 5 meals per 24 hours. All food was provided. Continuous glucose monitoring was used to assess diurnal glycemia. Results Maternal glucose levels were 51% lower on the low GL day with lower incremental area under the curve (iAUC±SEM 549±109 vs 1120±198 mmol/L min, p=0.015). Glycemic variability was significantly lower on the low GL day, as demonstrated by a lower average SD (0.7±0.1 vs 0.9±0.1, p<0.001) and lower mean amplitude of glycemic excursions (2.1±0.2 vs 2.7±0.2 mmol/L, p<0.001). Conclusions A lower GL meal plan in pregnancy acutely halves day-long maternal glucose levels and reduces glucose variability, providing further evidence to support the utility of a low GL diet in pregnancy. Gestational Diabetes MellitusContinuous Glucose MonitoringPregnancyGlycemic Index ==== Body Significance of this study What is already known about this subject? Glycemic variability, characterized by acute and chronic glucose excursions, is an important predictor of comorbidities in individuals with diabetes. Low glycemic index (GI) foods produce lower postprandial blood glucose spikes compared to high GI foods. In several randomized controlled trials (but not all), low GI diets in pregnancy have resulted in lower (normalized) birth weight, lower ponderal index improved glucose tolerance, lower gestational weight gain, and reduction in need for insulin therapy in women with gestational diabetes mellitus (GDM). What are the new findings? Low glycemic load meals improve diurnal glucose control and produce less glycemic variability than conventional diets in women at risk of GDM. How might these results change the focus of research or clinical practice? Modestly lower carbohydrate diets with an emphasis on low GI food sources can be recommended to women at risk of developing diabetes in pregnancy. Introduction Maternal glucose intolerance is an important predictor of adverse perinatal outcomes,1 including increased risk of offspring diabetes and obesity in later life.2–4 Importantly, the consequences of maternal glycemia on perinatal outcomes do not occur at a specific threshold but rather on a continuum.1 Recently, glycemic variability, characterized by acute and chronic glucose excursions, was associated with increased infant ponderal index in women with type 1 diabetes5 and impaired early-phase insulin secretion in women with gestational diabetes mellitus (GDM).6 Glucose oscillation has been found to induce greater oxidative stress and tissue damage than sustained high blood glucose levels (BGLs).7–9 Given the strong associations between maternal glycemia, perinatal outcomes and fetal programming, strategies to optimize maternal glycemic control are crucial for the long-term health of mother and offspring.1 6 10 Blood glucose concentrations are strongly influenced by the quality and quantity of carbohydrates in the diet. In healthy individuals and those with diabetes, low glycemic index (GI) foods produce lower postprandial blood glucose spikes compared to high GI foods11 and have been associated with improved diurnal glycemic profiles and lesser glycemic variability, compared to a high GI diet.12–15 Dietary glycemic load (GL), the product of the GI of a food and its carbohydrate content, has been found to be the best predictor of postprandial glycemia and insulinemia in healthy adults consuming a mixed diet.16 Thus, the combination of a modestly lower carbohydrate intake from low GI food sources with a reciprocal increase in fat and protein may be the simplest dietary strategy to optimize maternal glucose control. Until now, no study has explored the effect of diets of varying dietary GL on diurnal glucose levels and glycemic variability in pregnancy. The aim of this study was to use continuous glucose monitoring (CGM) to test the acute effect of two fiber-matched diets, one high GL and one low GL, 1-day each, on glycemic profiles and glycemic variability in third trimester pregnant women at risk of GDM. We hypothesized that the low GL meals would produce lower diurnal glucose levels and lesser glycemic variability, compared to the high GL meals. Research design and methods This was a within-subject randomized crossover trial comparing the effects of 1-day low GL meals and 1-day high GL meals, in women between 26 and 32 weeks of gestation at higher risk of GDM. Participants were recruited through the antenatal clinic at the Royal Prince Alfred Hospital, Camperdown, NSW, Australia. Women were prescreened at the time of their second obstetric appointment (19–23 weeks gestation) and invited to participate in the study. Those expressing interest were given details of the study and contacted again after routine testing for GDM at 26–28 weeks gestation. Women were eligible if they had at least one of the following risk factors for GDM: prepregnancy body mass index (BMI) ≥30 kg/m2, age ≥35 years, polycystic ovarian syndrome, history of GDM or glucose intolerance, history of a previous birth >4000 g, family history of type 2 diabetes (first-degree relative), belonging to an ethnic group with a high prevalence of GDM (Aboriginal, Torres Strait Islander, Polynesian, Middle Eastern, South Asian, South East Asian). Women with diagnosed GDM, pre-existing diabetes, multiple pregnancy or special dietary requirements (gluten-intolerant, vegetarian) were excluded. GDM diagnosis was based on the 1998 Australian Diabetes in Pregnancy Society (ADIPS) criteria until January 2015 (fasting BGL ≥5.5 mmol/L, 1 h BGL ≥10.0 mmol/L or 2 h BGL ≥8.0 mmol/L17) after which new diagnosis criteria were adopted recommending that fasting BGL >5.1 mmol/L be classified as GDM.18 In total, 146 women were approached between January 2014 and July 2015, of whom 70 expressed interest. Of these, 39 subsequently declined, 5 were lost to follow-up, and 7 were excluded following a diagnosis of GDM (figure 1). A total of 19 women met the inclusion criteria and started the trial. This study was approved by the Human Research Ethics Committee of the Sydney South West Area Health Service (RPAH Zone). All participants gave written informed consent. Figure 1 Participants flow diagram. Study protocol Continuous glucose monitors CGM systems (Medtronic MiniMed, Northridge, California) were used to electrochemically measure subcutaneous interstitial glucose concentrations every 5 min, generating 288 measurements per day. The monitors were calibrated four times a day against capillary blood glucose measurements, using blood glucose meters (Accu-Check Performa; Roche) as per the manufacturer's instructions. The monitors were placed on the lower back, on the right or left side, depending on the women's preference. Diet and protocol Women were provided with all the foods required for the 1-day low GL meals and 1-day high GL meals. Dietary macronutrient composition of the two test days were formulated by the dietitian (SB) to produce a twofold difference in GL by reducing carbohydrate content and GI, while matching the fiber content and energy (table 1). The mean GL of the high and low GL days were 144 and 70 units, over the 24 h period, respectively. A within-subject crossover trial was used, and the order of the low and high GL days was randomized using a computer-generated randomization program. On day 1 of CGM use, women were instructed to eat at their discretion. On days 2 and 3, women were instructed to consume the test foods provided. On day 4, the CGM was removed and data of days 2 and 3 (midnight to midnight) were extracted for analysis. Dietary compliance was assessed using the log book. Women were asked to record the time at which the meals were consumed and whether the meals were consumed entirely. In a small number of women, plate wastage was recorded. Table 1 Low and high GL diets Low GL meal High GL day Foods g kJ P Fat Sat f CHO Fib Sug GI GL Foods g kJ P Fat Sat f CHO Fib Sug GI GL Breakfast Breakfast Goodness Superfoods barley and oats* 35 500 5 3 1 16 6 1 53 9 Uncle Tobys High fibre oats* 40 563 4 3 1 21 6 4 75 16 Burgen wholegrain bread 80 803 10 5 1 25 6 2 39 10 Wholemeal bread 60 554 5 1 0 25 4 1 70 18 Vegemite spread 20 117 5 0 0 2 0 0 0 0 Vegemite spread 20 117 5 0 0 2 0 0 0 0 Margarine 14 407 0 11 2 0 0 0 0 0 Margarine 14 407 0 11 2 0 0 0 0 0 Wheat bran† 13 167 2 1 0 4 4 0 0 0  Total (g) 1827 20 19 4 43 12 3 45 18 Total (g) 1808 16 16 3 52 14 4 67 34 Lunch Lunch Burgen wholegrain bread 80 803 10 5 1 25 6 2 39 10 Wholemeal bread 90 866 8 2 0 38 6 2 70 26 Tuna canned in oil 80 559 22 5 1 0 0 0 0 0 Tuna canned in oil 80 559 22 5 1 0 0 0 0 0 Tomato (1 average) 80 40 0 0 0 2 1 0 0 0 Tomato (1 average) 80 40 0 0 0 2 1 0 0 0 Cucumber (1 average) 100 40 0 0 0 2 1 0 0 0 Cucumber (1 average) 100 40 0 0 0 2 1 0 0 0 Lettuce (1 cup) 35 8 0 0 0 0 1 0 0 0 Lettuce (1 cup) 35 8 0 0 0 0 1 0 0 0 Dressing (ready) 30 265 0 5 1 5 0 4 0 0 Dressing (ready) 30 265 0 5 1 5 0 4 0 0 Be Natural nut bar 40 870 8 15 2 12 3 8 29 3 Vita-weat biscuits 30 453 3 2 0 19 3 1 65 12  Total (g) 2585 40 30 5 46 12 2 32 13 Total (g) 2231 33 14 2 66 12 7 64 39 Afternoon tea Afternoon tea Activia yogurt 125 448 6 2 1 17 0 16 25 4 Sakata rice crackers 30 471 2 1 0 25 0 1 80 20 Carrot (1 average) 70 80 0 0 0 4 2 0 16 1 Philadelphia cream cheese 40 305 3 6 4 2 0 0 0 0  Total (g) 448 6 2 1 17 0 16 25 4 Total (g) 856 5 7 4 31 2 1 67 21 Dinner Dinner Edgell four bean mix 75 379 6 1 0 12 6 2 37 5 Mashed potato (dry)‡ 25 420 2 2 1 18 3 2 85 16 Inghams chicken breast 100 858 13 9 2 19 0 3 0 0 Inghams chicken breast 100 858 13 9 2 19 0 3 0 0 Tomatoes (2 average) 160 80 0 0 0 4 2 0 0 0 Tomatoes (2 average) 160 80 0 0 0 4 2 0 0 0 Cucumber (1 average) 100 40 0 0 0 2 1 0 0 0 Cucumber (1 average) 100 40 0 0 0 2 0 0 0 0 Lettuce (1 cup) 35 8 0 0 0 0 1 0 0 0 Lettuce (1 cup) 35 8 0 0 0 0 1 0 0 0 Sweet corn canned 72 231 2 1 0 9 2 1 55 5 Sweet corn canned 72 231 2 1 0 9 2 1 55 5 Dressing (ready) 30 265 0 5 1 5 0 4 0 0 Dressing (ready) 30 265 0 5 1 5 0 4 0 0 Wheat bran § 6 48 0 0 0 2 2 0 0 0  Total (g) 1861 21 16 3 51 12 10 41 19 Total (g) 1950 17 17 4 59 10 10 56 30 Supper Supper Belvita biscuits 50 964 4 8 1 36 3 11 45 16 SunRice rice cakes 40 663 3 4 1 28 2 0 78 22  Total (g) 964 4 8 1 36 3 11 45 16 Total (g) 663 3 4 1 28 2 0 78 22 Total (g) 7685 91 74 14 193 39 54 39 70 Total (g) 7508 74 58 14 236 41 23 65 144 Total % energy 20 36 7 40 4 11 Total % energy 17 29 7 50 4 5     Total % energy does not account for the energy provided by fiber (2 kcal/g). *Oats were made on water. †Wheat bran was added to the oats to increase the amount of fiber. ‡Mash was reconstituted from the powder using water. §Wheat bran was added to the mash to increase the amount of fiber. CHO, carbohydrate; Fib, fiber; GI, glycemic index; GL, glycemic load; kJ, kilojoules; P, protein; Sat f, saturated fat; Sug, sugar. Statistical analysis Interstitial glucose values were obtained by CGMs. Comparison between the high and low GL days was assessed by paired Student's t-test. The mean amplitude of glycemic excursion (MAGE) was determined by calculating the arithmetic mean of the difference between consecutive peaks and nadirs if the difference was >1 SD of the mean glucose,19 20 using the automated GlyCulator algorithm.21 The incremental area under the curve (iAUC) was calculated across the 24 h, using the trapezoidal rule. The glucose value at midnight was used as baseline, and all the interstitial glucose output from the CGMs were used. Any area below baseline concentration was ignored. All the other variables were generated by transferring the CGM stored data into the data management system. The sample size calculation was based on the mean±SEM of iAUC glucose reported by Solomon et al22 in eight participants consuming representative diets (low GI or high GI) over the course of the day. In our study, a sample of 12 participants provided >80% power to detect a difference greater than twofold between the high and low GL groups, with an α-value of 0.05. Results In total, 19 women started the trial. Data were not recorded for one participant because of CGM failure. One participant was admitted to hospital on day 2 for reasons unrelated to the study, and her data were excluded from the analysis. Results were available for the remaining 17 participants, including two full 24 hours (midnight to midnight) glucose monitoring days. Dietary compliance was higher on the low GL day compared to the high GL day. On the low GL day, one woman did not consume the supper (∼960 kJ deficit). On the high GL day, two women did not consume the supper (∼660 kJ deficit), two ate half of the supper (∼330 kJ deficit), one ate half of the afternoon tea (∼420 kJ deficit) and one ate half of the breakfast porridge (∼280 kJ deficit). The omission of these foods resulted in a reduction in the daily total energy content of between 4% and 12%. The mean±SD age was 34.8±4 years and prepregnancy BMI was 23.8±4.7 kg/m2. The mean gestational age at study entry was 29.3±1.3 weeks, and 71% of the participants were Caucasian. Dietary GL and diurnal glucose levels Diurnal glucose responses measured by CGMs on the high GL day and low GL day are shown in figure 2, and summary results are presented in table 2. Table 2 Average maternal glucose levels (n=17) for the low GL and high GL meals Low GL High GL p Value Highest glucose value (mmol/L) 7.1±0.2 7.6±0.2 0.026 Lowest glucose value (mmol/L) 3.9±0.2 3.7±0.2 0.237 Average (mmol/L) 5.3±0.1 5.4±0.2 0.352 SD average (mmol/L) 0.7±0.1 0.9±0.1 <0.001 MAGE (mmol/L) 2.1±0.2 2.7±0.2 <0.001 iAUC (mmol/L min) 549±109 1120±198 0.015 Proportion of time (%)  In target (3.9–7.8 mmol/L) 95.1±1.7 87.7±3.2 0.031  Above target 0.4±0.2 4.4±2.8 0.180  Below target 4.5±1.7 7.9±2.2 0.129 Mean±SEM (all such values). iAUC, incremental area under the curve (calculated across the 24 h, using the trapezoidal rule); MAGE, mean amplitude of glycemic excursions (determined by calculating the arithmetic mean of the difference between consecutive peaks and nadirs if the difference is >1 SD of the mean glucose). Figure 2 Mean (±SEM) diurnal glucose levels following low GL meals (gray dots) and high GL meals (black dots). Average mealtimes are represented by the black arrows. Diurnal glucose levels were significantly lower on the low GL day, as demonstrated by a 51% lower iAUC on the low GL day compared to the high GL day (iAUC±SEM 549±109 mmol/L min vs 1120±198 mmol/L min, p=0.015). The peak glucose concentration was also significantly lower on the low GL day (mean±SEM 7.1±0.2 mmol/L vs 7.6±0.2 mmol/L, p=0.026). Time spent within the normal (target) glucose range was significantly longer on the low GL day, resulting from fewer above and lesser below target values, hence demonstrating less glucose oscillation (low GL 95.1±1.7% vs high GL 87.7±3.2%, p=0.031). Glycemic variability was significantly lower on the low GL day, as demonstrated by a lower average SD and lower MAGE (SD average 0.7±0.1 vs 0.9±0.1, p<0.001; MAGE 2.1±0.2 mmol/L vs 2.7±0.2 mmol/L, p<0.001). Conclusions To the best of our knowledge, this is the first study to explore the effects of modestly higher protein/lower carbohydrate (lower GL) meals versus conventional meals on day-long glucose profiles in women at risk of GDM. Our findings show that the low GL diet achieved improved glycemic control as judged by ∼50% lower diurnal glucose levels, increased time within target glucose range and less glycemic oscillation than the conventional diet. Our findings have implications for improving the dietary management of pregnant women with overweight, obesity, and other risk factors for GDM. Increased maternal glycemia is associated with excessive growth and adiposity,1 23 poor vascular health,24 and increased risk of metabolic disorders and obesity in the offspring.10 25 Higher glycemic variability, even of a modest degree, has been linked to higher fetal ponderal index, independently of glycated haemoglobin, in pregnant women with type 1 diabetes.5 In the present study, MAGE and average SD were significantly lower on the low GL day, potentially reducing the risk of preeclampsia and neonatal complications associated with high MAGE.26 Currently, diet therapy aimed at lowering BGLs represents the first line of treatment in women with GDM.27 In a previous study, we demonstrated that a low GI diet reduced the need for insulin therapy, without increasing the risk of adverse pregnancy outcomes.28 In a self-selected subgroup of pregnant women at high risk of GDM, a low GI diet resulted in improved infant weight-for-age and length-for-age z-scores and thinner carotid intima-media thickness at 1 year of age.24 In women with a history of macrosomia, a low GI diet improved glucose tolerance and gestational weight gain, although not the risk of macrosomia.29 In this study, improvement in maternal glucose levels on the low GL diet was achieved by reducing the proportion of energy as carbohydrate (from 50% to 40%E), as well as by replacing high GI sources of carbohydrate with low GI sources. The reduction in carbohydrate energy was accompanied by a modest increase in protein energy (from 17% to 20%E) and fat energy (from 29% to 36%E). However, the meals with the largest difference in GL (lunch and afternoon tea) produced the largest difference in postprandial glycemia over the course of the day. Thus, carbohydrate and GI influence ambient glucose concentrations. Importantly, the sources of fat were chosen such that there was no increase in the proportion of energy as saturated fat (<10% in both diets). The low GL diet therefore resembled a modestly lower carbohydrate diet, similar to that associated with the lowest risk of metabolic syndrome in the PREDIMED study.30 A recent meta-analysis indicates that this macronutrient distribution is associated with better markers of glucose homeostasis and HbA1c than higher carbohydrate diets.31 Our findings should not be interpreted as a ‘green light’ for low or very low carbohydrate diets in women at risk of GDM. Severe restriction of carbohydrate markedly increases fatty acid oxidation and therefore ketone levels.32 In pregnancy, the risks associated with high maternal ketone concentration are currently unclear but may include adverse effects on offspring intelligence.33 Furthermore, carbohydrate-restricted diets may induce unbalanced macronutrient intake by increasing dietary fat.34 Data in pregnancy suggest a strong influence of maternal triglycerides and free-fatty acids (FFAs) on excessive fetal adiposity accretion.35 36 However, in GDM, diets with lower fat and higher ‘complex’ carbohydrate resulted in lower levels of maternal fasting glucose and FFAs, compared to a conventional diet.37 Outside of pregnancy, high fat diets have been shown to promote insulin resistance.38 The strengths of this study include the randomized controlled design, the use of CGM with frequent assessment of maternal glucose concentrations across the day and the provision of foods with known (tested) GI and nutrient composition and the crossover study design. To best represent typical lifestyle patterns within the study, participants were free-living and the meals were consumed according to the mother's schedule. Our study had some limitations, most notably the small sample size and the short duration of the study; only 24 h per GL diet of CGM data were analyzed instead of an average over >1 day which might not represent the true day-to-day variation in glycemic response. Only acute effects were reported, and there was no lead in diets and no washout day. More tightly controlled meal times may have increased the differences between the two diets. In conclusion, this study showed that, in third trimester pregnant women at risk of GDM, low GL meals improve diurnal glycemic control and glycemic variability, compared to high GL meals. This study adds to the evidence supporting the utility of a low GL diet in pregnancy to optimize glycemic control. The authors thank all the study participants. Contributors: JPB-M, NVK, AB, JG, GPR, and TPM conceived and designed the study; NVK and JG researched the data; NVK analyzed the data and wrote the manuscript; SB designed the diets; and FSA was involved in the design of the diets and generated the incremental area under the curve. All authors reviewed and approved the final manuscript. JPB-M is the guarantor of this work and takes responsibility for the contents of the article. Competing interests: JPB-M is the President of the Glycemic Index Foundation, Director of the Sydney University Glycemic Index Research Service and author of popular books about the glycemic index of foods. FSA is a director of the Glycemic Index Foundation, manages the Sydney University Glycemic Index Research Service and is a coauthor of popular books about the glycemic index of foods. No other potential conflicts of interest relevant to this article are declared. Ethics approval: Sydney South West Area Health Service (RPAH Zone). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Metzger BE , Lowe LP , Dyer AR Hyperglycemia and adverse pregnancy outcomes . N Engl J Med 2008 ;358 :1991 –2002 . 10.1056/NEJMoa0707943 18463375 2 Crume TL , Ogden L , Daniels S The impact of in utero exposure to diabetes on childhood body mass index growth trajectories: the EPOCH Study . J Pediatrics 2011 ;158 :941 –6 . 3 Catalano PM The impact of gestational diabetes and maternal obesity on the mother and her offspring . J Dev Orig Health Dis 2010 ;1 :208 –15 . 10.1017/S2040174410000115 25141869 4 Hillier TA , Pedula KL , Schmidt MM Childhood obesity and metabolic imprinting: the ongoing effects of maternal hyperglycemia . Diabetes Care 2007 ;30 :2287 –92 . 10.2337/dc06-2361 17519427 5 Dalfra MG , Sartore G , Di Cianni G Glucose variability in diabetic pregnancy . Diabetes Technol Ther 2011 ;13 :853 –9 . 10.1089/dia.2010.0145 21751862 6 Su JB , Wang XQ , Chen JF Glycemic variability in gestational diabetes mellitus and its association with beta cell function . Endocrine 2013 ;43 :370 –5 . 10.1007/s12020-012-9753-5 22815046 7 Chang CM , Hsieh CJ , Huang JC Acute and chronic fluctuations in blood glucose levels can increase oxidative stress in type 2 diabetes mellitus . Acta Diabetol 2012 ;49 (Suppl 1) :S171 –7 . 10.1007/s00592-012-0398-x 22547264 8 Monnier L , Mas E , Ginet C Activation of oxidative stress by acute glucose fluctuations compared with sustained chronic hyperglycemia in patients with type 2 diabetes . JAMA 2006 ;295 :1681 –7 . 10.1001/jama.295.14.1681 16609090 9 Ceriello A , Esposito K , Piconi L Oscillating glucose is more deleterious to endothelial function and oxidative stress than mean glucose in normal and type 2 diabetic patients . Diabetes 2008 ;57 :1349 –54 . 10.2337/db08-0063 18299315 10 Dabelea D , Crume T Maternal environment and the transgenerational cycle of obesity and diabetes . Diabetes 2011 ;60 :1849 –55 . 10.2337/db11-0400 21709280 11 Brand-Miller JC , Stockmann K , Atkinson F Glycemic index, postprandial glycemia, and the shape of the curve in healthy subjects: analysis of a database of more than 1,000 foods . Am J Clin Nutr 2009 ;89 :97 –105 . 10.3945/ajcn.2008.26354 19056599 12 Brynes AE , Adamson J , Dornhorst A The beneficial effect of a diet with low glycaemic index on 24 h glucose profiles in healthy young people as assessed by continuous glucose monitoring . Br J Nutr 2005 ;93 :179 –82 .15788110 13 Nansel TR , Gellar L , McGill A Effect of varying glycemic index meals on blood glucose control assessed with continuous glucose monitoring in youth with type 1 diabetes on basal-bolus insulin regimens . Diabetes Care 2008 ;31 :695 –7 . 10.2337/dc07-1879 18202243 14 Brynes AE , Lee JL , Brighton RE A low glycemic diet significantly improves the 24-h blood glucose profile in people with type 2 diabetes, as assessed using the continuous glucose MiniMed monitor . Diabetes Care 2003 ;26 :548 –9 .12547905 15 Fabricatore AN , Ebbeling CB , Wadden TA Continuous glucose monitoring to assess the ecologic validity of dietary glycemic index and glycemic load . Am J Clin Nutr 2011 ;94 :1519 –24 . 10.3945/ajcn.111.020354 22071699 16 Bao J , Atkinson F , Petocz P Prediction of postprandial glycemia and insulinemia in lean, young, healthy adults: glycemic load compared with carbohydrate content alone . Am J Clin Nutr 2011 ;93 :984 –96 . 10.3945/ajcn.110.005033 21325437 17 Hoffman L , Nolan C , Wilson JD Gestational diabetes mellitus—management guidelines. The Australasian Diabetes in Pregnancy Society . Med J Aust 1998 ;169 :93 –7 .9700346 18 Metzger BE , Gabbe SG , Persson B , International Association of Diabetes and Pregnancy Study Groups Consensus Panel . International association of diabetes and pregnancy study groups recommendations on the diagnosis and classification of hyperglycemia in pregnancy . Diabetes Care 2010 ;33 :676 –82 . 10.2337/dc09-1848 20190296 19 Service FJ Glucose variability . Diabetes 2013 ;62 :1398 –404 . 10.2337/db12-1396 23613565 20 Baghurst PA Calculating the mean amplitude of glycemic excursion from continuous glucose monitoring data: an automated algorithm . Diabetes Technol Ther 2011 ;13 :296 –302 . 10.1089/dia.2010.0090 21291334 21 Czerwoniuk D , Fendler W , Walenciak L GlyCulator: a glycemic variability calculation tool for continuous glucose monitoring data . J Diabetes Sci Technol 2011 ;5 :447 –51 . 10.1177/193229681100500236 21527118 22 Solomon TP , Haus JM , Kelly KR A low-glycemic index diet combined with exercise reduces insulin resistance, postprandial hyperinsulinemia, and glucose-dependent insulinotropic polypeptide responses in obese, prediabetic humans . Am J Clin Nutr 2010 ;92 :1359 –68 . 10.3945/ajcn.2010.29771 20980494 23 HAPO Study Cooperative Research Group . Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study: associations with neonatal anthropometrics . Diabetes 2009 ;58 :453 –9 . 10.2337/db08-1112 19011170 24 Kizirian NV , Kong Y , Muirhead R Effects of a low-glycemic index diet during pregnancy on offspring growth, body composition, and vascular health: a pilot randomized controlled trial . Am J Clin Nutr 2016 ;103 :1073 –82 . 10.3945/ajcn.115.123695 26936333 25 Ma RC , Tutino GE , Lillycrop KA Maternal diabetes, gestational diabetes and the role of epigenetics in their long term effects on offspring . Prog Biophys Mol Biol 2015 ;118 :55 –68 . 10.1016/j.pbiomolbio.2015.02.010 25792090 26 Yu F , Lv L , Liang Z Continuous glucose monitoring effects on maternal glycemic control and pregnancy outcomes in patients with gestational diabetes mellitus: a prospective cohort study . J Clin Endocrinol Metab 2014 ;99 :4674 –82 . 10.1210/jc.2013-4332 25057872 27 Blumer I , Hadar E , Hadden DR Diabetes and pregnancy: an endocrine society clinical practice guideline . J Clin Endocrinol Metab 2013 ;98 :4227 –49 . 10.1210/jc.2013-2465 24194617 28 Moses RG , Barker M , Winter M Can a low-glycemic index diet reduce the need for insulin in gestational diabetes mellitus? A randomized trial . Diabetes Care 2009 ;32 :996 –1000 . 10.2337/dc09-0007 19279301 29 Walsh JM , McGowan CA , Mahony R Low glycaemic index diet in pregnancy to prevent macrosomia (ROLO study): randomised control trial . BMJ 2012 ;345 :e5605 .22936795 30 Juanola-Falgarona M , Salas-Salvado J , Buil-Cosiales P Dietary glycemic index and glycemic load are positively associated with risk of developing metabolic syndrome in middle-aged and elderly adults . J Am Geriatr Soc 2015 ;63 :1991 –2000 . 10.1111/jgs.13668 26480969 31 Imamura F , Micha R , Wu JH Effects of saturated fat, polyunsaturated fat, monounsaturated fat, and carbohydrate on glucose-insulin homeostasis: a systematic review and meta-analysis of randomised controlled feeding trials . PLoS Med 2016 ;13 :e1002087 10.1371/journal.pmed.1002087 27434027 32 Cotter DG , Schugar RC , Crawford PA Ketone body metabolism and cardiovascular disease . Am J Physiol Heart Circ Physiol 2013 ;304 :H1060 –76 . 10.1152/ajpheart.00646.2012 23396451 33 Rizzo T , Metzger BE , Burns WJ Correlations between antepartum maternal metabolism and child intelligence . N Engl J Med 1991 ;325 :911 –6 . 10.1056/NEJM199109263251303 1881416 34 Hernandez TL Carbohydrate content in the GDM Diet: two views: view 1: nutrition therapy in gestational diabetes: the case for complex carbohydrates . Diabetes Spectr 2016 ;29 :82 –8 . 10.2337/diaspect.29.2.82 27182176 35 Harmon KA , Gerard L , Jensen DR Continuous glucose profiles in obese and normal-weight pregnant women on a controlled diet: metabolic determinants of fetal growth . Diabetes Care 2011 ;34 :2198 –204 . 10.2337/dc11-0723 21775754 36 Schaefer-Graf UM , Graf K , Kulbacka I Maternal lipids as strong determinants of fetal environment and growth in pregnancies with gestational diabetes mellitus . Diabetes Care 2008 ;31 :1858 –63 . 10.2337/dc08-0039 18606978 37 Hernandez TL , Van Pelt RE , Anderson MA Women with gestational diabetes randomized to a higher-complex carbohydrate/low-fat diet manifest lower adipose tissue insulin resistance, inflammation, glucose, and free fatty acids: a pilot study . Diabetes Care 2016 ;39 :39 –42 . 10.2337/dc15-0515 26223240 38 Lichtenstein AH , Schwab US Relationship of dietary fat to glucose metabolism . Atherosclerosis 2000 ;150 :227 –43 .10856515

PMC005xxxxxx/PMC5372137.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01474910.1136/bmjopen-2016-014749Evidence Based PracticeProtocol150616941692State of reporting of primary biomedical research: a scoping review protocol http://orcid.org/0000-0002-3472-8513Li Guowei 123Mbuagbaw Lawrence 12Samaan Zainab 124Jin Yanling 1Nwosu Ikunna 1Levine Mitchell A H 1234Adachi Jonathan D 24Thabane Lehana 12 1 Department of Clinical Epidemiology & Biostatistics, McMaster University, Hamilton, Ontario, Canada 2 St. Joseph's Healthcare Hamilton, McMaster University, Hamilton, Ontario, Canada 3 Centre for Evaluation of Medicines, Programs for Assessment of Technology in Health (PATH) Research Institute, McMaster University, Hamilton, Ontario, Canada 4 Department of Medicine, McMaster University, Hamilton, Ontario, CanadaCorrespondence to Dr Guowei Li; lig28@mcmaster.ca2017 29 3 2017 7 3 e01474914 10 2016 28 2 2017 3 3 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction Incomplete or inconsistent reporting remains a major concern in the biomedical literature. Incomplete or inconsistent reporting may yield the published findings unreliable, irreproducible or sometimes misleading. In this study based on evidence from systematic reviews and surveys that have evaluated the reporting issues in primary biomedical studies, we aim to conduct a scoping review with focuses on (1) the state-of-the-art extent of adherence to the emerging reporting guidelines in primary biomedical research, (2) the inconsistency between protocols or registrations and full reports and (3) the disagreement between abstracts and full-text articles. Methods and analyses We will use a comprehensive search strategy to retrieve all available and eligible systematic reviews and surveys in the literature. We will search the following electronic databases: Web of Science, Excerpta Medica Database (EMBASE), MEDLINE and Cumulative Index to Nursing and Allied Health Literature (CINAHL). Our outcomes are levels of adherence to reporting guidelines, levels of consistency between protocols or registrations and full reports and the agreement between abstracts and full reports, all of which will be expressed as percentages, quality scores or categorised rating (such as high, medium and low). No pooled analyses will be performed quantitatively given the heterogeneity of the included systematic reviews and surveys. Likewise, factors associated with improved completeness and consistency of reporting will be summarised qualitatively. The quality of the included systematic reviews will be evaluated using AMSTAR (a measurement tool to assess systematic reviews). Ethics and dissemination All findings will be published in peer-reviewed journals and relevant conferences. These results may advance our understanding of the extent of incomplete and inconsistent reporting, factors related to improved completeness and consistency of reporting and potential recommendations for various stakeholders in the biomedical community. incomplete reportingbiased reportingreporting guidelineconsistencyscoping review ==== Body Strengths and limitations of this study In this scoping review, we will assess the consistency and completeness of reporting in the biomedical literature with regards to adherence to reporting guidelines, consistency between protocols or registrations and full reports and agreement between abstracts and full-text articles. Results from our study will advance our understanding of the extent of incomplete and inconsistent reporting, factors related to improved completeness and consistency of reporting and potential recommendations for various stakeholders in the biomedical community. A potential limitation may be the small number of eligible studies for this scoping review. Introduction Primary research is generally defined as the empirical research studies with collection of original primary data.1 The current reporting in primary biomedical research remains an issue of concern in the literature.2 For instance, it is widely recognised that incomplete reporting is pervasive in biomedical research, leading to potential waste of resources, sceptical interpretation of findings and even scientific misconduct.2 One study showed that over 50% of research findings were not sufficiently or completely reported to make them usable or replicable, which represented a substantial waste of resources and efforts.3 Likewise, it is difficult to make an informed judgement about the risk of bias and credibility of findings in a study due to its incomplete reporting and lack of linkage to protocol or registration.4 Moreover, incomplete reporting can result in unnecessary exposure or harm to patients and lead to imprecise or biased treatment effect estimates to inform decision-making.2 5 To improve transparent and complete reporting in biomedical research, reporting guidelines have been developed and widely adopted by more and more journals. The EQUATOR (Enhancing Quality and Transparency of Health Research) network provides support for the dissemination of such guidelines including the CONSORT (Consolidated Standards of Reporting Trials) for clinical trials, STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) for observational studies, PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) for systematic reviews, STARD (Standards for Reporting Diagnostic accuracy studies) for diagnostic or prognostic studies and ARRIVE (Animal Research: Reporting In Vivo Experiments) for animal studies, among others.6 Evidence has shown that application of guidelines is associated with improved standards of reporting, and looking for missing items from guidelines of submissions in the peer review process can enhance the quality of peer reviews and the finalised publications.7–10 Despite the usefulness of reporting guidelines, adherence to such guidelines in the biomedical research remains unsatisfactorily low.4 11 12 Table 1 Summary of key factors for the three parts (guideline adherence, inconsistency between protocols/registrations and full reports and inconsistency between abstracts and full reports) included in the scoping review Key factor Guideline adherence Inconsistency between protocols or registrations and full reports Inconsistency between abstracts and full reports Primary objective Current state of reporting in primary biomedical research Secondary objective Factor associated with improved completeness or consistency of reporting Outcome Level of guideline adherence Level of (in)consistent reporting Comparator reference Reporting guidelines Protocols or registrations Full reports Main data collected Adherence to the items listed in guidelines Inconsistent reporting on study-validity-related factors* Data analysis Qualitative description summarised *Study-validity-related factors including research questions, study designs, study populations or sample sizes, interventions or exposures, time duration, comparators, statistical plan, result presentations and interpretations and conclusions or recommendations. Beyond poor adherence to reporting guidelines, inconsistent or biased reporting between protocols or registrations and fully published articles has also raised significant concerns.13–16 For instance, one study comparing protocols and full reports in clinical trials found that approximately two-thirds of full reports had at least planned primary outcome modified, introduced or omitted.17 Similarly, another study focusing on trials funded by Canadian Institutes of Health Research reported that 40% of the trials had a difference in primary outcomes between protocols and full reports.18 Furthermore, abstracts, as the generally most read and accessed section of a publication, were found to be distorted or overly-optimistic presentations of results than were shown in full reports.19 Discrepancy between abstracts and full reports deserves more intensive attention and stringent examination in biomedical research because (1) abstracts are usually prepared with the least care; (2) readers draw conclusions about a study mainly depending on abstracts and (3) audiences may make their decisions only based on abstracts especially when full reports are not accessible.4 19 20 Even though there is increasing evidence on incomplete and inconsistent reporting in different fields of biomedicine and for different guidelines, there is no overarching summary of the evidence with regards to (1) the state-of-the-art extent of adherence to the emerging reporting guidelines in primary biomedical research, (2) the inconsistency between protocols or registrations and full reports or (3) the disagreement between abstracts and full-text articles. Therefore, we aim to conduct a scoping review to explore the current state of incomplete and inconsistent reporting in primary biomedical research and to investigate factors associated with improved completeness and consistency of reporting, based on evidence from systematic reviews and surveys. While the existing systematic reviews and surveys generally evaluate a specific research area or a group of journals or diseases with quantitative syntheses conducted, our scoping review will differ from them in mapping literature and addressing the state of reporting in the overall primary biomedical community, comprehensively summarising the heterogeneous evidence with a qualitative description reported, and assessing evidence gaps and providing recommendations for future research.21 22 Methods In this scoping review, we will use a systematic and comprehensive approach to retrieve all available and eligible systematic reviews and surveys in the literature.23 Our study will be conducted and reported based on the PRISMA guideline.24 However, no risk-of-bias assessment in individual studies or quantitative synthesis will be performed because they are not relevant to this scoping review. Our results will be presented in three parts including (1) current adherence to reporting guidelines; (2) inconsistency between protocols or registrations and full reports and (3) discrepancy between abstracts and full reports. The outline of this scoping review is shown in figure 1. We also provide a summary table for these three parts (table 1). For the first part, we will build on previous work on adherence to reporting guidelines which was limited to six guidelines for human studies and up to 2012.11 Our previous work will be expanded, updated and included in this scoping review. Figure 1 Flow diagram showing the outline of this scoping review. Study eligibility Systematic reviews that include primary studies and evaluate incomplete or inconsistent reporting with a focus on adherence to guidelines, comparison between protocols or registrations and full reports or consistency between abstracts and full reports, will be eligible. For the purposes of this review, an eligible systematic review will be defined as study with predetermined objectives, eligibility criteria, at least one electronic database searched, data extraction and at least one study included. All the surveys that include primary studies and focus on specific research questions in primary biomedical research will be eligible for inclusion in this scoping review. Adherence to reporting guidelines: We will include systematic review and surveys of the following guidelines: CONSORT, PRISMA, STROBE, STARD, ARRIVE, QUOROM (Quality of Reporting of Meta-analysis), TREND (Transparent Reporting of Evaluations with Non-randomized Designs), MOOSE (Meta-analysis Of Observational Studies in Epidemiology), CARE (Case Report), SRQR (Standards for Reporting Qualitative Research), COREQ (Consolidated criteria for Reporting Qualitative research), TRIPOD (Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis), SQUIRE (Standards for QUality Improvement Reporting Excellence), CHEERES (Consolidated Health Economic Evaluation Reporting Standards), SPIRIT (Standard Protocol Items: Recommendations for Interventional Trials) and REMARK (Reporting Recommendations for Tumour Marker Prognostic Studies). Systematic reviews that do not evaluate adherence to any of the aforementioned guidelines will not be included in our study. Consistency between protocols/registration and full reports: Systematic reviews or surveys from all fields of biomedical research will be eligible if they included a study objective of comparing protocols or registrations with full reports and provided data on such comparison. Agreement between abstracts and full reports: Systematic reviews or surveys from all fields of biomedical research will be eligible if they included a study objective of comparing abstracts with full reports and provided data on such comparison. Furthermore, to expand the extent of this scoping review, we will also include systematic reviews or surveys that specifically investigated the incomplete or inconsistent reporting for study subgroups for all the three parts above (figure 1). Exclusion criteria For all the three parts, the systematic reviews or surveys will be excluded if (a) their objectives are not incomplete or inconsistent reporting, (b) they do not focus on primary biomedical research studies, (c) they only publish editorials, abstracts, letters or commentaries without full-length texts, (d) they are duplicates of the included systematic reviews or surveys, or (e) they do not provide data on incomplete or inconsistent reporting. Search strategy We will search the electronic databases including Web of Science, Excerpta Medica Database (EMBASE), MEDLINE and Cumulative Index to Nursing and Allied Health Literature (CINAHL), for relevant studies. The search will be limited between January 1996 and 30 September 2016 given that the CONSORT (Consolidated Standards of Reporting Trials) statement was the first reporting guideline in biomedical research and developed in 1996.25 The search strategy will be designed with the assistance of an experienced librarian. Key descriptors that include terms for systematic reviews or surveys, reporting, and guidelines or adherence or inconsistency or registrations or protocols or abstracts will be used for the search, for instance, (Systematic reviews OR surveys OR reviews) AND (quality of reporting OR completeness of reporting OR selective reporting OR consistency of reporting OR biased reporting OR subgroup) AND ((QUOROM OR TREND OR MOOSE OR CONSORT OR STROBE OR PRISMA OR CARE OR SRQR OR COREQ OR STARD OR TRIPOD OR SQUIRE OR CHEERES OR ARRIVE OR SPIRIT OR REMARK) OR (Adherence OR Consistency OR Protocol OR Registration OR Abstract)). Online supplemental table S1 shows the detailed search terms used in this scoping review. 10.1136/bmjopen-2016-014749.supp1supplemental table Ovid search terms modified for MEDLINE, EMBASE, CINAHL and Web of Science (from January 1996 and September 30th 2016) Study selection Titles and abstracts retrieved will be first screened for eligibility before full texts are thoroughly examined. Reasons will be documented for excluded studies when assessing full texts. All the reference lists from the included systematic reviews or surveys will be also reviewed to retrieve additional relevant studies. We will limit the search to English language because of the lack of resources for translation of other languages. All the search processes will be performed by two reviewers (YJ and IN) independently. Disagreement will be addressed by consensus after discussion, and a third reviewer (GL) will be consulted if no consensus is reached. The kappa statistic will be used to quantify the level of agreement previous to their consensus between the two reviewers (YJ and IN).26 Outcomes In this scoping review, our primary outcomes include levels of adherence to reporting guidelines, levels of consistency between protocols or registrations and full reports and the agreement between abstracts and full reports, all of which are expressed as percentages, quality scores or categorised rating (such as high, medium, low). Specifically, for the first part, incomplete reporting will be assessed by the levels of adherence to reporting guidelines and their checklists when available, for example, for the CONSORT guideline, the percentage of adopting the guideline and the rates/scores of adhering to the components (title and abstract, introduction, methods, results, discussion and other information) among the included primary studies in the systematic review or survey will be our outcomes of interest. Levels of consistency between protocols or registrations and full reports and between abstracts and full reports will be evaluated by the agreement on the study-validity-related factors including research questions, study designs, study samples, interventions or exposures, outcome measures, time duration, comparators, statistical plan, result presentations and interpretations and conclusions or recommendations. For instance, some studies may investigate the changes in the study-validity-related factors from the prespecified protocols that are identified in full reports; the percentages of such changes will be our outcomes collected. Our secondary outcomes are the factors associated with improved completeness and consistency of reporting as reported from the included systematic reviews and surveys. Data collection Two reviewers (YJ and IN) will independently collect data from the included systematic reviews or surveys using data extraction forms. The data extraction forms will be piloted and modified before its final version to be used. Specifically, we will extract the data as shown below: basic characteristics: authors, publication year, journal in which the study is published, field of study, study region, number of primary studies included, number of study samples (including animals and participants) and reporting guideline (or its extension or modification) assessed in the systematic review or survey; for the adherence to guidelines, we will gather the reported adherence to the items specified in the corresponding guideline; for the consistency between protocols or registrations and full reports, and the agreement between abstracts and full reports, data extracted include (dis)concordance for research question, study population or sample size, intervention (or exposure), comparator, outcome, time duration, study design, statistical plan, result presentations and interpretations, conclusion and other information specifically evaluated in the systematic review or survey; outcome measures presented as levels of adherence to reporting guidelines or levels of consistency will be collected for all the relevant items if provided; factors that are found to be related to improved completeness and consistency of reporting in the individual systematic reviews or survey; authors' overall conclusion in the systematic review or survey. Any disagreement will be resolved by the two reviewers' discussion and consensus. In addition, we will contact the authors of included systematic reviews to collect essential and relevant data if necessary. Data analysis The levels of adherence to guidelines and the levels of consistency will be described using medians and IQRs across all the included studies. The general characteristics of included studies, levels of adherence to reporting guidelines or levels of consistency between protocols or registrations and full reports and between abstracts and full reports, factors related to improved completeness and consistency of reporting and conclusions in the included studies will be summarised and discussed in our review. No pooled analyses or quantitative syntheses will be performed given the heterogeneity of the included systematic reviews and surveys. Likewise, factors associated with improved completeness and consistency of reporting will be summarised qualitatively. Quality assessment of included studies We will evaluate the quality of all the included systematic reviews, using the AMSTAR (a measurement tool to assess systematic reviews) criteria.27 The R(evised)-AMSTAR will not be used in our study, given its limited application and unknown measurement properties.28 However, some items of AMSTAR may not be applicable to all the included systematic reviews. For instance, the item 9 ‘were the methods used to combine the findings of studies appropriate’ (because not all the systematic reviews used a pooled estimate) is not relevant to some included studies, thereby being omitted from the quality evaluation. Likewise, we will not assess quality of the included surveys due to lack of relevant assessment tools or guidelines. Discussion Incomplete or inconsistent reporting remains a major concern in the biomedical literature including preclinical studies, diagnostic research, qualitative studies, economic studies, clinical trials and observational studies, among others.2 4 When the reporting is incomplete or inconsistent, the apparent methodological quality of published findings may not reveal the actual quality of the study as evaluated from the protocol or registration or abstracts, yielding the published findings unreliable, irreproducible or sometimes misleading.3 4 17 29 In this scoping review, we will assess the completeness of reporting in the literature and adherence to reporting guidelines, consistency between protocols or registrations and full reports and agreement between abstracts and full-text articles. We will present our results as three parts, where the first part of adherence to reporting guidelines is an updated and expanded research based on our previous work.11 In contrast, for the other parts of inconsistency between protocols or registrations and full reports and discrepancy between abstracts and full reports, no study summarising all the best current evidence in multidisciplines is available. Unlike the individual systematic review and survey that reports confirmatory point estimates in a specific area or disease or in a group of journals,13 16 17 30 our scoping review will show the general mapping for the state of reporting in the overall primary biomedical research. With the evidence gaps explored in this scoping review, findings may advance our understanding of the extent of incomplete and inconsistent reporting, factors related to improved completeness and consistency of reporting and potential recommendations for various stakeholders in the biomedical community. All findings will be published in peer-reviewed journals electronically and in print. Contributors: GL, LM, ZS and LT were responsible for the study conception and design. GL, YJ and LT were responsible for drafting the manuscript. LM, ZS, YJ, IN, MAHL and JDA made several revisions and provided professional support. All authors read and approved the final version of the manuscript. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Not applicable. We will use the data that are already published and publicly accessible. ==== Refs References 1 Glass GV Primary, secondary, and meta-analysis of research 1 . Educ Res 1976 ;5 :3 –8 . 10.3102/0013189X005010003 2 Dickersin K , Chalmers I Recognizing, investigating and dealing with incomplete and biased reporting of clinical research: from Francis Bacon to the WHO . J R Soc Med 2011 ;104 :532 –8 . 10.1258/jrsm.2011.11k042 22179297 3 Chalmers I , Glasziou P Avoidable waste in the production and reporting of research evidence . Lancet 2009 ;374 :86 –9 . 10.1016/S0140-6736(09)60329-9 19525005 4 Glasziou P , Altman DG , Bossuyt P Reducing waste from incomplete or unusable reports of biomedical research . Lancet 2014 ;383 :267 –76 . 10.1016/S0140-6736(13)62228-X 24411647 5 Simera I , Altman DG , Moher D Guidelines for reporting health research: the EQUATOR network's survey of guideline authors . PLoS Med 2008 ;5 :e139 10.1371/journal.pmed.0050139 18578566 6 Simera I , Altman DG Writing a research article that is “fit for purpose”: EQUATOR Network and reporting guidelines . Evid Based Med 2009 ;14 :132 –4 . 10.1136/ebm.14.5.132 19794009 7 Prady SL , Richmond SJ , Morton VM A systematic evaluation of the impact of STRICTA and CONSORT recommendations on quality of reporting for acupuncture trials . PLoS ONE 2008 ;3 :e1577 10.1371/journal.pone.0001577 18270568 8 Cobo E , Cortés J , Ribera JM , et al Effect of using reporting guidelines during peer review on quality of final manuscripts submitted to a biomedical journal: masked randomised trial . BMJ 2011 ;343 :d6783 10.1136/bmj.d6783 22108262 9 Tunis AS , McInnes MD , Hanna R Association of study quality with completeness of reporting: have completeness of reporting and quality of systematic reviews and meta-analyses in major radiology journals changed since publication of the PRISMA statement? Radiology 2013 ;269 (2):413 –26 . 10.1148/radiol.13130273 23824992 10 Smith BA , Lee HJ , Lee JH Quality of reporting randomized controlled trials (RCTs) in the nursing literature: application of the consolidated standards of reporting trials (CONSORT) . Nurs Outlook 2008 ;56 :31 –37. e33 . 10.1016/j.outlook.2007.09.002 18237622 11 Samaan Z , Mbuagbaw L , Kosa D A systematic scoping review of adherence to reporting guidelines in health care literature . J Multidiscip Healthc 2013 ;6 :169 –88 . 10.2147/JMDH.S43952 23671390 12 Turner L , Shamseer L , Altman DG Does use of the CONSORT statement impact the completeness of reporting of randomised controlled trials published in medical journals? A Cochrane review . Syst Rev 2012 ;1 :60 10.1186/2046-4053-1-60 23194585 13 Dwan K , Gamble C , Williamson PR , Systematic review of the empirical evidence of study publication bias and outcome reporting bias—an updated review . PLoS ONE 2013 ;8 :e66844 10.1371/journal.pone.0066844 23861749 14 Hopewell S , Clarke M , Askie L Reporting of trials presented in conference abstracts needs to be improved . J Clin Epidemiol 2006 ;59 :681 –4 . 10.1016/j.jclinepi.2005.09.016 16765270 15 Pitkin RM , Branagan MA , Burmeister LF Accuracy of data in abstracts of published research articles . JAMA 1999 ;281 :1110 –11 . 10.1001/jama.281.12.1110 10188662 16 Dwan K , Altman DG , Clarke M Evidence for the selective reporting of analyses and discrepancies in clinical trials: a systematic review of cohort studies of clinical trials . PLoS Med 2014 ;11 :e1001666 10.1371/journal.pmed.1001666 24959719 17 Chan AW , Hróbjartsson A , Haahr MT Empirical evidence for selective reporting of outcomes in randomized trials: comparison of protocols to published articles . JAMA 2004 ;291 :2457 –65 . 10.1001/jama.291.20.2457 15161896 18 Chan AW , Krleža-Jerić K , Schmid I Outcome reporting bias in randomized trials funded by the Canadian Institutes of Health Research . CMAJ 2004 ;171 :735 –40 . 10.1503/cmaj.1041086 15451835 19 Ochodo EA , de Haan MC , Reitsma JB Overinterpretation and misreporting of diagnostic accuracy studies: evidence of “spin” . Radiology 2013 ;267 :581 –8 . 10.1148/radiol.12120527 23360738 20 Hopewell S , Clarke M , Moher D CONSORT for reporting randomized controlled trials in journal and conference abstracts: explanation and elaboration . PLoS Med 2008 ;5 :e20 10.1371/journal.pmed.0050020 18215107 21 Peters MD , Godfrey CM , Khalil H Guidance for conducting systematic scoping reviews . Int J Evid Based Healthc 2015 ;13 :141 –6 . 10.1097/XEB.0000000000000050 26134548 22 Tricco AC , Lillie E , Zarin W A scoping review on the conduct and reporting of scoping reviews . BMC Med Res Methodol 2016 ;16 :15 10.1186/s12874-016-0116-4 26857112 23 Brien SE , Lorenzetti DL , Lewis S Overview of a formal scoping review on health system report cards . Implement Sci 2010 ;5 :2 10.1186/1748-5908-5-2 20205791 24 Moher D , Liberati A , Tetzlaff J Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement . BMJ 2009 ;339 :b2535 10.1136/bmj.b2535 19622551 25 Begg C , Cho M , Eastwood S Improving the quality of reporting of randomized controlled trials. The CONSORT statement . JAMA 1996 ;276 :637 –9 . 10.1001/jama.1996.03540080059030 8773637 26 Viera AJ , Garrett JM Understanding interobserver agreement: the kappa statistic . Fam Med 2005 ;37 :360 –3 .15883903 27 Shea BJ , Grimshaw JM , Wells GA Development of AMSTAR: a measurement tool to assess the methodological quality of systematic reviews . BMC Med Res Methodol 2007 ;7 :10 10.1186/1471-2288-7-10 17302989 28 Pieper D , Buechter RB , Li L Systematic review found AMSTAR, but not R(evised)-AMSTAR, to have good measurement properties . J Clin Epidemiol 2015 ;68 :574 –83 . 10.1016/j.jclinepi.2014.12.009 25638457 29 Mhaskar R , Djulbegovic B , Magazin A Published methodological quality of randomized controlled trials does not reflect the actual quality assessed in protocols . J Clin Epidemiol 2012 ;65 :602 –9 . 10.1016/j.jclinepi.2011.10.016 22424985 30 Bhandari M , Devereaux PJ , Guyatt GH An observational study of orthopaedic abstracts and subsequent full-text publications . J Bone Joint Surg Ame 2002 ;84-A (4 ):615 –21 .

PMC005xxxxxx/PMC5372138.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01502610.1136/bmjopen-2016-015026Nutrition and MetabolismResearch150617141843Cross-sectional associations between dietary intake and carotid intima media thickness in type 2 diabetes: baseline data from a randomised trial Chiavaroli Laura 12Mirrahimi Arash 23Ireland Christopher 12Mitchell Sandra 12Sahye-Pudaruth Sandhya 12Coveney Judy 2Olowoyeye Omodele 45Patel Darshna 12de Souza Russell J 267Augustin Livia S A 28Bashyam Balachandran 12Pichika Sathish Chandra 2Blanco Mejia Sonia 12Nishi Stephanie K 12Leiter Lawrence A 1279Josse Robert G 1279McKeown-Eyssen Gail E 110Moody Alan R 45Kendall Cyril W C 1211Sievenpiper John L 1279Jenkins David J A 1279 1 Department of Nutritional Sciences, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada 2 Clinical Nutrition and Risk Factor Modification Centre, St. Michael's Hospital, Toronto, Ontario, Canada 3 Faculty of Health Sciences, School of Medicine, Queen's University, Kingston, Ontario, Canada 4 Department of Medical Imaging, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada 5 Institute of Medical Science, University of Toronto, Toronto, Ontario, Canada 6 Department of Health Research Methods, Evidence & Impact, Faculty of Health Sciences, McMaster University, Hamilton, Ontario, Canada 7 Li Ka Shing Knowledge Institute, St. Michael's Hospital, Toronto, Ontario, Canada 8 National Cancer Institute ‘Fondazione G. Pascale’, Naples, Italy 9 Division of Endocrinology and Metabolism, St. Michael's Hospital, Toronto, Ontario, Canada 10 Dalla Lana School of Public Health, University of Toronto, Toronto, Ontario, Canada 11 College of Pharmacy and Nutrition, University of Saskatchewan, Saskatoon, Saskatchewan, CanadaCorrespondence to Dr David J A Jenkins; nutritionproject@smh.caA version of the abstract was published in the 76th Scientific Sessions Abstract Book and the June 2016 supplement to the journal Diabetes. 2017 22 3 2017 7 3 e0150263 11 2016 16 2 2017 27 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective To assess associations between dietary intake and carotid intima media thickness (CIMT) by carotid ultrasound (CUS), a surrogate marker of cardiovascular disease (CVD) risk, in those with type 2 diabetes. Design Cross-sectional analysis of baseline data from 325 participants from three randomised controlled trials collected in the same way. Setting Risk Factor Modification Centre, St. Michael's Hospital, Toronto, Canada. Participants 325 participants with type 2 diabetes, taking oral antidiabetic agents, with an HbA1c between 6.5% and 8.0% at screening, without a recent cardiovascular event. Main outcome measures CIMT by CUS and associations with dietary intake from 7-day food records, as well as anthropometric measures and fasting serum samples. Results CIMT was significantly inversely associated with dietary pulse intake (β=−0.019, p=0.009), available carbohydrate (β=−0.004, p=0.008), glycaemic load (β=−0.001, p=0.007) and starch (β=−0.126, p=0.010), and directly associated with total (β=0.004, p=0.028) and saturated (β=0.012, p=0.006) fat intake in multivariate regression models adjusted for age, smoking, previous CVD event, blood pressure medication, antidiabetic medication and ultrasonographer. Conclusions Lower CIMT was significantly associated with greater consumption of dietary pulses and carbohydrates and lower total and saturated fat intake, suggesting a potential role for diet in CVD risk management in type 2 diabetes. Randomised controlled trials are anticipated to explore these associations further. Trial registration number NCT01063374. NUTRITION & DIETETICSDIABETES & ENDOCRINOLOGY ==== Body Strengths and limitations of this study A strength of the study methods includes the use of carotid intima media thickness (CIMT) by carotid ultrasound, which is the recommended screening tool for assessing cardiovascular disease (CVD) risk by some CVD prevention clinical practice guidelines, considered a biomarker of atherosclerosis and is associated with overall CVD risk, particularly in those with type 2 diabetes. The CIMT scans were all performed by one of two highly trained ultrasonographers using the same scanner at the same site and using the same reading protocol. The dietary data are obtained from 7-day dietary food records in which participants were trained on how to complete by study dietitians and which at each visit were reviewed in detail by study dietitians in the presence of the participant. A limitation of the analyses includes that they are conducted using CIMT results from only one CIMT scan obtained at baseline; however, studies have demonstrated mean max CIMT measurements to have good reproducibility, particularly using the method we adopted where measurements are made of 12 segments of the carotid artery. Although at higher CVD risk, participants in this study had relatively well-controlled diabetes, blood pressure and low-density lipoprotein cholesterol, with the majority on blood pressure and cholesterol-lowering medication at baseline, therefore possibly limiting the ability to assess associations between risk factors for CVD and CIMT and dietary intake, as well as limiting application to those with uncontrolled risk factors. Introduction People with type 2 diabetes are at high risk of cardiovascular disease (CVD), the leading cause of death in this population.1 2 Risk assessment through atherosclerosis imaging includes the use of carotid intima media thickness (CIMT) by carotid ultrasound (CUS), the recommended screening tool for assessing CVD risk by some CVD prevention clinical practice guidelines.3 4 CIMT is considered a biomarker of atherosclerosis and is associated with overall CVD risk, particularly in those with type 2 diabetes.5 6 Observational studies have recently demonstrated that carotid atherosclerosis, assessed by CIMT, is associated with glycaemic status7–9 and intervention trials with antidiabetic agents have demonstrated reductions in CIMT both in those with and without diabetes.10–12 Some of these antidiabetic agents exert their effect postprandially, by reducing the postprandial blood glucose peak. Trials of insulin secretagogues (nateglinide and repaglinide) and the α-glucosidase inhibitor, acarbose, have demonstrated reductions in CIMT and identified markers of glycaemia (HbA1c and glucose peak) as determinants of changes in CIMT.10 13–15 In addition to antidiabetic drugs, dietary strategies continue to be sought as means to assist in diabetes management. However, few dietary intervention trials have explored the effect on CIMT. These have demonstrated regression of CIMT with a Mediterranean diet pattern.16 Dietary strategies that may be useful for diabetes management, including low-GI (low-glycaemic index) diets17 18 and those rich in dietary pulses,19 20 have demonstrated associations with improved CVD risk,21 thus warrant exploration of their effects on CIMT as a subclinical biomarker of CVD. The objective of the present study was to determine the associations between dietary intake variables, particularly GI, and risk of CVD assessed by CIMT in participants with type 2 diabetes. Methods Participants Details of the study protocol have been previously published.22 Participants recruited for a 3-year dietary intervention study had a diagnosis of type 2 diabetes >6 months prior to the start of the study, an HbA1c between 6.5% and 8.0% at screening, were on oral antidiabetic agents at a stable dose for ≥8 weeks, not on insulin, without gastrointestinal disease, clinically significant liver disease or history of cancer, except non-melanoma skin cancer, and had not had a major cardiovascular event or major surgery in the past 6 months. Participants also had a CUS scan at the Medical Imaging Department at Sunnybrook Health Sciences Centre to assess CIMT as part of screening criteria, where only those participants with a maximum CIMT ≥1.2 mm were eligible for the 3-year study. Those who did not meet this CIMT cut-point had the option to participate in one of two concurrent trials of shorter duration, the details of which have been published.23 24 This cross-sectional study was conducted on 325 study participants using baseline data from these three trials, all of which had the same inclusion criteria (with the exception of the additional criteria of a maximum CIMT ≥1.2 mm for the 3-year trial) and included 7-day food records, anthropometric measures, fasting blood samples and CIMT measures which were collected in the same way. Written consent was obtained from all participants. Design In this cross-sectional analysis, data were obtained from baseline measures of study participants including CIMT data obtained from CUS scans completed at the Medical Imaging Department at the Sunnybrook Health Sciences Centre MRI research unit. One of two highly trained and certified sonographers performed CIMT measures using a Philips iU22 Ultrasound system (Philips Healthcare, Andover, Massachusetts, USA) with standardised CUS scanning and reading protocols.25 26 CIMT was measured with the subject recumbent, with the neck extended and rotated away from the side of interest. As previously described,22 imaging was performed on the right and left carotid arteries, with identification of the near wall (closest to the skin surface) and the far wall (farthest from the skin surface) of three arterial segments: the proximal 8 mm of the internal carotid artery, the carotid bifurcation beginning at the tip of the flow divider (site of the division of flow between the external carotid artery and internal carotid artery) and extending 8 mm proximally, and the common carotid artery 8–16 mm proximal to the flow divider, and measurements included posterior wall plaque, if present. The mean value of the 12 maximal CIMT measurements (mean maximum CIMT) was used as the outcome measure, which has demonstrated good reproducibility (interclass correlation coefficient (ICC) >0.8).27 For participant baseline data on risk factors, all three studies included two baseline clinic visits at the Risk Factor Modification Centre, St. Michael's Hospital and occurred within 1 month of each other. At the first, baseline anthropometric and fasting blood measures were obtained. Participants were given detailed instruction on how to complete a 7-day food record which was returned at the next visit. At the second baseline clinic visit, anthropometric and fasting blood measures were again obtained and each participant was randomised. Anthropometric data included body weight, seated blood pressure measured as the mean of triplicate measures made with an automatic sphygmomanometer (Omron HEM 907 XL, OMRON Healthcare, Burlington, Ontario, Canada), and waist (at the umbilicus, 2 inches above and lying down) and hip circumference. Blood measures included HbA1c, fasting glucose and fasting lipids. Data were also obtained on demographics including age, sex, estimated duration of diabetes, smoking and medication use. For those who participated in more than one of the three studies (n=14), their first CUS and corresponding baseline study measurements were used in the present analyses. Biochemical and dietary analyses The HbA1c value was analysed within 24 hours using whole blood collected in EDTA Vacutainer tubes (Vacutainer; Becton, Dickinson and Co) in the hospital routine analytical laboratory by a turbidimetric inhibition latex immunoassay (TINIA Roche Diagnostics) with a coefficient of variation between assays of 3–4%. Blood glucose and serum lipid levels were also measured in the hospital routine analytical laboratory using a Random Access Analyzer and Beckman reagents (SYNCHRON LX Systems; Beckman Coulter), with a coefficient of variation of 1.6–2.3% for blood glucose level and 1.3–3.0% for total cholesterol (total-C), triglycerides, and high-density lipoprotein cholesterol (HDL-C) levels. The low-density lipoprotein cholesterol (LDL-C) level was calculated by the method of Friedewald et al28 (LDL-C level=total-C−[(triglycerides/5)×(HDL-C level)]). Dietary assessments using participant completed 7-day food records were analysed using a computer program (ESHA Food Processor SQL V.10.9; ESHA, Salem, Oregon, USA) based on the USDA database,29 supplemented with data from the Canada Nutrient File,30 and with GI values from international GI tables31 using the bread scale (where bread=100; for the glucose scale, multiply by 0.71). Glycaemic load (GL) was calculated as GI*available carbohydrate÷100. Product data were updated with manufacturers' nutrient information and relevant foods were analysed by Covance Laboratories (3301 Kinsman Blvd, Madison Wiscosin, USA). Owing to interest in GI, in addition to general dietary variables, data on particularly low-GI foods, which we and others had previously demonstrated to have benefit in diabetes management,23 32 33 were extracted from the food records, including dietary pulses, temperate climate fruit and nut intake. Statistical analyses These are post hoc analyses performed on all baseline study participants who had a CUS scan (n=325). The original power calculation was based on the main intervention trial. However, we performed post hoc sample size calculations to assess our ability to detect associations of GI and CIMT. Given the slope of −0.210 for GI and CIMT, the SD of our log transformed GI of 0.075, the SD of our Box-Cox transformed CIMT of 0.229, and with 80% power and α=0.05, to detect an association, we would need a sample size of 1640. This may be a reflection of our lack of range of exposure levels of GI at baseline (range: 58–98 and IQR: 76–83, bread scale). We have also calculated a post hoc sample size for dietary pulses, as a particularly low-GI food which we previously demonstrated to have benefit on CVD risk factors in diabetes,23 given a slope of −0.025 for dietary pulses and CIMT, and an SD of our log transformed dietary pulses of 1.592, 254 participants would have been sufficient to detect an association at α=0.05, 1−β=0.80. Data are expressed as means±SD unless otherwise indicated. Multivariate mixed-effects regression models were conducted using SAS software, V.9.4 (SAS Institute: SAS/STAT Proprietary Software 9.4. Cary, NC: SAS Institute; 2002–2012.) to assess the association between dietary intake and CIMT. Dietary variables were energy adjusted by expressing intake as g/1000 kcal or as a percentage of energy. To adjust for energy in GI and GL analyses, total caloric intake was added to the models as a continuous variable. All dietary variables were analysed as continuous variables. Sensitivity analyses were performed where energy was alternatively adjusted for using the residual method. Since there are multiple potential confounders with CIMT, including age, sex, smoking, prior CVD event, cholesterol medication use, blood pressure medication use, duration of diabetes, type of antidiabetic medication and waist circumference, we included only those that were significantly associated with CIMT in our data set (at p<0.1). Therefore, the multivariate model was adjusted for age, smoking, previous CVD event, blood pressure medication and type of antidiabetic medication. Smoking was defined using three categories: current (current smoker or quit within the past year), former (quit between 1 and 15 years ago), and non-smoker (never smoked, or quit over 15 years ago) according to the WHO definitions of coronary heart disease (CHD) risk.34 Previous CVD event was defined as yes or no. Use of blood pressure-lowering medication was defined as user or non-user based on the baseline visits. Type of antidiabetic medication was defined as user of a sulfonylurea or thiazolidinediones or non-user since these have been demonstrated to have potential negative associations with CVD risk.13 35 36 Of the 325 participants with a CIMT measurement and who were randomised into one of the three studies, one participant did not attend their week 0 visit, thus did not have dietary data. Overall 10-year CVD risk was calculated using the Framingham risk score according to the 2008 Framingham cardiovascular risk equation.37 Ultrasonographer, another potential confounder, was treated as a random effect throughout analyses. CIMT was non-normally distributed, therefore was transformed using the Box-Cox transformation.38 The λ for CIMT was −1, thus, (CIMT−1−1)/−1 was used for the transformation. Dietary variables were transformed using the natural logarithm in models where the model fit improved (adjusted R-squared increased) in the adjusted model when being regressed against transformed CIMT. The transformed dietary variables include: GI, dietary pulses, temperate climate fruit, dietary fibre, viscous fibre, cereal fibre, starch, monounsaturated fatty acids, omega 3 fatty acids, dietary cholesterol and vegetable protein. Thus, these dietary variables were transformed as ln[g/1000 kcal (or %kcal) +1]. Probability values <0.05 were considered statistically significant. Results Characteristics of the 325 study participants are presented in table 1. The average mean maximum CIMT was 1.0±0.3 mm and maximum CIMT was 2.0±0.9 mm. The mean age for all participants was 60.3±8.7 years, 56% were male, the mean body mass index (BMI) was 30.3±5.7 kg/m2, and the mean waist circumference was 105.5±15.0 cm in females and 104.1±12.3 cm in males. Fifteen participants (4.6%) had a previous CVD event, 84.6% non-smokers, 10.2% former smokers and 5.2% were current smokers. Of the participants, 72.3% were taking cholesterol-lowering and 66.5% blood pressure-lowering medications. Table 1 Participants characteristics Participant characteristics (n=325) Mean±SD Mean max bilateral CIMT, mm 1.0±0.3 Max CIMT, mm 2.0±0.9 Age, y 60.3±8.7 Sex, female/male 142/183 Estimated diabetes duration, y 8.2±6.1 Body weight, kg 84.3±18.2 BMI, kg/m2 30.3±5.7 Waist circumference, cm* 104.7±13.5 Waist:hip ratio 1.0±0.1 Systolic blood pressure, mm Hg 122.2±11.2 Diastolic blood pressure, mm Hg 71.6±8.2 Mean arterial pressure, mm Hg 88.5±8.2 Pulse, bpm 71.9±9.6 Pulse pressure, mm Hg 50.8±10.0 Fasting glucose, mmol/L 7.5±1.5 HbA1c, % 7.1±0.5 Total-C mmol/L 4.0±1.0 HDL-C mmol/L 1.2±0.3 LDL-C mmol/L 2.2±0.8 Serum triglycerides, mmol/L 1.5±0.9 Total:HDL-C ratio 3.6±1.0 Non-HDL-C, mmol/L 2.9±0.9 CVD risk, FRS 19.5±11.3 Previous CVD event, % 4.6 Smoking  Non-smoker, % 84.6  Former, % 10.2  Current, % 5.2 Diabetes meds—sulfonylurea or TZD % 37.5 Cholesterol meds, % 72.3 Blood pressure meds, % 66.5 Ethnicity, %  African 6.8  European 32.3  Far Eastern 8.9  Hispanic 1.9  Indian/South Asian 22.8  Other Caucasian 19.7  Other 20.0 *Measured at the umbilicus. BMI, body mass index; bmp, beats per minute; CIMT, carotid intima media thickness; CVD, cardiovascular disease; FRS, Framingham Risk Score; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; meds, medication use; Total-C, total cholesterol; TZD, thiazolidinediones; y, years. CIMT and baseline dietary intake The associations between CIMT and dietary variables using multivariate regression models are presented in table 2. CIMT was significantly inversely associated with GL (β=−0.001, p=0.007), dietary pulses (β=−0.019, p=0.009), available carbohydrate (β=−0.004, p=0.008) and starch (β=−0.126, p=0.010), and positively associated with total fat (β=0.004, p=0.028) and saturated fat (β=0.012, p=0.006), in multivariate models. Sensitivity analyses using the residual model for energy adjustment, revealed consistent results (data not shown). Table 2 Dietary intake and associations with carotid intima media thickness Dietary intake* and association with CIMT (n=324) Mean±SD Unadjusted† Age adjusted† Multivariate‡ β SE p Value β SE p Value β SE p Value GI§¶** 79.0±5.9 −0.210 0.172 0.222 −0.084 0.159 0.597 −0.138 0.157 0.381 GL**†† 149.9±46.8 −0.001 0.000 0.010 −0.001 0.000 0.015 −0.001 0.000 0.007 Dietary pulses, g/1000 kcal§ 18.7±29.9 −0.025 0.008 0.002 −0.019 0.008 0.014 −0.019 0.007 0.009 Nuts, g/1000 kcal 7.0±9.0 −0.001 0.001 0.507 −0.001 0.001 0.334 −0.001 0.001 0.313 Temperate climate fruit, g/1000 kcal§ 45.9±42.0 0.015 0.009 0.101 0.004 0.009 0.643 0.004 0.008 0.638 Available carbohydrates, % 42.7±7.1 −0.004 0.002 0.041 −0.004 0.002 0.010 −0.004 0.002 0.008  Fibre, g/1000 kcal§ 15.0±5.2 −0.008 0.042 0.853 −0.060 0.039 0.124 −0.048 0.038 0.211  Viscous fibre, g/1000 kcal§ 0.4±0.8 0.021 0.038 0.587 −0.012 0.035 0.725 −0.015 0.034 0.668  Cereal fibre, g/1000 kcal§ 3.6±3.0 0.016 0.020 0.430 0.004 0.018 0.835 0.011 0.018 0.540  Starch, %§ 28.1±6.6 −0.154 0.054 0.004 −0.114 0.050 0.025 −0.126 0.049 0.010  Sugar, % 14.6±4.8 0.002 0.003 0.389 −0.002 0.003 0.386 −0.002 0.003 0.521 Total fat, % 32.8±6.3 0.003 0.002 0.165 0.004 0.002 0.046 0.004 0.002 0.028  SFA, % 10.0±2.7 0.012 0.005 0.013 0.012 0.004 0.007 0.012 0.004 0.006  MUFA, %§ 12.9±3.4 0.037 0.053 0.487 0.057 0.052 0.273 0.059 0.048 0.223  PUFA, % 6.8±2.0 0.002 0.006 0.761 0.005 0.006 0.427 0.006 0.006 0.297  n-3, %§ 0.9±0.5 0.039 0.058 0.500 0.007 0.053 0.901 0.022 0.053 0.679  n-6, % 5.7±1.7 0.004 0.007 0.631 0.007 0.007 0.292 0.008 0.007 0.210 Diet cholesterol, mg/1000 kcal§ 139.9±57.1 0.026 0.021 0.234 0.022 0.020 0.255 0.020 0.019 0.310 Protein, % 18.9±3.2 0.005 0.004 0.182 0.005 0.004 0.187 0.005 0.004 0.174  Vegetable protein, %§ 7.4±1.9 −0.071 0.059 0.230 −0.089 0.055 0.103 −0.096 0.053 0.072  Animal protein, % 2.4±2.2 0.005 0.003 0.104 0.005 0.003 0.071 0.006 0.003 0.058 Energy, kcal 1781.3±451.3 0.000 0.000 0.112 0.000 0.000 0.671 0.000 0.000 0.662 Sodium, mg/1000 kcal 1385.2±336.2 0.000 0.000 0.098 0.000 0.000 0.087 0.000 0.000 0.191 Alcohol, % 1.8±3.2 0.003 0.004 0.395 0.003 0.004 0.417 0.003 0.004 0.488 Bold values are <0.05 and thus statistically significant. *Percentages represent the percentage of total calories. †Regression models assessing the association with CIMT where all dietary variables are analysed as continuous variables. ‡Adjusted for age, smoking, previous CVD event, blood pressure medication, antidiabetic medication and ultrasonographer. §Log transformed where the model fit improved when regressed against CIMT. ¶GI bread scale (to convert to glucose scale, multiply by 0.71); low GI ≤78, medium GI 78–99, high GI ≥100. **Total caloric intake added to full model as a continuous variable to adjust for energy. ††GI multiplied by the mean total available carbohydrate intake per day divided by 100. CIMT, carotid intima media thickness (mean max bilateral, average of the 12 segment measures); CVD, cardiovascular disease; GI, glycaemic index; GL, glycaemic load; MUFA, monounsaturated fatty acids; n-3, omega 3 fatty acids; n-6, omega 6 fatty acids; PUFA, polyunsaturated fatty acids; SFA, saturated fatty acids. CIMT and risk factors CIMT was significantly positively associated with age (β=0.011, p<0.001; unadjusted), waist-to-hip ratio (β=0.529, p=0.019), systolic (β=0.004, p<0.001) and diastolic blood pressure (β=0.003, p=0.035), mean arterial pressure (β=0.005, p=0.002), pulse pressure (β=0.005, p<0.001), total:HDL-C ratio (β=0.027, p=0.020) and FRS (β=0.005, p<0.001), and inversely with pulse (as beats per minute) (β=−0.002, p=0.047) (table 3) in multivariate models. Non-smoking was associated with significantly lower CIMT when compared to current smokers (β=−0.160, p=0.003) and to former smokers (β=−0.094, p=0.004) (table 3). Table 3 Participant characteristics and associations with carotid intima media thickness Participant characteristics (n=325) Unadjusted* Age adjusted* Multivariate† β SE p Value β SE p Value β SE p Value Age, y 0.011 0.001 <0.001   Sex (male vs female) 0.029 0.026 0.254 0.035 0.024 0.146 0.029 0.023 0.207 Estimated diabetes duration, y 0.002 0.002 0.322 −0.003 0.002 0.137 −0.003 0.002 0.117 Body weight, kg 0.000 0.001 0.960 0.001 0.001 0.088 0.001 0.001 0.155 BMI, kg/m2 −0.002 0.002 0.369 0.002 0.002 0.440 0.001 0.002 0.630 Waist circumference, cm‡ 0.000 0.001 0.797 0.001 0.001 0.531 0.000 0.001 0.816 Waist:hip ratio 0.604 0.244 0.014 0.608 0.223 0.007 0.529 0.225 0.019 Systolic blood pressure, mm Hg 0.005 0.001 <0.001 0.004 0.001 <0.001 0.004 0.001 <0.001 Diastolic blood pressure, mm Hg −0.002 0.002 0.323 0.002 0.002 0.165 0.003 0.001 0.035 Mean arterial pressure, mm Hg 0.002 0.002 0.123 0.004 0.001 0.006 0.005 0.001 0.002 Pulse, bpm −0.004 0.001 0.003 −0.002 0.001 0.043 −0.002 0.001 0.047 Pulse pressure, mm Hg 0.009 0.001 <0.001 0.005 0.001 <0.001 0.005 0.001 <0.001 Fasting glucose, mmol/L 0.001 0.009 0.934 0.006 0.008 0.488 0.004 0.008 0.573 HbA1c, % −0.021 0.025 0.414 −0.001 0.023 0.954 −0.003 0.023 0.905 Total-C mmol/L 0.012 0.013 0.357 0.013 0.012 0.248 0.017 0.012 0.155 HDL-C mmol/L 0.018 0.045 0.687 −0.065 0.042 0.126 −0.057 0.041 0.165 LDL-C mmol/L 0.015 0.015 0.337 0.019 0.014 0.185 0.021 0.014 0.136 Serum triglycerides, mmol/L 0.001 0.015 0.956 0.017 0.014 0.205 0.020 0.013 0.140 Total:HDL-C ratio 0.004 0.012 0.760 0.026 0.012 0.024 0.027 0.011 0.020 Non-HDL-C, mmol/L 0.012 0.014 0.387 0.021 0.012 0.088 0.025 0.013 0.052 CVD risk, FRS 0.008 0.001 <0.001 0.005 0.001 <0.001 0.005 0.001 <0.001 Previous CVD event −0.161 0.060 0.008 −0.104 0.056 0.064 Non-smokers vs current smokers −0.089 0.057 0.118 −0.160 0.053 0.003 Non-smokers vs ever smokers −0.072 0.035 0.042 −0.094 0.032 0.004 Not vs on sulfonylurea or TZD −0.043 0.026 0.105 −0.025 0.024 0.302 Not vs on cholesterol meds −0.041 0.028 0.152 −0.022 0.026 0.407 −0.022 0.026 0.416 Not vs on blood pressure meds −0.050 0.027 0.065 −0.013 0.025 0.592 Bold values are <0.05 and thus statistically significant. *Regression models assessing the association with CIMT, where p<0.05 is considered statistically significant. †Adjusted for age, smoking, previous CVD event, blood pressure medication, antidiabetic medication and ultrasonographer. ‡Measured at the umbilicus. BMI, body mass index; bmp, beats per minute; CIMT, carotid intima media thickness (mean max bilateral, average of the 12 segment measures); CVD, cardiovascular disease; FRS, Framingham Risk Score; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; meds, medication; Total-C, total cholesterol; TZD, thiazolidinediones; y, years. Post hoc explorations with CIMT Further explorations were conducted using the multivariate model to predict how CIMT in a person consuming one serving of dietary pulses per day would compare to a person not consuming any. Data points were obtained by taking mean values for log dietary pulse intake in the regression equation (at 0 and 0.5 increments up to 5) and then back transforming the β estimates for the response variable (transformed CIMT). The predicted model of the association between CIMT and dietary pulse intake revealed a logarithmic association (figure 1) where approximately one ¾ cup Canadian serving/day (∼132 g/day)39 was associated with a 7.5% lower CIMT compared with no intake (0 g/day) (0.078 mm CIMT difference). The same was conducted for saturated fat which demonstrated a linear association where for every 1% of total calorie increase in saturated fat, CIMT was about 0.011 mm greater (see online supplementary figure S1). Figure 1 Model of association between dietary pulse intake and carotid intima media thickness (CIMT). The multivariate regression model is adjusted for age, smoking, previous cardiovascular disease (CVD) event, blood pressure medication, antidiabetic medication and ultrasonographer. 10.1136/bmjopen-2016-015026.supp1supplementary data To explore the association with starch further, each major source of starch in the diet was assessed, including potato, pasta, rice, bread and pulses. Grams of starch from each carbohydrate source were calculated using the following foods: boiled white potato, cooked macaroni for pasta, long grain rice, white and whole wheat bread, and the average starch from three beans (chickpea, black bean and kidney bean) and lentils. Starch from each source was expressed as a percentage of total calories (%kcal) for each variable in the model. All other sources were pooled (other starch). All starch sources were added to the adjusted model and then removed as necessary using a backwards stepwise regression. Dietary pulse starch and rice starch were the only significant contributors to the negative association of starch with CIMT (p=0.048 and p=0.023, respectively) (table 4). Dietary pulse and rice intake were also highly positively correlated with each other (r=0.272, p<0.001). Table 4 Starch sources and association with carotid intima media thickness Effect on CIMT, n=324 β SE p Value* Full Model Potato starch† 0.026 0.019 0.168 Whole wheat bread starch 0.001 0.004 0.709 White bread starch −0.006 0.004 0.143 Rice starch −0.006 0.003 0.036 Pasta starch 0.006 0.005 0.253 Pulse starch† −0.039 0.022 0.077 Other starch† −0.045 0.035 0.194 Stepwise model‡  Rice starch −0.006 0.003 0.023  Pulse starch† −0.043 0.021 0.048 *Multivariate regression model adjusted for age, smoking, previous CVD event, blood pressure medication, antidiabetic medication and ultrasonographer; starch variables are expressed as percentages of total calories. †Log transformed. ‡Adjusted multivariate regression model where each starch variable was removed from the model one at a time based on least significance as per backward stepwise regression. CIMT, carotid intima media thickness (mean max bilateral, average of the 12 segment measures); CVD, cardiovascular disease. Post hoc explorations were also conducted to assess associations between the metabolic profile of participants and all dietary variables (see online supplementary table S1). Of the dietary variables significantly associated with CIMT, greater intake of dietary pulses, carbohydrates, starch and GL and lower intake of fat and saturated fat were generally associated with lower body weight (r=−0.129, p=0.021; r=−0.210, p<0.001; r=−0.175, p=0.002; r=−0.157, p=0.005; r=0.202, p<0.001; r=0.321, p<0.001, respectively), systolic blood pressure (r=−0.145, p=0.009; r=−0.141, p=0.012; r=−0.130, p=0.020; r=−0.143, p=0.011; r=0.063, p=0.260; r=0.030, p=0.592, respectively), diastolic blood pressure (r=−0.172, p=0.002; r=−0.123, p=0.028; r=−0.135, p=0.016; r=−0.109, p=0.053; r=0.094, p=0.095; r=0.114, p=0.042), total-C (r=−0.118, p=0.036; r=−0.129, p=0.021; r=−0.160, p=0.004; r=−0.132, p=0.019; r=0.119, p=0.034; r=0.092, p=0.101) and LDL-C (r=−0.109, p=0.052; r=−0.128, 0.022; r=−0.151, p=0.007; r=−0.116, p=0.039; r=0.131, p=0.019; r=0.059, p=0.296) in adjusted Pearson correlations. Discussion Using CIMT as a predictive marker of CVD risk, we evaluated the associations with dietary intake in participants with type 2 diabetes. The objective of the main study from which these baseline data are taken is to assess the effect of a low-GI diet on markers of macrovascular disease, however, in the present cross-sectional analysis, there was no significant association between GI and CIMT. This may be because we were underpowered due to the small variance of measured GI in our population, no participant had a high-GI diet and the average GI was at the low end of the medium GI range, thus limiting our ability to assess any association. Although we found no association, low-GI diets have been demonstrated in systematic reviews and meta-analyses of randomised controlled trials to significantly reduce both total-C and LDL-C compared to high-GI diets,40 as well as to reduce oxidative stress and inflammation.41 Interestingly, in a recent randomised controlled trial of those with the metabolic syndrome randomised to receive either metformin or a low-GI diet for 8 weeks, both groups demonstrated significant improvements in metabolic syndrome components including body weight, blood pressure, cholesterol and glycaemia.42 Furthermore, the antidiabetic drug acarbose, which effectively converts the diet into a low-GI diet by delaying dietary carbohydrate absorption, has been associated with a reduced incidence of hypertension and CHD events in a small number of prediabetic participants in the Study to Prevent Non-Insulin-Dependent Diabetes Mellitus (STOP NIDDM) trial.43 Therefore, low-GI diets may have the potential to reduce CVD risk, particularly in those at high risk. Thus, further exploration into the potential benefit of low-GI diets on CVD risk is needed, particularly to assess change due to low-GI interventions. GL was significantly associated negatively with CIMT which differs from what was expected. Systematic reviews and meta-analyses of previous studies have demonstrated a positive association between GL and CVD risk.17 18 However, GL is the product of GI and available carbohydrate. We found a strong significant negative association between available carbohydrates and CIMT, but no effect of GI. We found a strong positive association between total and saturated fat intake and CIMT, therefore the surprising negative association between GL and CIMT may just mean that a higher carbohydrate and GL diet is simply an indicator of a lower total and saturated fat diet that predictably was associated with lower CIMT. Of the specific low-GI foods of interest assessed, dietary pulse intake was significantly inversely associated with CIMT. In the predicted model (figure 1) approximately one ¾ cup serving/day39 was associated with a 7.5% lower CIMT compared to no intake (0 g/day). Although associations with CIMT have not been assessed previously, dietary pulse intake has been associated with reduced risk of CHD and CVD,19 20 which supports the association found in the current study between dietary pulse intake and CIMT as a subclinical marker of CVD risk. Additionally, dietary pulses are part of a Mediterranean diet and this diet has been associated with improved CIMT in a number of studies,16 44–46 as was highlighted in a recent systematic review of dietary factors and CIMT by Petersen et al.47 We also found dietary pulse intake to be significantly associated with lower body weight, systolic and diastolic blood pressure, mean arterial pressure and cholesterol (see online supplementary table S1). A series of systematic reviews and meta-analyses have found dietary pulse intake to significantly improve body weight, blood pressure and cholesterol,48–50 therefore supporting the associations observed in this study. Each of these potential pathways has been associated with lower CVD risk. Additionally, dietary pulses are high in fibre, potassium and vegetable protein, and low in saturated fat, each of which has been demonstrated to lower blood pressure23 51 52 and improve cholesterol.48 Furthermore, although not explored in the current study, dietary pulses may also act through reduced inflammation, as supported by another recent systematic review and meta-analysis,53 which may also affect carotid plaque burden, since inflammation within atherosclerotic lesions increases the risk of plaque rupture and subsequent thromboembolism.54 The Mediterranean diets in the Prevención con Dieta Mediterránea (PREDIMED) study, which have been shown to be lower GI diets,55 were also found to downregulate cellular and circulating adhesion molecules and other inflammatory biomarkers.56 Since, dietary pulses are particularly low-GI foods, consuming a lower GI diet may also be beneficial for CVD risk. Our recent randomised controlled clinical trial in those with type 2 diabetes demonstrated that a low-GI diet with a particular emphasis on dietary pulses significantly lowered systolic blood pressure and heart rate, relative to a wheat fibre diet, both of which are negatively associated with CVD risk.23 Although the recent systematic review of dietary factors and CIMT by Petersen et al47 did not reveal any studies with results on dietary pulses, they did conclude from the observational studies retrieved, that greater intake of fruit, whole grains and fibre and a lower intake of saturated fat was associated with lower CIMT. From our analyses, we find support for the association between lower saturated fat intake and lower CIMT. A previous study demonstrated that for every 10 g/day (about 5% of calories) increase in saturated fat, CIMT is 0.03 mm greater.57 This is comparable to our analysis in which we found that for every 1% of total calorie increase in saturated fat, CIMT is about 0.011 mm greater (see online supplementary figure S1). The greater difference in CIMT for every 1% increase in saturated fat intake in our population compared to the previous study may be because of the higher risk of our population since they all had type 2 diabetes, were of greater age (60 vs 48 years) and greater BMI (30 vs 25 kg/m2). In recent years, new evidence from prospective studies has suggested that not all types of saturated fats play the same role in CVD development, including evidence that saturated fats from dairy products may play a protective role whereas those from other food sources may increase risk.58 59 Unfortunately, we were unable to explore different sources of saturated fat from our data. We did have data on protein from dairy sources which we explored post hoc and found had a significant positive association with CIMT (β=0.075, p=0.006; multivariate model). Further exploration into effects of different sources of saturated fat is warranted. Although we did not analyse whole grains, we did find a significant negative association between dietary starch intake and CIMT. Furthermore, in the starch post hoc analyses exploring major sources of starch, dietary pulse starch and rice starch were the only significant contributors to the negative association of starch with CIMT (table 4) and since they were also highly positively correlated with each other, this may mean they are consumed together, for example, as lentils and rice, a common dish. This result further strengthens the findings for dietary pulses. We did not find a significant association with CIMT and dietary fibre in the present analysis (β=−0.048, p=0.211). However, the recent systematic review highlighted that the PREDIMED study found a significantly lower CIMT with low (<25 g/day) versus high (>35 g/day) fibre intakes (−0.051 mm, 95% CI −0.094 to −0.009).60 When we explored dietary fibre based on intakes <25, 25–35 and >35 g/day, there was a trend for lower CIMT with increasing fibre intake, but this trend was not significant (p=0.119). Furthermore, when we adjusted for energy using the residual method, similar to the approach used in the PREDIMED study, there was again an inverse trend, although the difference between the highest and lowest fibre groups did not reach statistical significance (p=0.088) (see online supplementary table S2). Further exploration into any possible association between dietary fibre and CIMT is warranted, particularly since many studies have demonstrated an inverse association between dietary fibre intake and cardiovascular risk.61 Taken together, the results demonstrate that a higher carbohydrate diet may benefit CIMT, a marker of CVD risk. This is especially true for a diet with high-quality carbohydrate where the starch comes largely from dietary pulses, a particularly low-GI food, and which is also low in total fat, particularly saturated fat. The associations between these specific dietary variables and the metabolic profile of the participants reveal that they may act through better control of body weight, blood pressure and cholesterol. Strengths and limitations of this study A strength of the analyses is that the CIMT scans were all performed using the same scanner at the same site and using the same reading protocol. A further strength is the method of collection of dietary data. Although the majority of participants were overweight and may have underreported their intake,62 we took three important steps to minimise the impact of misreporting. First, we used prospectively collected 7-day food records: widely cited as the ‘gold standard’ of dietary measurement.63 Second, these records were reviewed by the study dietitian at the time of collection and in the presence of the participant and details clarified (eg, nature of the margarine). Third, for this analysis, a priori, we aimed to exclude participants reporting intakes below 500 and 800 calories or above 3500 or 4000 calories, respectively for women and men.63 No participant reported levels outside these cut-points, therefore we included all dietary records in our analyses. Finally, we have adjusted for energy in the regression models to dampen the effect of misestimation. We also found a correlation between calories and body weight (r=0.31, p<0.001; unadjusted). Therefore, we believe the dietary data are of reasonable quality. Limitations of these analyses include that they are conducted using CIMT results from only one CIMT scan obtained at baseline. However, studies have demonstrated mean max CIMT measurements to have good reproducibility, particularly using the method we adopted where measurements are made of 12 segments of the carotid artery.27 Furthermore, although CIMT has been associated with CVD5 6 a major limitation is that carotid plaque is a stronger predictor of CVD,64–66 although we did include posterior wall plaques in the CIMT measures, if present. Another limitation is possible residual confounding due to unmeasured or uncontrolled variables, although CIMT confounders were adjusted for in the analyses. The participants were also at high cardiovascular risk, thus application to a healthier cohort is limited. Furthermore, although at higher CVD risk, participants had relatively well-controlled diabetes (HbA1c, mean 7.1±0.5%), blood pressure (mean 122±11/72±8 mm Hg) and LDL-C (mean 2.2±0.8 mmol/L), with 67% on blood pressure medication and 72% on cholesterol-lowering medication at baseline, therefore possibly limiting the ability to assess associations between risk factors for CVD and CIMT and dietary intake, as well as limiting application to those with uncontrolled risk factors. Also, due to the good glycaemic control of the participants, this may explain why there were no associations found with either HbA1c or glucose and CIMT. Importantly, the cross-sectional design of the study is a limitation to establish causality due to the possibility of reverse causation bias. Longitudinal studies and randomised controlled trials are needed to confirm the observed associations. Conclusion Overall, greater consumption of dietary pulses, which are particularly low-GI foods, and available carbohydrates and lower saturated fat were significantly associated with lower baseline CIMT, as well as body weight, blood pressure and cholesterol, suggesting a potential role for diet in CVD risk reduction in type 2 diabetes. Properly designed randomised controlled trials are necessary to confirm if these dietary factors, including increased intake of dietary pulses and a reduction in saturated fat intake, are potential strategies to reduce CVD risk in those with type 2 diabetes. Furthermore, these types of trials will also be necessary to better assess if there is any effect of GI, where a low-GI diet is the result of healthy low-GI dietary advice. Thus the results of the main trial underway, which will allow for the assessment of changes resulting from a low-GI intervention, are greatly anticipated. Contributors: DJAJ, LAL, RGJ, GEM, ARM, CWCK and JLS conceived the study and wrote the study protocol. LC, AM, CI, SM, SS-P, JC, OO, DP, RJdS, LSAA, BB, SBM, SKN, ARM and DJAJ contributed significantly to the acquisition of the data. LC, AM, CI, RGJ, SCP, GEM, ARM and DJAJ significantly contributed to the analyses of the data. LC and DJAJ drafted the manuscript and all authors revised the manuscript critically for important intellectual content. All authors reviewed and approved the final manuscript and agree to be accountable for all aspects of the work. DJAJ, LC, CI, RJD and SCP are the guarantors and take responsibility for the integrity of the data and the accuracy of the data analyses. The authors assume full responsibility for the accuracy and completeness of the ideas presented. Funding: This work was supported by the Canadian Institutes of Health Research (CIHR), as well as Loblaw, and Barilla. AM was supported by the Queen's University, School of Medicine JD Hatcher Summer Studentship Award. JLS was funded by a PSI Graham Farquharson Knowledge Translation Fellowship, Canadian Diabetes Association (CDA) Clinician Scientist award, CIHR INMD/CNS New Investigator Partnership Prize, and Banting & Best Diabetes Centre Sun Life Financial New Investigator Award. DJAJ was funded by the Government of Canada through the Canada Research Chair Endowment. Disclaimer: None of the funders had a role in the study design, collection, analysis or interpretation of the data; in the writing of the report; and in the decision to submit the paper for publication. The researchers acted independently from the funders. Competing interests: All authors have completed the ICMJE uniform disclosure form at http://www.icmje.org/coi_disclosure.pdf and declare the following: LC has received research support from the Canadian Institutes of Health Research (CIHR). LC is a clinical research coordinator at Glycaemic Index Laboratories, Toronto, Ontario, Canada. RJdS is funded by a Canadian Institutes of Health Research (CIHR) Postdoctoral Fellowship Award and has received research support from the CIHR, the Calorie Control Council, the Canadian Foundation for Dietetic Research, and The Coca-Cola Company (investigator initiated, unrestricted grant). RJdS has served as an external resource person to the World Health Organization's Nutrition Guidelines Advisory Group on trans fats, saturated fats, and polyunsaturated fats. The WHO paid for his travel and accommodation to attend meetings from 2012–2016 to present and discuss this work. He has also done contract research for the Canadian Institutes of Health Research's Institute of Nutrition, Metabolism, and Diabetes, Health Canada, and the World Health Organization for which he received remuneration. He has held a grant from the Canadian Foundation for Dietetic Research as a principal investigator, and is a co-investigator on several funded team grants from Canadian Institutes of Health Research. RJdS received compensation for a lecture on dietary fat given at McMaster Pediatric Nutrition Days in 2016. LSAA has received an honorarium from the Nutrition Foundation of Italy (NFI) to co-organize a glycaemic index summit. ARM is currently consultant for Jansen/Johnson and Johnson. JLS has received research support from the Canadian Institutes of health Research (CIHR), Canadian Diabetes Association (CDA), PSI Foundation, Calorie Control Council, Banting and Best Diabetes Centre (BBDC), American Society for Nutrition (ASN), INC International Nut and Dried Fruit Council Foundation, National Dried Fruit Trade Association, The Coca-Cola Company (investigator initiated, unrestricted), Dr. Pepper Snapple Group (investigator initiated, unrestricted), and The Tate and Lyle Nutritional Research Fund at the University of Toronto. He has received speaker fees and/or honoraria from the Canadian Diabetes Association (CDA), Canadian Nutrition Society (CNS), University of Alabama at Birmingham, Dr. Pepper Snapple Group, Dairy Farmers of Canada, Nutrition Foundation of Italy (NFI), C3 Collaborating for Health, Sprim Brasil, WhiteWave Foods, Rippe Lifestyle, mdBriefcase, Alberta Milk, FoodMinds LLC, Memac Ogilvy & Mather LLC, PepsiCo, and Pulse Canada. He has ad hoc consulting arrangements with Winston & Strawn LLP, Perkins Coie LLP, and Tate & Lyle. He is a member of the European Fruit Juice Association Scientific Expert Panel. He is on the Clinical Practice Guidelines Expert Committees of the Canadian Diabetes Association (CDA), European Association for the study of Diabetes (EASD), and Canadian Cardiovascular Society (CCS), as well as an expert writing panel of the American Society for Nutrition (ASN). He serves as an unpaid scientific advisor for the Food, Nutrition, and Safety Program (FNSP) and the Technical Committee on Carbohydrates of the International Life Science Institute (ILSI) North America. He is a member of the International Carbohydrate Quality Consortium (ICQC), Executive Board Member of the Diabetes and Nutrition Study Group (DNSG) of the EASD, and Director of the Toronto 3D Knowledge Synthesis and Clinical Trials foundation. His wife is an employee of Unilever Canada. CWCK has received research grants, travel funding, consultant fees, honoraria, or has served on the scientific advisory board for Abbott Laboratories, Advanced Food Materials Network, Agriculture and Agri-Food Canada (AAFC), Almond Board of California, American Peanut Council, American Pistachio Growers, Barilla, Bayer, California Strawberry Commission, Calorie Control Council, Canadian Institutes of Health Research (CIHR), Canola Council of Canada, The Coca-Cola Company, Danone, General Mills, Hain Celestial, International Nut and Dried Fruit Council, International Tree Nut Council Nutrition Research and Education Foundation, Kellogg, Kraft, Loblaw Brands, Nutrition Foundation of Italy, Oldways Preservation Trust, Orafti, Paramount Farms, Peanut Institute, PepsiCo, Pulse Canada, Sabra Dipping Co., Saskatchewan Pulse Growers, Solae, Sun-Maid, Tate & Lyle, Unilever and White Wave Foods. He is on the Dietary Guidelines Committee for the Diabetes Nutrition Study Group of the European Association for the Study of Diabetes. DJAJ has received research grants from Saskatchewan Pulse Growers, the Agricultural Bioproducts Innovation Program through the Pulse Research Network, the Advanced Foods and Material Network, Loblaw Companies, Unilever, Barilla, the Almond Board of California, Agriculture and Agri-Food Canada, Pulse Canada, Kellogg's Company, Canada, Quaker Oats, Canada, Procter & Gamble Technical Centre ., Bayer Consumer Care, Springfield, NJ, Pepsi/Quaker, International Nut and Dried Fruit, Soy Foods Association of North America, the Coca-Cola Company (investigator initiated, unrestricted grant), Solae, Haine Celestial, the Sanitarium Company, Orafti, the International Tree Nut Council Nutrition Research and Education Foundation, the Peanut Institute, the Canola and Flax Councils of Canada, the Calorie Control Council (CCC), the CIHR, the Canada Foundation for Innovation and the Ontario Research Fund. He has been on the speaker's panel, served on the scientific advisory board and/or received travel support and/or honoraria from the Almond Board of California, Canadian Agriculture Policy Institute, Loblaw Companies, the Griffin Hospital (for the development of the NuVal scoring system, the Coca-Cola Company, EPICURE, Danone, Diet Quality Photo Navigation (DQPN), Saskatchewan Pulse Growers, Sanitarium Company, Orafti, the Almond Board of California, the American Peanut Council, the International Tree Nut Council Nutrition Research and Education Foundation, the Peanut Institute, Herbalife International, Pacific Health Laboratories, Nutritional Fundamental for Health, Barilla, Metagenics, Bayer Consumer Care, Unilever Canada and the Netherlands, Solae, Kellogg, Quaker Oats, Procter & Gamble, the Coca-Cola Company, the Griffin Hospital, Abbott Laboratories, the Canola Council of Canada, Dean Foods, the California Strawberry Commission, Haine Celestial, PepsiCo, the Alpro Foundation, Pioneer Hi-Bred International, DuPont Nutrition and Health, Spherix Consulting and WhiteWave Foods, the Advanced Foods and Material Network, the Canola and Flax Councils of Canada, the Nutritional Fundamentals for Health, AgriCulture and Agri-Food Canada, the Canadian Agri-Food Policy Institute, Pulse Canada, the Saskatchewan Pulse Growers, the Soy Foods Association of North America, the Nutrition Foundation of Italy (NFI), Nutrasource Diagnostics, the McDougall Program, the Toronto Knowledge Translation Group (St. Michael's Hospital), the Canadian College of Naturopathic Medicine, The Hospital for Sick Children, the Canadian Nutrition Society (CNS), the American Society of Nutrition (ASN), Arizona State University, Paolo Sorbini Foundation and the Institute of Nutrition, Metabolism and Diabetes. He received an honorarium from the US Department of Agriculture to present the 2013 W.O. Atwater Memorial Lecture. He received the 2013 Award for Excellence in Research from the International Nut and Dried Fruit Council. He received funding and travel support from the Canadian Society of Endocrinology and Metabolism to produce mini cases for the Canadian Diabetes Association (CDA). He is a member of the International Carbohydrate Quality Consortium (ICQC). His wife, ALJ, is a director and partner of Glycaemic Index Laboratories, and his sister received funding through a grant from the St. Michael's Hospital Foundation to develop a cookbook for one of his studies. AM, SM, SSP, JC, DP, OO, SBM, BB, SCP, LAL, RGJ and GM have no relevant conflicts of interest to declare. Ethics approval: St. Michael's Hospital ethics review board approved and provided oversight of this study. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Requests can be made to the corresponding author at nutritionproject@smh.ca. ==== Refs References 1 Mozaffarian D , Benjamin EJ , Go AS Heart disease and stroke statistics—2015 update: a report from the American Heart Association . Circulation 2015 ;131 :e29 –322 . 10.1161/CIR.0000000000000152 25520374 2 Shah AD , Langenberg C , Rapsomaniki E Type 2 diabetes and incidence of cardiovascular diseases: a cohort study in 1.9 million people . Lancet Diabetes Endocrinol 2015 ;3 :105 –13 . 10.1016/S2213-8587(14)70219-0 25466521 3 Perk J , De Backer G , Gohlke H European Guidelines on cardiovascular disease prevention in clinical practice (version 2012). The Fifth Joint Task Force of the European Society of Cardiology and Other Societies on Cardiovascular Disease Prevention in Clinical Practice (constituted by representatives of nine societies and by invited experts) . Eur Heart J 2012 ;33 :1635 –701 . 10.1093/eurheartj/ehs092 22555213 4 Greenland P , Alpert JS , Beller GA 2010 ACCF/AHA guideline for assessment of cardiovascular risk in asymptomatic adults: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines . J Am Coll Cardiol 2010 ;56 :e50 –103 . 10.1016/j.jacc.2010.09.001 21144964 5 Yoshida M , Mita T , Yamamoto R Combination of the Framingham risk score and carotid intima-media thickness improves the prediction of cardiovascular events in patients with type 2 diabetes . Diabetes Care 2012 ;35 :178 –80 . 10.2337/dc11-1333 22028278 6 Lorenz MW , Markus HS , Bots ML Prediction of clinical cardiovascular events with carotid intima-media thickness: a systematic review and meta-analysis . Circulation 2007 ;115 :459 –67 . 10.1161/CIRCULATIONAHA.106.628875 17242284 7 Mostaza JM , Lahoz C , Salinero-Fort MA Carotid atherosclerosis severity in relation to glycemic status: a cross-sectional population study . Atherosclerosis 2015 ;242 :377 –82 . 10.1016/j.atherosclerosis.2015.07.028 26275375 8 Marini MA , Succurro E , Castaldo E Cardiometabolic risk profiles and carotid atherosclerosis in individuals with prediabetes identified by fasting glucose, postchallenge glucose, and hemoglobin A1c criteria . Diabetes Care 2012 ;35 :1144 –9 . 10.2337/dc11-2032 22399698 9 Bobbert T , Mai K , Fischer-Rosinsky A A1C is associated with intima-media thickness in individuals with normal glucose tolerance . Diabetes Care 2010 ;33 :203 –4 . 10.2337/dc09-1009 19808917 10 Esposito K , Giugliano D , Nappo F Campanian Postprandial Hyperglycemia Study Group . Regression of carotid atherosclerosis by control of postprandial hyperglycemia in type 2 diabetes mellitus . Circulation 2004 ;110 :214 –9 . 10.1161/01.CIR.0000134501.57864.66 15197140 11 Mita T , Katakami N , Yoshii H Alogliptin, a dipeptidyl peptidase 4 inhibitor, prevents the progression of carotid atherosclerosis in patients with type 2 diabetes: The Study of Preventive Effects of Alogliptin on Diabetic Atherosclerosis (SPEAD-A) . Diabetes Care 2016 ;39 :139 –48 . 10.2337/dc15-0781 26628419 12 Sidhu JS , Kaposzta Z , Markus HS Effect of rosiglitazone on common carotid intima-media thickness progression in coronary artery disease patients without diabetes mellitus . Arterioscler Thromb Vasc Biol 2004 ;24 :930 –4 . 10.1161/01.ATV.0000124890.40436.77 15001452 13 Nissen SE , Wolski K Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes . N Engl J Med 2007 ;356 :2457 –71 . 10.1056/NEJMoa072761 17517853 14 Hanefeld M , Chiasson JL , Koehler C Acarbose slows progression of intima-media thickness of the carotid arteries in subjects with impaired glucose tolerance . Stroke 2004 ;35 :1073 –8 . 10.1161/01.STR.0000125864.01546.f2 15073402 15 Yun P , Du AM , Chen XJ Effect of acarbose on long-term prognosis in acute coronary syndromes patients with newly diagnosed impaired glucose tolerance . J Diabetes Res 2016 ;2016 :1602083 10.1155/2016/1602083 26770983 16 Murie-Fernandez M , Irimia P , Toledo E Carotid intima-media thickness changes with Mediterranean diet: a randomized trial (PREDIMED-Navarra) . Atherosclerosis 2011 ;219 :158 –62 . 10.1016/j.atherosclerosis.2011.06.050 21802081 17 Ma XY , Liu JP , Song ZY Glycemic load, glycemic index and risk of cardiovascular diseases: meta-analyses of prospective studies . Atherosclerosis 2012 ;223 :491 –6 . 10.1016/j.atherosclerosis.2012.05.028 22727193 18 Mirrahimi A , de Souza RJ , Chiavaroli L Associations of glycemic index and load with coronary heart disease events: a systematic review and meta-analysis of prospective cohorts . J Am Heart Assoc 2012 ;1 :e000752 10.1161/JAHA.112.000752 23316283 19 Nouri F , Sarrafzadegan N , Mohammadifard N Intake of legumes and the risk of cardiovascular disease: frailty modeling of a prospective cohort study in the Iranian middle-aged and older population . Eur J Clin Nutr 2016 ;70 :217 –21 . 10.1038/ejcn.2015.153 26395434 20 Bazzano LA , He J , Ogden LG Legume consumption and risk of coronary heart disease in US men and women: NHANES I Epidemiologic Follow-up Study . Arch Intern Med 2001 ;161 :2573 –8 . 10.1001/archinte.161.21.2573 11718588 21 Rodriguez-Monforte M , Flores-Mateo G , Sanchez E Dietary patterns and CVD: a systematic review and meta-analysis of observational studies . Br J Nutr 2015 ;114 :1341 –59 . 10.1017/S0007114515003177 26344504 22 Chiavaroli L , Mirrahimi A , Ireland C Low-glycaemic index diet to improve glycaemic control and cardiovascular disease in type 2 diabetes: design and methods for a randomised, controlled, clinical trial . BMJ Open 2016 ;6 :e012220 10.1136/bmjopen-2016-012220 23 Jenkins DJ , Kendall CW , Augustin LS Effect of legumes as part of a low glycemic index diet on glycemic control and cardiovascular risk factors in type 2 diabetes mellitus: a randomized controlled trial . Arch Intern Med 2012 ;172 :1653 –60 . 10.1001/2013.jamainternmed.70 23089999 24 Jenkins DJ , Kendall CW , Vuksan V Effect of lowering the glycemic load with canola oil on glycemic control and cardiovascular risk factors: a randomized controlled trial . Diabetes Care 2014 ;37 :1806 –14 . 10.2337/dc13-2990 24929428 25 Pitt B , Byington RP , Furberg CD Effect of amlodipine on the progression of atherosclerosis and the occurrence of clinical events. PREVENT Investigators . Circulation 2000 ;102 :1503 –10 . 10.1161/01.CIR.102.13.1503 11004140 26 Riley WA , Barnes RW , Applegate WB Reproducibility of noninvasive ultrasonic measurement of carotid atherosclerosis. The asymptomatic carotid artery plaque study . Stroke 1992 ;23 :1062 –8 .1636178 27 Dogan S , Duivenvoorden R , Grobbee DE Ultrasound protocols to measure carotid intima-media thickness in trials; comparison of reproducibility, rate of progression, and effect of intervention in subjects with familial hypercholesterolemia and subjects with mixed dyslipidemia . Ann Med 2010 ;42 : 447 –64 . 10.3109/07853890.2010.499132 20645885 28 Friedewald WT , Levy RI , Fredrickson DS Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge . Clin Chem 1972 ;18 :499 –502 .4337382 29 US Department of Agriculture. Composition of foods, agriculture handbook No. 23: The agriculture research service. Washington, DC: US Dept of Agriculture, 2010. 30 Canadian Nutrient File. http://webprod3.hc-sc.gc.ca/cnf-fce/index-eng.jsp 31 Atkinson FS , Foster-Powell K , Brand-Miller JC International tables of glycemic index and glycemic load values: 2008 . Diabetes Care 2008 ;31 :2281 –3 . 10.2337/dc08-1239 18835944 32 Salas-Salvadó J, Bulló M, Babio N, et al. Reduction in the incidence of type 2 diabetes with the Mediterranean diet: results of the PREDIMED-Reus nutrition intervention randomized trial. Diabetes Care 2011;34:14–9 10.2337/dc11-0338 33 Jenkins DJ , Srichaikul K , Kendall CW The relation of low glycaemic index fruit consumption to glycaemic control and risk factors for coronary heart disease in type 2 diabetes . Diabetologia 2011 ;54 :271 –9 . 10.1007/s00125-010-1927-1 20978741 34 Tobacco Free Initiative (TFI) . Fact sheet about health benefits of smoking cessation 2016 http://www.who.int/tobacco/quitting/benefits/en/(accessed 20 Jan 2016 ). 35 Monami M , Genovese S , Mannucci E Cardiovascular safety of sulfonylureas: a meta-analysis of randomized clinical trials . Diabetes Obes Metab 2013 ;15 :938 –53 . 10.1111/dom.12116 23594109 36 Rosen CJ The rosiglitazone story--lessons from an FDA Advisory Committee meeting . N Engl J Med 2007 ;357 :844 –6 . 10.1056/NEJMp078167 17687124 37 D'Agostino RB Sr, Vasan RS , Pencina MJ General cardiovascular risk profile for use in primary care: the Framingham Heart Study . Circulation 2008 ;117 :743 –53 . 10.1161/CIRCULATIONAHA.107.699579 18212285 38 Box GEP , Cox DR An analysis of transformations . J R Stat Soc 1964 ;26 :211 –34 . 39 Health Canada. Eating well with Canada’s food guide. 2011. Retrieved from http://www.hc-sc.gc.ca/fn-an/food-guide-aliment/index-eng.php . 40 Goff LM , Cowland DE , Hooper L Low glycaemic index diets and blood lipids: a systematic review and meta-analysis of randomised controlled trials . Nutr Metab Cardiovasc Dis 2013 ;23 :1 –10 . 10.1016/j.numecd.2012.06.002 22841185 41 Buyken AE , Goletzke J , Joslowski G Association between carbohydrate quality and inflammatory markers: systematic review of observational and interventional studies . Am J Clin Nutr 2014 ;99 :813 –33 . 10.3945/ajcn.113.074252 24552752 42 Rajabi S , Mazloom Z , Zamani A Effect of low glycemic index diet versus metformin on metabolic syndrome . Int J Endocrinol Metab 2015 ;13 :e23091 10.5812/ijem.23091 26587028 43 Chiasson JL , Josse RG , Gomis R Acarbose treatment and the risk of cardiovascular disease and hypertension in patients with impaired glucose tolerance: the STOP-NIDDM trial . JAMA 2003 ;290 :486 –94 . 10.1001/jama.290.4.486 12876091 44 Giannini C , Diesse L , D'Adamo E Influence of the Mediterranean diet on carotid intima-media thickness in hypercholesterolaemic children: a 12-month intervention study . Nutr Metab Cardiovasc Dis 2014 ;24 :75 –82 . 10.1016/j.numecd.2013.04.005 23809150 45 Ellingsen I , Seljeflot I , Arnesen H Vitamin C consumption is associated with less progression in carotid intima media thickness in elderly men: A 3-year intervention study . Nutr Metab Cardiovasc Dis 2009 ;19 :8 –14 . 10.1016/j.numecd.2008.01.006 18472409 46 Sala-Vila A , Romero-Mamani ES , Gilabert R Changes in ultrasound-assessed carotid intima-media thickness and plaque with a Mediterranean diet: a substudy of the PREDIMED trial . Arterioscler Thromb Vasc Biol 2014 ;34 :439 –45 . 10.1161/ATVBAHA.113.302327 24285581 47 Petersen KS , Clifton PM , Keogh JB The association between carotid intima media thickness and individual dietary components and patterns . Nutr Metab Cardiovasc Dis 2014 ;24 :495 –502 . 10.1016/j.numecd.2013.10.024 24374005 48 Ha V , Sievenpiper JL , de Souza RJ Effect of dietary pulse intake on established therapeutic lipid targets for cardiovascular risk reduction: a systematic review and meta-analysis of randomized controlled trials . CMAJ 2014 ;186 :E252 –62 . 10.1503/cmaj.131727 24710915 49 Jayalath VH , de Souza RJ , Sievenpiper JL Effect of dietary pulses on blood pressure: a systematic review and meta-analysis of controlled feeding trials . Am J Hypertens 2014 ;27 :56 –64 . 10.1093/ajh/hpt155 24014659 50 Li SS , Kendall CW , de Souza RJ Dietary pulses, satiety and food intake: a systematic review and meta-analysis of acute feeding trials . Obesity (Silver Spring) 2014 ;22 :1773 –80 . 10.1002/oby.20782 24820437 51 He J , Whelton PK Effect of dietary fiber and protein intake on blood pressure: a review of epidemiologic evidence . Clin Exp Hypertens 1999 ;21 :785 –6 . 52 Lee YP , Puddey IB , Hodgson JM Protein, fibre and blood pressure: potential benefit of legumes . Clin Exp Pharmacol Physiol 2008 ;35 :473 –6 .18307744 53 Salehi-Abargouei A , Saraf-Bank S , Bellissimo N Effects of non-soy legume consumption on C-reactive protein: a systematic review and meta-analysis . Nutrition 2015 ;31 :631 –9 . 10.1016/j.nut.2014.10.018 25837205 54 Nighoghossian N , Derex L , Douek P The vulnerable carotid artery plaque: current imaging methods and new perspectives . Stroke 2005 ;36 :2764 –72 . 10.1161/01.STR.0000190895.51934.43 16282537 55 Rodríguez-Rejón AI , Castro-Quezada I , Ruano-Rodrígguez C Effect of a Mediterranean diet intervention on dietary glycemic load and dietary glycemic index: The PREDIMED Study . J Nutr Metab 2014 ;2014 :985373 10.1155/2014/985373 25295183 56 Mena MP , Sacanella E , Vazquez-Agell M Inhibition of circulating immune cell activation: a molecular antiinflammatory effect of the Mediterranean diet . Am J Clin Nutr 2009 ;89 :248 –56 . 10.3945/ajcn.2008.26094 19056596 57 Merchant AT , Kelemen LE , de Koning L Interrelation of saturated fat, trans fat, alcohol intake, and subclinical atherosclerosis . Am J Clin Nutr 2008 ;87 :168 –74 .18175752 58 Astrup A , Rice Bradley BH , Brenna JT Regular-fat dairy and human health: A synopsis of symposia presented in Europe and North America (2014–2015) . Nutrients 2016 ;8: pii: E463 10.3390/nu8080463 59 Soedamah-Muthu SS , Ding EL , Al-Delaimy WK Milk and dairy consumption and incidence of cardiovascular diseases and all-cause mortality: dose-response meta-analysis of prospective cohort studies . Am J Clin Nutr 2011 ;93 :158 –71 . 10.3945/ajcn.2010.29866 21068345 60 Buil-Cosiales P , Irimia P , Ros E Dietary fibre intake is inversely associated with carotid intima-media thickness: a cross-sectional assessment in the PREDIMED study . Eur J Clin Nutr 2009 ;63 :1213 –9 . 10.1038/ejcn.2009.45 19550433 61 Satija A , Hu FB Cardiovascular benefits of dietary fiber . Curr Atheroscler Rep 2012 ;14 :505 –14 . 10.1007/s11883-012-0275-7 22872372 62 Stanton RA Nutrition problems in an obesogenic environment . Med J Aust 2006 ;184 :76 –9 .16411873 63 Willett W Nutritional Epidemiology . New York, NY : Oxford University Press , 1998 . 64 Zhang Y , Guallar E , Qiao Y Is carotid intima-media thickness as predictive as other noninvasive techniques for the detection of coronary artery disease? Arterioscler Thromb Vasc Biol 2014 ;34 :1341 –5 . 10.1161/ATVBAHA.113.302075 24764454 65 Duivenvoorden R , de Groot E , Elsen BM In vivo quantification of carotid artery wall dimensions: 3.0-Tesla MRI versus B-mode ultrasound imaging . Circ Cardiovasc Imaging 2009 ;2 :235 –42 . 10.1161/CIRCIMAGING.108.788059 19808598 66 Harloff A , Zech T , Frydrychowicz A Carotid intima-media thickness and distensibility measured by MRI at 3 T versus high-resolution ultrasound . Eur Radiol 2009 ;19 :1470 –9 . 10.1007/s00330-009-1295-8 19214524

PMC005xxxxxx/PMC5372139.txt

==== Front Lupus Sci MedLupus Sci MedlupusscimedlupusLupus Science & Medicine2053-8790BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR lupus-2016-00017610.1136/lupus-2016-000176Brief Communication1506Use of SLICC criteria in a large, diverse lupus registry enables SLE classification of a subset of ACR-designated subjects with incomplete lupus Aberle Teresa 1Bourn Rebecka L 1Chen Hua 1Roberts Virginia C 1Guthridge Joel M 1Bean Krista 1Robertson Julie M 1Sivils Kathy L 1Rasmussen Astrid 1Liles Meghan 1Merrill Joan T 1Harley John B 2Olsen Nancy J 3Karp David R 4James Judith A 151 Department of Arthritis and Clinical Immunology, Oklahoma Medical Research Foundation, Oklahoma City, Oklahoma, USA2 Cincinnati Children's Hospital Medical Center and US Department of Veterans Affairs Medical Center, Cincinnati, Ohio, USA3 Division of Rheumatology, Penn State Milton S. Hershey Medical Center, University Drive, Hershey, Pennsylvania, USA4 Department of Medicine, University of Texas Southwestern Medical Center, Dallas, Texas, USA5 Departments of Medicine and Pathology, University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma, USACorrespondence to Dr Judith A James; judith-james@omrf.org2017 17 3 2017 4 1 e0001767 7 2016 26 9 2016 17 10 2016 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective SLE is traditionally classified using the American College of Rheumatology (ACR) criteria. The Systemic Lupus International Collaborating Clinics (SLICC) recently validated an alternative system. This study examined large cohorts of subjects with SLE and incomplete lupus erythematosus (ILE) to compare the impact of ACR and SLICC criteria. Methods Medical records of subjects in the Lupus Family Registry and Repository were reviewed for documentation of 1997 ACR classification criteria, SLICC classification criteria and medication usage. Autoantibodies were assessed by indirect immunofluorescence (ANA, antidouble-stranded DNA), precipitin (Sm) and ELISA (anticardiolipin). Other relevant autoantibodies were detected by precipitin and with a bead-based multiplex assay. Results Of 3575 subjects classified with SLE under at least one system, 3312 (92.6%) were classified as SLE by both systems (SLEboth), 85 only by ACR criteria (SLEACR-only) and 178 only by SLICC criteria (SLESLICC-only). Of 440 subjects meeting 3 ACR criteria, 33.9% (149/440) were SLESLICC-only, while 66.1% (n=291, designated ILE) did not meet the SLICC classification criteria. Under the SLICC system, the complement criterion and the individual autoantibody criteria enabled SLE classification of SLESLICC-only subjects, while SLEACR-only subjects failed to meet SLICC classification due to the combined acute/subacute cutaneous criterion. The SLICC criteria classified more African-American subjects by the leucopenia/lymphopenia criterion than did ACR criteria. Compared with SLEACR-only subjects, SLESLICC-only subjects exhibited similar numbers of affected organ systems, rates of major organ system involvement (∼30%: pulmonary, cardiovascular, renal, neurological) and medication history. Conclusions The SLICC criteria classify more subjects with SLE than ACR criteria; however, individuals with incomplete lupus still exist under SLICC criteria. Subjects who gain SLE classification through SLICC criteria exhibit heterogeneous disease, including potential major organ involvement. These results provide supportive evidence that SLICC criteria may be more inclusive of SLE subjects for clinical studies. Systemic Lupus ErythematosusClassificationIncomplete Lupus Erythematosus (ILE)DiagnosisNational Heart, Lung, and Blood Institutehttp://dx.doi.org/10.13039/100000050R24HL105333National Human Genome Research Institutehttp://dx.doi.org/10.13039/100000051U01HG008666National Institute of Diabetes and Digestive and Kidney Diseaseshttp://dx.doi.org/10.13039/100000062R01DK107502U.S. Department of Veterans Affairshttp://dx.doi.org/10.13039/100000738I01BX001834National Institute of Allergy and Infectious Diseaseshttp://dx.doi.org/10.13039/100000060R37AI24717U01AI101934U19AI082714National Institute of General Medical Scienceshttp://dx.doi.org/10.13039/100000057P30GM103510U54GM104938National Institute of Arthritis and Musculoskeletal and Skin Diseaseshttp://dx.doi.org/10.13039/100000069P30AR053483P30AR070549 ==== Body Introduction The clinical and immunological heterogeneity of patients with SLE hinders timely diagnosis, effective management and treatment development. Clinical trials of SLE typically select subjects based on the American College of Rheumatology (ACR) classification criteria,1 which require meeting ≥4 of 11 clinical and/or serological criteria. Although the ACR criteria remain a historical standard for identifying patients with SLE, individuals diagnosed with lupus by expert rheumatologists may not meet these criteria, while some who do meet the criteria have minimal disease. Therefore, ongoing efforts have sought more sensitive and specific criteria to identify patients with significant lupus.2 In 2012, the Systemic Lupus International Collaborating Clinics (SLICC) validated new SLE classification criteria through a series of consensus exercises using symptomatology and laboratory results drawn from real rheumatologic cases.3 SLE classification using SLICC criteria requires either meeting ≥4 of 17 criteria, including at least one clinical and one immunological criterion, or demonstrating biopsy-proven lupus nephritis with positive ANA or antidouble-stranded (ds)DNA.3 Because SLICC criteria emphasise immunological and haematological lupus manifestations, it has been proposed that subjects who gain SLE classification through SLICC criteria may be less likely to exhibit clinically significant organ involvement compared with subjects classified through ACR criteria.4 To address this question, the current study compared subjects who were classified by SLICC criteria with other subjects with SLE and incomplete lupus erythematosus (ILE) in a large, well-characterised, racially and geographically diverse cohort. Methods Study subjects This study was performed in accordance with the principles of the Declaration of Helsinki and approved by the Oklahoma Medical Research Foundation (OMRF) Institutional Review Board. Study participants were previously enrolled to the Lupus Family Registry and Repository (LFRR)5 and provided written informed consent, detailed clinical questionnaire information, connective tissue disease screening questionnaire responses,6 demographic information, blood samples and medical records, which were reviewed for ACR1 and SLICC3 criteria and for medication history (see online supplementary methods, supplementary figure 1). 10.1136/lupus-2016-000176.supp1supplementary data Autoantibody detection Autoantibodies were analysed by the College of American Pathologists-certified OMRF Clinical Immunology Laboratory. ANA and anti-dsDNA were analysed by indirect immunofluorescence, extractable nuclear antibodies by immunodiffusion and anticardiolipin by ELISA.7 Autoantibody specificities were compared using a multiplexed, bead-based assay (BioPlex 2200, Bio-Rad, Hercules, California, USA) that simultaneously detects dsDNA, chromatin, ribosomal P, Ro/SSA (60 and 52 kDa), La/SSB, Sm, SmRNP complex, RNP, centromere B, Scl-70 and Jo-1 autoantibodies.8 Anti-dsDNA is reported in IU/mL with a manufacturer-specified positive cut-off of 10.0 IU/mL, and other specificities as an Antibody Index (AI) value (range 0–8) based on the fluorescence intensity of each of the other autoantibody specificities, with a manufacturer-recommended positive cut-off of AI=1.0.8 Statistical analyses In R V.3.3.0 (R Foundation, https://www.r-project.org/), we compared means by unpaired t-test, medians by Mann-Whitney U test and proportions by either logistic regression using SLESLICC-only as the reference group or Fisher's exact test for comparisons with an observed value of 0. Two-sided p<0.05 was considered to be statistically significant. Results Approximately one-third of subjects with 3 ACR criteria are classified with SLE by SLICC criteria Medical record review of subjects in the LFRR identified 3397 subjects with SLE classified using ACR criteria. Of these, 3312 (97.5%) also reached SLICC classification (SLEboth), while 85 reached only ACR classification (SLEACR-only). An additional 178 reached only SLICC classification, but not ACR classification (SLESLICC-only). Approximately one-third of subjects with only three ACR criteria (149/440; 33.9%) met SLE classification by SLICC criteria. The other 291 subjects with three ACR criteria were not classified by SLICC criteria. These subjects, designated ILE, served as a comparison group expected to have more limited disease. Demographics were similar across the three SLE groups, while subjects with ILE were slightly older (table 1). Table 1 Demographics of subjects with SLE and ILE based on 2012 SLICC and 1997 ACR criteria SLESLICC-only* (n=178) SLEACR-only† (n=85) SLEboth‡ (n=3312) ILE§ (n=291) Sex  Female, n (%) 160 (89.9) 74 (87.0) p=0.494 2976 (89.9) p=0.989 255 (87.6) p=0.458 Age, years  Average (range) 43.7 (10–81) 45.4 (12–79) p=0.345 42.0 (8–82) p=0.118 47.5 (9–80) p=0.002 Race, n (%)  European American 89 (50.0) 52 (61.2) p=0.090 1466 (44.3) p=0.134 165 (56.7) p=0.158  African-American 44 (24.7) 14 (16.5) p=0.133 1079 (32.6) p=0.030 69 (23.7) p=0.804  Hispanic 12 (6.7) 5 (5.9) p=0.791 239 (7.2) p=0.811 12 (4.1) p=0.216  Asian 10 (5.6) 2 (2.4) p=0.250 128 (3.9) p=0.245 10 (3.4) p=0.261  American Indian 2 (1.1) 5 (5.9) p=0.044 99 (3.0) p=0.165 10 (3.4) p=0.144  Mixed 21 (11.8) 7 (8.2) p=0.383 276 (8.3) p=0.109 22 (7.6) p=0.126  Other¶ 0 (0.0) 0 (0.0) p=1.000 25 (0.8) p=0.985 3 (1.0) p=0.985 Bold p values are significant (p<0.05) for comparison with SLESLICC-only. *SLESLICC-only were classified with SLE by SLICC criteria, but not ACR criteria. †SLEACR-only were classified with SLE by ACR criteria, but not SLICC criteria. ‡SLEboth were classified with SLE by both SLICC and ACR criteria. §Patients with ILE met three ACR criteria and were not classified with SLE by SLICC standards. ¶Other includes Pacific Islander and unknown. ACR, American College of Rheumatology; ILE, incomplete lupus erythematosus; SLICC, Systemic Lupus International Collaborating Clinics. Subjects who do not meet ACR classification criteria gain SLE classification through SLICC haematological, immunological and alopecia criteria Two SLESLICC-only subjects (1.1%) were classified by SLICC criteria through biopsy-proven lupus nephritis with positive ANA or anti-dsDNA (figure 1A, bottom). The remaining 176 (98.9%) had one to four more SLICC criteria than ACR criteria. SLESLICC-only subjects gained criteria through low complement levels (81/178, 45.5%) and the separation of ACR immunological subcriteria into separate SLICC criteria (69/178, 38.8%), but African-American SLESLICC-only subjects most often gained criteria through the less stringent definition of leucopenia/lymphopenia (16/44, 36%) (figure 1D–E). Other than maculopapular rash, leading to a new criterion in 38 SLESLICC-only subjects (21.3%), and sensory neuropathy (14 SLESLICC-only subjects; 7.9%), new SLICC subcriteria made little contribution to additional individuals reaching SLE classification (see online supplementary figure 3). Figure 1 Subjects gain SLE classification through Systemic Lupus International Collaborating Clinics (SLICC) criteria of low complement, immunological manifestations and leucopenia/lymphopenia. (A) Medical record review identified subjects classified with SLE by American College of Rheumatology (ACR) criteria only (n=85; top, grey) or by SLICC criteria only (n=178; bottom, black). Labelled dots indicate the number of subjects who satisfied a given number of ACR criteria (y-axis) and SLICC criteria (x-axis). Criteria lost (B, C) or gained (D, E) under the SLICC system compared with the ACR system were evaluated in all SLESLICC-only (black) and SLEACR-only (grey) subjects (B, D) or in subjects self-reporting African-American race (C, E). See online supplementary figure 2 for criteria gained and lost in European American and other subjects. AA, African American; dsDNA, anti-double-stranded DNA; LN, lupus nephritis; SA, subacute. Of the 85 SLEACR-only subjects, 76 (89.4%) met <4 SLICC criteria (figure 1A, top). Nine (10.6%) met ≥4 SLICC clinical criteria, but were excluded by SLICC criteria due to an absence of immunological criteria. Loss of SLE classification by SLICC criteria was primarily due to the combination of malar rash and photosensitivity into a single SLICC criterion (53/85; 62.4% of SLEACR-only; figure 1B, see online supplementary figure 1). However, among African-American SLEACR-only subjects, the majority lost a criterion due to the stricter threshold for anticardiolipin positivity (figure 1C). Subjects classified with SLE only by SLICC criteria share clinical and immunological features with other subjects with SLE, including major organ involvement Acute/subacute cutaneous rashes, arthritis and leucopenia/lymphopenia were the most common SLICC clinical criteria in all groups (table 2). SLESLICC-only subjects exhibited relatively low prevalence of acute/subacute cutaneous rashes and arthritis, but higher prevalence of alopecia, leucopenia/lymphopenia and thrombocytopenia. SLESLICC-only and SLEboth subjects exhibited similar prevalence of multiple SLICC immunological criteria and had more SLICC immunological criteria than SLEACR-only or ILE (table 2). SLESLICC-only sera displayed significantly more autoantibody specificities and higher prevalence of lupus-associated specificities than SLEACR-only or ILE. SLEboth displayed the highest number and prevalence of lupus-associated specificities. Autoantibodies not specifically associated with lupus (anticentromere B, anti Scl-70 and anti Jo-1), were observed at low frequencies in all groups. The rate of major clinical involvement (serositis, renal or neurological) did not differ between SLESLICC-only and SLEACR-only (48/178, 27.0% vs 19/85, 22.4%; p=0.422), but was significantly lower in SLESLICC-only compared with SLEboth (2098/3312, 63.3%; p<0.0001) and higher compared with ILE (38/291, 11.3%; p<0.0001; see online supplementary table S1). Table 2 SLICC criteria, autoantibody specificities and medication history in patients with SLE and ILE based on SLICC and 1997 ACR criteria SLESLICC-only† (n=178) SLEACR-only‡ (n=85) SLEboth§ (n=3312) ILE¶ (n=291) SLICC clinical criteria  Number positive, mean 2.06 2.27 p=0.244 4.15 p<0.0001 1.45 p<0.0001  Acute/subacute cutaneous rashes, n (%) 76 (42.7) 71 (83.5) p<0.0001 2514 (75.9) p<0.0001 124 (42.6) p=0.986  Chronic cutaneous rashes, n (%) 10 (5.6) 6 (7.1) p=0.648 562 (17.0) p<0.001 26 (8.9) p=0.194  Oral/nasal ulcers, n (%) 11 (6.2) 13 (15.3) p=0.020 934 (28.2) p<0.0001 31 (10.6) p=0.104  Alopecia, n (%) 46 (25.8) 6 (7.1) p=0.001 1248 (37.7) p=0.002 2 (0.7) p<0.0001  Arthritis, n (%) 67 (37.6) 46 (54.1) p=0.012 2344 (70.8) p<0.0001 131 (45.0) p=0.117  Serositis, n (%) 10 (5.6) 9 (10.6) p=0.152 1198 (36.2) p<0.0001 17 (5.8) p=0.920  Renal, n (%) 23 (12.9) 9 (10.6) p=0.589 1262 (38.1) p<0.0001 13 (4.5) p=0.001  Neurological, n (%) 22 (12.4) 5 (5.9) p=0.113 585 (17.7) p=0.071 4 (1.4) p<0.001  Anaemia, n (%)* 3 (1.7) 0 (0.0) p=0.553 253 (7.6) p=0.007 1 (0.3) p=0.166  Leucopenia/lymphopenia, n (%) 83 (46.6) 26 (30.6) p=0.014 2339 (70.6) p<0.0001 67 (23.0) p<0.0001  Thrombocytopenia, n (%) 16 (9.0) 2 (2.4) p=0.064 498 (15.0) p=0.029 5 (1.7) p=0.001 SLICC immunological criteria  Number positive, mean 2.54 0.90 p<0.0001 2.86 p<0.001 1.25 p<0.0001  ANA, n (%) 176 (98.9) 74 (87.1) p=0.001 3299 (99.6) p=0.165 280 (96.2) p=0.109  Anti-dsDNA, n (%) 93 (52.2) 2 (2.4) p<0.0001 2128 (64.3) p=0.001 34 (11.7) p<0.0001  Anti-Sm, n (%) 32 (18.0) 1 (1.2) p=0.004 807 (24.4) p=0.053 8 (2.8) p<0.0001  Antiphospholipid, n (%)* 67 (37.6) 0 (0.0) p<0.0001 1016 (30.7) p=0.051 39 (13.4) p<0.0001  Complement, n (%)* 81 (45.5) 0 (0.0) p<0.0001 1884 (56.9) p=0.003 3 (1.0) p<0.0001  Coombs, n (%)* 3 (1.7) 0 (0.0) p=0.553 323 (9.8) p=0.002 0 (0.0) p=0.054 Autoantibody specificities**  Number positive, median 1 0 p=0.004 2 p<0.0001 1 p<0.0001  dsDNA, n (%) 37 (22.7) 1 (1.6) p=0.005 803 (30.2) p=0.043 13 (4.5) p<0.0001  Chromatin, n (%) 62 (38.0) 12 (19.0) p=0008 1433 (53.9) p=0.0001 47 (16.4) p<0.0001  Ribosomal P, n (%) 9 (5.5) 2 (3.2) p=0.468 355 (13.4) p=0.005 3 (1.0) p=0.011  Ro/SSA, n (%) 48 (29.4) 16 (25.4) p=0.545 1049 (39.5) p=0.011 64 (22.4) p=0.097  La/SSB, n (%) 17 (10.4) 4 (6.3) p=0.348 388 (14.6) p=0.142 24 (8.4) p=0.472  Sm, n (%) 24 (14.7) 4 (6.3) p=0.096 726 (27.3) p=0.0005 17 (5.9) p=0.003  SmRNP, n (%) 45 (27.6) 11 (17.5) p=0.116 1056 (39.7) p=0.002 35 (12.2) p<0.0001  RNP, n (%) 44 (27.0) 12 (19.0) p=0.217 954 (35.9) p=0.022 45 (15.7) p=0.004  Centromere B, n (%) 6 (3.7) 2 (3.2) p=0.853 100 (3.8) p=0.957 19 (6.6) p=0.194  Scl-70, n (%) 6 (3.7) 2 (3.2) p=0.854 72 (2.7) p=0.465 6 (2.1) p=0.323  Jo-1, n (%)* 0 (0.0) 0 (0.0) p=1.000 8 (0.3) p=1.000 2 (0.7) p=0.537 Medications used  Number, median 2 2 p=0.212 3 p<0.0001 2 p<0.0001  None, n (%) 12 (6.7) 4 (4.7) p=0.052 34 (1.0) p<0.0001 45 (15.5) p=0.006  Hydroxychloroquine, n (%) 133 (74.7) 66 (77.6) p=0.605 2755 (83.2) p=0.004 173 (59.4) p<0.0001  Steroids, n (%) 147 (82.6) 67 (78.8) p=0.464 3105 (93.8) p<0.0001 187 (64.3) p<0.0001  Immunosuppressants, n (%) 60 (33.7) 25 (29.4) p=0.486 1683 (50.8) p<0.0001 80 (27.5) p=0.154  Major immunosuppressants, n (%) 41 (23.0) 11 (12.9) p=0.058 1309 (39.5) p<0.0001 30 (10.3) p<0.001 Bold p values are significant (p<0.05) for comparison with SLESLICC-only by logistic regression or by Fisher's exact test where indicated (*) due to a 0 value. Note that power may be inadequate to detect differences when events are rare, particularly when the total n is also low, as for SLEACR-only. †SLESLICC-only were classified with SLE by SLICC criteria, but not ACR criteria. ‡SLEACR-only were classified with SLE by ACR criteria, but not SLICC criteria. §SLEboth were classified with SLE by both SLICC and ACR criteria. ¶Patients with ILE met three ACR criteria and were not classified with SLE by SLICC criteria. **Determined by in-house, multiplex, bead-based assay. ACR, American College of Rheumatology; ILE, incomplete lupus erythematosus; SLICC, Systemic Lupus International Collaborating Clinics. Subjects classified with SLE by only SLICC or only ACR criteria demonstrate similar medication histories Nearly all subjects had used at least one lupus-related medication type, including hydroxychloroquine, steroids, immunosuppressants (methotrexate, azathioprine and sulfasalazine) and/or major immunosuppressants (mycophenolate mofetil, cyclophosphamide) (table 2). Neither the number of medication types used nor the use of each medication type differed significantly between SLESLICC-only and SLEACR-only. Major immunosuppressant use was slightly more common among SLESLICC-only subjects compared with SLEACR-only subjects, but this difference was non-significant. Medication use was greatest in SLEboth and lowest in ILE. Discussion In a heterogeneous disease, optimised classification criteria would maximise inclusion of patients with clinically significant disease and exclude those without it. Although classification criteria are in many ways more restrictive than diagnostic criteria, classification criteria may directly impact patient access to new biologics; belimumab was approved only for patients meeting SLE classification criteria, since the trials excluded all others. While limited by retrospective design using community-based medical records from clinical care, lack of follow-up data and relatively small number of SLEACR-only subjects, this study provides new insights to patients identified by ACR and SLICC classification criteria. The most ill patients with obvious, multiple-organ SLE are classified by both ACR and SLICC criteria. Therefore, we compared these criteria in a large collection of patients with partial lupus syndromes. Twice as many subjects met only SLICC criteria (SLESLICC-only) as met only ACR criteria (SLEACR-only), consistent with previous reports suggesting increased sensitivity of SLICC compared with ACR criteria.3 9–13 However, SLICC criteria did exclude many subjects with clinically suggestive features of lupus. Despite a relatively low prevalence of acute/subacute cutaneous rashes and arthritis, SLESLICC-only subjects displayed a phenotypic range similar to other patients with SLE and distinct from ILE, including haematological, immunological and major organ system (serositis, renal or neurological) involvement. They were also younger than subjects with ILE, supporting the probability of a defined connective tissue disease.14 Consistent with previous studies,11 12 SLEACR-only subjects primarily lost SLE classification under SLICC criteria due to the combination of malar rash and photosensitivity; SLESLICC-only subjects primarily gained a criterion through low complement. African-Americans comprised >30% of our registry and primarily gained classification through the SLICC leucopenia/lymphopenia criterion or lost classification due to the stricter SLICC antiphospholipid criterion. In the absence of racially informed reference values, the leucopenia/lymphopenia criterion may lead to misclassification of patients with benign leucopenia of ethnicity; this highlights the need to consider racial diversity when developing and applying SLE classification criteria.15 Disease severity did not differ between SLESLICC-only and SLEACR-only subjects, based on major organ system involvement and medication history. Along with a trend for increased major immunosuppressant use, SLESLICC-only subjects presented several features associated with increased risk for morbidity and mortality, including a marginally higher proportion of minority subjects and increased prevalence of thrombocytopenia, anti-dsDNA and anticardiolipin responses compared with SLEACR-only.16 17 Therefore, although they lack ACR classification, patients who gain classification under SLICC criteria appear to have significant disease, and prospective study is warranted. Additionally, immunological and haematological similarities between SLESLICC-only and SLEboth subjects suggest that these patients might benefit from the same mechanistically targeted treatments and could be included in the same trials. The authors thank the personnel and participants of the Lupus Family Registry and Repository. The authors thank Cathy Velte, Camille Anderson, Sandy Long and Tim Gross for technical assistance and Miles Smith for scientific editing. Contributors: TA, RLB, JTM, JBH, NJO, DRK and JAJ designed the study. TA, VCR, JMG, JMR, KLS, AR, JTM, JBH, NJO, DRK and JAJ participated in data acquisition. TA, RLB, HC, JMG, KB, JMR, ML, JTM, JBH, NJO, DRK and JAJ participated in data analysis and/or interpretation. All authors assisted with the development of the manuscript and approved the final version to be published. JAJ had final responsibility for the decision to submit for publication. Funding: Research reported in this publication was supported by the US NIH through the National Institute of Allergy and Infectious Disease (U19AI082714, U01AI101934 and R37AI24717), Institutional Development Awards (IDeA) from the National Institute of General Medical Sciences (P30GM103510 and U54GM104938), the National Institute of Arthritis and Musculoskeletal and Skin Diseases (P30AR053483, P30AR070549), the National Human Genome Research Institute (U01HG008666), the National Heart, Lung, and Blood Institute (R24HL105333) and the National Institute of Diabetes and Digestive and Kidney Diseases (R01DK107502). This work was also supported by the US Department of Veterans Affairs (I01BX001834). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health, the Department of Veterans Affairs or the US government. The study sponsors had no role in the study design; in the collection, analysis and interpretation of the data; in the writing of the report or in the decision to submit the paper for publication. NJO reports grants from Mallinckrodt Pharmaceuticals, Resolve Therapeutics, Horizon Pharmaceuticals, Roche/Genentech and Aurinia Pharmaceuticals outside the submitted work. All other authors declare no conflicts of interest. Competing interests: None declared. Ethics approval: Oklahoma Medical Research Foundation Institutional Review Board. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: All relevant data for this study are being published. ==== Refs References 1 Hochberg MC Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus . Arthritis Rheum 1997 ;40 :1725 doi:10.1002/1529-0131(199709)40:9<1725::AID-ART29>3.0.CO;2-Y 2 Petri M Review of classification criteria for systemic lupus erythematosus . Rheum Dis Clin North Am 2005 ;31 :245 –54 , vi doi:10.1016/j.rdc.2005.01.00915922144 3 Petri M , Orbai AM , Alarcón GS Derivation and validation of the Systemic Lupus International Collaborating Clinics classification criteria for systemic lupus erythematosus . Arthritis Rheum 2012 ;64 :2677 –86 . doi:10.1002/art.3447322553077 4 Amezcua-Guerra LM , Higuera-Ortiz V , Arteaga-Garcia U Performance of the 2012 SLICC and the 1997 ACR classification criteria for systemic lupus erythematosus in a real-life scenario . Arthritis Care Res (Hoboken) 2015 ;67 :437 –41 .25073545 5 Rasmussen A , Sevier S , Kelly JA The lupus family registry and repository . Rheumatology (Oxford) 2011 ;50 :47 –59 . doi:10.1093/rheumatology/keq30220864496 6 Walitt BT , Constantinescu F , Katz JD Validation of self-report of rheumatoid arthritis and systemic lupus erythematosus: The Women's Health Initiative . J Rheumatol 2008 ;35 :811 –8 .18398940 7 Heinlen LD , McClain MT , Merrill J Clinical criteria for systemic lupus erythematosus precede diagnosis, and associated autoantibodies are present before clinical symptoms . Arthritis Rheum 2007 ;56 :2344 –51 . doi:10.1002/art.2266517599763 8 Bruner BF , Guthridge JM , Lu R Comparison of autoantibody specificities between traditional and bead-based assays in a large, diverse collection of patients with systemic lupus erythematosus and family members . Arthritis Rheum 2012 ;64 :3677 –86 . doi:10.1002/art.3465123112091 9 Ighe A , Dahlstrom O , Skogh T Application of the 2012 Systemic Lupus International Collaborating Clinics classification criteria to patients in a regional Swedish systemic lupus erythematosus register . Arthritis Res Ther 2015 ;17 :3 doi:10.1186/s13075-015-0521-925575961 10 Sag E , Tartaglione A , Batu ED Performance of the new SLICC classification criteria in childhood systemic lupus erythematosus: a multicentre study . Clin Exp Rheumatol 2014 ;32 :440 –4 .24642380 11 Ines L , Silva C , Galindo M Classification of systemic lupus erythematosus: systemic lupus international collaborating clinics versus American College of Rheumatology criteria . Arthritis Care Res (Hoboken) 2015 ;67 :1180 –5 . doi:10.1002/acr.2253925581417 12 Pons-Estel GJ , Wojdyla D , McGwin G Jr The American College of Rheumatology and the Systemic Lupus International Collaborating Clinics classification criteria for systemic lupus erythematosus in two multiethnic cohorts: a commentary . Lupus 2014 ;23 :3 –9 . doi:10.1177/096120331351288324285098 13 Ungprasert P , Sagar V , Crowson CS Incidence of systemic lupus erythematosus in a population-based cohort using revised 1997 American College of Rheumatology and the 2012 Systemic Lupus International Collaborating Clinics classification criteria . Lupus 2016 ;26:240–7 . 14 Calvo-Alen J , Alarcon GS , Burgard SL Systemic lupus erythematosus: predictors of its occurrence among a cohort of patients with early undifferentiated connective tissue disease: multivariate analyses and identification of risk factors . J Rheumatol 1996 ;23 :469 –75 .8832985 15 Williams EM , Bruner L , Adkins A I too, am America: a review of research on systemic lupus erythematosus in African-Americans . Lupus Sci Med 2016 ;3 :e000144 doi:10.1136/lupus-2015-00014427651918 16 Ziakas PD , Dafni UG , Giannouli S Thrombocytopenia in lupus as a marker of adverse outcome—seeking Ariadne's thread . Rheumatol 2006 ;45 :1261 –5 . 17 Petri M , Purvey S , Fang H Predictors of organ damage in systemic lupus erythematosus: the Hopkins Lupus Cohort . Arthritis Rheum 2012 ;64 :4021 –8 . doi:10.1002/art.3467222932985

PMC005xxxxxx/PMC5372140.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01365510.1136/bmjopen-2016-013655Mental HealthProtocol150617121712Meta-analysis of the effectiveness of psychological and medical treatments for binge-eating disorder (MetaBED): study protocol Hilbert Anja 1Petroff David 2Herpertz Stephan 3Kersting Anette 4Pietrowsky Reinhard 5Tuschen-Caffier Brunna 6Vocks Silja 7Schmidt Ricarda 1 1 Department of Medical Psychology and Medical Sociology, Department of Psychosomatic Medicine and Psychotherapy, University of Leipzig Medical Center, Leipzig, Germany 2 Clinical Trial Center Leipzig, University of Leipzig, Leipzig, Germany 3 Department of Psychosomatic Medicine, LWL-University Clinic, Ruhr-University Bochum, Bochum, Germany 4 Department of Psychosomatic Medicine and Psychotherapy, University of Leipzig Medical Center, Leipzig, Germany 5 Department of Clinical Psychology, Institute of Experimental Psychology, University of Düsseldorf, Düsseldorf, Germany 6 Department of Psychology, University of Freiburg, Freiburg, Germany 7 Department of Clinical Psychology and Psychotherapy, Institute of Psychology, University of Osnabrück, Osnabrück, GermanyCorrespondence to Professor Anja Hilbert; anja.hilbert@medizin.uni-leipzig.de2017 29 3 2017 7 3 e01365528 7 2016 23 1 2017 31 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction Binge-eating disorder (BED) was included as its own diagnostic entity in the Fifth Edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5). An increasing number of treatment studies have been published, but an up-to-date comprehensive meta-analysis on diverse treatment approaches for BED is lacking. In an updated and extension of a previous meta-analysis, the goals of this study are to assess the short-term and long-term effectiveness of psychological and medical treatments for BED. Methods and analysis We will search bibliographic databases and study registries, including manual searches for studies published before January 2016. The search strategy will include terms relating to binge eating and diverse forms of psychological and medical interventions. Language will be restricted to English. The studies included will be treatment studies, that is, randomised-controlled trials, and non-randomised and non-controlled studies, for individuals with BED (DSM-IV or DSM-5), and studies that provided a pre-treatment and at least one post-treatment or follow-up assessment of binge eating. The primary outcomes will be the number of binge-eating episodes, abstinence from binge eating and diagnosis of BED at post-treatment and/or follow-up(s), and changes from pre-treatment to post-treatment and/or follow-up(s). Likewise, as secondary outcomes, eating disorder and general psychopathology, quality of life, and body weight will be analysed and adverse events and treatment drop-out will be examined. Study search, selection and data extraction, including risk of bias assessment, will be independently performed by 2 reviewers and consensus will be sought. Moderator analyses will be conducted, and equity aspects will be considered. Sensitivity analyses will be conducted to determine the robustness of the results. Ethics and dissemination Ethical approval is not required for this meta-analysis. Published in a peer-reviewed journal and disseminated electronically and in print, this meta-analysis will form the basis of the renewal of the German evidence-based S3 Guidelines of Diagnosis and Treatment of Eating Disorders, specifically BED. Trial registration number CRD42016043604. meta-analysisbinge-eating disordertreatmentpsychotherapypharmacotherapybariatric surgery ==== Body Strengths and limitations of this study Strengths of this study are the provision of a comprehensive meta-analysis of the efficacy and effectiveness of psychological and medical treatments for binge-eating disorder, including identification of moderators of outcome. The PRISMA-P and MOOSE guidelines for meta-analysis protocols are followed. GRADE factors will be considered for evaluating the quality of evidence. A limitation is that economic aspects will not be considered. Background Binge-eating disorder (BED), characterised by recurrent binge eating that occurs in the absence of regular inappropriate compensatory behaviours, was first included as its own diagnostic entity in the Fifth Edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5).1 Compared with its provisional inclusion as an eating disorder diagnosis in need of further study in the preceding version of the DSM (DSM-IV),2 the DSM-5 requires binge-eating episodes to occur once per week over 3 months, lowering the diagnostic threshold of the DSM-IV, which required at least 2 days of binge eating per week over 6 months. For both diagnostic classifications, extant literature has indicated BED to be associated with severe health impairments, including increased eating disorder and general psychopathology, mental comorbidity, obesity and associated medical sequelae, and impaired quality of life. With a lifetime prevalence rate of 1.9%, BED is the most common eating disorder, with peak onset in adolescence or early adulthood.3 4 An increasing number of clinical studies investigating the efficacy (explanatory trials to determine whether an intervention produces the expected effect under ideal, highly controlled circumstances) and effectiveness (pragmatic trials to measure the effect under ‘real world’, less controlled conditions)5 of diverse treatment approaches to BED has been published and compiled in meta-analyses6–13 and systematic6 8 14–19 as well as narrative reviews.20–22 Among the meta-analyses, which in contrast to reviews allow a quantitative comparison of effects, two comprehensive reports addressed the efficacy and effectiveness of several broader treatment categories of BED, for example, psychotherapy, self-help treatment, weight loss treatment, pharmacotherapy and combined treatments. Vocks et al13 analysed 38 treatment studies for individuals with BED (randomised-controlled trials (RCTs), and non-randomised and non-controlled studies), until June 2006. In examining the RCTs, they found that psychotherapy and structured self-help, both mostly based on cognitive–behavioural therapy, significantly reduced binge eating (large effects) and eating disorder psychopathology (medium-to-large effects) at post-treatment when compared with untreated control groups. RCTs on pharmacotherapy, mainly using antidepressants, significantly reduced binge eating at post-treatment (medium effect), but not the associated eating disorder psychopathology (less than small effects), in comparison with untreated control groups. Both psychotherapy and pharmacotherapy significantly reduced post-treatment depressive symptoms (medium effects). Uncontrolled studies on conservative weight loss treatment demonstrated a significant post-treatment reduction of binge eating (medium effect), and a non-significant reduction of depressive symptoms. Combination treatments did not yield larger effects than monotherapeutic approaches. Except for weight loss treatment (medium effect), none of the interventions resulted in a significant weight reduction. In the RCTs, drop-out rates were non-significant in psychotherapy, self-help and pharmacotherapy when compared with untreated control groups. Important limitations of this early meta-analysis consist of a lack of examination of adverse effects and safety, quality of life, risk of bias and a lack of systematic documentation of the search process. Recently, a comprehensive systematic review and meta-analysis of 34 psychological and pharmacological RCTs for BED searched up to November 2015 (for MEDLINE up to May 2016) provided several meta-analytic comparisons with regard to waiting list or placebo control conditions, showing that therapist-led cognitive–behavioural therapy led to a greater post-treatment abstinence from binge eating when compared with waiting list. Furthermore, second-generation antidepressants were superior to placebo at post-treatment in bringing about abstinence from binge eating and reducing binge-eating episodes, eating disorder psychopathology and depression, but not weight.6 8 The stimulant lisdexamfetamine showed significantly greater odds of achieving post-treatment abstinence from binge eating and a greater reduction of binge-eating episodes and eating disorder psychopathology when compared with placebo. Serious adverse events were documented for lisdexamfetamine. Further meta-analytic comparisons on other treatment approaches, for example, structured self-help or weight loss treatment, were not conducted. With a concentration on RCTs, the effectiveness of treatments under ‘real-world’ conditions was not addressed. Other meta-analyses specifically focused on one category of treatment only, for example, pharmacotherapy,12 specifically antidepressants11 or lisdexamfetamine,23 combination treatments,24 or manualised self-help,7 while providing detailed results for BED alone. A further meta-analysis on RCTs in bulimia nervosa and recurrent binge eating,9 10 however, did not specify the results for BED as its own diagnosis. Since Vocks et al13 conducted their meta-analysis (search until June 2006), a number of large-scale clinical studies of BED and/or their long-term follow-ups have been published.25–27 Adding these studies would increase the precision of effect estimation; allow for the examination of longer term follow-up effects in order to determine the maintenance of therapeutic gains over time and facilitate moderator analyses with sufficient power. Further planned extensions of Vocks et al13 include an examination of adverse effects associated with treatment in order to evaluate safety; evaluation of quality of life; use of standard tools of risk of bias assessment and inclusion of other treatments with measured effect on BED symptomatology (eg, bariatric surgery). In the planned meta-analysis, a comprehensive search strategy based on the largest databases and clinical registries and a systematic documentation of the search in a PRISMA flow diagram will be performed. The PRISMA-P28 and MOOSE29 guidelines for reporting of the meta-analysis will be followed. Objectives The objectives of this meta-analysis are: to assess the short-term and long-term effectiveness of psychological and medical treatments for patients with BED in randomised-controlled, non-randomised and non-controlled treatment studies with inactive or active control groups, regarding binge eating, eating disorder and general psychopathology, quality of life, and body weight; to determine adverse events and treatment drop-out; and to examine equity aspects and risk of bias (PICO framework; Population, Intervention, Comparison, Outcome). Following this procedure, this report is intended to form the basis of the renewal of the German evidence-based S3 Guidelines of Diagnosis and Treatment of Eating Disorders,30 specifically BED, aimed at guiding the translation of clinical research into practice. Methods and analysis Study designs Randomised-controlled treatment studies will be included, complemented by non-randomised and non-controlled studies for increasing external validity. Case reports and studies with a sample size smaller than n=10 will be excluded. Participants Patients with a pre-treatment diagnosis of BED according to DSM-IV or DSM-5, including BED of low frequency and/or limited duration, will be included. Studies examining multiple patient groups will be included in case of separate data reports for patients with BED. Experimental interventions Psychological (eg, psychotherapy, self-help, conservative weight loss treatment) and medical treatments (eg, pharmacotherapy, bariatric surgery) applied to individuals with BED are included in this study. Treatments are defined to be attempts to relieve or cure BED or a highly related comorbidity (eg, bariatric surgery for treatment of obesity). Control conditions Inactive control groups (eg, no treatment, wait-list, placebo) and active control groups (eg, time/attention control, usual care, other treatment) will be considered as control conditions. Outcomes The core pathology of BED (binge-eating episodes or days, abstinence from binge eating, and/or diagnosis of BED) is to be assessed as an outcome measure. Sufficient data are required to allow the calculation of effect sizes (eg, M, SD and/or n, % at pre-treatment and post-treatment or follow-up(s)). Thus, a pre-treatment and at least one post-treatment or follow-up assessment are necessary. Primary outcomes: As primary outcomes, the number of binge-eating episodes, abstinence from binge eating and diagnosis of BED will be considered, forming the main outcome criteria in most treatment studies of BED.13 Binge-eating episodes commonly involve eating an amount of food that is definitely larger than what other people would eat under similar circumstances, associated with a sense of loss of control (LOC) over eating (objective binge-eating episodes).1 31 As bariatric surgery limits the possible amount of food intake,32 episodes of LOC eating which include both objective and subjective binge-eating episodes (eating an objectively or subjectively large amount of food, associated with a sense of LOC over eating), are considered where appropriate. Abstinence from binge eating will be defined as zero binge-eating episodes over a specified time frame, and presence or absence of a diagnosis of BED will be based on DSM-IV or DSM-5. Secondary outcomes: As secondary outcomes, associated eating disorder psychopathology will be operationalised through attitudes regarding eating behaviour and body image, for example, as determined through the Eating Disorder Examination or the Eating Disorder Inventory. Further, general, non-eating disorder psychopathology will be operationalised through measures of depression, for example, the Beck Depression Inventory or Hamilton Rating Scale for Depression. Quality of life will be operationalised through measures focusing on mental and/or physical well-being, for example, the Short Form Health Survey or the Impact of Weight on Quality of Life Scale. Body weight, body mass index (BMI, kg/m2) or other measures of excess weight will be considered if derived from objective measurement of body weight and height. Potential adverse consequences of the treatment will be operationalised as total adverse events and adverse events related to the intervention. Drop-out from treatment will be determined as a proxy of compliance with treatment. Search methods The search strategy will include terms related to binge eating and diverse forms of psychological and medical interventions in title, abstract and keywords (or full texts): (binge eat*) AND (efficac* OR effect* OR outcome OR counsel* OR interven* OR pharmaco* OR drug OR psychoanaly* OR psychotherap* OR therap* OR treat* OR train* OR weight loss OR weight reduction OR self-help OR bariatric surg* OR weight loss surg* OR weight reduction surg* OR obesity surg*; for a full search example see online supplementary file). 10.1136/bmjopen-2016-013655.supp1supplementary file Language restrictions will refer to English. Studies published or accepted for publication from inception to January 2016 will be sought. The searches will be rerun before the final analyses and further studies retrieved for inclusion prior to submission of the meta-analytical report. A new search for the total publication time period will be carried out because of greater availability of databases and registries when compared with Vocks et al,13 and because of the necessity to systematically document the search in a PRISMA flow diagram. The following data sources will be used: Electronic bibliographic databases: MEDLINE, PubMED, PsycINFO, EMBASE, PUBPSYCH, LILACS, CINAHL, AMED, Web of Science, DARE, ANNUAL REVIEWS, NIHR Centre for Reviews and Dissemination, CDSR, Clinical Psychology Review. Study registries: PROSPERO, CENTRAL, International Clinical Trials Registry Platform, ClincalTrials.gov, EU Clinical Trials Register, ISRCTN Trial Registry, Deutsches Register Klinischer Studien. Authors of ongoing studies will be contacted. Manual searches: reference lists of included studies and review articles identified during the search. Publications in the International Journal of Eating Disorders from 1990 to January 2016. Study selection Two psychologists (MSc level) will independently review all abstracts and titles for eligibility. If deemed eligible or where eligibility is unclear, full-text papers will be obtained. Disagreement will be resolved through consensus. The two psychologists will independently assess all full-text papers for inclusion. Where unclear because of a lack of information, study authors will be contacted. Disagreement will be resolved through consensus and under supervision of AH. In case of additional publications referred to in the primary included paper (eg, study protocol, intervention description) these publications will be obtained and considered as part of the included study. Multiple reports within the framework of one study will be assembled in order to form one unit of analysis in the review. Double publications of the same trial will be excluded, based on automatic and manual screening. Data extraction The standardised coding scheme and handbook used by Vocks et al13 with evidence of good inter-rater reliability will be extended and updated with regard to this meta-analysis' goals. The handbook provides definitions, coding instructions, examples and an overview of data management. Data extraction will be performed independently by two psychologists (MSc level). While one rating will include the entire study information, the duplicate rating will concentrate on outcome variables and further essential information to be used in meta-analytic comparisons as well as risk of bias assessment.33 Following training, a pilot procedure will be conducted using a representative sample of 10 studies to be reviewed independently. Disagreement between raters will be resolved through consensus and in consultation with AH. Data collection refers to: eligibility, study design, inclusion and exclusion criteria, participant characteristics (eg, sociodemographic, anthropometric), time points of assessment, sample size, intervention characteristics (eg, setting, duration, compliance, integrity), outcomes, drop-out, and adverse events, and risk of bias (eg, blinding, sequence generation, allocation sequence concealment). Data will be recorded for pretreatment, post-treatment and/or follow-up, where available. In order to facilitate comparisons with Vocks et al,13 the previous ratings will be compared with the current ratings by determining inter-rater reliability for the main outcome variables. Risk of bias assessment The Cochrane Collaboration's Risk of Bias Tool33 and the Effective Practice and Organization of Care Risk of Bias Tool34 will be used to assess the risk of bias in randomised-controlled studies as well as non-randomised studies, with additional items for assessing the risk of bias in uncontrolled studies, for example, considering the newly published Risk of Bias in Non-randomized Studies of Interventions (ROBINS-I).35 Summary assessments of low, medium or high risk of bias will be performed as recommended.33 The risk of bias will be examined during data extraction, following the procedure of independent coding and disagreement resolution described above, and will be used to interpret outcome data. Equity Participant characteristics known to be important from an equity perspective will be recorded (PROGRESS framework; Place, Race, Occupation, Gender, Religion, Education, Socioeconomic status, Social status).36 In addition, information will be collected on whether or not interventions included particular strategies to address diverse or disadvantaged populations. If indicated, these characteristics will be used in moderator analyses (see Moderation analyses section). Statistical models and measures of treatment effect Two types of analyses will be performed. First, the treatment effect will be compared for treatment versus non-interventional and placebo controls. These analyses will be performed for each treatment category separately (see below), and only RCTs will be included. Second, the treatment effect will be estimated within each treatment category, where all types of studies will be included. Random-effects models (general and generalised linear mixed models) will be used and compared with fixed-effects models as a sensitivity analysis. Primary outcomes: The number of binge-eating episodes (or days) over the past 28 days will be coded. If the time period differs from 28 days, the number of episodes will be scaled appropriately. The treatment effect will be measured as a standardised mean difference between the number of binge-eating episodes at post-treatment and/or follow-up(s) and the number at baseline. The CI associated with the treatment effect relies on the SD of this difference. For estimation of this SD, the following procedure will be used: If available, SD will be taken directly from the article; otherwise, If a test without covariates for the difference in question was performed, then SD will be determined from the test statistic or p value; otherwise, The SD for the difference in means will be estimated from the SD at baseline and the SD at post-treatment, assuming a correlation between the two based on available data. If none of the above is available, the trial will not be included in the analysis of this primary outcome. If only median and IQR are provided, mean and SD will not be estimated, since high abstinence rates at post-treatment may result in a median of 0, making such estimates unreliable. The above outcome will be complemented by the difference in means, since binge eating episodes are all measured on the same scale and the result can be easier to interpret. Abstinence from binge eating and diagnosis of BED are binary outcomes and will be measured as rates at post-treatment (after attempting to account for different observational times as necessary). The Wilson CI will be used to quantify uncertainty on the estimate and rates will be analysed for each treatment category separately. A comparison of intervention to control groups for RCTs will use ORs. Secondary outcomes: Continuous and binary outcomes will be treated along similar lines as specified for the primary outcomes above. If different scales are used, standardised means will be computed. For count data, if counts are small enough that treating them as proportions (ie, a continuous variable) may be inappropriate, then tables containing the count (numerator) and the number of patients (denominator) will be provided. Unit of analysis Treatment groups within publications will be the units of analysis both for randomised and non-randomised trials. Dealing with missing data As a sensitivity analysis, the subset of trials will be analysed where there has been appropriate treatment of missing values (eg, multiple imputations). Such a sensitivity analysis may not be needed if the subset differs only from the entire set or is very small. Assessment of heterogeneity The standard statistic in meta-analyses, Q, will be used to quantify and test how much individual trials deviate from the grand mean. Moreover, the variance τ2 from the random-effects model will be provided. Assessment of reporting biases Funnel plots with standardised mean differences on the horizontal and SE on the vertical axis will be inspected. As sensitivity analyses, trim and fill procedures and the fail-safe N will be used to quantify how unpublished negative outcome trials might affect results. Moderation analyses A power analysis is planned. Various parameters can be expected to affect outcomes and will be treated with meta-regression if the number of trials permits. Characteristics describing the trials will be presented in tables. The following variables will be extracted: Treatment characteristics: format (eg, individual, group), treatment category (eg, psychotherapy, self-help treatment, weight loss treatment, pharmacotherapy, bariatric surgery and combined treatments), single treatments (eg, cognitive-behavioural therapy, specific drugs), length of treatment. Participant characteristics: recruitment, age, sex, ethnicity, education, pre-treatment values (eg, baseline number of binge-eating episodes, BMI), duration of BED, diagnosis of BED. Study design: RCT, non-randomised trial with comparison group, non-randomised trial without comparison group, type of blinding (eg, patient, outcome assessors). Analysis method: for example, intention-to-treat analyses, completer analyses. Manualisation: for example, manual used, no manual use reported. Treatment integrity check: for example, integrity check by means of therapist supervision, no integrity check used. Therapist training: for example, general training of therapists, specific training for the study intervention, no training of therapists. Interview-based assessment of binge eating and/or of BED: for example, structured clinical interview (eg, Eating Disorder Examination), no structured clinical interview. Strength of evidence An evaluation of the strength of evidence will be provided with consideration of the GRADE recommendations. Study record and data management Records retrieved from electronic database and manual searches will be managed using EndNote X7 (Thomson Reuters, 2013). Data extraction of selected studies will be performed using a predefined Microsoft Excel (Microsoft Office Professional Plus 2010, V.14.0.7015.1000) coding form based on the coding scheme (see Data extraction section). All data will be analysed using R V.3.3.0 (R Core Team. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing 2016. http://www.R-project.org/). Dissemination The study results will be disseminated through peer-reviewed publication (electronic and in print) and conference presentations to the scientific community, and through further publications and presentations to the public and healthcare professionals. This meta-analysis is intended to form the basis of the renewal of the German evidence-based S3 Guidelines of Diagnosis and Treatment of Eating Disorders,30 specifically BED, guiding the dissemination of evidence-based treatments into clinical practice. No restrictions on publication exist. Authorship will follow the Rules of Good Scientific Practice of the German Research Foundation, and no professional writers will be used. The authors are grateful to Robert Richter, MSc, for his help with study selection, and to Hans-Christian Puls, MSc, for his help with coding. They are further grateful to Anja Sembill, for providing the codes from the Vocks et al13 meta-analysis for determining inter-rater reliability. They acknowledge support from the German Research Foundation (DFG) and University of Leipzig within the program of Open Access Publishing. Contributors: AH, DP, RS, SH, AK, RP, BT-C and SV were involved in protocol writing. RS and AH were involved in study selection, data extraction and disagreement resolution. RS was involved in data entry. DP was involved in data analysis. AH, DP and RS were involved in reporting. AH, SH, AK, RP, BT-C and SV are forming the BED Clinical Guidelines Group, under the auspices of which this meta-analysis is conducted. Funding: This work was supported by funds from the Christina Barz Foundation. AH, DP and RS were supported by grant 01EO1501 from the German Federal Ministry of Education and Research. Competing interests: The authors will give presentations on this topic, with honoraria and travel costs covered by organisers. Provenance and peer review: Not commissioned; externally peer reviewed. ==== Refs References 1 American Psychiatric Association . Diagnostic and Statistical Manual of Mental Disorders 5th Edition (DSM-5) . Arlington, VA : American Psychiatric Association , 2013 . 2 American Psychiatric Association . Diagnostic and Statistical Manual of Mental Disorders 4th Edition (DSM-IV) . Washington DC : American Psychiatric Association , 1994 . 3 Kessler RC , Berglund PA , Chiu WT The prevalence and correlates of binge eating disorder in the World Health Organization World Mental Health Surveys . Biol Psychiatry 2013 ;73 :904 –14 . 10.1016/j.biopsych.2012.11.020 23290497 4 Swanson SA , Crow SJ , Le Grange D Prevalence and correlates of eating disorders in adolescents: results from the National Comorbidity Survey Replication Adolescent Supplement . Arch Gen Psychiatry 2011 ;68 :714 –23 . 10.1001/archgenpsychiatry.2011.22 21383252 5 Gartlehner G , Hansen RA , Nissman D A simple and valid tool distinguished efficacy from effectiveness studies . J Clin Epidemiol 2006 ;59 :1040 –8 . 10.1016/j.jclinepi.2006.01.011 16980143 6 Berkman ND , Brownley KA , Peat CM Management and outcomes of binge-eating disorder. Comparative effectiveness review No. 160 . Rockville, MD : Agency for Healthcare Research and Quality , 2015 http://www.effectivehealthcare.ahrq.gov/reports/final.cfm (accessed 28 Jul, 2016 ). 7 Beintner I , Jacobi C , Schmidt UH Participation and outcome in manualized self-help for bulimia nervosa and binge eating disorder—a systematic review and metaregression analysis . Clin Psychol Rev 2014 ;34 :158 –76 . 10.1016/j.cpr.2014.01.003 24508686 8 Brownley KA , Berkman ND , Peat CM Binge-eating disorder in adults: a systematic review and meta-analysis . Ann Intern Med 2016 ;165 :409 –20 . 10.7326/M15-2455 27367316 9 Hay PJ , Bacaltchuk J , Stefano S Psychotherapy for bulimia nervosa and binging . Cochrane Database Syst Rev 2004 ;(3) :CD000562 10.1002/14651858.CD000562.pub2 15266434 10 Hay PPJ , Bacaltchuk J , Stefano S Psychological treatments for bulimia nervosa and binging . Cochrane Database Syst Rev 2009 ;(4) :CD000562 10.1002/14651858.CD000562.pub3 .19821271 11 Stefano SC , Bacaltchuk J , Blay SL Antidepressants in short-term treatment of binge eating disorder: systematic review and meta-analysis . Eat Behav 2008 ;9 :129 –36 . 10.1016/j.eatbeh.2007.03.006 18329590 12 Reas DL , Grilo CM Review and meta-analysis of pharmacotherapy for binge-eating disorder . Obesity (Silver Spring) 2008 ;16 :2024 –38 . 10.1038/oby.2008.333 19186327 13 Vocks S , Tuschen-Caffier B , Pietrowsky R Meta-analysis of the effectiveness of psychological and pharmacological treatments for binge eating disorder . Int J Eat Disord 2010 ;43 :205 –17 . 10.1002/eat.20696 19402028 14 Brownley KA , Berkman ND , Sedway JA Binge eating disorder treatment: a systematic review of randomized controlled trials . Int J Eat Disord 2007 ;40 :337 –48 . 10.1002/eat.20370 17370289 15 da Luz FQ , Hay P , Gibson AA Does severe dietary energy restriction increase binge eating in overweight or obese individuals? A systematic review . Obes Rev 2015 ;16 :652 –65 . 10.1111/obr.12295 26094791 16 Hay PJ , Claudino AM Clinical psychopharmacology of eating disorders: a research update . Int J Neuropsychopharmacol 2012 ;15 :209 –22 . 10.1017/S1461145711000460 21439105 17 Hay P A systematic review of evidence for psychological treatments in eating disorders: 2005–2012 . Int J Eat Disord 2013 ;46 :462 –9 . 10.1002/eat.22103 23658093 18 McElroy SL , Guerdjikova AI , Mori N Pharmacological management of binge eating disorder: current and emerging treatment options . Ther Clin Risk Manag 2012 ;8 :219 –41 . 10.2147/TCRM.S25574 22654518 19 Reas DL , Grilo CM Pharmacological treatment of binge eating disorder: update review and synthesis . Expert Opin Pharmacother 2015 ;16 :1463 –78 . 10.1517/14656566.2015.1053465 26044518 20 Iacovino JM , Gredysa DM , Altman M Psychological treatments for binge eating disorder . Curr Psychiatry Rep 2012 ;14 :432 –46 . 10.1007/s11920-012-0277-8 22707016 21 Wilson GT Treatment of binge eating disorder . Psychiatr Clin North Am 2011 ;34 :773 –83 . 10.1016/j.psc.2011.08.011 22098803 22 Wilson GT , Zandberg LJ Cognitive–behavioral guided self-help for eating disorders: effectiveness and scalability . Clin Psychol Rev 2012 ;32 :343 –57 . 10.1016/j.cpr.2012.03.001 22504491 23 Fornaro M , Solmi M , Perna G Lisdexamfetamine in the treatment of moderate-to-severe binge eating disorder in adults: systematic review and exploratory meta-analysis of publicly available placebo-controlled, randomized clinical trials . Neuropsychiatr Dis Treat 2016 ;12 :1827 –36 . 10.2147/NDT.S109637 27524900 24 Grilo CM , Reas DL , Mitchell JE Combining pharmacological and psychological treatments for binge eating disorder: current status, limitations, and future directions . Curr Psychiatry Rep 2016 ;18 :55 10.1007/s11920-016-0696-z 27086316 25 Wilson GT , Wilfley DE , Agras WS Psychological treatments of binge eating disorder . Arch Gen Psychiatry 2010 ;67 :94 –101 . 10.1001/archgenpsychiatry.2009.170 20048227 26 Grilo CM , Crosby RD , Wilson GT 12-month follow-up of fluoxetine and cognitive behavioral therapy for binge eating disorder . J Consult Clin Psychol 2012 ;80 :1108 –13 . 10.1037/a0030061 22985205 27 Hilbert A , Bishop ME , Stein RI Long-term efficacy of psychological treatments for binge eating disorder . Br J Psychiatry 2012 ;200 :232 –7 . 10.1192/bjp.bp.110.089664 22282429 28 Moher D , Shamseer L , Clarke M Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement . Syst Rev 2015 ;4 :1 10.1186/2046-4053-4-1 25554246 29 Stroup DF , Berlin JA , Morton SC Meta-analysis of observational studies in epidemiology. A proposal for reporting . JAMA 2000 ;283 :2008 –12 . 10.1001/jama.283.15.2008 10789670 30 Association of the Scientific Medical Societies . [Arbeitsgemeinschaft der Wissenschaftlichen Medizinischen Fachgesellschaften, AWMF] (2010). S3 Leitlinie Diagnostik und Therapie von Essstörungen [S3 Guidelines of Diagnosis and Treatment of Eating Disorders]. http://www.awmf.org/leitlinien/detail/ll/051-026.html (accessed 28 Jul 2016 ). 31 Wilson GT , Fairburn CG Cognitive treatments for eating disorders . J Consult Clin Psychol 1993 ;61 :261 –9 . 10.1037/0022-006X.61.2.261 8473579 32 Meany G , Conceição E , Mitchell JE Binge eating, binge eating disorder and loss of control eating: effects on weight outcomes after bariatric surgery . Eur Eat Disord Rev 2014 ;22 :87 –91 . 10.1002/erv.2273 24347539 33 Higgins JPT , Green S , eds. Cochrane handbook for systematic reviews of interventions version 5.1.0 . [updated March 2011] The Cochrane Collaboration , 2011 http://www.handbook.cochrane.org (accessed 28 Jul, 2016 ). 34 Effective Practice and Organisation of Care (EPOC) . Suggested risk of bias criteria for EPOC reviews. EPOC resources for review authors . Oslo : Norwegian Knowledge Centre for the Health Services , 2015 http://epoc.cochrane.org/epoc-specific-resources-review-authors (accessed 28 Jul 2016 ). 35 Sterne JA , Hernán MA , Reeves BC ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions . BMJ 2016 ;355 :i4919 10.1136/bmj.i4919 27733354 36 O'Neill J , Tabish H , Welch V Applying an equity lens to interventions: using PROGRESS ensures consideration of socially stratifying factors to illuminate inequities in health . J Clin Epidemiol 2014 ;67 :56 –64 . 10.1016/j.jclinepi.2013.08.005 24189091

PMC005xxxxxx/PMC5372141.txt

==== Front RMD OpenRMD OpenrmdopenrmdopenRMD Open2056-5933BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR 28405472rmdopen-2016-00036610.1136/rmdopen-2016-000366Spondyloarthritis1506Original articleCourse of patients with juvenile spondyloarthritis during 4 years of observation, juvenile part of GESPIC Weiß Anja 1Minden Kirsten 12Listing Joachim 1Foeldvari Ivan 3Sieper Joachim 2Rudwaleit Martin 456 1 German Rheumatism Research Center, Berlin, Germany 2 Charité University Medicine Berlin, Berlin, Germany 3 Hamburg Center for Pediatric Rheumatology, Hamburg, Germany 4 Klinikum Bielefeld, Bielefeld, Germany 5 Charite University, Medicine, Berlin, Germany 6 Ghent University, Ghent, BelgiumCorrespondence to Anja Weiß; weiss@drfz.de2017 21 3 2017 3 1 e00036622 9 2016 3 2 2017 7 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective To describe the course and the 4-year outcome of juvenile spondyloarthritis (jSpA). Methods Patients with a diagnosis of jSpA and an age at onset ≤16 years were included in the German Spondyloarthritis Inception cohort (GESPIC) and followed up prospectively for 4 years. Results 118 patients (73% men, 66% HLA-B27 positive, mean age 13.5 years, mean symptom duration 2.2 years) were enrolled in 2 study centres: 52% of patients with jSpA were captured by the enthesitis-related arthritis subgroup of the International League of Associations for Rheumatology classification criteria. At inclusion, the majority of patients had active peripheral arthritis (75.4%), followed by inflammatory back pain (IBP) (19.5%) and enthesitis (16.1%). There was a significant improvement in clinical manifestations and in patient-reported outcomes over time. During the 4-year follow-up, 85% of the patients had at least 1 period of remission on drug ≥6 months, and 46% of the patients achieved remission ≥12 months without medication, of whom 68% kept this status and 32% worsened. At the end of 4 years of observation, 23% of the patients were in remission without medication, but 57% still suffered from active disease. Patients with peripheral arthritis had a likelihood of 29% for having peripheral arthritis after 4 years, whereas the likelihood of IBP persistence was 53% for those with IBP at enrolment. Conclusions Although 1 quarter of patients with jSpA achieved remission off medication after 4 years, the likelihood of having recurrent or persistent disease into adulthood is substantial, particularly for jSpA with IBP. Trial registration number NCT 01277419. SpondyloarthritisAnkylosing SpondylitisArthritisBundesministerium für Bildung und Forschunghttp://dx.doi.org/10.13039/501100002347FKZ 01G19946 ==== Body Key messages Our findings extend the knowledge on the disease course and outcome of juvenile spondyloarthritis under treatment with non-steroidal anti-inflammatory drugs (NSAIDs) and/or disease-modifying antirheumatic drugs (DMARDs) within a specialised care. In patients already treated at study entry, we observed an additional improvement over time for peripheral arthritis, uveitis, the Bath Ankylosing Spondylitis Disease Activity Index and physician’s global assessment of disease activity but not for pain, inflammatory back pain and function. The likelihood of reaching at least one episode of remission on medication was 85% and of reaching the status of remission off medication at least once was 46%. Our results provide guidance to physicians in counselling patients and their parents on the expected outcome of certain disease manifestations 4 years after treatment with NSAIDS or conventional DMARDs. Introduction Juvenile spondyloarthritis (jSpA) is a group of chronic inflammatory disease with symptom onset at 16 years of age or younger. In contrast to adult-onset spondyloarthritis (SpA), dominant clinical signs in jSpA are enthesitis and peripheral arthritis rather than inflammatory back pain (IBP).1 Frequently, knee(s), ankle(s) and hips are affected. Boys are more frequently affected than girls, and HLA-B27 positivity, a positive family history and uveitis are also characteristics of jSpA. A minority of these patients develop ankylosing spondylitis (AS) prior to the age of 16, which is defined as juvenile-onset AS. Juvenile-onset AS differs in the first symptoms, the clinical picture, the frequency of manifestations at onset and severity in course of disease from adult-onset AS.1–4 The International League of Associations for Rheumatology (ILAR) developed classification criteria for juvenile idiopathic arthritis (JIA).5 Within the JIA classification criteria, jSpA was not considered as one disease entity.1 Most patients with jSpA fulfil criteria of enthesitis-related arthritis (ERA). However, within the ILAR system, patients with psoriasis or a positive family history for psoriasis are excluded from the ERA group but are considered to belong to the jSpA spectrum.1 4 For this reason and to take the concept of SpA as a group of inter-related disorders with shared clinical features into account, the European Spondyloarthropathy Study Group (ESSG) criteria6 for adult SpA were applied in this study. The ESSG criteria have previously been shown to perform well also in juveniles with SpA.7 8 The knowledge on the course and the outcome of jSpA from published studies is limited.9–22 Available studies were mainly based on JIA cohorts, and therefore, the subgroup of ERA or jSpA constituted often small subgroups in these JIA cohorts. An unfavourable outcome was reported for ERA or jSpA in most of these studies.10 11 13 15 17 22 According to Minden et al,10 only 16% of patients with jSpA achieved remission during 5 years of follow-up and more than half of the patients had active disease after ∼15 years of follow-up.11 17 Early diagnosis and treatment is important because jSpA can be a progressive disease.23 We used data from the juvenile arm of the German Spondyloarthritis Inception Cohort (GESPIC)24 (clinical trials.gov NCT 01277419) to describe the course and the 4-year outcome in a larger group of patients with jSpA. GESPIC was initiated in the prebiological era to investigate prospectively the long-term outcome of early stages of AS and SpA. Juveniles and adults were included in two different subcohorts. Here, we first report on the juvenile part and describe to what degree clinical signs of disease and patient-reported outcomes improved or worsened, the proportion of patients who still have an active disease after 4 years of observation, and the probability of an arthritis, enthesitis or IBP-free outcome in patients who suffered from these symptoms. Patients and methods Study design GESPIC is a prospective longitudinal cohort of patients with early SpA conducted in various centres across Germany. For the juvenile arm of GESPIC, patients had to have a diagnosis of juvenile SpA according to the rheumatologist's judgement and, similar to inclusion criteria among adult patients with SpA, patients with jSpA should additionally fulfil either the modified New York criteria for AS or the ESSG criteria for SpA, the latter with minor modifications: HLA-B27, dactylitis and acute anterior uveitis were added to the list of parameters of which at least one parameter must be present.6 25 In addition, patients had to have IBP or peripheral synovitis at study entry. Patient’s age at symptoms onset had to be ≤16 years and age at study start had to be <18 years. There were no restrictions regarding disease duration or type of treatment. Patients were enrolled consecutively in 2 out of 15 GESPIC study centres between February 2002 and December 2003. These 2 centres were specialised in paediatric rheumatology only, while the remaining 13 GESPIC centres cared for adult patients only. Clinical status was assessed at baseline and every 6 months for 4 years. Informed consent was obtained from all patients and their parents. The study was performed in accordance with the Declaration of Helsinki. Ethical approval was obtained by the Ethics Commission of the Freie Universität Berlin. Outcome assessments Outcome assessments are based on questionnaires for patients with juvenile SpA that were completed by physicians and patients or their parents, respectively. Peripheral arthritis was defined as a joint with swelling not due to deformity or joints with loss of motion plus pain and/or tenderness. Enthesitis was assessed clinically at the iliac crest, greater trochanter, medial condyle, lateral condyle, achilles tendon and plantar fascia; additional enthesitic sites could be documented as well. Physician’s global assessment of disease activity, patient’s global assessment and global pain were assessed on 0–10 numerical rating scales in which higher scores indicate higher disease activity or more severe pain, respectively. The Bath Ankylosing Spondylitis Disease Activity Index (BASDAI)26 was applied to measure disease activity. In a post hoc analysis, we furthermore calculated the Disease Activity Score for AS (ASDAS) with erythrocyte sedimentation rate (ESR)27 which was developed after the start of this study. Physical function was measured by the Bath Ankylosing Spondylitis Functional Index (BASFI, scale 0–10)26 and by the Childhood Health Assessment Questionnaire (CHAQ)28 ranging from 0 (no disability) to 3 (very severe disability). The Questionnaire for Measuring Health-Related Quality of Life in children and adolescents (KINDL)29 is a German self-reported instrument. It can be analysed individually in its six dimensions (physical well-being, emotional well-being, self-esteem, family, friends and everyday functioning in school), and disease as an optional subscale, or in a total score. The KINDL ranges from 0 to 100. Higher scores indicate a better quality of life. Remission was defined based on the Wallace criteria.30 To consider the specific characteristics of jSpA, the criteria were modified. For inactive disease, six of the following had to be fulfilled: physician's global assessment of disease activity score of 0, no joints with peripheral arthritis, no uveitis or enthesitis, morning stiffness <15 min, no IBP and either ESR ≤20 mm/hour or C reactive protein (CRP) level ≤5 mg/L. By definition, patients who did not fulfil these criteria had active disease. We distinguished two types of remission: clinical remission on medication (defined by inactive disease for at least 6 months on medication) and remission off medication (defined by inactive disease for at least 12 months and without any medication). Assessment of SpondyloArthritis international Society (ASAS) classification criteria were assessed.31 In patients with IBP, imaging was intended but was rarely performed and therefore excluded from further analysis. Statistical methods All patients enrolled were included in the analysis. Owing to a high portion of patients with undifferentiated SpA (84%), no stratification according to specific SpA subgroups was performed. To test whether there were changes in the clinical status over time, linear mixed models and generalised estimation equations (GEE) were used: mixed models were applied to estimate mean changes, and GEE models to estimate changes in the percentages of patients with specific clinical characteristics. For the analysis of remission, Breslow-type estimates of likelihood of achieving remission were used. The t-test and non-parametric Wilcoxon test were applied to compare HLA-B27-positive and HLA-B27-negative patients at baseline. Patients lost to follow-up were compared with completers based on clinical data (BASDAI, pain, peripheral arthritis (yes/no), enthesitis (yes/no)) assessed at the last study visit of the dropouts and the corresponding visit of the completers. Since these comparisons did not result in significant differences (data not shown), an adjustment for a possible confounding by patients lost to follow-up was not performed. p Values of <0.05 were considered to be statistically significant. Results SpA subgroups and classification In total, 118 patients with jSpA were enrolled. Ninety-nine patients (84%) were diagnosed as having undifferentiated SpA, nine (8%) with reactive arthritis (ReA), seven (6%) with psoriatic-SpA (Pso-SpA) and three (2.5%) with juvenile AS. All patients fulfilled the modified ESSG criteria and three patients the modified New York criteria for AS either at enrolment or before. The classification of the patients with jSpA according to the ILAR5 criteria system was as follows: 61 patients (52%) fulfilled the criteria of ERA, 11 (9%) for psoriatic arthritis, 10 (9%) for rheumatoid factor (RF)-negative polyarthritis, 4 (3%) for RF-positive polyarthritis, 7 (6%) for oligoarthritis and 18 (15%) were classified as cases of undifferentiated arthritis. There were seven (6%) patients with missing ILAR classification. One hundred and fifteen (97%) patients fulfilled the ASAS classification criteria. Disease demographics and characteristics prior to and at enrolment Seventy-three per cent of the patients were men, and 66% were HLA-B27 positive. Prior to study enrolment, nearly all of the patients ever had peripheral arthritis (96%). The most frequently affected joints were knees (77%), ankles (40%), hips (38%), toes (27%) and fingers (25%). Signs of symmetrical arthritis were observed in approximately half of the patients who had arthritis of the feet or fingers. Forty-four per cent of the patients ever suffered from enthesitis and 6–13% from psoriasis, uveitis, tarsitis or dactylitis (table 1). At enrolment, 75% of the patients suffered from peripheral arthritis, 16% from enthesitis and 20% from IBP (table 1). Table 1 Clinical characteristics ever and at study enrolment   Juvenile SpA (n=118) Clinical manifestations ever before study enrolment* Clinical manifestations/findings at study enrolment Peripheral arthritis, n (%) 113 (95.8) 89 (75.4) Arthritis joint count (0–64), mean (SD) 5.1 (5.7) 2.0 (2.3) Arthritis joint count (0–64), n (%)  0 5 (4.2) 27 (23.3)  1 15 (12.7) 34 (29.3)  2–4 54 (45.8) 45 (38.8)  5 or more 44 (37.3) 10 (8.6) Enthesitis, n (%) 52 (44.1) 19 (16.1) Enthesitis count (0–12), mean (SD) 0.9 (1.4) 0.3 (0.8) Enthesitis count (0–12), n (%)  0 66 (56.9) 95 (84.1)  1 22 (19) 11 (9.7)  2–4 25 (21.6) 6 (5.3)  5 or more 3 (2.6) 1 (0.9) Inflammatory back pain, n (%) 38 (32.2) 23 (19.5) Dactylitis, n (%) 15 (12.7) 3 (2.5) Tarsitis, n (%) 11 (9.3) 10 (8.5) Psoriasis, n (%) 7 (5.9) 7 (5.9) Uveitis, n (%) 8 (6.8) 3 (2.5) CRP (mg/L), mean (SD) – 9 (14.9) CRP ≥5 mg/L, n (%) – 48 (50.5) ESR (mm/hour), mean (SD) – 16.2 (16.4) ESR >20 mm/hour, n (%) – 27 (22.9) *According to medical charts. Disease manifestations in relation to HLA-B27 status HLA-B27-positive and HLA-B27-negative patients differed in physician’s global assessments of disease activity (mean (SD): 2.2 (1.5) vs 1.5 (1.2); p=0.007), age, family history, CRP and ESR (table 2). In other physician and patient-reported outcomes, no significant differences were observed between HLA-B27-positive and HLA-B27-negative patients (data not shown). Table 2 Patient characteristics at study enrolment by HLA-B27 status HLA-B27 positive, n=78 HLA-B27 negative, n=40 p Value All patients, n=118 Age at study start, mean (SD) 13.9 (2.5) 12.6 (2.8) 0.008 13.5 (2.7) Male, n (%) 60 (77) 26 (65) 0.168 86 (73) Symptom duration (years), mean (SD) 2.3 (2.1) 2.1 (1.5) 0.551 2.2 (1.9) Duration since diagnosis (years), mean (SD) 1.6 (1.8) 1.4 (1.5) 0.788 1.5 (1.7) Positive family history for SpA, n (%) 29 (37) 31 (78) 0.0001 60 (51) CRP (mg/L), mean (SD) 9.8 (15.9) 7.7 (13.1) 0.001 9 (14.9) CRP ≥5 mg/L, n (%) 40 (51) 8 (20) 0.0001 48 (50.5) ESR (mm/hour), mean (SD) 19 (18.3) 10.9 (9.9) 0.002 16.2 (16.4) ESR >20 mm/hour, n (%) 20 (25) 7 (18) 0.319 27 (23) CRP, C reactive protein; ESR, erythrocyte sedimentation rate; SpA, spondyloarthritis. Treatment and disease activity during follow-up At enrolment, 96% of the patients with jSpA received medical therapy. There was a strong decrease in the portion of patients treated with non-steroidal anti-inflammatory drugs (NSAIDs) and a smaller decrease in disease-modifying antirheumatic drug (DMARD) treatment over time (table 3). This resulted in a significant and clear increase in the percentage of patients who did not receive any treatment with NSAIDs, DMARDs or glucocorticoids. Table 3 Treatment at baseline reflecting the past 6 months and at follow-up Therapy At study enrolment n=118 Month 6 n=117 Year 1 n=114 Year 2 n=104 Year 3 n=93 Year 4 n=68 No therapy, n (%) 5 (4.2) 31 (27.4) 37 (35.9) 47 (53.4) 39 (56.5) 41 (69.5) NSAID, n (%) 101 (85.6) 78 (66.7) 60 (52.6) 38 (35.5) 25 (26.9) 16 (23.5) csDMARD, n (%) 48 (40.7) 51 (43.6) 45 (39.5) 27 (25.2) 17 (18.3) 12 (17.6)  Methotrexate, n (%) 26 (22) 25 (21.4) 22 (19.3) 11 (10.3) 11 (11.8) 8 (11.8)  Sulfasalazine, n (%) 17 (14.4) 22 (18.8) 18 (15.8) 12 (11.2) 5 (5.4) 4 (5.9) bDMARD (TNFi), n (%) 5 (4.2) 5 (4.3) 4 (3.5) 6 (5.6) 5 (5.4) 4 (5.9) Combination NSAID and DMARDs, n (%) 36 (30.5) 31 (26.5) 29 (25.4) 18 (16.8) 8 (8.6) 5 (7.4) Glucocorticoids >0.2 mg/kg, n (%) 10 (8.5) 1 (0.9) 1 (0.9) 1 (0.9) 1 (1.1) 0 (0) Glucocorticoids ≤0.2 mg/kg, n (%) 8 (6.8) 6 (5.1) 2 (1.8) 2 (1.9) 1 (1.1) 1 (1.5) Glucocorticoids, intra-articular, n (%) 18 (15.3) 8 (6.8) 4 (3.5) 3 (2.8) 2 (2.2) 0 (0) bDMARD, biological disease-modifying antirheumatic drug; csDMARD, conventional synthetic disease-modifying antirheumatic drug; NSAID, non-steroidal anti-inflammatory drug; TNFi, tumour necrosis factor inhibitor. Although disease activity at enrolment was an assessment under treatment in the vast majority of the patients with jSpA (96%), further significant improvements in arthritis, enthesitis, mean number of joints with limited range of motion, BASDAI, BASFI, CHAQ, patient’s and physician’s global assessment of disease activity were observed during 4 years of observation (table 4). Table 4 Clinical manifestations and assessments from study enrolment until year 4 Clinical parameter* At study enrolment n=118 Month 6 n=117 Year 1 n=114 Year 2 n=104 Year 3 n=93 Year 4 n=68 p Value Current status  Arthritis, n (%) 89 (75.4) 60 (51.3) 44 (38.6) 19 (17.8) 13 (14) 12 (17.6) 0.0001  Arthritis joint count (0–64), mean (SD) 2.0 (2.3) 1.3 (2.1) 1.0 (1.8) 0.4 (1.1) 0.3 (0.7) 0.5 (1.1) 0.0001  Joints with limited range of motion, n (%) 70 (59.3) 63 (53.8) 57 (50) 33 (30.8) 23 (24.7) 28 (41.2) 0.615  Joints with limited range of motion, mean (SD) 1.5 (1.4) 1.4 (1.6) 1.3 (1.4) 0.9 (1.4) 0.9 (1.6) 1.1 (1.2) 0.015  Enthesitis, n (%) 19 (16.1) 11 (9.4) 12 (10.5) 6 (5.6) 2 (2.2) 2 (2.9) 0.079  Enthesitis count (0–12), mean (SD) 0.3 (0.8) 0.1 (0.5) 0.2 (0.6) 0.1 (0.5) 0.02 (0.1) 0.1 (0.6) 0.513  Inflammatory back pain, n (%) 23 (19.7) 16 (14.5) 11 (10.9) 14 (16.9) 8 (11.9) 7 (13.7) 0.58  Patient's global assessment (0–10 NRS), mean (SD) n.d. 2.9 (2.6) 2.1 (2.2) 2.2 (2.7) 1.8 (2.3) 2.3 (2.9) 0.006  Pain (0–10 NRS), mean (SD) 2.5 (2.3) 1.9 (2.1) 1.5 (1.8) 1.6 (2.3) 1.4 (2.1) 1.3 (1.9) 0.244  ASDAS (0–10), mean (SD) n.d. 1.6 (0.7) 1.4 (0.7) 1.4 (0.8) 1.4 (0.8) 1.3 (0.6) 0.092  BASDAI (0–10), mean (SD) 2.0 (1.7) 1.6 (1.5) 1.3 (1.3) 1.4 (1.4) 1.1 (1.1) 1.0 (1.1) 0.002  BASFI (0–10), mean (SD) 0.7 (1.0) 0.6 (1.0) 0.4 (0.7) 0.4 (0.9) 0.3 (0.6) 0.3 (0.5) 0.002  CHAQ (0–3), mean (SD) 0.2 (0.3) 0.2 (0.4) 0.1 (0.2) 0.1 (0.3) 0.1 (0.2) 0.1 (0.2) 0.0001  KINDL (0–100), mean (SD) 76 (12.3) 77 (12.3) 79 (11.8) 80 (11.8) 81 (13.0) 82 (11.9) 0.165 Current or within the past 6 months  Arthritis, n (%) 113 (95.8) 83 (70.9) 60 (52.6) 30 (28) 17 (18.3) 15 (22.1) 0.0001  Enthesitis, n (%) 52 (44.1) 28 (23.9) 20 (17.5) 10 (9.3) 4 (4.3) 2 (2.9) 0.0001  Inflammatory back pain, n (%) 38 (32.2) 20 (18.2) 14 (13.9) 19 (22.9) 11 (16.4) 9 (16.7) 0.495  Physician’s global assessment (0–10 NRS), mean (SD) 2.0 (1.5) 1.6 (1.3) 1.4 (1.3) 1.3 (1.3) 0.9 (1.0) 1.0 (1.1) 0.0003  Uveitis, n (%) 8 (6.8) 2 (1.7) 3 (2.6) 1 (0.9) 1 (1.1) 4 (5.9) 0.07  Remission on medication, n (%) 0 (0) 6 (5.6) 10 (10.4) 21 (25.6) 23 (34.8) 23 (43.4) 0.0001  Remission off medication, n (%) – 0 (0) 2 (2.1) 3 (3.7) 7 (10.6) 12 (22.6) 0.0004 *In the case of mean changes, mixed linear models were used to test changes over time. Generalised equation models were used to analyse changes over time for percentages. BASDAI, Bath Ankylosing Spondylitis Disease Activity Index; BASFI, Ankylosing Spondylitis Functional Index; CHAQ, Childhood Health Assessment Questionnaire; KINDL, Questionnaire for Measuring Health-Related Quality of Life in children and adolescents; n.d., not done; NRS, numerical rating scale. After 4 years, active disease was present in 57% of the patients: 18% had peripheral arthritis, 14% had IBP, 3% suffered from enthesitis and 9% had an ESR >20 mm/hour. Other SpA manifestations occurred in one-fifth of the patients during the 4 years of follow-up: 13% reported dactylitis, and 6% reported psoriasis at study enrolment, which resolved completely after 4 years of follow-up. A new onset of dactylitis and tarsitis was observed in five and four patients, respectively. Psoriasis and uveitis were reported newly by three patients each. Urethritis and chronic inflammatory bowel disease were observed in one patient, and diarrhoea (more than once a month) in four patients during 4 years of follow-up. Three patients had a diagnosis of juvenile AS at enrolment and three further patients developed juvenile AS. Quality of life Quality of life was assessed by the total KINDL score and its subscales which range from 0 to 100. Patients in this cohort started with a KINDL mean total score of 76 (SD=12.3). Considering the KINDL subscales at study enrolment, we found that ‘family’ had the highest value (mean (SD): 84 (14.3)) and ‘disease’ the lowest value (mean (SD): 57 (12.3)). Significant improvement from study start until year 4 was observed for the subscales ‘physical wellbeing’ (mean (SD) from 80 (13.7) to 85 (12.3); p=0.036) and ‘friends’ (from 75 (17.6) to 84 (11.7); p=0.013). Achievement of remission The cumulative portion of patients achieving remission on and off medication is shown in figure 1A. By the end of 4 years, 85% of the patients had at least one period of remission on medication. In 49% of the remission periods on medication, treatment was terminated and remission off medication was maintained 6 months later (figure 1B), whereas 13% remained in the status of remission on medication, and 38% flared. The likelihood of achieving a status of remission off medication at least once during observation was 46% (figure 1A). The probability of remaining in this status was 68%, and that of a flare 32% (figure 1B). Figure 1 (A) Cumulative likelihood of achieving remission on and off medication. (B) Transition probabilities: left: transition probability of a status in remission on medication; right: transition probability of a status in remission off medication. Outcome at 4 years of follow-up Using GEE models, we estimated the outcome (remission or active disease) after 4 years for all patients enrolled (n=118) and the outcome of certain disease manifestations in patients with or without such manifestations at baseline (table 5). The likelihood of having active disease after 4 years has also been estimated according to HLA-B27 status, gender and family history (table 5). Table 5 Likelihood of having peripheral arthritis, enthesitis, inflammatory back pain or active disease at year 4 depending on status at study entry Status at study entry Status after 4 years* Peripheral arthritis present Peripheral arthritis present (n=82) 29% (17%; 44%) No peripheral arthritis (n=26) 8% (4%; 18%) Enthesitis present Enthesitis present (n=18) 8% (2%; 32%) No enthesitis (n=88) 3% (1%; 12%) IBP present IBP present (n=23) 53% (31%; 74%) No IBP (n=94) 5% (3%; 15%) Active disease present All patients 54% (41%; 66%) HLA-B27 positive (n=73) 59% (45%; 72%) HLA-B27 negative (n=37) 42% (26%; 59%) Male (n=79) 50% (37%; 63%) Female (n=31) 63% (44%; 79%) Positive SpA family history (n=57) 39% (25%; 56%) No SpA family history (n=41) 66% (50%; 80%) *Likelihood of a certain disease status is expressed in per cent; 95% confidence limits are shown in parentheses. Accordingly, we calculated a 71% chance of being arthritis-free after 4 years for patients with peripheral arthritis at baseline. The percentage of an enthesitis-free outcome for patients with enthesitis at study entry was even higher (92%). Inversely, for patients with IBP at study enrolment, there was still a 53% risk that symptoms of IBP sustained. This approach confirms that almost half of the patients (54%) still had an active disease 4 years after enrolment. The likelihood of having active disease was higher in HLA-B27-positive patients, women and patients without a family history of SpA. Discussion Our prospective, 4-year observational study of patients with jSpA clearly extends the knowledge currently available from small groups of patients with jSpA as a subgroup of JIA cohorts10 16 32 and one larger study, which, however, had to deal with a high loss to follow-up.18 In our cohort, 85% of the patients with jSpA reached at least one episode of remission on medication during 4 years of follow-up in specialised paediatric rheumatology care. The likelihood of reaching a status of remission off medication at least once was 46%. However, these remission episodes were not stable. After 4 years of follow-up, 23% of the patients were in remission without medication; yet nearly one out of five patients had peripheral arthritis and a similar portion had IBP. In total, half of the patients had an active disease. Guzman et al18 found cumulative probabilities of attaining inactive disease in patients with ERA within 4 years of 93% which is even higher than our remission on medication rate of 85%, but our modification of the Wallace criteria was stricter, too. Our findings agree with those of others who also found high rates of active disease in the long-term follow-up of patients with ERA or jSpA.10 13 17 Remission rates in patients with ERA were found to be significantly lower than in patients with oligoarthritis,10 15 17 and also lower than in patients with polyarticular JIA.15 17 18 The low remission rates correspond to poorer physical function, more bodily pain and reduced spinal flexion in the long term in patients with ERA compared with those with persistent oligoarticular or polyarticular JIA.17 These remission rates may have changed recently since tumour necrosis factor inhibitors are now increasingly used in the treatment of jSpA. Treatment with these biologics played a negligible role in our inception cohort as well as in previous other studies. In one study, male sex and a positive family history were found to be associated with a poorer outcome.15 Our data did not confirm these findings. In JIA in general, HLA-B27 positivity was found to be associated with a poorer outcome.10 15 A similar, yet statistically non-significant, trend was seen even within the patients with jSpA in our cohort as well as in another ERA cohort.15 Flato et al33 found that a long disease duration before first admission is a predictor of a progressive disease, a worse functional outcome and disease persistence into adulthood.17 33 We were not able to evaluate this aspect, since we did not collect the respective information and nearly all of our patients were already treated at enrolment. We asked whether the clinical status achieved in specialised care could further improve during the following 3–4 years, and if treatment with NSAIDs and/or DMARDs is being tapered whenever appropriate. We observed an additional improvement over time for peripheral arthritis, uveitis, BASDAI and physician’s global assessment of disease activity but not for pain, IBP and BASFI. Selvaag et al32 found similar results of a significant improvement in disease activity and health status but not in pain for patients with juvenile rheumatoid arthritis and jSpA during 3 years of observation. Oen et al14 analysed patients with JIA within 6 months after diagnosis who were treated with a limited number of therapies, and found improvements until 6 months follow-up in peripheral arthritis, physician global assessment, patient’s global assessment and CHAQ in all analysed subtypes. Our results may provide some guidance to physicians in counselling patients and their parents on the expected outcome of certain disease manifestations 4 years after treatment with NSAIDs or conventional DMARDs. We found rather high chances for patients suffering from enthesitis or arthritis at baseline to achieve an enthesitis-free or arthritis-free status at follow-up. In contrast, there was a 53% risk that symptoms of IBP were still present 4 years later. Klotsche et al34 found that an increase in health-related quality of life on therapy with etanercept is correlated to parameters of disease activity such as pain, painful and swollen joints and ESR. To assess health-related quality of life, we used the KINDL, a German self-reported generic instrument which is accepted internationally.35 36 The KINDL mean score at baseline was comparable to healthy adolescents.37 During follow-up, the subscales ‘physical wellbeing’ and ‘friends’ further improved significantly, and only the subscale ‘disease’ worsened. The strength of our study is the prospective study design and the large number of patients with jSpA (n=118), which is larger than those in the study by Flato et al (n=55)17 and Minden et al (n=28),10 and only slightly smaller than that in the study by Guzman et al (n=144).18 The retention rate after 4 years was 58% in our study, which can be considered as a weakness, yet our retention rate was much higher than the retention rate of 12% in the Guzman et al study. We furthermore possibly underestimated the percentage of patients who developed juvenile AS at follow-up due to missing imaging data. Conclusion The results of this 4-year observation cohort in juveniles with SpA describe the disease course under conventional treatment with NSAIDs and/or DMRADs in jSpA. Major clinical manifestations such as peripheral arthritis, enthesitis or extra-articular manifestations could successfully be improved. Despite early and effective treatment, there is still a rather high risk of a (residual) active disease or disease flares during 4 years of follow-up. This risk is especially high for IBP, and much lower for peripheral arthritis and enthesitis. For these reasons, our results from the prebiological era confirm the need for new treatment options also in jSpA and the necessity for a successful transition of patients with active disease to adult rheumatology care. The authors are grateful to Mrs Beate Buss for data monitoring, and to all patients and their parents who voluntarily participated in this cohort. The authors also like to thank Dr U Ravens-Sieberer for providing the Kiddo Kindl questionnaire on quality of life. Contributors: AW performed data analysis, statistical analysis, prepared tables and figures and wrote the draft manuscript. KM took care of study patients, was involved in data interpretation and manuscript preparation. JL performed statistical analysis, data interpretation and was involved in manuscript preparation. IF took care of after study patients. JS was responsible for getting funding of the trial, designed the study and was involved in manuscript preparation. MR was responsible for getting funding of the trial, designed the study and was involved in data interpretation and manuscript preparation. All authors read and approved the final manuscript. Funding: As part of the German Competence Network in Rheumatology (Kompetenznetz Rheuma), GESPIC has been financially supported by the Bundesministerium für Bildung und Forschung (BMBF), FKZ 01G19946. Competing interests: None declared. Patient consent: Obtained. Ethics approval: Ethics Commission of the Freie Universität Berlin. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Tse SM , Laxer RM New advances in juvenile spondyloarthritis . Nat Rev Rheumatol 2012 ;8 :269 –79 . 10.1038/nrrheum.2012.37 22487801 2 Burgos-Vargas R The assessment of the spondyloarthritis international society concept and criteria for the classification of axial spondyloarthritis and peripheral spondyloarthritis: a critical appraisal for the pediatric rheumatologist . Pediatr Rheumatol Online J 2012 ;10 :14 10.1186/1546-0096-10-14 22650358 3 O'Shea FD , Boyle E , Riarh R Comparison of clinical and radiographic severity of juvenile-onset versus adult-onset ankylosing spondylitis . Ann Rheum Dis 2009 ;68 :1407 –12 . 10.1136/ard.2008.092304 18782793 4 Burgos-Vargas R , Rudwaleit M , Sieper J The place of juvenile onset spondyloarthropathies in the Durban 1997 ILAR classification criteria of juvenile idiopathic arthritis. International League of Associations for Rheumatology . J Rheumatol 2002 ;29 :869 –74 .12022342 5 Petty RE , Southwood TR , Manners P International League of Associations for Rheumatology classification of juvenile idiopathic arthritis: second revision, Edmonton, 2001 . J Rheumatol 2004 ;31 :390 –2 .14760812 6 Dougados M , van der Linden S , Juhlin R The European Spondylarthropathy Study Group preliminary criteria for the classification of spondylarthropathy . Arthritis Rheum 1991 ;34 :1218 –27 . 10.1002/art.1780341003 1930310 7 Prieur AM , Listrat V , Dougados M [Criteria for classification of spondylarthropathies in children] . Arch Fr Pediatr 1993 ;50 :379 –85 .8239888 8 Prieur AM Evaluation of 4 sets of criteria for spondylarthropathy in children—a multicenter European cross-sectional study . Arthritis Rheum 1992 ;35 :S190 10.1002/art.1780350210 9 Horneff G , Fitter S , Foeldvari I Double-blind, placebo-controlled randomized trial with adalimumab for treatment of juvenile onset ankylosing spondylitis (JoAS): significant short term improvement . Arthritis Res Ther 2012 ;14 :R230 10.1186/ar4072 23095307 10 Minden K , Kiessling U , Listing J Prognosis of patients with juvenile chronic arthritis and juvenile spondyloarthropathy . J Rheumatol 2000 ;27 :2256 –63 .10990244 11 Minden K , Niewerth M , Listing J Long-term outcome in patients with juvenile idiopathic arthritis . Arthritis Rheum 2002 ;46 :2392 –401 . 10.1002/art.10444 12355487 12 Adib N , Hyrich K , Thornton J Association between duration of symptoms and severity of disease at first presentation to paediatric rheumatology: results from the Childhood Arthritis Prospective Study . Rheumatology (Oxford) 2008 ;47 :991 –5 . 10.1093/rheumatology/ken085 18417527 13 Nordal E , Zak M , Aalto K Ongoing disease activity and changing categories in a long-term nordic cohort study of juvenile idiopathic arthritis . Arthritis Rheum 2011 ;63 :2809 –18 . 10.1002/art.30426 21560116 14 Oen K , Duffy CM , Tse SM Early outcomes and improvement of patients with juvenile idiopathic arthritis enrolled in a Canadian multicenter inception cohort . Arthritis Care Res (Hoboken) 2010 ;62 :527 –36 . 10.1002/acr.20044 20391508 15 Berntson L , Nordal E , Aalto K HLA-B27 predicts a more chronic disease course in an 8-year followup cohort of patients with juvenile idiopathic arthritis . J Rheumatol 2013 ;40 :725 –31 . 10.3899/jrheum.121257 23547219 16 Flato B , Aasland A , Vinje O Outcome and predictive factors in juvenile rheumatoid arthritis and juvenile spondyloarthropathy . J Rheumatol 1998 ;25 :366 –75 .9489836 17 Flato B , Hoffmann-Vold AM , Reiff A Long-term outcome and prognostic factors in enthesitis-related arthritis: a case-control study . Arthritis Rheum 2006 ;54 :3573 –82 . 10.1002/art.22181 17075863 18 Guzman J , Oen K , Tucker LB The outcomes of juvenile idiopathic arthritis in children managed with contemporary treatments: results from the ReACCh-Out cohort . Ann Rheum Dis 2015 ;74 :1854 –60 . 10.1136/annrheumdis-2014-205372 24842571 19 Hyrich KL , Lal SD , Foster HE Disease activity and disability in children with juvenile idiopathic arthritis one year following presentation to paediatric rheumatology. Results from the Childhood Arthritis Prospective Study . Rheumatology (Oxford) 2010 ;49 :116 –22 . 10.1093/rheumatology/kep352 19926670 20 Olivieri I , Cutro MS , D'Angelo S Low frequency of axial involvement in southern Italian Caucasian children with HLA-B27 positive juvenile onset undifferentiated spondyloarthritis . Clin Exp Rheumatol 2012 ;30 :290 –6 .22325303 21 Rosenberg AM , Petty RE A syndrome of seronegative enthesopathy and arthropathy in children . Arthritis Rheum 1982 ;25 :1041 –7 . 10.1002/art.1780250902 7126288 22 Weiss PF , Xiao R , Biko DM Assessment of sacroiliitis at diagnosis of juvenile spondyloarthritis by radiography, magnetic resonance imaging, and clinical examination . Arthritis Care Res (Hoboken) 2016 ;68 :187 –94 . 10.1002/acr.22665 26212574 23 Stone M , Warren RW , Bruckel J Juvenile-onset ankylosing spondylitis is associated with worse functional outcomes than adult-onset ankylosing spondylitis . Arthritis Rheum 2005 ;53 : 445 –51 . 10.1002/art.21174 15934110 24 Rudwaleit M , Haibel H , Baraliakos X The early disease stage in axial spondylarthritis: results from the German Spondyloarthritis Inception Cohort . Arthritis Rheum 2009 ;60 :717 –27 . 10.1002/art.24483 19248087 25 van der Linden S , Valkenburg HA , Cats A Evaluation of diagnostic criteria for ankylosing spondylitis. A proposal for modification of the New York criteria . Arthritis Rheum 1984 ;27 :361 –8 . 10.1002/art.1780270401 6231933 26 Calin A , Garrett S , Whitelock H A new approach to defining functional ability in ankylosing spondylitis: the development of the Bath Ankylosing Spondylitis Functional Index . J Rheumatol 1994 ;21 :2281 –5 .7699629 27 Lukas C , Landewe R , Sieper J Development of an ASAS-endorsed disease activity score (ASDAS) in patients with ankylosing spondylitis . Ann Rheum Dis 2009 ;68 :18 –24 . 10.1136/ard.2008.094870 18625618 28 Singh G , Athreya BH , Fries JF Measurement of health status in children with juvenile rheumatoid arthritis . Arthritis Rheum 1994 ;37 :1761 –9 . 10.1002/art.1780371209 7986222 29 Ravens-Sieberer U , Bullinger M Assessing health-related quality of life in chronically ill children with the German KINDL: first psychometric and content analytical results . Qual Life Res 1998 ;7 :399 –407 . 10.1023/A:1008853819715 9691720 30 Wallace CA , Giannini EH , Huang B , Childhood Arthritis Rheumatology Research A American College of Rheumatology provisional criteria for defining clinical inactive disease in select categories of juvenile idiopathic arthritis . Arthritis Care Res (Hoboken) 2011 ;63 :929 –36 . 10.1002/acr.20497 21717596 31 Rudwaleit M , Landewe R , van der Heijde D The development of Assessment of SpondyloArthritis international Society classification criteria for axial spondyloarthritis (part I): classification of paper patients by expert opinion including uncertainty appraisal . Ann Rheum Dis 2009 ;68 :770 –6 . 10.1136/ard.2009.108217 19297345 32 Selvaag AM , Lien G , Sorskaar D Early disease course and predictors of disability in juvenile rheumatoid arthritis and juvenile spondyloarthropathy: a 3 year prospective study . J Rheumatol 2005 ;32 :1122 –30 .15940778 33 Flato B , Lien G , Smerdel A Prognostic factors in juvenile rheumatoid arthritis: a case-control study revealing early predictors and outcome after 14.9 years . J Rheumatol 2003 ;30 :386 –93 .12563700 34 Klotsche J , Minden K , Thon A Improvement in health-related quality of life for children with juvenile idiopathic arthritis after start of treatment with etanercept . Arthritis Care Res (Hoboken) 2014 ;66 :253 –62 . 10.1002/acr.22112 23983081 35 Rajmil L , Serra-Sutton V , Fernandez-Lopez JA [The Spanish version of the German health-related quality of life questionnaire for children and adolescents: the Kindl] . An Pediatr (Barc) 2004 ;60 :514 –21 . 10.1016/S1695-4033(04)78320-4 15207162 36 Helseth S , Lund T Assessing health-related quality of life in adolescents: some psychometric properties of the first Norwegian version of KINDL . Scand J Caring Sci 2005 ;19 :102 –9 . 10.1111/j.1471-6712.2005.00326.x 15877635 37 Muller-Godeffroy E , Lehmann H , Kuster RM [Quality of life and psychosocial adaptation in children and adolescents with juvenile idiopathic arthritis and reactive arthritis] . Z Rheumatol 2005 ;64 :177 –87 . 10.1007/s00393-005-0652-5 15868335

PMC005xxxxxx/PMC5372142.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01320610.1136/bmjopen-2016-013206Gastroenterology and HepatologyResearch150616951689Comparison of the prognostic values of inflammation markers in patients with acute pancreatitis: a retrospective cohort study Li Yuanyuan Zhao Ying Feng Limin Guo Renyong Department of Laboratory Medicine, The First Affiliated Hospital, College of Medicine, Zhejiang University, Hangzhou, ChinaCorrespondence to Dr Renyong Guo; guorenyongzyyy@126.com2017 27 3 2017 7 3 e01320626 6 2016 19 2 2017 6 3 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives Inflammation-based prognostic markers (neutrophil–lymphocyte ratio (NLR), prognostic nutritional index (PNI), red cell distribution width (RDW) and lymphocyte–monocyte ratio (LMR)) are associated with overall survival in some diseases. This study assessed their prognostic value in mortality and severity in acute pancreatitis (AP). Design A retrospective cohort study. Setting Patients with AP were recruited from the emergency department at our hospital. Participants A total of 359 patients with AP (31 non-survivors) were enrolled. Primary and secondary outcome measures Mortality and severity of AP were the primary and secondary outcome measures, respectively. Biochemistry and haematology results of the first test after admission were collected. Independent relationships between severe AP (SAP) and markers were assessed using multivariate logistic regression models. Mortality prediction ability was evaluated using receiver operating characteristic (ROC) curves. Overall survival was evaluated using the Kaplan-Meier method, with differences compared using the log-rank test. Independent relationships between mortality and each predictor were estimated using the Cox proportional hazard models. Results Compared with survivors of AP, non-survivors had higher RDW (p<0.001), higher NLR (p<0.001), lower LMR (p<0.001) and lower PNI (p<0.001) at baseline. C reactive protein (CRP; OR=8.251, p<0.001), RDW (OR=2.533, p=0.003) and PNI (OR=7.753, p<0.001) were independently associated with the occurrence of SAP. For predicting mortality, NLR had the largest area under the ROC curve (0.804, p<0.001), with a 16.64 cut-off value, 82.4% sensitivity and 75.6% specificity. RDW was a reliable marker for excluding death owing to its lowest negative likelihood ratio (0.11). NLR (HR=4.726, p=0.004), CRP (HR=3.503, p=0.003), RDW (HR=3.139, p=0.013) and PNI (HR=2.641, p=0.011) were independently associated with mortality of AP. Conclusions NLR was the most powerful marker of overall survival in this patient series. acute pancreatitismortalityneutrophil-lymphocyte ratioprognostic nutritional index ==== Body Strengths and limitations of this study Compared with survivors of acute pancreatitis (AP), non-survivors had higher red cell distribution width (RDW) and neutrophil–lymphocyte ratio (NLR), and lower lymphocyte–monocyte ratio (LMR) and prognostic nutritional index (PNI) at baseline. NLR exhibited a higher area under the receiver operating characteristic curve for the prediction of mortality compared with other markers. RDW was suitable as a reliable marker to exclude death. NLR, PNI, C reactive protein and RDW were independently associated with overall survival of AP. This was a retrospective cohort analysis. Introduction Acute pancreatitis (AP) is rapid-onset inflammation of the pancreas that varies in severity from a self-limiting mild illness to rapidly progressive multiple organ failure. Statistics suggest that 10–20% of patients with AP develop severe AP (SAP),1 which usually has an unfavourable disease progression and is associated with a poor prognosis.2 3 Prediction of disease severity can guide the management of patients with AP and improve the outcome. Organ failure and infected pancreatic necrosis are common causes of mortality in such patients,4 and a new international multidisciplinary classification of SAP incorporates both events as determinants of severity.5 The predictive values of various markers, such as Acute Physiology and Chronic Health Evaluation II (APACHE II) and Bedside Index of Severity in Acute Pancreatitis scores, C reactive protein (CRP) and procalcitonin, have been previously assessed.6–8 A systematic review concluded that it was justifiable to use blood urea nitrogen after 48 hours of hospital admission for predicting persistent organ failure.9 In clinical studies, most studies have focused on disease severity, and only a few have directly investigated the relationship between predictors and mortality of AP. Furthermore, no reliable predictor of persistent organ failure within 48 hours of admission has been identified.9 There is increasing evidence that the presence of a systemic inflammatory response is associated with poor survival in patients with various aetiologies, including malignancy.10–17 Many direct or combined markers of systemic inflammation are based on routine, inexpensive and readily available laboratory tests. Red cell distribution width (RDW),10 neutrophil–lymphocyte ratio (NLR), prognostic nutritional index (PNI)11 and lymphocyte–monocyte ratio (LMR)12 have been used to predict the prognosis of disease. RDW was found to be an independent marker of short-term and long-term prognosis in intensive care units.10 NLR at admission served as an independent predictor of 3-month mortality rates in patients with acute-on-chronic liver failure.13 Increased pretreatment LMR was associated with a significantly more favourable prognosis in patients with solid tumours.12 Despite this evidence, very few studies have focused on the direct relationship between inflammation-based prognostic markers and mortality of AP. A cross-sectional study found a significant association between RDW and mortality in patients with AP.18 Another study investigated the prognostic value of NLR in AP and determined an optimal ratio for prediction of severity.19 To the best of our knowledge, the current study is the first to simultaneously compare the prognostic value of these inflammation-based prognostic markers (NLR, PNI, CRP, RDW and LMR) of mortality in patients with AP. Materials and methods Participants This retrospective cohort analysis consecutively enrolled a series of patients with AP who were admitted to the emergency department at our hospital between 1 July 2013 and 18 August 2015. A diagnosis of AP required two of three features: (1) prolonged abdominal pain characteristic of AP, (2) threefold elevation of serum amylase and/or lipase levels above the normal range, and (3) characteristic findings of AP on abdominal ultrasonography and/or CT scan.1 Mild AP (MAP) was defined as an absence of organ failure and an absence of local or systemic complications.1 Moderately SAP (MSAP) was defined as no evidence of persistent organ failure, but the presence of local or systemic complications and/or organ failure that resolved within 48 hours. SAP was defined as persistent organ failure (>48 hours).1 Patients with recurrent pancreatitis were enrolled only at first admission. Patients with traumatic pancreatitis, autoimmune pancreatitis, diabetes mellitus, tumour or liver failure were excluded. The prognostic information we focused on included overall survival and the severity of the disease. All enrolled patients were followed for 100 days or until death. All clinical data were retrieved from medical records. For patients with AP, 100 days of prognostic information (survival or non-survival) was obtained by checking medical records or by contacting the patients' family members. Ethics statement Each participant provided written informed consent after being provided with an explanation of the study by phone, letter or email. The study was conducted in accordance with the ethical principles contained within the Declaration of Helsinki. Demographic information and laboratory analysis Demographic information, including age, sex, aetiology and complication, was collected from medical records. Pretreatment laboratory data, including complete blood counts, serum CRP, albumin and amylase, were obtained during the emergency visit. An XE-2100 haematology autoanalyser (Sysmex Corp, Kobe, Japan), a Hitachi 7600 chemistry analyser (Hitachi High-Technologies, Tokyo, Japan) and Roche reagents (Roche Diagnostics, Indianapolis, Indiana, USA) were used in the laboratory. We assessed the prognostic value of general inflammation-based prognostic markers (NLR, CRP, RDW, PNI and LMR) for predicting the mortality of AP. Additionally, their ability to predict the severity of AP (SAP or not SAP) was assessed. NLR and LMR were ratios of two types of blood cell. PNI=albumin (g/L)+5×total lymphocyte count (109/L). Statistical analysis Variables are expressed as mean±SD or median (range) and categorical data as percentages, as appropriate. Differences between the two groups were assessed using an independent sample t-test, Mann-Whitney U test or χ2 test, as appropriate. Multiple comparisons were performed by one-way analysis of variance or Kruskal-Wallis H tests, as appropriate. The Bonferroni method was used to adjust for multiple comparisons. Multivariate logistic regression analyses were used to assess whether the inflammation markers were independent factors for predicting SAP in patients with AP by unadjusted and adjusted models successively. Patients with AP were randomly divided into estimation and validation cohorts by random number generators. The accuracy of each marker to predict mortality was assessed using receiver operating characteristic (ROC) curves. The sensitivity, specificity, positive likelihood ratio (+LR) and negative likelihood ratio (−LR) were calculated. +LR represents the ratio of the true-positive rate to the false-positive rate. −LR represents the ratio of the false-negative rate to the true-negative rate. These two parameters, which are not influenced by prevalence rate, are stable and objective for assessing diagnostic value. Combination models were developed using binary logistic regression analyses. Overall survival curves were calculated using the Kaplan-Meier method, and differences in survival rates were compared using the log-rank test. Univariate and multivariate Cox proportional hazard models were used to estimate the significance and independence of the relationship of each marker and mortality. The variables with a p value <0.1 in univariate analysis were included in a multivariate Cox proportional hazard regression model. A p value <0.05 was considered statistically significant. Statistical analyses were performed with SPSS V.19.0 (SPSS, Chicago, Illinois, USA). Results Patient characteristics A total of 359 patients with AP (197 MAP, 76 MSAP and 86 SAP) were enrolled in the study. The predefined probability of type I error was 0.05 (α=0.05), and the sample size was large enough to guarantee 0.90 of test power (β=0.1). Forty-five patients were excluded from the analysis, including those with traumatic pancreatitis (n=1), autoimmune pancreatitis (n=5), diabetes mellitus (n=7), tumour (n=7), liver failure (n=2) or incomplete medical records or who were lost to follow-up (n=23). Tables 1 and 2 show the baseline characteristics of the patients. There were no significant differences in age (p=0.352), aetiology (p=0.875) or sex (p=0.919) among the three groups (MAP, MSAP and SAP). As the illness worsened, CRP, RDW and NLR gradually increased, but PNI decreased (all p<0.05; table 1). LMR decreased significantly (p<0.001) in patients with MSAP compared with patients with MAP, but there was no significant difference between patients with MSAP and patients with SAP (p=0.883). Table 1 Demographics and laboratory findings in patients with acute pancreatitis Variables (1) MAP (n=197) (2) MSAP (n=76) (3) SAP (n=86) p Value All groups 1 vs 2 2 vs 3 Age (years) 51.43±16.00 48.47±13.28 50.69±14.61 0.352 0.446 1.000 Male (%) 108 (54.8%) 41 (53.9%) 49 (57.0%) 0.919 0.896 0.699 Aetiology (1/2/3/4)% 52%/12%/11%/25% 51%/16%/13%/20% 47%/15%/14%/24% 0.875 0.664 0.892 WCC (×109/L) 11.5 (3.1–32.0) 14.1 (4.5–36.8) 16.05 (5.9–38.4) <0.001 <0.001 0.278 Lymphocyte (×109/L) 1.1 (0.2–9.4) 1.0 (0.2–2.6) 0.80 (0.2–2.9) <0.001 0.004 0.089 Platelet (×109/L) 202 (21–502) 193 (58–548) 163 (27–540) 0.004 0.376 0.046 Albumin (g/L) 38.29±5.07 34.38±6.39 29.99±5.35 <0.001 <0.001 <0.001 CRP (mg/L) 53.9 (0.7–386) 133.6 (3.2–436.5) 196.1 (27.1–426.7) <0.001 <0.001 <0.001 Amylase (U/L) 398 (13–5191) 222 (27–3845) 581 (16–2377) 0.141 0.083 0.056 RDW (%) 12.8 (11.4–19.2) 13.0 (11.3–16.3) 13.7 (11.7–23.6) <0.001 0.013 0.014 NLR 8.46 (1.33–55) 14.60 (1.73–60) 19.65 (3.57–53.67) <0.001 <0.001 0.020 LMR 1.88 (0.28–13.33) 1.03 (0.29–5.33) 1.14 (0.22–6.32) <0.001 <0.001 0.883 PNI 44.53±6.63 39.36±6.71 34.55±6.02 <0.001 <0.001 <0.001 Mortality (%) 0 (0%) 0 (0%) 31 (36.0%) <0.001 – <0.001 Continuous variables are presented as mean±SD or median (range). Aetiology (1/2/3/4)%, 1, 2, 3 and 4 represent gallstone, alcohol, hypertriglyceridaemia and other aetiologies, respectively. 1 vs 2, MAP group versus MSAP group; 2 vs 3, MSAP group versus SAP group. CRP, C reactive protein; LMR, lymphocyte–monocyte ratio; MAP, mild acute pancreatitis; MSAP, moderately severe acute pancreatitis; NLR, neutrophil–lymphocyte ratio; PNI, prognostic nutritional index; RDW, red cell distribution width; SAP, severe acute pancreatitis; WCC, white cell count. Table 2 Demographics and laboratory findings in survivors and non-survivors of acute pancreatitis Variables Survivors (n=328) Non-survivors (n=31) p Value Age (years) 49.84±14.88 58.90±15.60 0.001 Male (%) 179 (54.6%) 19 (61.3%) 0.472 Aetiology (1/2/3/4)% 50%/13%/12%/25% 58%/19%/10%/13% 0.346 WCC (×109/L) 12.85 (3.1–38.4) 18.5 (6.5–29.3) 0.001 Lymphocytes (×109/L) 1.08 (0.17–9.40) 0.60 (0.30–1.60) <0.001 Platelet (×109/L) 197 (21–548) 159 (27–376) 0.001 Albumin (g/L) 35.95±6.30 30.44±5.54 <0.001 CRP (mg/L) 98.6 (0.7–436.5) 239.2 (27.1–398.2) <0.001 Amylase (U/L) 343.5 (13–5191) 909 (16–2377) 0.010 RDW (%) 13 (11.3–19.2) 13.8 (12.6–23.6) <0.001 NLR 10.47 (1.33–60.0) 25.0 (8.67–53.67) <0.001 PNI 41.71±7.50 34.00±6.35 <0.001 LMR 1.51 (0.22–13.33) 1.13 (0.24–2.26) <0.001 Continuous variables are presented as mean±SD or median (range). Aetiology (1/2/3/4)%, 1, 2, 3 and 4 represent gallstone, alcohol, hypertriglyceridaemia and other aetiologies, respectively. CRP, C reactive protein; LMR, lymphocyte–monocyte ratio; NLR, neutrophil–lymphocyte ratio; PNI, prognostic nutritional index; RDW, red cell distribution width; WCC, white cell count. Compared with survivors of AP, non-survivors were older (p=0.001) and had higher CRP (p<0.001), amylase (p=0.010), RDW (p<0.001) and NLR (p<0.001). Conversely, lymphocyte count (p<0.001), platelets (p=0.001), albumin (p<0.001), LMR (p<0.001) and PNI (p<0.001) were lower in non-survivors than in survivors (table 2). The relationship between markers and severity of AP The multivariate logistic regression models revealed that high CRP (>110 vs ≤110 mg/L, adjusted OR=8.251, 95% CI 3.897 to 17.468, p<0.001), RDW (>13.0% vs ≤13.0%, adjusted OR=2.533, 95% CI 1.365 to 4.702, p=0.003) and low PNI (<41.1 vs ≥41.1, adjusted OR=7.753, 95% CI 3.400 to 17.680, p<0.001) were independent factors for predicting SAP in patients with AP (table 3). Table 3 ORs of prognostic factors for predicting SAP in patients with AP Model 1 Model 2 Model 3 Factors OR (95% CI) p Value OR (95% CI) p Value OR (95% CI) p Value NLR (>11.36 vs ≤11.36) 3.707 (2.173 to 6.326) <0.001 3.578 (2.082 to 6.149) <0.001 1.463 (0.711 to 3.010) 0.301 CRP(>110 vs ≤110 mg/L) 9.867 (5.116 to 19.030) <0.001 12.609 (6.304 to 25.218) <0.001 8.251 (3.897 to 17.468) <0.001 RDW (>13.0% vs ≤13.0%) 3.368 (2.003 to 5.663) <0.001 3.529 (2.076 to 5.998) <0.001 2.533 (1.365 to 4.702) 0.003 PNI (<41.1 vs ≥41.1) 9.951 (5.055 to 19.589) <0.001 11.356 (5.665 to 22.766) <0.001 7.753 (3.400 to 17.680) <0.001 LMR (<1.43 vs ≥1.43) 2.564 (1.539 to 4.271) <0.001 2.552 (1.524 to 4.274) <0.001 0.722 (0.355 to 1.471) 0.370 Model 1: unadjusted model. Model 2: adjusted for age, gender and amylase. Model 3: NLR was adjusted for age, gender, amylase, CRP, RDW, PNI and LMR; CRP was adjusted for age, gender, amylase, NLR, RDW, PNI and LMR; RDW was adjusted for age, gender, amylase, CRP, NLR, PNI and LMR; PNI was adjusted for age, gender, amylase, NLR, CRP, RDW and LMR; LMR was adjusted for age, gender, amylase, NLR, CRP, RDW and PNI. AP, acute pancreatitis; CRP, C reactive protein; LMR, lymphocyte–monocyte ratio; NLR, neutrophil–lymphocyte ratio; PNI, prognostic nutritional index; RDW, red cell distribution width; SAP, severe acute pancreatitis. The markers' power for predicting 100 days mortality The enrolled 359 patients with AP were randomly grouped into two cohorts: the estimation cohort (n=181) and the validation cohort (n=178). No significant difference was observed between the estimation and the validation cohorts in all characteristics (see online supplementary table S1). ROC curves of the estimation cohort were constructed to evaluate the ability of each marker to predict 100 days mortality in AP. Table 4 shows the area under the ROC curves (AUC) and optimal cut-off values. The ability of NLR to predict mortality (AUC=0.804, p<0.001) was good; those of PNI (AUC=0.769, p<0.001), CRP (AUC=0.774, p<0.001), RDW (AUC=0.769, p<0.001) and LMR (AUC=0.744, p<0.001) were fair. The NLR had the largest AUC, and RDW and PNI had the highest sensitivity and specificity, respectively. Therefore, these three markers were selected for combination. The AUC for NLR+PNI, NLR+RDW and PNI+RDW were 0.825 (95% CI 0.761 to 0.877), 0.854 (95% CI 0.794 to 0.902) and 0.806 (95% CI 0.741 to 0.861), respectively (figure 1). There were no significant differences in AUC for combined index and NLR (p=0.699, p=0.167 and p=0.975, respectively). Table 4 Discriminatory ability of inflammation-based markers for predicting mortality in patients with AP Index AUC (95% CI) p Value* Cut-off† Sensitivity (%) Specificity (%) +LR −LR Training cohort NLR 0.804 (0.738 to 0.859) <0.001 16.64 82.4 75.6 3.38 0.23 CRP 0.774 (0.706 to 0.833) <0.001 162.2 mg/L 76.5 73.8 2.92 0.32 RDW 0.769 (0.700 to 0.828) <0.001 13.0% 94.1 54.3 2.06 0.11 PNI 0.769 (0.701 to 0.828) <0.001 33.1 58.8 88.4 5.08 0.47 LMR 0.744 (0.674 to 0.806) <0.001 1.40 82.4 57.3 1.93 0.31 Validation cohort NLR 0.851 (0.790 to 0.900) <0.001 16.64 85.7 73.8 3.27 0.19 CRP 0.753 (0.683 to 0.815) <0.001 162.2 mg/L 71.4 65.2 2.06 0.44 RDW 0.708 (0.635 to 0.773) 0.001 13.0% 85.7 50.0 1.71 0.29 PNI 0.791 (0.724 to 0.848) <0.001 33.1 42.9 88.4 3.70 0.65 LMR 0.677 (0.603 to 0.745) 0.015 1.40 78.6 49.4 1.55 0.43 Overall NLR 0.823 (0.780 to 0.861) <0.001 16.64 83.9 74.4 3.27 0.22 CRP 0.762 (0.714 to 0.805) <0.001 162.2 mg/L 74.2 69.8 2.46 0.37 RDW 0.742 (0.693 to 0.786) <0.001 13.0% 90.3 49.7 1.80 0.19 PNI 0.781 (0.734 to 0.822) <0.001 33.1 51.6 88.4 4.46 0.55 LMR 0.710 (0.660 to 0.757) <0.001 1.40 77.4 54.0 1.68 0.42 *The p value is comparing the AUC with 0.5. †The cut-off values were derived from a training cohort. −LR, negative likelihood ratio; +LR, positive likelihood ratio; AP, acute pancreatitis; AUC, area under the receiver operating characteristic curve; CRP, C reactive protein; LMR, lymphocyte–monocyte ratio; NLR, neutrophil–lymphocyte ratio; PNI, prognostic nutritional index; RDW, red cell distribution width. Figure 1 ROC curves analysis for predicting mortality by NLR and combined markers in the estimation cohort. ROC, receiver operating characteristic; NLR, neutrophil–lymphocyte ratio; PNI, prognostic nutritional index; RDW, red cell distribution width. 10.1136/bmjopen-2016-013206.supp1supplementary table Demographics and laboratory findings in estimation and validation cohorts For NLR, the optimal cut-off value for mortality prediction was 16.64, with a sensitivity of 82.4% and specificity of 75.6%. RDW had the highest sensitivity (94.1%) and lowest −LR (0.11), so it was a reliable predictive index for excluding mortality in patients with AP. PNI had the highest specificity (88.4%) and +LR (5.08), so it was most suitable for use as a confirmed index among the indexes assessed. In the validation cohort, AUCs for NLR, CRP, RDW, PNI and LMR were 0.851 (95% CI 0.790 to 0.900), 0.753 (95% CI 0.683 to 0.815), 0.708 (95% CI 0.635 to 0.773), 0.791 (95% CI 0.724 to 0.848) and 0.677 (95% CI 0.603 to 0.745), respectively. There were no significant differences in AUC for NLR, CRP, RDW, PNI and LMR between the estimation and validation cohorts (p=0.477, p=0.809, p=0.437, p=0.782 and p=0.455, respectively). Survival analysis Patients with AP were stratified into groups by cut-off values. Kaplan-Meier survival curves demonstrate the relationships between inflammation-based prognostic markers and overall survival of patients with AP (figure 2A−E). Elevated NLR (p<0.001), CRP (p<0.001) and RDW (p<0.001) were associated with increased probability of death. Conversely, decreased PNI (p<0.001) and LMR (p=0.001) were associated with decreased overall survival. Figure 2 Relationship between inflammation-based prognostic markers and overall survival in patients with acute pancreatitis. A, B, C, D and E show the relationship between NLR, CRP, RDW, PNI and LMR, and overall survival in patients with acute pancreatitis, respectively. CRP, C reactive protein; LMR, lymphocyte–monocyte ratio; NLR, neutrophil–lymphocyte ratio; PNI, prognostic nutritional index; RDW, red cell distribution width. According to the cut-off values for the factors, low NLR (≤16.64), low CRP (≤162.2 mg/L), low RDW (≤13.0%), high PNI (>33.1) and high LMR (>1.40) were selected as references. Univariate analysis and Cox regression revealed that age (p<0.001), amylase (p=0.001), NLR (p<0.001), PNI (p<0.001), CRP (p<0.001), RDW (p<0.001) and LMR (p=0.002) were associated with AP mortality (table 5). These factors were evaluated using multivariate Cox regression. Age (HR=4.039, 95% CI 1.873 to 8.713, p<0.001), NLR (HR=4.726, 95% CI 1.627 to 13.726, p=0.004), CRP (HR=3.503, 95% CI 1.534 to 7.999, p=0.003), RDW (HR=3.139, 95% CI 1.277 to 7.714, p=0.013) and PNI (HR=2.641, 95% CI 1.248 to 5.590, p=0.011) were independently associated with mortality of AP (table 5). Table 5 Prognostic factors of overall survival in patients with acute pancreatitis by univariate and multivariate analyses Univariate analysis Multivariate analysis Factors HR (95% CI) p Value HR (95% CI) p Value Age (>63 vs ≤63 years) 5.384 (2.653 to 10.925) <0.001 4.039 (1.873 to 8.713) <0.001 Gender (female vs male) 0.767 (0.372 to 1.579) 0.471 Amylase (>618 vs ≤618 U/L) 3.544 (1.699 to 7.526) 0.001 2.173 (0.965 to 4.891) 0.061 NLR (>16.64 vs ≤16.64) 13.130 (5.041 to 34.205) <0.001 4.726 (1.627 to 13.726) 0.004 CRP (>162.2 vs ≤162.2 mg/L) 6.127 (2.740 to 13.701) <0.001 3.503 (1.534 to 7.999) 0.003 RDW (>13.0% vs ≤13.0%) 4.929 (2.022 to 12.017) <0.001 3.139 (1.277 to 7.714) 0.013 PNI (≤33.1 vs >33.1) 6.912 (3.414 to 13.991) <0.001 2.641 (1.248 to 5.590) 0.011 LMR (≤1.40 vs >1.40) 3.797 (1.636 to 8.813) 0.002 1.036 (0.403 to 2.659) 0.942 CRP, C reactive protein; LMR, lymphocyte–monocyte ratio; NLR, neutrophil–lymphocyte ratio; PNI, prognostic nutritional index; RDW, red cell distribution width. Discussion AP is an inflammatory disease, with mortality arising mainly from organ failure or infected pancreatic necrosis.4 Our study estimated the prognostic value of various inflammation-based prognostic markers for predicting mortality of AP. According to classifications of AUC,20 21 the ability of the NLR to predict mortality was good, while those of PNI, CRP, RDW and LMR were fair. Cox regression analysis revealed that age, NLR, PNI, CRP and RDW were independently associated with mortality of AP. Additionally, PNI, CRP and RDW were independently associated with the occurrence of SAP in patients with AP. NLR, CRP, RDW and PNI are inexpensive, convenient and readily available in clinical settings. From examination of AUC, NLR had the best performance. With an NLR>16.64 at the time of admission, the risk of dying increased 3.726-fold compared with NLR≤16.64. RDW was the most reliable marker for excluding death in patients with AP, owing to its lowest −LR (0.11). PNI had the highest specificity (88.4%) and +LR (5.08), so it was most suitable to be a confirmed index among the indexes assessed. However, fluctuations in the NLR and CRP can be influenced by the use of antibiotics; therefore, NLR and CRP are not suitable for patients undergoing intensive use of antibiotics. Similarly, blood transfusion and parenteral nutrition may affect RDW and PNI, respectively, so the predictive value of RDW and PNI in these patients was discounted. In AP, inflammation propagates and promotes tissue destruction via activation of a cascade of inflammatory cytokines, proteolytic enzymes and oxygen-free radicals.19 22 Neutrophils, lymphocytes and monocytes are the three main types of white cell counts (WCC). Neutrophils play a key role in the development of local tissue destruction and systemic complications of SAP.23 Depletion of neutrophils has been associated with an improved prognosis of AP.23 The percentage of immature neutrophilic granulocytes might be used clinically as a simple early predictor of an adverse outcome in SAP.24 Additionally, recent studies revealed that the extent of lymphopaenia was associated with disease severity.25–27 Lymphopaenia has been reported to have independent prognostic value for some diseases,19 26–29 including AP. Takeyama et al28 found that impairment of cellular immunity caused by peripheral lymphocyte apoptosis was linked to the subsequent development of infectious complications in AP. Monocytes produce various cytokines and inflammatory mediators that further amplify inflammatory cell recruitment into the pancreas as well as distant organs such as the lungs.30 Similar to neutrophils, a protective effect was also found by depleting macrophages in a mouse model of AP.31 Theoretically, NLR and LMR, which combine two opposing parameters, should be more accurate than either parameter alone. We found that the NLR had the greatest prognostic value of all the factors we evaluated. It is, however, important to apply the NLR with caution in clinical settings. Broad-spectrum antibiotics with good tissue penetration, which are essential medicines in the treatment of SAP, can affect WCC by reducing inflammation. Thus, the prognostic value of NLR in AP is uncertain if the effect of antibiotic treatment is not taken into account.32 For this reason, the neutrophil and lymphocyte counts used in this study were from the first complete blood cell count conducted during the emergency visit. We confirmed that the enrolled patients were untreated at that time; consequently, our results are most likely applicable to untreated patients. Unlike for the NLR, the predictive ability of the LMR was only fair, and was not independently associated with overall survival in AP. Serum albumin is a negative acute phase response reactant, and reflects the body's nutritional status. Albumin <25 g/L was an independent prognostic factor related to a poor prognosis of AP.33 Variation of albumin within 24 hours has been identified as a risk factor for a poor prognosis of critically ill patients in the early stages of SAP.34 The PNI, which includes serum albumin and lymphocyte count, is an independent predictor of poor overall survival in patients with hepatocellular carcinoma.35 To the best of our knowledge, few studies have reported on the application of PNI for predicting mortality of AP, but we found that it was an independent prognostic factor, and was suitable as a confirmed marker. Numerous studies have reported RDW as a strong independent prognostic factor in various diseases and conditions, such as cardiovascular diseases, rheumatoid arthritis, cancer and critical illnesses.18 36–38 Our results are consistent with the study by Yao and Lv,18 who reported a significant association between RDW and mortality of patients with AP. Additionally, we found that RDW was most suitable as a reliable excluding marker among the markers we assessed. The mechanisms underlying the association between RDW and mortality in AP remain unclear. The obvious metabolic abnormalities in non-survivors of AP, including inflammation, oxidative stress, poor nutritional status and persistent organ failure, lead to deregulation of red blood cell homoeostasis involving both impaired erythropoiesis and abnormal red blood cell survival.38 RDW reflects these impairments in homoeostasis, but only further research can confirm this speculation. The prognostic markers evaluated in this study are direct or combined markers of systemic inflammation that are based on routine, inexpensive and readily available laboratory tests. To the best of our knowledge, this is the first study to compare the prognostic value of these markers for predicting mortality in patients with AP simultaneously. Additionally, suitable excluding and identifying markers were found. Some potential limitations of the study should be noted. Although we have taken special care to avoid sources of bias and confounding, some potential bias may still exist in this retrospective, single-centre study. Information available at the beginning of the study may have affected the selection of the study participants, although the medical records and laboratory data were collected separately by two people. The reasons for incomplete medical records or why patients were lost to follow-up (n=23) are not known. These patients were excluded from the analyses. As a result, a larger prospective study is needed to validate the results. Second, only the first set of admission blood results were investigated. Since factors change with time, they should be surveyed in the future because of the rapid onset of inflammation. Third, the typical prediction models, such as the APACHE II score, should be included in future research. Fourth, for better validity, +LR should be near 10, and −LR should be 0.2. Unfortunately, no marker examined had perfect +LR and −LR simultaneously. Therefore, the markers should be selected with caution based on particular needs. The marker with the higher +LR was more suitable to confirm death, while the marker with the lower −LR was more suitable to exclude death. Appropriate judgement based on the available information will be more reliable; however, these markers are still valuable based on their acceptable AUC. Finally, we only described the association of each of the predictors with mortality of AP; the underlying mechanisms need to be investigated. In conclusion, we found that age, NLR, PNI, CRP and RDW were independently associated with overall survival of AP. NLR had the best overall performance, RDW was suitable as a reliable marker to exclude death, and PNI was a good predictive marker for death. When applying these markers, any possible influence from therapy should be considered. The authors thank Edanz Group for helping edit the English of the final manuscript. Contributors: RG and YL designed the experiments. RG and YL contributed to the data collection. YZ conducted the data analysis. YL, RG and LF wrote the manuscript. All authors reviewed the manuscript. Funding: This work was financially supported by grants from the Zhejiang Provincial Natural Science Foundation of China (LY15H190002) and the Department of Education Foundation of Zhejiang Province, China (Y201330146). Competing interests: None declared. Patient consent: Obtained. Ethics approval: This study was approved by the Ethics Committee of the First Affiliated Hospital of Zhejiang University School of Medicine, China. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Banks PA , Bollen TL , Dervenis C Classification of acute pancreatitis—2012: revision. The diagnosis of acute pancreatitis requires two of the following three features: and definitions by international consensus . Gut 2013 ;62 :102 –11 . 10.1136/gutjnl-2012-302779 23100216 2 Maheshwari R , Subramanian RM Severe acute pancreatitis and necrotizing pancreatitis . Crit Care Clin 2016 ;32 :279 –90 . 10.1016/j.ccc.2015.12.006 27016168 3 Bugiantella W , Rondelli F , Boni M Necrotizing pancreatitis: a review of the interventions . Int J Surg 2016 ;28 (Suppl 1 ):S163 –71 . 10.1016/j.ijsu.2015.12.038 26708848 4 Petrov MS , Shanbhag S , Chakraborty M Organ failure and infection of pancreatic necrosis as determinants of mortality in patients with acute pancreatitis . Gastroenterology 2010 ;139 :813 –20 . 10.1053/j.gastro.2010.06.010 20540942 5 Dellinger EP , Forsmark CE , Layer P Determinant-based classification of acute pancreatitis severity: an international multidisciplinary consultation . Ann Surg 2012 ;256 :875 –80 . 10.1097/SLA.0b013e318256f778 22735715 6 Papachristou GI , Muddana V , Yadav D Comparison of BISAP, Ranson's, APACHE-II, and CTSI scores in predicting organ failure, complications, and mortality in acute pancreatitis . Am J Gastroenterol 2010 ;105 :435 –41 ; quiz 42 10.1038/ajg.2009.622 19861954 7 Mounzer R , Langmead CJ , Wu BU Comparison of existing clinical scoring systems to predict persistent organ failure in patients with acute pancreatitis . Gastroenterology 2012 ;142 :1476 –82 ; quiz e15–6 10.1053/j.gastro.2012.03.005 22425589 8 Bezmarevic M , Mirkovic D , Soldatovic I Correlation between procalcitonin and intra-abdominal pressure and their role in prediction of the severity of acute pancreatitis . Pancreatology 2012 ;12 :337 –43 . 10.1016/j.pan.2012.05.007 22898635 9 Yang CJ , Chen J , Phillips AR Predictors of severe and critical acute pancreatitis: a systematic review . Dig Liver Dis 2014 ;46 :446 –51 . 10.1016/j.dld.2014.01.158 24646880 10 Hunziker S , Celi LA , Lee J Red cell distribution width improves the simplified acute physiology score for risk prediction in unselected critically ill patients . Crit Care 2012 ;16 :R89 10.1186/cc11351 22607685 11 Kinoshita A , Onoda H , Imai N Comparison of the prognostic value of inflammation-based prognostic scores in patients with hepatocellular carcinoma . Br J Cancer 2012 ;107 :988 –93 . 10.1038/bjc.2012.354 22878374 12 Teng JJ , Zhang J , Zhang TY Prognostic value of peripheral blood lymphocyte-to-monocyte ratio in patients with solid tumors: a meta-analysis . Onco Targets Ther 2016 ;9 :37 –47 . 10.2147/OTT.S94458 26730202 13 Chen L , Lou Y , Chen Y Prognostic value of the neutrophil-to-lymphocyte ratio in patients with acute-on-chronic liver failure . Int J Clin Pract 2014 ;68 :1034 –40 . 10.1111/ijcp.12408 24666824 14 Smith RA , Bosonnet L , Raraty M Preoperative platelet-lymphocyte ratio is an independent significant prognostic marker in resected pancreatic ductal adenocarcinoma . Am J Surg 2009 ;197 :466 –72 . 10.1016/j.amjsurg.2007.12.057 18639229 15 Proctor MJ , Morrison DS , Talwar D An inflammation-based prognostic score (mGPS) predicts cancer survival independent of tumour site: a Glasgow Inflammation Outcome Study . Br J Cancer 2011 ;104 :726 –34 . 10.1038/sj.bjc.6606087 21266974 16 Proctor MJ , Morrison DS , Talwar D A comparison of inflammation-based prognostic scores in patients with cancer. A Glasgow Inflammation Outcome Study . Eur J Cancer 2011 ;47 :2633 –41 . 10.1016/j.ejca.2011.03.028 21724383 17 Zampieri FG , Ranzani OT , Sabatoski V An increase in mean platelet volume after admission is associated with higher mortality in critically ill patients . Ann Intensive Care 2014 ;4 :20 10.1186/s13613-014-0020-1 25520853 18 Yao J , Lv G Association between red cell distribution width and acute pancreatitis: a cross-sectional study . BMJ Open 2014 ;4 :e004721 10.1136/bmjopen-2013-004721 19 Suppiah A , Malde D , Arab T The prognostic value of the neutrophil-lymphocyte ratio (NLR) in acute pancreatitis: identification of an optimal NLR . J Gastrointest Surg 2013 ;17 :675 –81 . 10.1007/s11605-012-2121-1 23371356 20 Zweig MH , Campbell G Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine . Clin Chem 1993 ;39 :561 –77 .8472349 21 Altman DG , Bland JM Diagnostic tests 3: receiver operating characteristic plots . BMJ 1994 ;309 :188 .8044101 22 Felderbauer P , Muller C , Bulut K Pathophysiology and treatment of acute pancreatitis: new therapeutic targets--a ray of hope? Basic Clin Pharmacol Toxicol 2005 ;97 :342 –50 . 10.1111/j.1742-7843.2005.pto_274.x 16364048 23 Xue J , Sharma V , Habtezion A Immune cells and immune-based therapy in pancreatitis . Immunol Res 2014 ;58 :378 –86 . 10.1007/s12026-014-8504-5 24710635 24 Zhu L , Chen G , Xia Q Use of band cell percentage as an early predictor of death and ICU admission in severe acute pancreatitis . Hepatogastroenterology 2010 ;57 :1543 –8 .21443118 25 Zahorec R Ratio of neutrophil to lymphocyte counts—rapid and simple parameter of systemic inflammation and stress in critically ill . Bratisl Lek Listy 2001 ;102 :5 –14 .11723675 26 de Jager CP , van Wijk PT , Mathoera RB Lymphocytopenia and neutrophil-lymphocyte count ratio predict bacteremia better than conventional infection markers in an emergency care unit . Crit Care 2010 ;14 :R192 10.1186/cc9309 21034463 27 Le Tulzo Y , Pangault C , Gacouin A Early circulating lymphocyte apoptosis in human septic shock is associated with poor outcome . Shock 2002 ;18 :487 –94 .12462554 28 Takeyama Y , Takas K , Ueda T Peripheral lymphocyte reduction in severe acute pancreatitis is caused by apoptotic cell death . J Gastrointest Surg 2000 ;4 :379 –87 .11058856 29 Pezzilli R , Billi P , Beltrandi E Circulating lymphocyte subsets in human acute pancreatitis . Pancreas 1995 ;11 :95 –100 .7667248 30 McKay C , Imrie CW , Baxter JN Mononuclear phagocyte activation and acute pancreatitis . Scand J Gastroenterol Suppl 1996 ;219 :32 –6 .8865469 31 Saeki K , Kanai T , Nakano M CCL2-induced migration and SOCS3-mediated activation of macrophages are involved in cerulein-induced pancreatitis in mice . Gastroenterology 2012 ;142 :1010 –20.e9 . 10.1053/j.gastro.2011.12.054 22248664 32 Binnetoglu E , Akbal E , Gunes F The prognostic value of neutrophil-lymphocyte ratio in acute pancreatitis is controversial . J Gastrointest Surg 2014 ;18 :885 10.1007/s11605-013-2435-7 24395068 33 Gonzalvez-Gasch A , de Casasola GG , Martin RB A simple prognostic score for risk assessment in patients with acute pancreatitis . Eur J Intern Med 2009 ;20 :e43 –8 . 10.1016/j.ejim.2008.09.014 19393477 34 Chen Y , Zhang ZW , Wang B Relationship between early serum albumin variation and prognosis in patients with severe acute pancreatitis treated in ICU . Sichuan Da Xue Xue Bao Yi Xue Ban 2013 ;44 :237 –41 .23745263 35 Pinato DJ , North BV , Sharma R A novel, externally validated inflammation-based prognostic algorithm in hepatocellular carcinoma: the prognostic nutritional index (PNI) . Br J Cancer 2012 ;106 :1439 –45 . 10.1038/bjc.2012.92 22433965 36 Lappegard J , Ellingsen TS , Vik A Red cell distribution width and carotid atherosclerosis progression. The Tromso Study . Thromb Haemost 2015 ;113 :649 –54 .25631329 37 Bekler A , Tenekecioglu E , Erbag G Relationship between red cell distribution width and long-term mortality in patients with non-ST elevation acute coronary syndrome . Anatol J Cardiol 2015 ;15 :634 –9 . 10.5152/akd.2014.5645 25550178 38 Salvagno GL , Sanchis-Gomar F , Picanza A Red blood cell distribution width: A simple parameter with multiple clinical applications . Crit Rev Clin Lab Sci 2015 ;52 :86 –105 . 10.3109/10408363.2014.992064 25535770

PMC005xxxxxx/PMC5372143.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01385010.1136/bmjopen-2016-013850Respiratory MedicineResearch150617311723RCT of the effect of berryfruit polyphenolic cultivar extract in mild steroid-naive asthma: a cross-over, placebo-controlled study Power Sharon 12Williams Mathew 1Semprini Alex 123Munro Claire 4Caswell-Smith Rachel 123Pilcher Janine 123Holliday Mark 1Fingleton James 123Harper Jacquie 5Hurst Roger 6Weatherall Mark 24Beasley Richard 12http://orcid.org/0000-0001-5327-3027Braithwaite Irene 123 1 Medical Research Institute of New Zealand, Wellington, New Zealand 2 Capital and Coast District Health Board, Wellington, New Zealand 3 Victoria University of Wellington, Wellington, New Zealand 4 University of Otago, Wellington, New Zealand 5 Malaghan Institute of Medical Research, Wellington, New Zealand 6 The New Zealand Institute for Plant & Food Research, Palmerston North, New ZealandCorrespondence to Dr Irene Braithwaite; Irene.Braithwaite@mrinz.ac.nz2017 20 3 2017 7 3 e01385011 8 2016 31 1 2017 2 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective There is preclinical evidence that consumption of berryfruit extract may reduce chronic airways inflammation and modify airway remodelling in allergen-induced models of lung inflammation. We investigated the effect of berryfruit extract on the fractional expired nitric oxide (FeNO), a biomarker of eosinophilic airways inflammation, in adults with steroid-naïve asthma. Design Randomised placebo-controlled cross-over double-blind trial. Setting Single-centre community-based trial. Participants 28 steroid-naïve mild asthmatics with Feno >40 ppb, of whom 25 completed both study interventions. Interventions Participants were randomised to receive, according to the cross-over design, 100 mg berryfruit polyphenolic extract (BFPE) or placebo for 4 weeks, with a 4-week washout period between the interventions. Primary outcome measure The primary outcome variable was FeNO at 4 weeks, analysed by a mixed linear model, with a random effect for participant and baseline FeNo as a covariate. Results The mean (SD) natural logarithm transformed (ln) FeNO after 4 weeks of treatment for the BFPE and placebo groups was 4.28 (0.47) and 4.22 (0.47), respectively. The paired change from baseline mean (SD) BFPE minus placebo ln FeNO was −0.03 (0.39), N=25. The mixed linear model estimate, with baseline covariate adjustment, difference in ln FeNO, was −0.002 (95% CI −0.15 to 0.14), p=0.98. This is equivalent to a ratio of geometric mean FeNO of 1.0 (95% CI 0.86 to 1.15). Conclusions In steroid-naïve participants with mild asthma and elevated FeNO, there was no effect of BFPE on FeNO, a biomarker of eosinophilic airways inflammation. Caution is required in the extrapolation of apparent benefit in murine models of lung eosinophilia to clinical efficacy in patients with asthma. Trial registration number ANZCTR: 12613000451707; Results. Animal modelsClinical respiratory medicine ==== Body Strengths and limitations of this study This was a cross-over, double-blind, randomised, placebo-controlled trial. Participants were steroid-naive patients with asthma with elevated fractional expired nitric oxide, maximising the likelihood of a response to the properties of berryfruit polyphenolic extract. The study participants' asthma may not have been severe enough to detect clinically significant improvements in lung function or quality of life Introduction The prevalence of asthma has markedly increased in many countries throughout the world over the past 50 years leading to asthma becoming one of the most common non-communicable diseases in children and adults.1 2 The reasons for these global trends are poorly understood. Furthermore, there are no primary prevention strategies that have undergone scrutiny in randomised controlled trials that have provided sufficient evidence to lead to widespread implementation in clinical practice.3 This has led to the consideration of novel strategies, emanating from epidemiological observations and animal experiments, from which proof-of-concept clinical trials can be undertaken in humans.4 In particular, dietary factors may contribute to the pathogenesis of asthma. There are a number of non-experimental studies exploring the association between fruit consumption and respiratory symptoms and incidence of lung diseases such as asthma in non-experimental studies.5–10 Consumption of kiwifruit and citrus fruits is associated with a lower prevalence of asthma in children11 and fruit and vegetable intake is also positively associated with reduced risk of asthma in adults.12 13 High fruit juice consumption has been associated with an increased risk of asthma in children.8 This association may be due to high excess fructose fruit juices such as apple, rather than other juices low in excess fructose such as orange, which may be protective.9 A possible mechanism for the variation in these findings is that consumption of fruits high in excess free fructose leads to the formation of proinflammatory advanced glycation end products (AGE) in the intestine, which are absorbed into the circulation, causing a systemic immune response. Since the lungs have a high concentration of AGE receptors,14 pulmonary tissue may be more affected by this mechanism than other organs. Until now, results of more robust study designs, such as randomised controlled trials, offer limited support to the hypothesis that specific fruit supplementation may be effective for asthma.15–19 Asthma is a chronic inflammatory disease of the airways20 and fruit-derived polyphenolic flavonoid compounds such as anthocyanins and proanthocyanins attenuate lung inflammation in preclinical studies21–23 and have immune modulatory actions.24 Berryfruits, as well as being low in excess free fructose,25 contain high amounts of anthocyanins and proanthocyanins, and in vitro studies indicate that they have the potential to reduce eosinophilic inflammation in asthma.26 In vivo feeding studies using ovalbumin-induced chronic lung inflammation models in mice show that consumption of a specific berry variety decreases numbers of mucus-producing cells and reduces collagen deposition, suggesting an effect on airway remodelling.27 As many as 40% of asthmatics use complementary and alternative medicine treatments.28 Research that identifies evidence of efficacy or lack of efficacy of these products is of interest to those with and those treating asthma. The objective of this study was to assess the effect of a berryfruit polyphenolic extract (BFPE) in mild asthma. Our hypothesis in this study is that BFPE reduces fractional expired nitric oxide (FeNO), a biomarker of eosinophilic airway inflammation, which is elevated in patients with asthma and is highly sensitive to anti-inflammatory inhaled corticosteroid treatment.29–34 We also investigated possible effects on airways resistance, serum eosinophils and asthma control over the same time period. Methods Participants and methods This was a cross-over, double-blind, randomised, cross-over, placebo-controlled trial in 28 participants with mild asthma to compare FeNO between use of BFPE and a placebo. Eligible participants were aged between 18 and 75 years, with a doctor's diagnosis of asthma, were steroid-naïve (no inhaled or oral corticosteroid in the past 90 days), and had an FeNO of >40 ppb. Participants were excluded if they were unable to provide informed consent, unable or unwilling to comply with study procedures (including not consuming confounding foodstuffs during study periods), pregnant, trying to conceive, or had a known hypersensitivity to berryfruits. Participants were recruited from an existing database within the Medical Research Institute of New Zealand (MRINZ). Contact was made via telephone and participants were invited to attend the MRINZ for screening (visit 1). The trial was approved by the New Zealand Health and Disability Ethics Committee (Wellington) and written informed consent was obtained from all the trial participants. Randomised treatments Participants were randomised to receive either encapsulated BFPE 1 000 mg daily or matching placebo capsules for 4 weeks, followed by a 4-week washout period, and then the other randomised treatment for 4 weeks. The BFPE and matching placebo tablets were supplied by Plant & Food Research. Randomisation The randomisation schedule was generated by the study statistician, independently of the investigators undertaking the visits, using a computer-generated random number sequence. To maintain blinding, bottles of tablets were placed in numbered opaque bags by non-clinical investigators. On randomisation, the bagged tablets were given to the participant by clinical investigators according to the randomisation schedule. Compliance counts were undertaken by non-clinical investigators. Study design The study comprised four visits (V1–V4) followed by a final phone call 1 week after completion of the second intervention (figure 1). At V1, participants were randomised in a ratio of 1:1 to the order in which they received 28 days of treatment with BFPE or placebo. There was a washout period between the two interventions of 28 days. This washout period was chosen as it is double the length of time for FeNO to return to baseline levels after cessation of inhaled corticosteroid treatment.35 Participants were requested to refrain from consuming berry-containing foods and drinks for the entire duration of the study. Figure 1 Study format. BFPE, berryfruit polyphenolic extract. During each 4-week intervention period, participants self-completed a questionnaire of their compliance in terms of consumption of foods and drinks which they had been asked to avoid, which were provided on a list (see online supplementary appendix 1). The amount, date and food type were documented and returned on subsequent visits (V2/V4). Investigational product and placebo compliance were captured via a capsule count performed at the end of each intervention period (V2/V4). 10.1136/bmjopen-2016-013850.supp1supplementary appendix Measurements At each visit, measurements were made of the FeNO, lung function and peripheral blood eosinophil count, and the asthma control questionnaire (ACQ)36 was administered. The FeNO concentration was measured as per the American Thoracic Society (ATS) guidelines37 via an online nitric oxide monitor (NIOX, Aerocrine AB, Solna, Sweden). Forced expiratory volume in 1 s (FEV1), specific airways conductance and resistance (sRaw/sGaw, measures of large airway function) and maximum mid-expiratory flow 25–75 (MMEF25–75, a measure of small airway function) were measured by body plethysmograph (Masterlab 4.5 and 4.6 Erich-Jaeger, Wurzberg, Germany) according to ATS guidelines.38 Venepuncture was carried out to obtain a full blood count for measurement of peripheral blood eosinophil count. All the procedures were undertaken at baseline and repeated at each visit at time points of 4, 8 and 12 weeks. Outcome variables The primary outcome variable was the FeNO. Secondary outcomes included ACQ-5 score, FEV1, MMEF25–75, sRaw, sGaw and peripheral blood eosinophil count. Statistical analysis The primary analysis was a mixed linear model with the FeNO as a response variable, the treatments, BFPE and placebo, and baseline FeNO, before each cross-over treatment, as fixed effects. To account for the cross-over design, participants were treated as random effects. The FeNO was natural logarithm transformed (ln) for analysis purposes. Exponentiation of the difference in ln FeNO is interpreted as the ratio of geometric mean FeNO between two randomised groups. SAS V.9.2 was used and the statistical model is available in the online supplementary material. Sample size The sample size calculation was based on a paired t-test because we could identify no publications that reported variance components for a mixed linear model that was planned for the primary analysis. A sample size of 24 had 80% power, α 5%, to detect a difference in ln FeNO of 0.227, equivalent to a ratio of geometric mean values of 1.25, based on the paired SD of 0.38 for the difference in ln FeNO at two visits in a group of steroid-naïve asthmatics taking a placebo in a study recently conducted by the MRINZ.39 This is slightly larger than the ATS-defined minimally important decrease of 20% in FeNO for individuals with an FeNO 50 ppb or greater.34 The recruitment target was 28 participants, allowing for a 15% withdrawal rate. Results Between May and August 2013, 220 patients were approached for inclusion in the study. Following the exclusion of 97 (44%) who declined participation, 41 (19%) who were treated with steroids, 54 (25%) with comorbidities and other exclusion criteria, there were 28 (13%) participants who were randomised. There were three withdrawals, two due to worsening asthma (one during the placebo intervention and one during washout after BFPE treatment), and one because of renal calculi requiring surgical intervention (during the washout after BFPE treatment; figure 2). Figure 2 Participant inclusion/exclusion pathway. BFPE, berryfruit polyphenolic extract. The characteristics of the 28 randomised participants are shown in table 1. The participants had a mean age of 42, were predominantly male (61%) and had a mean baseline FeNO concentration of 76.9 ppb. The mean FEV1 was 100% predicted and mean ACQ-5 was 0.65. Table 1 Baseline characteristics of participants Variable N=28 Mean (SD) N (%) Age (years) 42.3 (11.8) BMI (kg/m2) 26.7 (5.1) Male sex 17 (61%) Ethnicity  European 23 (82%)  Māori 1 (4%)  Other 4 (14%) FeNO (ppb) 76.9 (41.48) FEV1% predicted 100.2 (15.7) MMEF25–75 (L/min) 156.3 (60.5) sRaw (kPa/s) 1.16 (0.42) sGaw (kPa/s) 0.96 (0.29) ACQ-5 0.65 (0.58) Eosinophils (×10^9L) 0.33 (0.19) ACQ-5, asthma control questionnaire-5; BMI, body mass index; FeNO, fraction of exhaled nitric oxide; FEV1, forced expiratory volume in 1 s; MMEF25–75, maximal mid expiratory flow; sGaw, specific airway conductance; sRaw, specific airway resistance. Primary outcome variable FeNO had a skewed distribution and the natural logarithm transformation was used in the analysis. The mean (SD) In Feno after 4 weeks of treatment for the BFPE (N=28) and placebo (N=25) groups was 4.28 (0.47) and 4.22 (0.47), respectively (table 2, figure 3). The paired change from baseline mean (SD) BFPE minus placebo (In) Feno was −0.03 (0.39), N=25. The mixed linear model estimate, with baseline covariate adjustment, difference in In Feno was −0.002 (95% CI −0.15 to 0.14), p=0.98. This is equivalent to a ratio of geometric mean FeNO of 1.0 (95% CI 0.86 to 1.15). The median values for FeNO at baseline and 4 weeks for BFPE were 68.0 and 73.5 ppb and for placebo 63.0 and 67.0 ppb. Table 2 FeNO measurements (median (IQR))* before and after 4 weeks of treatment with BFPE and placebo BFPE Placebo Difference in mean In FeNO, BFPE minus placebo† Ratio mean In FeNO, BFPE vs placebo†‡ p Value Baseline (n=28) ppm 4 weeks (n=28) ppm Baseline (n=26) ppm 4 weeks (n=25) ppm Median 68.0 73.5 63.0 67.0 IQR 50.5 to 96.5 49.5 to 112.0 52.0 to 98.0 53.0 to 94.0 Ln§ mean 4.27 4.28 4.24 4.22 −0.002 1.0 0.98 (SD) (0.44) (0.47) (0.38) (0.47) (95% CI) (−0.15 to 0.14) (0.86 to 1.15) FeNO had a skewed distribution and the natural logarithm transformation (ln) was used in the analysis of comparison between treatments. *IQR. †Mixed linear model with baseline covariate. ‡Exponent of the difference in natural logarithms. §Ln: natural logarithm. BFPE, berryfruit polyphenolic extract; FeNO, fraction of exhaled nitric oxide. Figure 3 Logarithm FeNO values for each participant throughout both intervention arms. FeNO, fraction of exhaled nitric oxide. Secondary outcome variables The estimates and associated CIs for the treatment-associated differences, BFPE versus placebo, in the secondary outcome variables: FEV1, MMEF25–75, sRaw and sGaw after 4 weeks of treatment are shown in table 3. MMEF25–75 was the only measure with a statistically significant difference and this favoured placebo. Asthma control remained stable throughout the intervention and washout periods with no significant difference in ACQ-5 between randomised treatment groups. There was no change from baseline in peripheral blood eosinophil count with 4 weeks of treatment of BFPE and no difference between BFPE and placebo treatments. Table 3 Secondary outcome measures (mean (SD)) before and after 4 weeks treatment with BFPE or placebo BFPE Placebo Variable Baseline (n=28) 4 weeks (n=28) Baseline (n=26) Post 4 weeks (n=25) Difference, BFPE minus placebo* (95% CI), p Value FEV1 (L) 3.61 (1.06) 3.54 (0.98) 3.56 (1.00) 3.53 (1.06) −0.07 (−0.20 to 0.06), p=0.26 MMEF25–75 (L/min) 160.5 (63.0) 152.8 (60.6) 155.5 (60.1) 159.4 (67.3) −11.8 (−22.6 to −1.0), p=0.034* sRaw (kPa/s) 1.12 (0.41) 1.22 (0.38) 1.21 (0.40) 1.18 (0.44) 0.09 (−0.06 to 0.24), p=0.22 sGaw (kPa/s) 0.99 (0.30) 0.90 (0.25) 0.91 (0.27) 0.94 (0.28) −0.07 (−0.18 to 0.03), p=0.17 ACQ-5 0.74 (0.61) 0.77 (0.61) 0.62 (0.57) 0.70 (0.76) −0.06 (−0.37 to 0.25), p=0.72 Eosinophils (×10^9 L) 0.31 (0.14) 0.32 (0.18)† 0.33 (0.21) 0.30 (0.16) 0.01 (−0.05 to 0.08), p=0.69 †N=27 as no blood sample was available. *Mixed linear model with baseline covariate. ACQ-5, asthma control questionnaire; BFPE, berryfruit polyphenolic extract; FEV1, forced expiratory volume in 1 s; MMEF25–75, maximal mid expiratory flow; sGAW, specific airway conductance; sRAW, specific airway resistance. There was a high rate of recorded medication compliance with 23/28 (82%) participants >95% compliant with BFPE and 19/25 (75%) >95% compliant with the placebo. There was no significant difference in compliance between the two treatment groups, p=0.09. Breaches in specified non-allowed food and drink consumption throughout the study were deemed as minor, with 21/28 (75%) patients consuming small amounts of berry-containing foods during the period of the study. The highest amount was three blueberry muffins consumed over the duration of the entire study by one participant. The most frequent adverse event was headache, which occurred in half of the participants. There were 20 reports of headaches, 6 occurring during BFPE consumption, 6 during placebo and 8 during the washout phase. Discussion This randomised placebo-controlled cross-over trial did not find evidence for an effect of BFPE on FeNO, a measure of eosinophilic airways inflammation, in mild steroid-naïve asthma. As a result, the findings of chronic anti-inflammatory and remodelling effects of BFPE in animal in vivo trials were not confirmed in this human study in asthma.27 There are a number of methodological issues that are relevant to the interpretation of the study findings. First, FeNO was the primary outcome variable as it is a validated biomarker of eosinophilic airways inflammation mediated by the T helper 2 (Th2) pathway.29 32 34 In steroid-naïve patients with mild asthma, FeNO is correlated with bronchial hyper-responsiveness, a fundamental physiological measure of asthma severity.40 FeNO is highly sensitive to the effects of Inhaled corticosteroid (ICS) and therefore highly responsive to anti-inflammatory therapy.33 34 For this reason, we studied ICS-naïve participants who had not taken ICS or oral steroids in the previous 90 days. Participants were also required to have a raised FeNO at baseline, to ensure that active airways inflammation was present, and therefore potentially amenable to anti-inflammatory treatment. FeNO measurements are also simple and non-invasive and, as shown in this study, are stable over long periods in mild asthma, providing a stable baseline. However, the drawback of the inclusion of steroid-naïve participants with mild asthma is that it limited the ability to investigate the effects of BFPE on lung function and ACQ, as most of the participants had normal lung function and well-controlled asthma. The mean baseline FEV1 was 100% predicted and the mean ACQ-5 score was 0.65, indicating that the study participants' asthma may not have been severe enough to detect clinically significant improvements in lung function or quality of life. Our study does not rule out an effect of BFPE on lung function and asthma control in patients with moderate-to-severe asthma. The study was placebo-controlled with the two treatments prepared in identical capsules. The 1000 mg dose of BFPE was based on animal data in which 10 mg/kg berryfruit extract was effective in reducing chronic inflammatory changes in the airways.27 Treatments were continued for 4 weeks based on the animal data showing effects in a chronic model with 30 days of dosing.27 The cross-over design allowed participants to serve as their own controls, minimising variability and increasing power to detect any changes in the primary outcome measure. There was also good integrity of the data set, as depicted by the closely matched baseline variables at the start of each intervention period, indicating good reproducibility. This study was designed partly as a mechanistic one addressing whether airways inflammation might be reduced by BPFE and the relevant variable for this is the FeNO. The CI for the change in ratio of geometric mean FeNO was well inside the prespecified clinical difference.34 For the secondary outcome variables, the confidence limits for change in FEV1 and ACQ were well within previous established clinically important differences,36 41 making type II errors unlikely. Clinically important changes in the remaining secondary outcome variables are not well established. There was some evidence of a slightly lower MMEF25–75 after BFPE compared with the placebo; however, this result should be interpreted with caution as we did not control for multiple measurements and this could be the result of type 1 error. To aid compliance and reduce introduction of other polyphenolic-containing foods and drinks, the study was performed outside the berryfruit season. This appeared to assist in maintaining a very low reported non-approved consumption of polyphenolic-containing foods and drinks. The animal in vivo boysenberry feeding studies on which this trial was based conferred some benefit with reduction in airways inflammation and modulation of the Th2 pathway in chronic allergen-driven models of airways inflammation in mice.27 There was evidence of decreased collagen deposition, suggesting an effect on airways remodelling. It has been proposed that these effects may be due to modulation of the functional phenotype of alternatively activated macrophages in the lung, and associated increased expression of profibrolytic matrix metalloproteinase-9 protein. Furthermore, in vitro studies of lung alveolar epithelial cells have shown that blackcurrant polyphenolic extracts suppress secretion of eotaxin 3, an important chemoattractant responsible for eosinophil recruitment into airways.26 However, our data did not support the replication of these effects to human individuals with asthma. This poor translation of positive outcomes from laboratory-based and animal models to human disease is not uncommon, particularly surrounding the extrapolation of apparent benefit in murine models of lung eosinophilia to clinical outcomes in patients with asthma.42–46 This is not surprising as chronic asthma is a disease unique to humans and there are fundamental anatomical, physiological and immunological differences between the species. In contrast to humans, mice are born with fully developed lungs, have far fewer orders of airways branching, a large airway epithelium that is not fully stratified, lack submucosal glands beyond the trachea and most mouse airways do not contain smooth muscle bundles. Murine airways are also smaller than humans in absolute terms but are relatively wider, providing challenges to the accurate measurement and definition of physiological parameters of disease and response to treatment. The allergic inflammation in mouse airways is predominantly parenchymal and vascular, rather than being generally restricted to the conducting airways as in human disease and chronic allergen models do not lead to increases in airway smooth muscle, a hallmark of asthma. As a result, the reasons underlying the failure to translate the murine findings24 for this berryfruit supplement into this human randomised controlled trial are likely to be multifactorial and support recommendations to improve murine systems to more closely reflect the disease for which they are used as models.42–46 In conclusion, in steroid-naïve participants with mild asthma and elevated FeNO, there was no effect of BFPE on FeNO, a biomarker of eosinophilic airways inflammation. Contributors: RB, IB, JF, JH, MH, RH, MWe and MWi were involved in study concept and design. IB, RC-S, CM, JP, SP and MWi were involved in acquisition of data. RB, IB, SP and AS were involved in drafting of the manuscript. All authors were involved in critical revision of the manuscript for important intellectual content. MWe was involved in statistical analysis. All authors provided administrative, technical and material support. SP and MWi were involved in study supervision.. Funding: This research was funded by a New Zealand Ministry of Business, Innovation and Employment Grant (Contract C11X1002), awarded to the New Zealand Institute of Plant and Food Research. The Medical Research Institute of New Zealand is supported by Health Research Council of New Zealand Independent Research Organisation funding (HRC 14/1002). Competing interests: None declared. Ethics approval: NZ Health and Disability Committee. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: Patient-level data are available from the corresponding author on request. The presented data are anonymised and the risk of identification is low. ==== Refs References 1 Eder W , Ege MJ , von Mutius E The asthma epidemic . N Engl J Med 2006 ;355 :2226 –35 . 10.1056/NEJMra054308 17124020 2 Masoli M , Fabian D , Holt S The global burden of asthma: executive summary of the GINA dissemination committee report . Allergy 2004 ;59 :469 –78 . 10.1111/j.1398-9995.2004.00526.x 15080825 3 Beasley R , Semprini A , Mitchell EA Risk factors for asthma: is prevention possible? Lancet 2015 ;386 :1075 –85 . 10.1016/S0140-6736(15)00156-7 26382999 4 Beasley R , Crane J , Lai CK Prevalence and etiology of asthma . J Allergy Clin Immunol 2000 ;105 :S466 –72 . 10.1016/S0091-6749(00)90044-7 10669525 5 Ellwood P , Asher MI , Björkstén B Diet and asthma, allergic rhinoconjunctivitis and atopic eczema symptom prevalence: an ecological analysis of the International Study of Asthma and Allergies in Childhood (ISAAC) data . Eur Respir J 2001 ;17 :436 –43 . 10.1183/09031936.01.17304360 11405522 6 Nagel G , Weinmayr G , Kleiner A Effect of diet on asthma and allergic sensitisation in the International Study on Allergies and Asthma in Childhood (ISAAC) Phase Two . Thorax 2010 ;65 :516 –22 . 10.1136/thx.2009.128256 20522849 7 Rosenlund H , Kull I , Pershagen G Fruit and vegetable consumption in relation to allergy: disease-related modification of consumption? J Allergy Clin Immunol 2011 ;127 :1219 –25 . 10.1016/j.jaci.2010.11.019 21215999 8 Berentzen NE , Van Stokkom VL , Gehring U Associations of sugar-containing beverages with asthma prevalence in 11-year-old children: the PIAMA birth cohort . Eur J Clin Nutr 2015 ;69 :303 –8 . 10.1038/ejcn.2014.153 25117998 9 DeChristopher LR , Uribarri J , Tucker KL Intakes of apple juice, fruit drinks and soda are associated with prevalent asthma in US children aged 2–9 years . Public Health Nutr 2016 ;19 :123 –30 . 10.1017/S1368980015000865 25857343 10 Uribarri J , Woodruff S , Goodman S Advanced glycation end products in foods and a practical guide to their reduction in the diet . J Am Diet Assoc 2010 ;110 :911 –16.e12 . 10.1016/j.jada.2010.03.018 20497781 11 Forastiere F , Pistelli R , Sestini P Consumption of fresh fruit rich in vitamin C and wheezing symptoms in children. SIDRIA Collaborative Group, Italy (Italian Studies on Respiratory Disorders in Children and the Environment) . Thorax 2000 ;55 :283 –8 . 10.1136/thorax.55.4.283 10722767 12 Shaheen SO , Sterne JAC , Thompson RL Dietary antioxidants and asthma in adults—population-based case–control study . Am J Respir Crit Care Med 2001 ;164 :1823 –8 . 10.1164/ajrccm.164.10.2104061 11734430 13 Romieu I , Varraso R Avenel V Fruit and vegetable intakes and asthma in the E3N study . Thorax 2006 ;61 :209 –15 . 10.1136/thx.2004.039123 16396945 14 Buckley ST , Ehrhardt C The receptor for advanced glycation end products (RAGE) and the lung . BioMed Research International , 2010 . 15 McKeever TM , Britton J Diet and asthma . Am J Respir Crit Care Med 2004 ;170 :725 –9 . 10.1164/rccm.200405-611PP 15256393 16 Fogarty AW , Antoniak M , Venn AJ A natural experiment on the impact of fruit supplementation on asthma symptoms in children . Eur Respir J 2009 ;33 :481 –5 . 10.1183/09031936.00162907 19213783 17 Allen S , Britton JR , Leonardi-Bee JA Association between antioxidant vitamins and asthma outcome measures: systematic review and meta-analysis . Thorax 2009 ;64 :610 –19 . 10.1136/thx.2008.101469 19406861 18 Lee SC , Yang YH , Chuang SY Reduced medication use and improved pulmonary function with supplements containing vegetable and fruit concentrate, fish oil and probiotics in asthmatic school children: a randomised controlled trial . Br J Nutr 2013 ;110 :145 –55 . 10.1017/S0007114512004692 23211647 19 Watson RR , Zibadi S , Rafatpanah H Oral administration of the purple passion fruit peel extract reduces wheeze and cough and improves shortness of breath in adults with asthma . Nutr Res 2008 ;28 :166 –71 . 10.1016/j.nutres.2008.01.003 19083404 20 Djukanović R , Roche WR , Wilson JW Mucosal inflammation in asthma . Am Rev Respir Dis 1990 ;142 :434 –57 . 10.1164/ajrccm/142.2.434 2200318 21 Garbacki N , Kinet M , Nusgens B Proanthocyanidins, from Ribes nigrum leaves, reduce endothelial adhesion molecules ICAM-1 and VCAM-1 . J Inflamm (Lond) 2005 ;2 :9 10.1186/1476-9255-2-9 16091140 22 Hurst SM , McGhie TK , Cooney JM Blackcurrant proanthocyanidins augment IFN-gamma-induced suppression of IL-4 stimulated CCL26 secretion in alveolar epithelial cells . Mol Nutr Food Res 2010 ;54 (Suppl 2 ):S159 –70 . 10.1002/mnfr.200900094 20229526 23 Park SJ , Shin WH , Seo JW Anthocyanins inhibit airway inflammation and hyperresponsiveness in a murine asthma model . Food Chem Toxicol 2007 ;45 :1459 –67 . 10.1016/j.fct.2007.02.013 17512652 24 Kim HP , Mani I , Iversen L Effects of naturally-occurring flavonoids and biflavonoids on epidermal cyclooxygenase and lipoxygenase from guinea-pigs . Prostaglandins Leukot Essent Fat Acids 1998 ;58 :17 –24 . 10.1016/S0952-3278(98)90125-9 25 Shepherd SJ , Gibson PR Fructose malabsorption and symptoms of irritable bowel syndrome: guidelines for effective dietary management . J Am Diet Assoc 2006 ;106 :1631 –9 . 10.1016/j.jada.2006.07.010 17000196 26 Nyanhanda T , Gould EM , McGhie T Blackcurrant cultivar polyphenolic extracts suppress CCL26 secretion from alveolar epithelial cells . Food Funct 2014 ;5 :671 –7 . 10.1039/c3fo60568j 24526266 27 Shaw OM , Hurst RD , Harper JL. Boysenberry ingestion supports fibrolytic macrophages with the capacity to ameliorate chronic lung remodelling . Am J Physiol Lung Cell Mol Physiol 2016 ;311 :L628 –38 . (L-00309-2015R2 in press) 10.1152/ajplung.00309.2015 .27371734 28 Engler RJM , With CM , Gregory PJ Complementary and alternative medicine for the allergist-immunologist: where do I start? J Allergy Clin Immunol 2009 ;123 :309 –16 . 10.1016/j.jaci.2009.01.001 19203654 29 Taylor DR Using biomarkers in the assessment of airways disease . J Allergy Clin Immunol 2011 ;128 :927 –34 . 10.1016/j.jaci.2011.03.051 21621830 30 Alving K , Weitzberg E , Lundberg JM Increased amount of nitric oxide in exhaled air of asthmatics . Eur Respir J 1993 ;6 :1368 –70 .7507065 31 Kharitonov SA , Yates D , Robbins RA Increased nitric oxide in exhaled air of asthmatic patients . Lancet 1994 ;343 :133 –5 . 10.1016/S0140-6736(94)90931-8 7904001 32 Jatakanon A , Lim S , Kharitonov SA Correlation between exhaled nitric oxide, sputum eosinophils, and methacholine responsiveness in patients with mild asthma . Thorax 1998 ;53 :91 –5 . 10.1136/thx.53.2.91 9624291 33 Silkoff PE , McClean P , Spino M Dose-response relationship and reproducibility of the fall in exhaled nitric oxide after inhaled beclomethasone dipropionate therapy in asthma patients . Chest 2001 ;119 :1322 –8 . 10.1378/chest.119.5.1322 11348935 34 Dweik RA , Boggs PB , Erzurum SC An official ATS clinical practice guideline: interpretation of exhaled nitric oxide levels (FENO) for clinical applications . Am J Respir Crit Care Med 2011 ;184 :602 –15 . 10.1164/rccm.9120-11ST 21885636 35 van Rensen EL , Straathof KC , Veselic-Charvat MA Effect of inhaled steroids on airway hyperresponsiveness, sputum eosinophils, and exhaled nitric oxide levels in patients with asthma . Thorax 1999 ;54 :403 –8 . 10.1136/thx.54.5.403 10212103 36 Juniper EF , O'Byrne PM , Guyatt GH Development and validation of a questionnaire to measure asthma control . Eur Respir J 1999 ;14 :902 –7 . 10.1034/j.1399-3003.1999.14d29.x 10573240 37 Cockcroft D , Killian D , Mellon J , Hargreave F . ATS/ERS recommendations for standardised procedures for the online and offline measurement of exhaled lower respiratory nitric oxide and nasal nitric oxide 2005 . Am J Respir Crit Care Med 2005 ;171 :912 –30 . 10.1164/rccm.200406-710ST 15817806 38 Miller MR , Hankinson J , Brusasco V ATS/ERS Task Force . Standardisation of spirometry . Eur Respir J 2005 ;26 :319 –38 . 10.1183/09031936.05.00034805 16055882 39 Marsh S , Aldington S , Williams M Complete reference ranges for pulmonary function tests from a single New Zealand population . N Z Med J 2006 ;119 :U2281 .17072356 40 Dupont LJ , Rochette F , Demedts MG Exhaled nitric oxide correlates with airway hyperresponsiveness in steroid-naive patients with mild asthma . Am J Respir Crit Care Med 1998 ;157 :894 –8 . 10.1164/ajrccm.157.3.9709064 9517608 41 Santanello NC , Zhang J , Seidenberg B What are minimal important changes for asthma measures in a clinical trial? Eur Respir J 1999 ;14 :23 –7 . 10.1034/j.1399-3003.1999.14a06.x 10489824 42 Wenzel S , Holgate ST The mouse trap: It still yields few answers in asthma . Am J Respir Crit Care Med 2006 ;174 :1173 –6 ; discussion 1176–8 10.1164/rccm.2609002 17110654 43 Finkelman FD , Wills-Karp M Usefulness and optimization of mouse models of allergic airway disease . J Allergy Clin Immunol 2008 ;121 :603 –6 . 10.1016/j.jaci.2008.01.008 18328889 44 Mercer PF , Abbott-Banner K , Adcock IM Translational models of lung disease . Clin Sci 2015 ;128 :235 –56 . 10.1042/CS20140373 25328010 45 Persson CGA , Erjefält JS , Korsgren M The mouse trap . Trends Pharmacol Sci 1997 ;18 :465 –7 . 10.1016/S0165-6147(97)90684-2 9458694 46 Kumar RK , Foster PS Are mouse models of asthma appropriate for investigating the pathogenesis of airway hyper-responsiveness? Front Physiol 2012 ;3 :312 10.3389/fphys.2012.00312 23060800

PMC005xxxxxx/PMC5372144.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01453810.1136/bmjopen-2016-014538Infectious DiseasesProtocol150617061724Surveillance Systems from Public Health Institutions and Scientific Societies for Antimicrobial Resistance and Healthcare-Associated Infections in Europe (SUSPIRE): protocol for a systematic review Núñez-Núñez María 123Navarro María Dolores 12Gkolia Panagiota 4Babu Rajendran Nithya 4del Toro María Dolores 12Voss Andreas 5Sharland Mike 6Sifakis Frangiscos 7Tacconelli Evelina 4Rodríguez-Baño Jesús 12and Members of the EPI-NET group on behalf of COMBACTE-MAGNET Consortium 1 Unit of Infectious Disease, Microbiology, and Preventive Medicine, Institute of Biomedicine of Seville (IBIS)/University Hospital Virgen Macarena-Virgen del Rocío/Spanish National Research Council (CSIC), Seville, Spain 2 Departamento de Medicina, Universidad de Sevilla, Seville, Spain 3 Department of Pharmacy, University Hospital Virgen Macarena, Seville, Spain 4 Infectious Diseases, Internal Medicine 1, DZIF Center, Tuebingen University Hospital, Tuebingen, Germany 5 Department of Medical Microbiology, Radboud University Medical Centre, Nijmegen, The Netherlands 6 Paediatric Infectious Diseases Research Group, St George's University London, London, UK 7 AstraZeneca LP, Gaithersburg, Maryland, USACorrespondence to María Núñez-Núñez; mnunez.pharm@gmail.com2017 27 3 2017 7 3 e01453810 10 2016 30 1 2017 31 1 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction The worldwide spread of antimicrobial resistance is now recognised as a global public health threat. Owing to the geographical heterogeneity, complexity and continuously evolving dynamics of resistant organisms and genes, surveillance is a key tool for understanding, measuring and informing actions in the fight against this problem. To date there is no harmonisation of key indicators or of methodologies used to obtain them. Methods and analysis The main objective of this project is to systematically review and analyse the current publicly available surveillance activities on antimicrobial resistance and healthcare-associated infections in Europe. Eligible activities are those endorsed by regional, national or transnational health organisations and scientific societies providing data on a periodic basis. Grey and peer-reviewed literature will be searched with no language restrictions. Three independent reviewers will perform a two-step selection process using a previously piloted, tailored electronic data extraction form. Descriptive summaries and tables of all relevant findings will be performed and reported according to PRISMA guidelines. Ethics and dissemination We did not seek ethical approval for this study because the data to be collected are not linked to individuals. Data will be presented at international conferences and published in peer-reviewed journals. Trial registration number CRD42016033867. Antimicrobial resistancehealthcare-associated infectionssurveillance ==== Body Strengths and limitations of this study Wide scope search strategy including peer-reviewed and grey literature. No language restriction and study protocol designed by a multidisciplinary team within a multinational collaborative consortium. Surveillance systems for which the methodology is not publicly available will be missed. The review will rely on the data provided in the accessible sources, while some surveillance activities might have changed after their publication. Despite following a carefully prepared search strategy, some surveillance systems or specific information from them may not be found. However, this will indicate a problem in the accessibility to the information. Introduction The worldwide spread of antimicrobial resistance (AMR) is now recognised as a global public health threat.1 2 Owing to the geographical heterogeneity, complexity and continuously evolving dynamics of resistant organisms and genes, surveillance is a key tool for understanding, measuring and informing actions in the fight against this problem. Indeed, surveillance systems for AMR have been developed by most national health systems and transnational organisations.3–5 Adequate characterisation of the burden of disease caused by resistant pathogens, their impact on patient outcomes and the areas and patient populations with the highest incidences are critical for identifying medical needs, establishing treatment protocols and efficiency of design for randomised controlled trials. The situation with healthcare-associated infections (HAI), which are often closely associated with AMR, is similar. An important problem for the surveillance of AMR and HAI is heterogeneity of the scope, focus, objectives, methodology, resources and reporting across the different regions and countries, despite the efforts of institutions such as the European Centre for Disease Prevention and Control (ECDC), Centers for Disease Control and Prevention (CDC) or the WHO.3–5 The ECDC conducted a comprehensive review of the current situation and implications of HAIs across Europe; these authors highlighted that while some progress has been achieved in recent years, there are still substantial intercountry and intracountry differences in surveillance methods and concluded that greater emphasis be placed on harmonisation.5 The objectives of SUSPIRE are to systematically review and analyse the surveillance activities endorsed by national or transnational health organisations and scientific societies that are performed in the European Union and European Economic Area (EU/EAA) and to provide regular data on HAI and AMR. The final objective of this effort is to provide recommendations that can be used for the harmonisation of surveillance systems. Methods Eligibility criteria SUSPIRE targets surveillance systems endorsed by regional, national or transnational health organisations or scientific societies in EU/EEA countries that are in place for the purpose of providing regular data on AMR and/or HAI. Information sources and search strategy A comprehensive search strategy for searching grey and peer-reviewed literature will be performed until 30 November 2016, using the Google search engine for the first task, and combining the following terms in the local languages of the various European countries or regions (depending on whether the public health providers are centralised/national or decentralised): ‘Antimicrobial resistance’ AND/OR ‘Hospital-associated’ OR ‘Hospital-acquired’ OR ‘Nosocomial’ AND ‘Surveillance’ AND ‘epidemiology’ OR ‘prevalence’ OR ‘incidence’. Sources of peer-reviewed literature to be searched include: PubMed, Embase, Scopus (Elsevier Science) and Web of Science (Thomson, ISI), using the same subject headings (MeSH). A draft PubMed search strategy is included in online supplementary appendix 1. 10.1136/bmjopen-2016-014538.supp1supplementary appendix Articles available until 30 June 2016 will be included. ‘Zoonotic’ OR ‘food-borne’ OR ‘outbreak’ will be excluded in all cases, and ‘human data’ will be imposed on the search whenever possible. We will also review the official health-related websites of national and transnational health organisations and scientific societies, such as the ECDC, WHO Europe, regional and national European governments, and of relevant scientific societies, such as ESCMID (European Society of Clinical Microbiology and Infectious Diseases; ISID (International Society for Infectious Diseases); IEA (International Epidemiological Association); ESICM (European Society of Intensive Care Medicine) or ERS (European Respiratory Society), among others. Additional data will be obtained from references in retrieved articles and data provided by stakeholders following the request to complete the catalogue of surveillance systems and reports on AMR and HAI. Inclusion and exclusion criteria The documents retrieved will be assessed for their content and the inclusion and exclusion criteria shown in box 1 applied for those documents to be finally included in the review. Box 1 Eligibility criteria for surveillance systems Inclusion criteria European surveillance systems, or including European data Epidemiological surveillance studies collecting data for at least five consecutive years from three or more sites, and with at least one data collection year within the last 10 years Promoted or endorsed by national or transnational health organisations and scientific societies Providing, or with the intention to provide, data on a periodic basis Providing data on at least one of the following: objective, scope (eg, hospital and healthcare-associated infections/pathogens), design and methodological issues Published scientific and ‘grey’ literature Human data No age/language restriction Exclusion criteria Animal, environmental or food data only Epidemiological studies or reports whose main objective is different from that of providing surveillance data Epidemiological studies or surveillance data promoted by private companies Outbreak reports Systems that are ‘inactive’ (not providing any information) for the last 10 years No language restrictions will apply; local experts, identified by the European Committee on Infection Control (EUCIC) of the European Society of Clinical Microbiology and Infectious Diseases, will be consulted. Data management, selection process and data extraction The literature search results will be uploaded to EndNote X7 software. Two independent reviewers will perform a two-step selection process. Titles and abstracts (if available) of the retrieved documents will be initially assessed and non-relevant documents excluded. The full text of potentially eligible documents will then be obtained and assessed for relevance or duplication against predefined selection criteria. Data will be extracted independently by two reviewers, using a tailored electronic data extraction form to be piloted beforehand on a representative sample. The same core data will be extracted from systems and studies including specific details about the following elements: programme, population, microbiology, methodology, quality indicators and data. A detailed list of variables is presented in table 1. Table 1 Summary list of variables collected Core element Variable Program System name & acronym* Title & author/s name† Location and magnitude Status: Active/inactive* & completed/ongoing† Coordinating organization and resources Focus: Type of activity (HAI, AMR) Link: System website/journal article Update year Population Demographics of population covered (age, gender) Comorbidities and risk factors Inclusion/exclusion criteria† Type and details of healthcare facilities Microbiology Specimen type and carriers tested Duplicates policy Clinical value of the sample Pathogens, sources and acquisition Microbiological methods for the identification and characterization of mechanisms of resistanceAntibiotics tested and resistance mechanisms Methodology and indicators used Clinical criteria and microbiological definitions used, Structure, process and outcome indicators reported, Measurement frequency of relevant indicators, Reporting of source data (volunteer/compulsory) Data analysis, reporting and dissemination Source of data and data collection systems, Quality assessments*, Stratification of reported data, Reporting type and frequency, Dissemination of data* *Surveillance systems. †Epidemiological studies. Disagreements between reviewers will be resolved by consultation with a third reviewer. Quality assessment There are several generic guidelines for evaluating human public health surveillance, and these typically include assessing a series of attributes, such as flexibility, acceptability and timeliness, using a combination of quantitative and qualitative techniques. The quality of the surveillance systems included in our review will be assessed on the basis of the attributes recommended in the Centres for Disease Prevention and Control guidelines for evaluating public health surveillance systems.6 The protocol was developed following the recently released PRISMA-P guidelines, and the review will be reported in accordance with the PRISMA statement.7 Data synthesis and descriptive analysis The data synthesis phase will involve collating and summarising the results in the form of a table that indicates the core characteristics of the systems and studies: type of activity, isolate source, population, phenotypes and mechanisms of resistance, definitions of acquisition, indicators and quality assessments. Frequency distributions expressed as percentages (%) will be calculated for each variable and displayed graphically. Analysis will be stratified by country, surveillance type (system/study) and activity (HAI/AMR). Dissemination Data will be presented at international conferences and published in peer-reviewed journals. Discussion Following the original SENIC studies in the 1970s, surveillance has been recognised as a key component of quality assurance in general and of infection control in particular.8 Since then, surveillance procedures and systems have evolved in accordance with increased awareness of the complexity and importance of HAIs as a patient safety concern9 and the rise of AMR. Most countries started surveillance systems for HAIs during the 1980s and 1990s. A series of articles published in 2001 showed how heterogeneous the surveillance activities performed in European countries at that time were.10 Over recent decades, the ECDC has been extraordinarily active in working towards the homogenisation of definitions, procedures and systems developed.3 6 In 2008, the ECDC conducted a review highlighting that there were still significant limitations to surveillance systems across European countries.11 The aim of SUSPIRE is to provide an update of the situation, with the additional purpose of specifically evaluating particular aspects of the definitions, methods and quality assessment, as well as the infections and microorganisms/resistances. AMR surveillance has traditionally focused on either the percentage of particular pathogens that are resistant to certain antimicrobial agents and/or rates of isolation of relevant resistant bacteria. However, some potentially important aspects, such as the specific populations or patients in which these infections occur, types of infection and their outcomes have not usually been considered because of the additional complexity and resources needed. Nevertheless, these are key aspects for informing management decisions about the prioritisation and provision of resources required to address the problem and determining the most urgent areas for research. Furthermore, the way that national surveillance systems are conceptually set up at present may be too slow to provide useful information for immediate global action in situations of the emergence of new resistant pathogens or the further spread of previously known ones, as is shown by the recent spread of carbapenemase-producing Enterobacteriaceae in many European countries and across borders.12 Therefore, specific evaluation of AMR surveillance systems and the gaps that exist is required in order to detect areas for improvement. The objectives of the COMBACTE-MAGNET consortium, funded by the European Union and, in kind, by the European Federation of Pharmaceutical Industries Association (EFPIA) through the Innovative Medicines Initiative, include building a European network by engaging representatives from the major EU financed projects, stakeholder experts and industry in order to develop a consensus programme and a homogeneous approach to current and future epidemiological surveillance strategies in Europe. The SUSPIRE project will work to achieve this objective. The research leading to these results was conducted as part of the COMBACTE-MAGNET consortium. For EPINET members and further information, please refer to http://www.COMBACTE.com. The authors thank Virginia Palomo for her administrative assistance in the development of the protocol. Contributors: JR-B conceived the study, led the development of the protocol, provided supervision and mentorship to MN-N who wrote the first draft, coordinated and integrated comments from coauthors and together with JR-B is the guarantor of the manuscript. MDN contributed with the selection of variables, and NBR contributed with the development of search strategy. MS and MDdT provided specific expertise on paediatric and surgical site infections variables, respectively. AV contributed with the selection of variables, and FS contributed with development of the protocol and variables selection. ET and PG also provided specific expertise on epidemiology and contributed to the development of search strategy. All authors critically reviewed successive drafts of the manuscript, provided important intellectual input and approved the final version for publication. Funding: This research project receives support from the Innovative Medicines Initiative Joint Undertaking under grant agreement no. 115737 resources of which are composed of financial contribution from the European Union Seventh Framework Programme (FP7/2007/2013) and EFPIA companies in kind contribution. JRB also received funding for research from Ministerio de Economía y Competitividad, Instituto de Salud Carlos III—cofinanced by European Development Regional Fund ‘A way to achieve Europe’ ERDF, Spanish Network for the Research in Infectious Diseases (REIPI RD12/0015). Competing interests: FS is an employee of AstraZeneca/Medimmune, an EFPIA (European Federation of Pharmaceutical Industries and Association) member company in the IMI JU. Costs related to research contributions by FS are borne by AstraZeneca/Medimmune and considered in-kind contribution under the IMI JU scheme. Rest of authors has declared no conflict of interest related to this paper. Provenance and peer review: Not commissioned; externally peer reviewed. ==== Refs References 1 European commission . Action plan against the rising threats from Antimicrobial Resistance. Communication from the commission to the European parliament and the council . Brussels , September 2011. http://ec.europa.eu/dgs/health_consumer/docs/communication_amr_2011_748_es.pdf 2 ECDC/EMEA Joint technical report . The bacterial challenge: time to react. Ref. EMEA/576176/2009 . Stockholm , September 2009. ISBN 978-92-9193-193-4 10.2900/2518 3 European Centre for Disease Prevention and Control (ECDC), website available on. http://ecdc.europa.eu/en/activities/surveillance/Pages/index.aspx (date last accessed September 2016 ). 4 Centers for Disease Control and Prevention (CDC). http://www.cdc.gov/nhsn (date last accessed September 2016) . 5 World Health Organization (WHO). http://www.who.int/topics/public_health_surveillance/en/ (date last accessed September 2016 ). 6 European Centre for Disease Prevention and Control (ECDC) . Data quality monitoring and surveillance system evaluation—A handbook of methods and applications . Stockholm , September 2014. http://ecdc.europa.eu/en/publications/publications/data-quality-monitoringsurveillance-system-evaluation-sept-2014.pdf 7 Shamseer L , Moher D , Clarke M , PRISMA-P Group . Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015: elaboration and explanation . BMJ 2015 ;349 :g7647 10.1136/bmj.g7647 8 Haley RW , Culver DH , White JW The efficacy of infection surveillance and control programs in preventing nosocomial infections in US hospitals . Am J Epidemiol 1985 ;121 :182 –205 . 10.1093/oxfordjournals.aje.a113990 4014115 9 Burke JP Infection control—a problem for patient safety . N Engl J Med 2003 ;348 :651 –6 . 10.1056/NEJMhpr020557 12584377 10 Gastmeier P European perspective on surveillance . J Hosp Infect 2007 ;65 (Suppl 2 ):159 –64 . 10.1016/S0195-6701(07)60036-X 17540263 11 European Centre for Disease Prevention and Control (ECDC) . Annual Epidemiological Report on Communicable Diseases in Europe 2008 . Stockholm , March 2008. http://ecdc.europa.eu/en/publications/publications/0812_sur_annual_epidemiological _report_2008.pdf 12 Glasner C , Albiger B , Buist G The European Survey on Carbapenemase-Producing Enterobacteriaceae (EuSCAPE) working group. Carbapenemase-producing Enterobacteriaceae in Europe: a survey among national experts from 39 countries, February 2013 . Euro Surveill 2013 ;18 (28 ):pii=20525 doi:10.2807/1560-7917.ES2013.18.28.2052510.2807/1560-7917.ES2013.18.28.20525

PMC005xxxxxx/PMC5372145.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01426310.1136/bmjopen-2016-014263Smoking and TobaccoResearch1506173417241703Elucidating challenges that electronic cigarettes pose to tobacco control in Asia: a population-based national survey in Taiwan Chang Hui-Chin 1Tsai Yi-Wen 1Shiu Ming-Neng 2Wang Ying-Ting 1Chang Po-Yin 3 1 Institute of Health and Welfare Policy, National Yang-Ming University, Taipei, Taiwan 2 Faculty of Pharmacy, National Yang-Ming University, Taipei, Taiwan 3 Department of Medicine, Stanford University School of Medicine, Palo Alto, California, USACorrespondence to Dr Po-Yin Chang; poyin@stanford.edu2017 27 3 2017 7 3 e01426311 10 2016 30 1 2017 27 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives This study investigated the prevalence and correlates of electronic cigarettes (e-cigarettes) use in Taiwan. Design and setting We studied a nationally representative random sample in the 2015 Taiwan Adult Smoking Behavior Survey. Participants This study included 26 021 participants aged 15 years or older (51% women, 79% non-smokers, 16% aged 15–24 years), after excluding 31 persons (0.1%) who had missing information on e-cigarette use. Primary outcome measures The prevalence of ever having used e-cigarettes was calculated in the overall sample and by smoking status (current, former and never) or age (15–24, 25–44 and ≥45 years). We performed multivariable log-binomial regression to assess correlates of ever having used e-cigarettes among all participants and separately for subgroups by smoking status and age. Results Approximately 3% of all participants had ever used e-cigarettes. The prevalence of ever having used e-cigarettes was high in current smokers (14%) and people aged 18–24 years (7%). E-cigarette use was particularly common in people aged 15–24 years who were current (49–52%) or former (22–39%) smokers. Ever having used e-cigarettes was positively associated with tobacco smoking (adjusted prevalence ratio (aPR): 21.5, 95% CI 15.4 to 29.8, current smokers; aPR: 8.3, 95% CI 15.2 to 13.1, former smokers), younger age and high socioeconomic status. Age remained a significant factor of ever having used e-cigarettes across smoking status groups. Among non-smokers, men had a 2.4-fold (95% CI 1.5 to 3.8) greater prevalence of e-cigarette use than women. Conclusions E-cigarette use was uncommon in the general population in Taiwan, but prevalence was high among smokers and young people. This study highlights challenges that e-cigarettes pose to tobacco control, which warrant high priority action by policymakers and public health professionals. E-cigarette regulations should focus on young people. Electronic cigarettesSmoking and tobaccoYoung adultEPIDEMIOLOGYPUBLIC HEALTHTaiwan ==== Body Strengths and limitations of this study This is the first study with a nationally representative sample in Taiwan investigating the prevalence and correlates of e-cigarette use. This was a cross-sectional survey with only one question that assessed ever having used e-cigarettes. Respondents may have under-reported e-cigarette use. Information about specific e-cigarette products, such as their flavour or design, was unavailable. The time trend and change of e-cigarette use warrant further research. The relatively small sample size of adolescents and young adults may restrict the application of our findings. Introduction Electronic cigarettes (e-cigarettes or vapour) have changed tobacco smoking behaviours and contemporary concepts of tobacco control. E-cigarettes deliver a nicotine aerosol by heating a solution that usually contains nicotine, propylene glycol, glycerol and flavouring agents.1 2 E-cigarette use has increased substantially,3 4 especially among young adults in most,4–11 but not all, countries.12 Marketed as alternatives to conventional cigarettes, e-cigarettes lower users' exposure to toxins by delivering nicotine without burning tobacco.1 13 14 However, data on the long-term health effects of chemicals in e-cigarette aerosols are lacking.2 13 15 16 Furthermore, concerns have been raised about diacetyl and acetyl propionyl in solutions of certain e-cigarette products.17 E-cigarettes may hinder tobacco control efforts. E-cigarette use has been associated with a lower likelihood of quitting smoking among cigarette smokers,18 and might be a gateway for non-smokers to initiate tobacco smoking,1 particularly among adolescents or young adults.1 19–21 The prevalence of e-cigarette use may differ between smokers and non-smokers and between younger and older adults. A meta-analysis study has demonstrated that smoking status is a determinant of e-cigarette use,22 and the behaviour of e-cigarette use is also greatly influenced by age.10 22 It is essential for public health professionals and policymakers to understand the epidemiology of e-cigarette use and relevant correlates within each subgroup of age and smoking status. E-cigarette use is increasing in the USA,6 Canada5 and Europe.7 Epidemiological studies in Asia have suggested a relatively low prevalence of e-cigarette use: prevalence in Hong Kong was 2.3% in 2014,11 and <1% in Indonesia and Malaysia in 2011.12 No laws in Taiwan are specifically designed to regulate e-cigarettes and the Taiwan government has not yet approved any nicotine-containing e-cigarettes. Only nicotine-free products can be sold legally, but these nicotine-free products cannot claim to be smoking cessation aids, and cannot be sold in a form resembling tobacco products. There are no published studies on e-cigarette use in Taiwan. Such research is essential for tobacco control and policy design. The current study used a 2015 Taiwanese national survey of individuals aged 15 years and older to:  (1) describe the epidemiology of e-cigarette use overall and by smoking status; and (2) investigate the prevalence of e-cigarette use in relation to socioeconomic characteristics in each subgroup of smoking status and age. Methods Taiwan Adult Smoking Behaviour Survey data The Adult Smoking Behavior Survey, 2015 was an annual computer-assisted telephone surveillance conducted by the Health Promotion Administration in Taiwan, to survey a nationally representative sample of non-institutionalised individuals aged 15 years or older.23 The sampling frame was the residential telephone directory of 25 administrative districts. The survey sample size for each district was probability proportional to the district population size. Within each district, the random digit dialling principle was followed to select the telephone number; the prefix of the telephone numbers was selected first, and the last two digits of the selected area codes were randomly selected. Once the call was picked up, an individual aged 15 years or older was randomly selected to answer the survey. A computer-assisted telephone interviewing system was used to elicit participants' demographic data, as well as information about the smoking behaviours, environmental tobacco smoke exposure and awareness of tobacco control interventions. At least 1068 telephone interviews were conducted for each administrative district except for Lianjiang County, where only 300 interviews were conducted due to its small population size. The response rate of the 2015 Adult Smoking Behavior Survey was 70.7%, resulting in 26 052 respondents who participated in the survey. Nationally representative estimates were produced using sampling weight, based on the sex, age, education level and population size of each district in 2014. Informed consent was waived. Variables Ever having used e-cigarettes Ever versus never having used e-cigarettes was assessed by the question: ‘Have you ever used an e-cigarette?’. Of 26 052 respondents, 31 (0.1%) who provided the answer ‘I don't know’ to the question were excluded, resulting in 26 021 respondents eligible for calculation of the prevalence of e-cigarette use. Smoking status Smoking status categories included: never smoker, former smoker and current smoker. Current smokers were defined as those who had smoked at least 100 cigarettes and reported smoking every day, or some days, in the 30 days preceding the interview. People who had smoked at least 100 cigarettes but had not smoked in the past 30 days were designated as former smokers. Never smokers comprised people who had never smoked or had smoked fewer than 100 cigarettes. Ever smokers included former and current smokers. Demographic and socioeconomic characteristics Demographic variables included age in category (15–17, 18–24, 25–44, 45–64 and ≥65 years), sex, marital status and residing in municipal areas (six major cities: Kaohsiung, New Taipei, Taichung, Tainan, Taipei and Taoyuan). Socioeconomic variables included education level, employment status and monthly household income (in New Taiwan Dollar, NTD). Statistical analysis We described demographic characteristics for the overall population and for three smoking subgroups: current, former and never smokers. Distributions of characteristics by smoking status were compared by χ2 test. We estimated the prevalence and 95% CIs of ever using e-cigarettes by demographic and socioeconomic variables for the overall population and for each smoking subgroup. Prevalence ratios and 95% CIs were estimated using multivariable log-binomial regression models for prevalence of ever using e-cigarettes in relation to sociodemographic factors in the overall population and by three smoking subgroups. These regression models included smoking status (for the entire population only), age groups, sex, residing in municipal area, marital status, education level and monthly household income. We also performed stratified analyses by smoking status (ever or never smokers) and age groups to investigate factors associated with e-cigarette use for each subgroup. The final sample for regression modelling included 25 112 respondents after we excluded 909 respondents who had missing information on age (n=766), marital status (n=159) or education level (n=106). Analyses were conducted in STATA V.12.1 (StataCorp. 2011, College Station, Texas, USA). The ‘svy’ command was used for sampling weight and ‘svy: proportion’ command was used to estimate 95% CIs of e-cigarette use prevalence. Results Sample characteristics The study sample included 26 021 respondents after excluding 31 persons (0.1%) who had missing information on e-cigarette use. Among the study sample, approximately half were women, and the majority were aged 25–64 years (67.7%), resided in municipalities, were married, were employed and had an education of high school or above (table 1). Approximately one-third of respondents self-reported a monthly household income between NTD 20 001 and 60 000 and one-third NTD 60 001 or more. Demographics of never smokers differed from current and former smokers. Compared with current or former smokers, never smokers were more likely to be women (63.0% vs 8.4–10.1%), younger than 25 years (18.9% vs 2.0–6.6%), reside municipally (43.5% vs 38.4–40.1%), be unmarried (44.3% vs 24.2–41.6%) and have a college or higher education level (46.0% vs 26.6–31.7%) (all p<0.001). Table 1 Demographics of respondents in Taiwan Adult Smoking Behavior Survey, 2015, overall and by smoking status Smoking status Entire population (n=26 021) Current smokers (n=2573) Former smokers (n=2817) Never smokers (n=20 621) Variable n (%) n (%) n (%) n (%) p Value Age (years) <0.001  15–17 958 (3.7) 13 (0.4) 3 (0.1) 942 (4.7)  18–24 1894 (12.0) 107 (6.2) 24 (1.9) 1763 (14.2)  25–44 6376 (33.4) 799 (44.5) 337 (20.3) 5239 (33.3)  45–64 10 792 (34.3) 1218 (39.4) 1406 (50.3) 8166 (31.5)  ≥65 5235 (13.6) 404 (8.3) 988 (24.9) 3839 (13.0)  Missing 766 (2.9) 32 (1.2) 59 (2.5) 672 (3.3) Sex <0.001  Female 14 398 (50.9) 281 (10.1) 204 (8.4) 13 907 (63.0)  Male 11 623 (49.2) 2292 (89.9) 2613 (91.6) 6714 (37.0) Residing in municipal area <0.001  No 16 369 (57.5) 1698 (61.6) 1830 (59.9) 12 834 (56.5)  Yes 9652 (42.5) 875 (38.4) 987 (40.1) 7787 (43.5) Married <0.001  No 9238 (42.0) 955 (41.6) 593 (24.2) 7685 (44.3)  Yes 16 624 (57.4) 1609 (58.2) 2206 (75.2) 12 805 (55.0)  Missing 159 (0.6) 9 (0.2) 18 (0.6) 131 (0.7) Education level <0.001  Middle school or below 7468 (27.0) 808 (32.5) 980 (35.1) 5674 (25.0)  High school 7728 (30.5) 976 (40.8) 833 (32.5) 5914 (28.6)  College or above 10 719 (42.1) 785 (26.6) 989 (31.7) 8944 (46.0)  Missing 106 (0.4) 4 (0.1) 15 (0.6) 86 (0.4) Employed <0.001  No 12 947 (44.4) 837 (26.0) 1405 (41.4) 10 700 (47.8)  Yes 13 074 (55.6) 1736 (74.0) 1412 (58.6) 9921 (52.2) Monthly household income (NTD) <0.001  ≤20 000 3882 (11.9) 385 (11.1) 540 (15.9) 2954 (11.6)  20 001–60 000 8785 (33.7) 951 (38.7) 905 (33.4) 6926 (33.0)  60 001–100 000 4899 (19.9) 475 (18.8) 515 (19.2) 3909 (20.2)  ≥100 001 3584 (13.8) 382 (15.7) 431 (15.4) 2771 (13.3)  Unknown 4871 (20.7) 380 (15.8) 426 (16.1) 4061 (22.0) 　 E-cigarette use in the overall population The prevalence of ever using e-cigarettes in the total population was 2.7% (95% CI 2.5% to 3.0%; table 2): highest among current smokers (14.2%, 95% CI 12.4% to 16.0%) and below 1.0% among never smokers. Participants aged 18–24 years had the highest prevalence (6.5%, 95% CI 5.2% to 7.9%) compared with people of other age groups. Overall, men (4.6%) had a greater prevalence than women (0.9%); however, among current smokers, men and women appeared to have a similar prevalence of e-cigarette use (14.1% in men and 15.3% in women). Participants with a monthly household income above NTD 100 001 (4.5%) also had a higher prevalence of e-cigarette use, compared with those with a lower monthly household income (0.8–3.6%). Table 2 Prevalence (95% CI) of ever having used e-cigarettes, overall and by smoking status, among participants in Taiwan Adult Smoking Behavior Survey, 2015 Entire population (n=26 021) Current smokers (n=2573) Former smokers (n=2817) Never smokers (n=20 621) Prevalence (95% CI) Prevalence (95% CI) Prevalence (95% CI) Prevalence (95% CI) Ever having used e-cigarettes 2.7 (2.40 to 2.99) 14.2 (12.38 to 16.02) 3.2 (2.23 to 4.19) 0.8 (0.59 to 0.92) Age (years)  15–17 2.2 (1.15 to 3.19) 52.4 (21.99 to 82.87) 22.4 (0.00 to 64.28) 1.5 (0.67 to 2.32)  18–24 6.5 (5.16 to 7.91) 49.3 (38.52 to 60.11) 39.1 (17.20 to 61.03) 2.9 (1.99 to 3.91)  25–44 3.8 (3.12 to 4.39) 16.8 (13.64 to 19.92) 7.3 (3.90 to 10.74) 0.6 (0.39 to 0.88)  45–64 1.5 (1.20 to 1.82) 8.1 (6.23 to 10.01) 1.6 (0.81 to 2.36) 0.1 (0.06 to 0.24)  ≥65 0.4 (0.21 to 0.58) 3.1 (1.16 to 4.95) 0.6 (0.15 to 1.06) 0.1 (0.00 to 0.16)  Missing 0.2 (0.00 to 0.49) 2.6 (0.00 to 7.75) 0.8 (0.00 to 2.30) n/a Sex  Female 0.9 (0.66 to 1.09) 15.3 (10.05 to 20.58) 7.7 (2.78 to 12.67) 0.4 (0.24 to 0.52)  Male 4.6 (4.03 to 5.13) 14.1 (12.14 to 16.01) 2.8 (1.83 to 3.76) 1.4 (1.01 to 1.78) Residing in municipal area  No 2.4 (2.09 to 2.80) 12.3 (10.17 to 14.35) 3.1 (1.88 to 4.27) 0.6 (0.42 to 0.81)  Yes 3.0 (2.54 to 3.54) 17.3 (14.00 to 20.62) 3.4 (1.75 to 5.07) 0.9 (0.64 to 1.24) Married  No 4.0 (3.44 to 4.60) 19.1 (15.79 to 22.49) 8.3 (4.94 to 11.64) 1.4 (1.05 to 1.76)  Yes 1.7 (1.46 to 2.03) 10.6 (8.68-12.58) 1.6 (0.92 to 2.28) 0.2 (0.14 to 0.34)  Missing 1.0 (0.00 to 2.40) 21.8 (0.00 to 49.30) n/a n/a Education level  Middle school or below 1.2 (0.76 to 1.65) 6.6 (3.94 to 9.25) 0.6 (0.00 to 1.33) 0.2 (0.02 to 0.34)  High school 3.7 (3.04 to 4.26) 16.9 (13.84 to 19.95) 3.7 (1.90 to 5.49) 0.6 (0.33 to 0.80)  College or above 3.0 (2.52 to 3.44) 19.4 (15.72 to 23.08) 5.7 (3.37 to 7.96) 1.2 (0.87 to 1.52)  Missing n/a n/a n/a   n/a   Employed  No 1.6 (1.26 to 1.91) 10.8 (7.92 to 13.68) 2.5 (1.16 to 3.79) 0.7 (0.43 to 0.91)  Yes 3.6 (3.12 to 4.04) 15.4 (13.15 to 17.62) 3.7 (2.34 to 5.12) 0.8 (0.60 to 1.07) Monthly household income (NTD)  ≤20 000 0.8 (0.39 to 1.19) 5.4 (2.48 to 8.37) 0.8 (0.00 to 2.00) 0.1 (0.00 to 0.19)  20 001–60 000 2.3 (1.79 to 2.73) 10.9 (8.22 to 13.55) 2.6 (1.20 to 4.00) 0.6 (0.31 to 0.82)  60 001–100 000 3.6 (2.87 to 4.32) 18.8 (14.44 to 23.10) 4.1 (1.67 to 6.57) 1.2 (0.71 to 1.73)  ≥100 001 4.5 (3.50 to 5.57) 23.3 (17.49 to 29.07) 2.1 (0.54 to 3.73) 1.3 (0.75 to 1.88)  Unknown 2.4 (1.76 to 3.05) 14.0 (9.23 to 18.82) 6.8 (2.87 to 10.73) 0.6 (0.32 to 0.95) Current and former smokers were more likely than never smokers to report ever having used e-cigarettes (table 3). Participants younger than 45 years had a greater prevalence ratio of ever using e-cigarettes relative to the 45–64 years age group, after adjusting for smoking status, sex, residing in municipal areas and other socioeconomic variables. Ever using e-cigarette was also associated with not being currently married, having a college or higher education and monthly household income ≥NTD60 001. Sex, employment status or residing in municipal areas was not associated with ever having used e-cigarettes. Table 3 Adjusted prevalence ratios for ever having used e-cigarettes, overall and by tobacco smoking status, in Taiwan, 2015 Entire population (n=25 112) Ever smokers (n=5276) Never smokers (n=19 843) Adjusted PR (95% CI) Adjusted PR (95% CI) Adjusted PR (95% CI) Smoking status  Never smokers 1.0 (Reference) n/a n/a  Former smokers 8.3 (5.21 to 13.1) n/a n/a  Current smokers 21.5 (15.4 to 29.8) n/a n/a Age (years)  15–17 4.4 (2.53 to 7.68) 7.0 (3.71 to 13.4) 8.5 (2.96 to 24.1)  18–24 5.2 (3.73 to 7.26) 6.1 (4.24 to 8.67) 10.7 (4.11 to 28.1)  25–44 2.1 (1.57 to 2.81) 2.4 (1.75 to 3.23) 2.7 (1.21 to 5.80)  45–64 1.0 (Reference) 1.0 (Reference) 1.0 (Reference)  ≥65 0.4 (0.22 to 0.66) 0.3 (0.18 to 0.56) 0.5 (0.09 to 3.03) Sex  Female 1.0 (Reference) 1.0 (Reference) 1.0 (Reference)  Male 1.3 (0.96 to 1.87) 1.0 (0.70 to 1.36) 2.4 (1.48 to 3.80) Residing in municipal area  No 1.0 (Reference) 1.0 (Reference) 1.0 (Reference)  Yes 1.2 (0.94 to 1.41) 1.1 (0.90 to 1.45) 1.4 (0.91 to 2.19) Married  No 1.0 (Reference) 1.0 (Reference) 1.0 (Reference)  Yes 0.7 (0.54 to 0.89) 0.7 (0.50 to 0.84) 0.6 (0.32 to 1.13) Education level  Middle school or below 1.0 (Reference) 1.0 (Reference) 1.0 (Reference)  High school 1.5 (0.97 to 2.30) 1.5 (0.97 to 2.42) 0.7 (0.19 to 2.28)  College or above 1.6 (1.04 to 2.51) 1.4 (0.90 to 2.31) 1.1 (0.32 to 4.05) Employed  No 1.0 (Reference) 1.0 (Reference) 1.0 (Reference)  Yes 0.9 (0.70 to 1.17) 1.0 (0.72 to 1.29) 1.3 (0.73 to 2.25) Monthly household income (NTD)  ≤60 000 1.0 (Reference) 1.0 (Reference) 1.0 (Reference)  ≥60 001 1.6 (1.22 to 1.97) 1.5 (1.12 to 1.94) 1.8 (1.09 to 3.09)  Unknown 1.1 (0.81 to 1.50) 1.2 (0.86 to 1.73) 0.9 (0.45 to 1.70) PR, prevalence ratio. E-cigarette use by smoking status Stratified by smoking status, age remained associated with ever having used e-cigarettes, with a high prevalence among people aged 24 years or younger. For example, the prevalence of ever having used e-cigarettes was 49.3% (95% CI 38.5% to 60.1%) among current smokers aged 18–24 years and 39% among former smokers of the same age (table 2). Compared with people aged 45–64 years, those aged 18–24 years had a 6-fold to 10-fold higher prevalence of ever having used e-cigarettes, irrespective of smoking status (table 3). Among current and former smokers, men and women had a similar prevalence ratio of ever having used e-cigarettes; yet among never smokers, men were more than twice as likely as women to self-report ever having used e-cigarettes. Associations between ever having used e-cigarettes and socioeconomic variables were not observed in smoking subgroups, except that education in current or former smokers and monthly household income in current smokers were each associated with ever having used e-cigarettes. Associations of e-cigarette with other characteristics by age and different smoking status Ever having used e-cigarettes in association with age and smoking status prompted us to perform analyses stratified by three age groups (18–24, 25–44, ≥45 years) in ever smokers and in never smokers (figures 1 and 2). Figure 1 Prevalence ratio (95% CI) for e-cigarette use among ever smokers aged 25–44 years and 45 years or older. Figure 2 Prevalence ratio (95% CI) for e-cigarette use among never smokers aged 18–24 and 25–44 years. Among ever smokers aged 25–44 years, marriage was associated with a lower prevalence of ever using e-cigarettes while those who lived in a municipal area had a greater prevalence (figure 1). Ever smoking men aged 25–44 years were less likely to self-report ever using e-cigarettes, which was not observed in other subgroups. Among never smokers aged 18–24 years, having a higher education level was associated with a decreased prevalence of ever using e-cigarettes, while living in a municipal area and having a higher household income were associated with increased prevalence (figure 2). Non-smoking men aged 25–44 years were 2.71 times more likely than women aged 25–44 years to ever use e-cigarettes. We did not perform the stratified analysis for ever smokers aged 18–24 years due to the small sample size. Results of never smokers aged 45 years or older were all non-significant (see online supplementary table S1). 10.1136/bmjopen-2016-014263.supp1supplementary table Prevalence of e-cigarette use among age groups with different smoking status Discussion In 2015, 2.7% of a nationally representative sample of Taiwan residents aged 15 years or older reported that they had ever used e-cigarettes. This prevalence was similar to that of Hong Kong (2.3% in 2014)11 and other Asian countries,12 but lower than the rate of 12.6% in the USA in 2014.6 Smoking status was the main determinant of ever having used e-cigarettes. Similar to countries where e-cigarette use is common,6 7 the prevalence of ever having used e-cigarettes was <5% among individuals who were never and former smokers. In current smokers, the prevalence of ever having used e-cigarettes was 14.2% in Taiwan, lower than the prevalence among smokers in Europe (20.3%, 2012),7 Australia (23.7%, 2013) and the UK (43.3%, 2013).8 The low prevalence of ever having used e-cigarettes in Taiwan could be partly due to the relatively low tobacco smoking rates (30% in men and 4% in women, 2015)24 and a restricted e-cigarette market. Of note, the prevalence among young smokers in Taiwan was as high as that in the western countries: 38–50% of former and current smokers aged 15–24 years had ever used e-cigarettes, suggesting that young adults in Taiwan who have ever smoked tend to experiment with e-cigarettes.6 7 The high prevalence in the young population signals the need for regulations in Taiwan and worldwide. E-cigarette may create nicotine dependency and initiate tobacco smoking among individuals who never smoked, particularly in adolescents.21 Among individuals aged 15–17 years in this study, up to 50% of smokers and 1.5% of individuals who never smoked had ever used e-cigarettes. These findings suggested that ∼49 000 smoking and 12 100 non-smoking adolescents have ever used e-cigarettes in Taiwan, given the reported smoking rate of 10.4%23 among 900 000 adolescents aged 15–17 years in 2015.25 More importantly, a prospective cohort of non-smoking adolescents has shown a 16% increase in initiation of tobacco smoking after 12 months among individuals who have ever used e-cigarettes (25% vs 9%).19 We estimate that 2000 adolescents in Taiwan may potentially initiate tobacco smoking. E-cigarette regulations and public education regarding nicotine dependency, particularly targeting adolescents, will be essential for tobacco control. E-cigarette policies and regulations vary around the world. According to the Institute for Global Tobacco Control, e-cigarette sales are banned in 26 countries and there are age or other restrictions on their purchase in 38 countries.26 In Taiwan, several existing laws apply to e-cigarettes based on their nicotine content or shape; however, regulations are not applicable to e-cigarettes that contain no nicotine, claim no therapeutic properties or do not resemble cigarettes. Some concerns have been raised. First, the safety of chemicals in aerosols or solutions remains to be established.2 13 15 Second, e-cigarette use in public may weaken the effect of smoke-free policies,1 20 decrease the quitting motivation of smokers1 and potentially diminish the denormalisation of smoking.1 Impacts of e-cigarette use on tobacco smoking in populations with a low smoking prevalence, such as Taiwan, warrant prospective investigation. E-cigarette use is more prevalent among people with high socioeconomic status in Asia,12 but not in Europe.7 Likewise, people in Taiwan with higher education levels or high household incomes, especially former and current cigarette smokers, are more likely to self-report ever using e-cigarettes. This pattern is consistent with the observation that people with high socioeconomic status tend to adopt new technology or knowledge quickly. In addition, our study suggests that women and men who have ever smoked had a similar likelihood of trying e-cigarettes, although men had a higher rate than women overall. This study reported the prevalence of ever having used e-cigarettes in a nationally representative sample in Taiwan. Previous studies have mostly reported the prevalence of e-cigarette use within specific regions. With a larger sample size, we were able to report prevalence stratified by smoking status and age groups. This study has some notable limitations. First, information about specific e-cigarette products, such as their flavour, content, solution or design, was unavailable. Second, the relatively small sample size of adolescents and young adults may limit the generalisation of these findings. Third, this was a cross-sectional survey with only one question that assessed ever having used e-cigarettes. Respondents may have under-reported e-cigarette use. E-cigarettes may pose some challenges to tobacco control. For example, the correlates of ever having used e-cigarettes differed by smoking status and age and seemed to vary by country. Regulations of e-cigarettes should be based on country-specific research. Furthermore, e-cigarettes may serve as a gateway to initiating tobacco smoking in the young population.19 In this study, 1.5% of never smokers aged 15–17 years in Taiwan, ∼12 100 adolescents, had ever used e-cigarettes. They may be susceptible to tobacco smoking. Conclusion and public health implications This national representative study in Taiwan suggests that people with higher education level and income are more likely to use e-cigarettes. Among smokers, women have a prevalence of using e-cigarettes similar to or even greater than that of men. The overall rate of ever having used e-cigarettes in Taiwan is lower than that in western countries; however, adolescents and young adults have the highest prevalence of ever using e-cigarettes—almost half of the ever-smokers aged 15–24 years reported having tried e-cigarettes. The emergence of e-cigarettes may have raised new challenges for tobacco control. Public health administrators and policy developers need to implement preventive measures against nicotine dependence among e-cigarette users, especially in the younger population. Future investigations of regular e-cigarette use, changes in tobacco smoking behaviours among e-cigarette users as well as underlying reasons for high prevalence of e-cigarette use among young adults are needed to guide public policy. Contributors: All authors contributed to drafts and approved the final version of the paper. H-CC and P-YC conceived the study and were in charge of the major work of manuscript writing. Y-WT and M-NS contributed to data acquisition, led the study design and revised the article for important intellectual content. Y-TW undertook the data analysis and contributed to results interpretation. H-CC, P-YC and Y-TW wrote the article. All authors had full access to all of the data and take responsibility for its integrity and accuracy in the analysis; P-YC is the guarantor of this manuscript on behalf of his coauthors. Imad Sawaya, MS, provided language editing and proofreading. Funding: National Yang-Ming University received research funding from the tobacco health and welfare surcharges by the Health Promotion Administration for the project titled “International Collaborative Project for the Evaluation of Medical Services for Smoking Cessation (G1031227-105)”. Disclaimer: The research presented in this paper is that of the authors and does not reflect the official policy of the Health Promotion Administration, Ministry of Health and Welfare, Taiwan. The study sponsor had no role in the study design, data analysis or interpretation and the researchers were independent of the sponsor in writing the manuscript. Competing interests: None declared. Ethics approval: The Institutional Review Board of National Yang-Ming University. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 World Health Organization . Electronic nicotine delivery systems|Report by WHO . Conference of the Parties to the WHO Framework Convention on Tobacco Control. Sixth session Moscow : Russian Federation , 13–18 October 2014. http://apps.who.int/gb/fctc/PDF/cop6/FCTC_COP6_10-en.pdf. (accessed 17 Aug 2016 ). 2 Grana R , Benowitz N , Glantz SA E-cigarettes: a scientific review . Circulation 2014 ;129 :1972 –86 . 10.1161/CIRCULATIONAHA.114.007667 24821826 3 King BA , Patel R , Nguyen KH Trends in awareness and use of electronic cigarettes among US adults, 2010–2013 . Nicotine Tob Res 2015 ;17 :219 –27 . 10.1093/ntr/ntu191 25239961 4 McMillen RC , Gottlieb MA , Shaefer RM Trends in electronic cigarette use among U.S. adults: use is increasing in both smokers and nonsmokers . Nicotine Tob Res 2015 ;17 :1195 –202 . 10.1093/ntr/ntu213 25381306 5 Reid JL , Rynard VL , Czoli CD Who is using e-cigarettes in Canada? Nationally representative data on the prevalence of e-cigarette use among Canadians . Prev Med 2015 ;81 :180 –3 . 10.1016/j.ypmed.2015.08.019 26348453 6 Delnevo CD , Giovenco DP , Steinberg MB Patterns of electronic cigarette use among adults in the United States . Nicotine Tob Res 2016 ;18 :715 –19 . 10.1093/ntr/ntv237 26525063 7 Vardavas CI , Filippidis FT , Agaku IT Determinants and prevalence of e-cigarette use throughout the European Union: a secondary analysis of 26 566 youth and adults from 27 Countries . Tob Control 2015 ;24 :442 –8 . 10.1136/tobaccocontrol-2013-051394 24935441 8 Yong HH , Borland R , Balmford J Trends in E-Cigarette awareness, trial, and use under the different regulatory environments of Australia and the United Kingdom . Nicotine Tob Res 2015 ;17 :1203 –11 . 10.1093/ntr/ntu231 25358657 9 Kinnunen JM , Ollila H , El-Amin Sel T Awareness and determinants of electronic cigarette use among Finnish adolescents in 2013: a population-based study . Tob Control 2015 ;24 :e264 –70 . 10.1136/tobaccocontrol-2013-051512 24827977 10 Carroll Chapman SL , Wu LT E-cigarette prevalence and correlates of use among adolescents versus adults: a review and comparison . J Psychiatr Res 2014 ;54 :43 –54 . 10.1016/j.jpsychires.2014.03.005 24680203 11 Jiang N , Chen J , Wang MP Electronic cigarette awareness and use among adults in Hong Kong . Addict Behav 2016 ;52 :34 –8 . 10.1016/j.addbeh.2015.08.008 26348280 12 Palipudi KM , Mbulo L , Morton J Awareness and current use of electronic cigarettes in Indonesia, Malaysia, Qatar, and Greece: findings from 2011–2013 Global adult tobacco surveys . Nicotine Tob Res 2016 ;18 :501 –7 . 10.1093/ntr/ntv081 25895951 13 Cheng T Chemical evaluation of electronic cigarettes . Tob Control 2014 ;23 (Suppl 2 ):ii11 –17 . 10.1136/tobaccocontrol-2013-051482 24732157 14 Hitchman SC , Brose LS , Brown J Associations between E-Cigarette type, frequency of use, and quitting smoking: findings from a longitudinal online panel survey in Great Britain . Nicotine Tob Res 2015 ;17 :1187 –94 . 10.1093/ntr/ntv078 25896067 15 Brown CJ , Cheng JM Electronic cigarettes: product characterisation and design considerations . Tob Control 2014 ;23 (Suppl 2 ):ii4 –10 . 10.1136/tobaccocontrol-2013-051476 24732162 16 Etter JF , Bullen C A longitudinal study of electronic cigarette users . Addict Behav 2014 ;39 :491 –4 . 10.1016/j.addbeh.2013.10.028 24229843 17 Farsalinos KE , Kistler KA , Gillman G Evaluation of electronic cigarette liquids and aerosol for the presence of selected inhalation toxins . Nicotine Tob Res 2015 ;17 :168 –74 . 10.1093/ntr/ntu176 25180080 18 Kalkhoran S , Glantz SA E-cigarettes and smoking cessation in real-world and clinical settings: a systematic review and meta-analysis . Lancet Respir Med 2016 ;4 :116 –28 . 10.1016/S2213-2600(15)00521-4 26776875 19 Leventhal AM , Strong DR , Kirkpatrick MG Association of electronic cigarette use with initiation of combustible tobacco product smoking in early adolescence . JAMA 2015 ;314 :700 –7 . 10.1001/jama.2015.8950 26284721 20 Bostean G , Trinidad DR , McCarthy WJ E-Cigarette use among never-smoking California students . Am J Pub Health 2015 ;105 :2423 –5 . 10.2105/AJPH.2015.302899 26469671 21 Coleman BN , Apelberg BJ , Ambrose BK Association between electronic cigarette use and openness to cigarette smoking among US young adults . Nicotine Tob Res 2015 ;17 :212 –18 . 10.1093/ntr/ntu211 25378683 22 Wang M , Wang JW , Cao SS Cigarette smoking and electronic cigarettes use: a meta-analysis . Int J Environ Res Public Health 2016 ;13 :1 10.1155/2016/7014857 23 Health Promotion Administration, Ministry of Health and Welfare R.O.C. (Taiwan) . Global Youth Tobacco Survey (GYTS) 2015. Taipei Taiwan. 2016. http://tobacco.hpa.gov.tw/Content.aspx?MenuId=580&CID=29701 (accessed 16 Aug 2016). 24 Health Promotion Administration, Ministry of Health and Welfare R.O.C. (Taiwan) . Taiwan Tobacco Control Annual Report 2015, Taipei, Taiwan, 2016. http://health99.hpa.gov.tw/media/public/pdf/21806.pdf (accessed 16 Aug 2016 ). 25 Department of Statistics, Ministry of the Interior R.O.C. (Taiwan) . Statistical yearbook of interior 2015 Yearly Bulletin of Interior Statistics. Taipei Taiwan. 2016 http://statis.moi.gov.tw/micst/stmain.jsp?sys=100 (accessed 16 Aug 2016) . 26 Institute for Global Tobacco Control . Country laws regulating e-cigarettes: a policy scan . Baltimore, MD : Johns Hopkins Bloomberg School of Public Health , 2016 http://globaltobaccocontrol.org/node/14052 (accessed 16 Aug 2016 ).

PMC005xxxxxx/PMC5372146.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01429310.1136/bmjopen-2016-014293HIV/AIDSResearch15061842169916921712Effect of alcohol consumption and psychosocial stressors on preterm and small-for-gestational-age births in HIV-infected women in South Africa: a cohort study Sania Ayesha 12Brittain Kirsty 34Phillips Tamsin K 34Zerbe Allison 1Ronan Agnes 34Myer Landon 34Abrams Elaine J 15 1 ICAP, Columbia University, Mailman School of Public Health, New York, USA 2 Department of Epidemiology, Mailman School of Public Health, Columbia University, New York, USA 3 Division of Epidemiology & Biostatistics, School of Public Health & Family Medicine, University of Cape Town, Cape Town, South Africa 4 Centre for Infectious Disease Epidemiology and Research, University of Cape Town, Cape Town, South Africa 5 College of Physicians and Surgeons, Columbia University, New York, USACorrespondence to Dr Ayesha Sania; ays328@mail.harvard.edu2017 20 3 2017 7 3 e01429317 9 2016 15 2 2017 24 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objectives Psychosocial stressors such as depression and stress, intimate partner violence (IPV) and alcohol use have been linked to preterm and small-for-gestational-age (SGA) births in general populations. The prevalence of psychosocial stressors and alcohol abuse is high in many HIV-infected (HIV+) populations. Our objective was to evaluate the effects of psychosocial stressors and alcohol abuse on birth outcomes in HIV-infected women. Methods Antenatal depression and non-specific psychological distress, periconception IPV and alcohol consumption were measured during the second trimester among HIV+ women initiating antiretroviral treatment with efavirenz + emtricitibine + tenofovir in Cape Town, South Africa. Log binomial regression models were used to estimate the risk ratios (RR) and 95% CIs of the effects of psychosocial stressors and periconception alcohol consumption on birth outcomes: SGA (birth weight <10th centile for gestational age) and preterm (<37 weeks) births. Results Of the 571 mother–infant pairs, 26% of women reported hazardous alcohol consumption (Alcohol Use Disorders Identification Test-C score ≥3) periconception periods, 11% reported depressive symptoms, 7% reported non-specific psychological distress and 15% reported experiencing physical or psychological IPV. 14% of infants were born preterm and 12% were SGA. Infants born to women reporting hazardous drinking were twice (adjusted RR 2.00 (95% CI 1.13 to 3.54)) as likely to be SGA compared with women reporting low alcohol intake. Alcohol consumption did not have a significant effect on the incidence of preterm birth. Depressive symptoms, non-specific psychological distress, physical and psychological IPV did not increase the risk of SGA or preterm birth significantly. Conclusions The observed elevated risk of SGA births associated with periconception alcohol consumption underscores the urgent need to reduce alcohol consumption among women of childbearing age. Interventions targeting modifiable risk factors of adverse birth outcomes need to be integrated into HIV prevention and maternal child health programmes to improve the long-term health of HIV-exposed children. Trial registration number NCT01933477; Pre-results. birth weightpretermSGAdepressionalcoholstress ==== Body Strengths and limitations of this study There are few data evaluating the effects of maternal psychosocial stressors and alcohol consumption on birth outcomes of HIV-infected pregnant women. Psychosocial stressors, alcohol consumption and birth outcomes were measured prospectively as part of a rigorously conducted randomised trial. Maternal self-report on psychosocial stressors and alcohol consumption is amenable to measurement error. The findings of this study will be relevant in designing interventions to reduce adverse birth outcomes in HIV-infected women but may not be generalisable to other populations. Introduction The burden of adverse pregnancy outcomes such as preterm and small-for-gestational-age (SGA) births is high among HIV-infected women.1 2 Infants born preterm and SGA face a significantly higher risk of growth faltering and mortality in childhood.3 4 These infants also experience the poorest neurodevelopmental outcomes.5 6 The high incidence of adverse pregnancy outcomes in HIV-infected women is partly attributed to HIV infection and the use of antiretroviral treatment (ART) in pregnancy to prevent mother-to-child transmission.7 8 To develop targeted interventions for reduction of adverse birth outcomes in HIV-infected populations, in addition to understanding the effects of ART on intrauterine growth and length of gestation, the role of classic risk factors for preterm and SGA births needs to be elucidated. Research in the general population reported harmful effects of prenatal depression, stress, intimate partner violence (IPV) and alcohol consumption during pregnancy on birth outcomes.9–11 Alcohol passed through the placenta is known to cause oxidative stress in the developing fetus leading to growth restrictions.12 Depression, stress and IPV can directly increase adverse birth outcomes by dysregulation of the hypothalamic–pituitary–adrenocortical axis, increasing levels of stress hormones which cause placental hypoperfusion leading to growth restriction and shortening of gestational age.13 14 Antenatal depression, stress and IPV can also influence birth outcomes indirectly via poor nutrition, hygiene and healthcare seeking (figure 1).15 16 Figure 1 Hypothesised direct and indirect pathways of effects of maternal mental health stressors on birth outcomes. (Figure drawn based on mechanisms described by Diego et al,13 Wadhwa et al,14 Neggers et al15 and Zuckerman et al16). ART, antiretroviral treatment; IPV, intimate partner violence; SGA, small for gestational age. Recent studies in HIV-infected populations reported antenatal depression and stress to be associated with poor adherence to ART, and poor HIV-related clinical, immunologic and virologic outcomes, which can potentially increase the incidence of adverse pregnancy outcomes.17–20 Although the prevalence of psychosocial stressors and alcohol abuse is high in many HIV-infected populations,21–23 their effects on birth outcomes have not yet been studied. In the present study, we examined the association of antenatal depression, non-specific psychological distress, IPV and alcohol abuse with preterm and SGA births in a cohort of HIV-infected pregnant women, the majority of whom initiated ART under the Option B+ guidelines in South Africa. Methods This prospective cohort study is part of the larger Maternal and Child Health (MCH)-ART study, an ongoing trial evaluating optimal strategies for providing ART services during the postpartum period.24 From March 2013 to June 2014, a cohort of 628 HIV-positive (HIV+) pregnant women were enrolled from the Gugulethu Midwife Obstetric Unit (MOU) in Cape Town, South Africa, and attended up to three antenatal study visits. Pregnant women, aged 18 years or older, were recruited in the parent study if they were eligible to initiate ART with efavirenz + emtricitibine + tenofovir based on the current national guidelines. Until June 2013, ART eligibility was assessed based on CD4 cell count or clinical staging. After June 2013, all women were considered eligible for lifelong ART irrespective of their CD4 cell count or clinical stage (Option B+). Women were eligible to participate in the study if they planned to stay in the city until delivery and during the postpartum period. Women received clinical care during pregnancy through routine HIV services delivered by the clinic staff at the MOU. Data and blood samples for research were collected by study staff at study visits separate from routine care. Information on socio-demographic characteristics, and blood samples to measure CD4 count, viral load and haemoglobin level were collected at the first antenatal visit. At all antenatal visits, a nurse midwife collected information on medical and obstetric history and conducted physical examinations. Data on maternal psychosocial stressors: depression, non-specific psychological distress, IPV and alcohol consumption were collected at the first antenatal study visit following enrolment. Birth weight was measured by the nurse midwife attending women at delivery and was abstracted from medical records by research staff. Gestational age was calculated based on ultrasonogram assessments, except for a few subjects (1%) for whom reported date of last menstrual period at first antenatal visit was used. Using the recently published Intergrowth-21st standard,25 newborns were grouped based on their weight for gestational age into SGA and appropriate for gestational age (AGA). SGA was defined as birth weight <10th centile for gestational age; AGA was defined as birth weight ≥10th centile for gestational age. Preterm birth was defined as birth before 37 completed weeks. Depressive symptoms during the week preceding the interview were measured with the Edinburgh Postnatal Depression Scale (EPDS).26 The 10-item questionnaire was summed to a continuous score where higher scores indicated greater frequency of depressive symptoms. A dichotomous depression symptoms (probable and major) variable was defined as EPDS score ≥13. The Kessler-10 scale (K-10) was used to measure non-specific psychological distress during the month preceding the interview.27 A continuous K-10 score was calculated by summing individual-item responses such that a higher score indicated greater frequency of symptoms of psychological distress. Ordinal categories of the K-10 scores were used to define well (<20), mild (20–24), moderate (25–29) and severe (≥30) psychological distress. Alcohol consumption during the year preceding pregnancy recognition was assessed using the Alcohol Use Disorders Identification Test (AUDIT).28 Alcohol consumption during pregnancy was assessed using a later AUDIT administered in the third trimester. AUDIT-C scores were calculated by summing individual item responses across the first three AUDIT items, where higher scores indicate more problematic alcohol use. Hazardous alcohol consumption was defined as AUDIT-C score ≥3.29 Information on IPV during the year preceding pregnancy recognition was collected using the violence against women (VAW) tool developed by the WHO.30 We used log binomial regression models to estimate the risk ratios (RR) and 95% CIs of the effects of maternal stressors on adverse pregnancy outcomes.31 Estimates of the effects of each stressor on SGA and preterm birth were obtained from separate models. To control for confounding, multivariable models were adjusted for socioeconomic status, maternal age, height, haemoglobin level, parity, CD4 cell count and HIV RNA viral load at study entry. We used missing indicator terms in the multivariate models for covariates with missing data.32 All statistical analyses were performed using SAS V.9.4 (SAS Institute, Cary, North Carolina, USA). All mothers provided written informed consent to enrol themselves and their infants in the study. The trial was registered at ClinicalTrials.gov as NCT01933477. Results Of the 628 women enrolled in the study, data were available on 601 live births and a total of 571 birth outcomes were included in the analyses. We excluded 10 infants (1.7%) due to missing birth weight data, 5 (1%) due to missing gestational age data, 14 (2.3%) due to missing gender data and 1 infant with implausible gestational age and birth weight combinations. Table 1 shows the maternal characteristics at study entry. The mean age of pregnant women was 28 years and the majority (82%) had had at least one prior pregnancy and 40% were married or cohabiting at study enrolment. About 26% of the mothers had completed secondary education and 38% were employed. More than half of the pregnant women had viral load <3 log10 copies/mL and about 47% had a CD4 cell count >350 cells/µL at enrolment. The majority (73%) of the pregnant women initiated ART under Option B+ guidelines at a median gestational age of 21 weeks (IQR 16–26 weeks). Between study enrolment and the first postpartum visit, 70% of the women reported no missed doses of ART and the majority (26%) of the remaining women reported up to three missed doses of ART per month. Table 1 Characteristics of study participants at enrolment (N=571) N* (%) or mean (SD) Maternal characteristics Age, years 28.34 (5.33) Marital status  Married/cohabiting 229 (40.11)  Unmarried 342 (59.89) Education  Did not finish secondary 420 (73.56)  Completed secondary 151 (26.44) Employment status  Unemployed 354 (61.89)  Employed 217 (38.11) BMI, kg/m2  <18.5 3 (0.64)  18.5 to <25 106 (22.70)  25 to <30 149 (31.91)  ≥30 209 (44.75) Height, cm  <145 13 (2.78)  145–149 36 (7.69)  150–154 104 (22.22)  ≥155 315 (67.31) Parity  0 102 (17.86)  1 235 (41.16)  2 165 (28.90)  ≥3 69 (12.08) Haemoglobin (g/L)  Normal (≥110) 283 (51.18)  Mild anaemia (100–109) 223 (40.33)  Moderate anaemia (70–99) 45 (8.14)  Severe anaemia (<70) 2 (0.36) CD4 cell count (cells/µL)  ≤200 103 (18.56)  201–350 192 (34.59)  351–500 123 (22.16)  >500 137 (24.68) Viral load, log10 copies/mL  <3.0 315 (56.25)  >3.0–4.0 31 (5.54)  >4.0–5.0 15 (2.68)  >5.0 199 (35.54) ART regimen  Option A 156 (27.45)  Option B+ 415 (72.55)  Gestational age at ART initiation in weeks, median (IQR) 21 (16–26) Adherence to ART, missed doses/month  0 395 (69.27)  1–3 148 (25.91)  >3 28 (4.90) Violence against women  Physical 82 (14.36)  Sexual 13 (2.28)  Psychological 81 (14.19)  Any violence 115 (20.14) Depressive symptoms: Edinburgh Postnatal Depression Scale  Normal: score <13 509 (89.14)  Probable depression: score 13–14 19 (3.33)  Major depression: score >15 43 (7.53) Non-specific psychological distress, Kessler-10  Well: <20 25 (91.94)  Mild distress: 20–24 27 (4.73)  Moderate distress: 25–29 9 (1.58)  Severe distress: ≥30 10 (1.75)  Heavy/hazardous drinking in women (AUDIT-C score ≥3) 149 (26.09) Infant characteristics Normal birth weight (≥2500 g) 476 (83.36) Moderate low birth weight (2000–2500 g) 54 (9.46) Very low birth weight (<2000 g) 41 (7.18) Gestational age in weeks  Term (≥37 weeks) 490 (85.81)  Late preterm (34–37 weeks) 44 (7.71)  Early preterm (<34 weeks) 37 (6.48) Weight for gestational age  Appropriate for gestational age (AGA) 503 (88.10)  Small for gestational age (SGA) 68 (11.90) *Totals may not add up to 571 due to missing values. ART, antiretroviral treatment; AUDIT, Alcohol Use Disorders Identification Test; BMI, body mass index; CD4, cluster of differentiation 4. Of the 571 infants included in the analyses, 81 (14%) were preterm and 68 (12%) were SGA (table 1). Seventeen per cent (95) of the infants were born with low birth weight (LBW) (<2500 g). Of this group, the majority (55, 10% of total) were preterm-AGA and 26 (4.5% of total) were term-SGA. Of the 430 (75%) normal birthweight infants, 3% were preterm-AGA and 5% were term-SGA (figure 2). Figure 2 Distribution of preterm and SGA in low birthweight and normal birthweight infants. AGA, appropriate for gestational age; SGA, small for gestational age. Figure 3 shows the prevalence of psychosocial stressors experienced by the pregnant women included in our study. More than a quarter of the women (26%) reported hazardous alcohol drinking during the year prior to pregnancy recognition. Of them, the majority (65%) reported having ≥5 standard drinks/day and 32% reported having ≥3 standard drinks/day (data not shown). Fewer women (9%) reported hazardous alcohol consumption in a third trimester interview which asked about alcohol consumption following recognition of pregnancy (data not shown). On the basis of the EPDS scale, 62 women (11%) had reported depressive symptoms (probable and major) during the preceding week; of them, 7% had symptoms consistent with major depression. Only 7% of the pregnant women reported non-specific psychological distress (mild, moderate and severe) during the preceding month on the K-10 scale. About 15% of the pregnant women reported having experienced physical or psychological IPV and 2% reported having experienced sexual IPV during the year prior to pregnancy recognition. Figure 3 Prevalence of psychosocial stressors and alcohol consumption in study population. The RRs for SGA and preterm birth are given in table 2. Infants born to pregnant women who reported hazardous drinking were twice as likely to be SGA compared with women reporting low alcohol intake in the periconception period (adjusted RR 2.00 (95% CI 1.13 to 3.54)). Hazardous alcohol drinking did not increase the incidence of preterm birth (adjusted RR 0.89 (95% CI 0.51 to 1.53)). Women reporting hazardous drinking during pregnancy assessed in the third trimester also experienced elevated risk of SGA (adjusted RR 1.62 (95% CI 0.68 to 3.84)) and preterm birth (adjusted RR 1.13 (95% CI 0.48 to 2.65)) but these risk elevations were not statistically significant (see online supplementary table S1). Table 2 Crude and adjusted estimates of effect of psychosocial factors and alcohol use on small-for-gestational-age (SGA) and preterm births   Small for gestational age (SGA) Preterm birth Risk ratio (95% CI) Risk ratio (95% CI) No. of live births No. of SGA Crude Adjusted* No. of preterm Crude Adjusted* Depressive symptoms  Normal 514 51 1.00 (Reference) 1.00 (Reference) 56 1.00 (Reference) 1.00 (Reference)  Depression 60 11 1.71 (0.84 to 3.47) 1.52 (0.70 to 3.29) 6 0.58 (0.24 to 1.38) 0.46 (0.18 to 1.15) Intimate partner violence  No violence 456 52 1.00 (Reference) 1.00 (Reference) 70 1.00 (Reference) 1.00 (Reference)  Physical 84 9 0.90 (0.43 to 1.89) 0.85 (0.39 to 1.89) 15 1.18 (0.64 to 2.17) 1.17 (0.62 to 2.20)  Sexual 13 2 1.36 (0.29 to 6.25) 1.57 (0.31 to 7.85) 0 –† –†  Psychological 84 10 1.05 (0.51 to 2.15) 1.04 (0.49 to 2.21) 12 0.91 (0.47 to 1.75) 0.95 (0.48 to 1.89)  Any violence 115 16 1.25 (0.68 to 2.29) 1.20 (0.63 to 2.27) 15 0.88 (0.48 to 1.62) 0.89 (0.47 to 1.66) Non-specific psychological distress  Well 552 65 1.00 (Reference) 1.00 (Reference) 83 1.00 (Reference) 1.00 (Reference)  Mild distress 26 2 0.57 (0.13 to 2.45) 0.69 (0.15 to 3.15) 2 0.45 (0.10 to 1.92) 0.44 (0.10 to 1.96)  Moderate distress 9 1 0.88 (0.11 to 7.19) 0.46 (0.05 to 4.02) 3 2.39 (0.61 to 9.43) 1.35 (0.30 to 6.01)  Severe distress 11 0 –† –† 0 –† –† Alcohol use  Low 441 41 1.00 (Reference) 1.00 (Reference) 66 1.00 (Reference) 1.00 (Reference)  Hazardous 158 27 2.06 (1.21 to 3.48) 2.00 (1.13 to 3.54) 22 0.93 (0.55 to 1.56) 0.89 (0.51 to 1.53) *Adjusted for maternal age, socio-economic status, BMI, hemoglobin level, parity, cd4 count at enrolment, and viral load at enrolment. †Model does not converge, few or no outcome in the exposure category. 10.1136/bmjopen-2016-014293.supp1supplementary table Crude and adjusted estimates of effect of intimate partner violence and alcohol use on SGA and preterm births. (exposures measured during third trimester) Compared with pregnant women reporting no depressive symptoms, women reporting depressive symptoms in the second trimester had a risk ratio of 1.52 (95% CI 0.70 to 3.29) for SGA birth. Depressive symptoms in the second trimester also did not increase the incidence of preterm birth. Women reporting non-specific psychological distress on the K-10 scale were not at a higher risk of delivering an SGA baby. A statistically non-significant elevation of risk of preterm birth was observed among women reporting moderate distress. A small statistically non-significant elevation of risk of SGA was observed among women reporting physical and psychological IPV in the year preceding pregnancy recognition. Discussion Our study evaluated the associations of maternal psychosocial stressors and alcohol abuse with preterm and SGA births among infants born to HIV-infected pregnant women initiating ART, in a semiurban South African setting with high prevalence of alcohol use and moderate prevalence of mental health stressors.33 34 Pregnant women in our study experienced a high incidence of preterm and moderate incidence of SGA births. We noted a twofold increase in the risk of SGA births among women reporting hazardous alcohol use during the periconception period. We did not observe any significantly increased risk of preterm and SGA births associated with maternal psychosocial stressors. Our study sample had a lower incidence of SGA (12 vs 23%) and a higher incidence of preterm births, <37 weeks (14 vs 8%) compared with the national average for South Africa.35 Low incidence of SGA in a population with high incidence of preterm births is expected. Since the period of rapid intrauterine growth is cut short for preterm infants, they are less likely to be growth restricted and diagnosed SGA using a conventional birth weight standard.36 The high preterm rate in our study population has been found to be associated with HIV infection but not with the timing or duration of ART use during pregnancy.37 Pregnant women in our study did not receive any protease inhibitor as part of their ART regimen which has been consistently linked with elevated risk of preterm birth in previous studies.38 39 Our observation that psychosocial stressors do not increase preterm and SGA births conforms with several prior studies reporting no effects of psychosocial stressors on adverse birth outcomes.40 The majority of the studies that demonstrated significant effects reported negative effects of depression and IPV on SGA and negative effects of stress on preterm births.41 42 Due to the low incidence of SGA in our study population, our analyses may have been underpowered to detect any small effects of stressors on SGA. In addition, our study sample was part of an intervention study, and thus may have more homogeneous healthcare utilisation and adherence to ART during pregnancy. Owing to the homogeneity of the potential mediating factors, we may not have observed effects of stressors on birth outcomes via the indirect pathway (figure 1). A meta-analysis of studies of effects of depression on LBW and preterm births found significant effects in the studies conducted in low-income countries and among women from low socioeconomic background in developed countries.10 This observation supports the hypothesis that the negative effects of depression and other mental health stressors reported in prior studies are primarily mediated via indirect pathways. The reported negative effects in some studies may also be confounded to some degree as women from low socioeconomic conditions are prone to experiencing depressive symptoms, stress and IPV.40 43 Although the potential for residual confounding remains in observational studies, biases in our estimates are likely to be minimal, as they are adjusted for the known confounders including socioeconomic status. Our observation of significant negative effects of periconception alcohol consumption on incidence of SGA births and no effect on incidence of preterm birth is in agreement with reports from developed countries44 with a few exceptions which reported protective effects of mild to moderate prenatal alcohol consumption on preterm birth.45 46 While the harmful role of alcohol consumption in pregnancy on fetal growth and the nervous system is well documented, data on the effects of periconception consumption is limited.47 48 Periconception alcohol consumption was demonstrated to be sufficient to programme growth restriction in late gestation and changes in placental structure in animal studies.49 A recent study conducted among a cohort of British women reported significant negative effects of periconception alcohol consumption on continuous birth weight centiles,50 which together with our findings on periconception consumption suggest that although many women modify their consumption behaviour following pregnancy recognition, the consequences of prior consumption may be persistent and irreversible. In places with a high prevalence of alcohol abuse and unplanned pregnancies (70% in our study population), the relatively small effect size observed magnified by a large number of pregnant women abusing alcohol can have considerable implications for the health of HIV-exposed uninfected infants via SGA births. Our estimates for the effects of periconception alcohol consumption on SGA and preterm births may be an underestimate due to the measurement error introduced by the long exposure-measurement window (1 year preceding pregnancy recognition) and also the choice of reference category for the analyses. We chose women with AUDIT-C score <3 as the reference category in our analyses which included women reporting low alcohol consumption. We were not able to account for the variability in type of alcohol consumed, as the AUDIT does not include questions on alcohol type. We may have introduced selection bias as we excluded infants with missing gestational age, birth weight and gender data, leading to an underestimate of the true effect sizes for all exposure categories. Since the frequencies of outcome were small in subgroups, we were unable to stratify the analyses by LBW and estimate the exposure effects among infants who were preterm and SGA. Also, we could not assess effect modification by maternal HIV status, as we did not have comparable data from HIV-negative pregnant women. Our study population has a high participation rate (1% of eligible women refused after screening), which originates from a poverty-stricken township of Cape Town with a high burden of HIV. While the findings of our study are relevant in designing interventions to reduce adverse birth outcomes in many similar HIV-infected populations, these results may not be generalisable to other populations. Our study is the first to evaluate the effects of maternal psychosocial stressors and alcohol consumption on birth outcomes of HIV-infected pregnant women using detailed and prospectively collected data. A large body of evidence supports that depression, stress and IPV affect the mother's ability to care for herself which has detrimental effects on the physical and mental development of infants.9 51–53 Although we did not observe any significant effects of maternal psychosocial stressors on birth outcomes, it is still important to screen and treat women for these conditions. Our finding on the effects of periconception alcohol consumption underscores the urgent need to reduce alcohol consumption among women of childbearing age. While efforts to prevent infant HIV infections have been increasingly successful, the high prevalence of adverse birth outcomes continues to threaten the health of the HIV-exposed children. Interventions targeting modifiable risk factors such as alcohol consumption incorporated into existing HIV prevention initiatives and maternal child health programmes are needed to further improve the survival and development potentials of these children. We thank the mothers and infants, as well as the field research team including nurses, midwives, supervisors and laboratory personnel, who made this study possible. Contributors: AS conceptualised and conducted the data analysis, interpreted the results and wrote the first draft of the manuscript. EJA and LM are the principal investigators of the parent trial and contributed to the study design, implementation, analysis and interpretation of the data. KB and TKP participated in study implementation and data collection, contributed in data analysis and data interpretation and provided critical inputs on the manuscript. AZ and AR participated in the study implementation and supervision of data collection. All authors read and approved the final manuscript. Funding: This research was supported by the President's Emergency Plan for AIDS Relief (PEPFAR) through the National Institute of Child Health and Human Development (NICHD), grant number 1R01HD074558. Additional funding comes from the Elizabeth Glaser Pediatric AIDS Foundation. AS was supported by the National Institute of Allergy & Infectious Diseases of the National Institutes of Health under award number T32AI114398. Competing interests: None declared. Patient consent: Obtained. Ethics approval: Human Research Ethics Committee of the University of Cape Town, Faculty of Health Sciences and the Columbia University Medical Centre Institutional Review Board. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available. ==== Refs References 1 Antoniou T , Zagorski B , Macdonald EM Trends in live birth rates and adverse neonatal outcomes among HIV-positive women in Ontario, Canada, 2002–2009: a descriptive population-based study . Int J STD AIDS 2014 ;25 :960 –6 . 10.1177/0956462414526861 24648318 2 Brocklehurst P , French R The association between maternal HIV infection and perinatal outcome: a systematic review of the literature and meta-analysis . Br J Obstet Gynaecol 1998 ;105 :836 –48 . 10.1111/j.1471-0528.1998.tb10227.x 9746375 3 Katz J , Lee AC , Kozuki N Mortality risk in preterm and small-for-gestational-age infants in low-income and middle-income countries: a pooled country analysis . Lancet 2013 ;382 :417 –25 . 10.1016/S0140-6736(13)60993-9 23746775 4 Christian P , Lee SE , Donahue Angel M Risk of childhood undernutrition related to small-for-gestational age and preterm birth in low- and middle-income countries . Int J Epidemiol 2013 ;42 :1340 –55 . 10.1093/ije/dyt109 23920141 5 McGowan JE , Alderdice FA , Doran J Impact of neonatal intensive care on late preterm infants: developmental outcomes at 3 years . Pediatrics 2012 ;130 :e1105 –12 . 10.1542/peds.2012-0745 23027176 6 Arcangeli T , Thilaganathan B , Hooper R Neurodevelopmental delay in small babies at term: a systematic review . Ultrasound Obstet Gynecol 2012 ;40 :267 –75 . 10.1002/uog.11112 22302630 7 Chen JY , Ribaudo HJ , Souda S Highly active antiretroviral therapy and adverse birth outcomes among HIV-infected women in Botswana . J Infect Dis 2012 ;206 :1695 –705 . 10.1093/infdis/jis553 23066160 8 Li N , Sando MM , Spiegelman D Antiretroviral therapy in relation to birth outcomes among HIV-infected women: a cohort study . J Infect Dis 2016 ;213 :1057 –64 . 10.1093/infdis/jiv389 26265780 9 Surkan PJ , Patel SA , Rahman A Preventing infant and child morbidity and mortality due to maternal depression . Best Pract Res Clin Obstet Gynaecol 2016 ;36 :156 –68 . 10.1016/j.bpobgyn.2016.05.007 27422745 10 Grote NK , Bridge JA , Gavin AR A meta-analysis of depression during pregnancy and the risk of preterm birth, low birth weight, and intrauterine growth restriction . Arch Gen Psychiatry 2010 ;67 : 1012 –24 . 10.1001/archgenpsychiatry.2010.111 20921117 11 Patra J , Bakker R , Irving H Dose-response relationship between alcohol consumption before and during pregnancy and the risks of low birthweight, preterm birth and small for gestational age (SGA)-a systematic review and meta-analyses . BJOG 2011 ;118 :1411 –21 . 10.1111/j.1471-0528.2011.03050.x 21729235 12 Kay HH , Tsoi S , Grindle K Markers of oxidative stress in placental villi exposed to ethanol . J Soc Gynecol Investig 2006 ;13 :118 –21 . 10.1016/j.jsgi.2005.11.007 13 Diego MA , Field T , Hernandez-Reif M Prenatal depression restricts fetal growth . Early Hum Dev 2009 ;85 :65 –70 . 10.1016/j.earlhumdev.2008.07.002 18723301 14 Wadhwa PD , Entringer S , Buss C The contribution of maternal stress to preterm birth: issues and considerations . Clin Perinatol 2011 ;38 :351 –84 . 10.1016/j.clp.2011.06.007 21890014 15 Neggers Y , Goldenberg R , Cliver S The relationship between psychosocial profile, health practices, and pregnancy outcomes . Acta Obstet Gynecol Scand 2006 ;85 :277 –85 . 10.1080/00016340600566121 16553174 16 Zuckerman B , Amaro H , Bauchner H Depressive symptoms during pregnancy: relationship to poor health behaviors . Am J Obstet Gynecol 1989 ;160 (Pt 1 ):1107 –11 . 10.1016/0002-9378(89)90170-1 2729387 17 Alhusen JL , Bullock L , Sharps P Intimate partner violence during pregnancy and adverse neonatal outcomes in low-income women . J Womens Health (Larchmt) 2014 ;23 :920 –6 . 10.1089/jwh.2014.4862 25290007 18 Ammassari A , Antinori A , Aloisi MS Depressive symptoms, neurocognitive impairment, and adherence to highly active antiretroviral therapy among HIV-infected persons . Psychosomatics 2004 ;45 :394 –402 . 10.1176/appi.psy.45.5.394 15345784 19 Ickovics JR , Hamburger ME , Vlahov D , HIV Epidemiology Research Study Group . Mortality, CD4 cell count decline, and depressive symptoms among HIV-seropositive women: longitudinal analysis from the HIV epidemiology research study . JAMA 2001 ;285 :1466 –74 . 10.1001/jama.285.11.1466 11255423 20 Starace F , Ammassari A , Trotta MP Depression is a risk factor for suboptimal adherence to highly active antiretroviral therapy . J Acquir Immune Defic Syndr 2002 ;31 (Suppl 3 ):S136 –9 . 10.1097/00126334-200212153-00010 12562037 21 Collins PY , Holman AR , Freeman MC What is the relevance of mental health to HIV/AIDS care and treatment programs in developing countries? A systematic review . Aids 2006 ;20 : 1571 –82 .16868437 22 Hatcher AM , Tsai AC , Kumbakumba E Sexual relationship power and depression among HIV-infected women in rural Uganda . PLoS ONE 2012 ;7 :e49821 10.1371/journal.pone.0049821 23300519 23 Kinyanda E , Woodburn P , Tugumisirize J Poverty, life events and the risk for depression in Uganda . Soc Psychiatry Psychiatr Epidemiol 2011 ;46 :35 –44 . 10.1007/s00127-009-0164-8 19916062 24 Myer L , Phillips TK , Zerbe A Optimizing antiretroviral therapy (ART) for maternal and child health (MCH): rationale and design of the MCH-ART study . J Acquir Immune Defic Syndr 2016 ;72 (Suppl 2 ):S189 –96 . 10.1097/QAI.0000000000001056 27355508 25 Papageorghiou AT , Ohuma EO , Altman DG International standards for fetal growth based on serial ultrasound measurements: the Fetal Growth Longitudinal Study of the INTERGROWTH-21st Project . Lancet 2014 ;384 :869 –79 . 10.1016/S0140-6736(14)61490-2 25209488 26 Cox JL , Holden JM , Sagovsky R Detection of postnatal depression. Development of the 10-item Edinburgh Postnatal Depression Scale . Br J Psychiatry 1987 ;150 :782 –6 .3651732 27 Kessler RC , Andrews G , Colpe LJ Short screening scales to monitor population prevalences and trends in non-specific psychological distress . Psychol Med 2002 ;32 :959 –76 . 10.1017/S0033291702006074 12214795 28 Saunders JB , Aasland OG , Babor TF Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO collaborative project on early detection of persons with harmful alcohol consumption--II . Addiction 1993 ;88 :791 –804 . 10.1111/j.1360-0443.1993.tb02093.x 8329970 29 Bush K , Kivlahan DR , McDonell MB The AUDIT alcohol consumption questions (AUDIT-C): an effective brief screening test for problem drinking. Ambulatory Care Quality Improvement Project (ACQUIP). Alcohol Use Disorders Identification Test . Arch Intern Med 1998 ;158 :1789 –95 .9738608 30 Garcia-Moreno C , Heise L , Jansen HA WHO multi-country study on women's health and domestic violence against women . Geneva : World Health Organization , 2005 . 31 Spiegelman D , Hertzmark E Easy SAS calculations for risk or prevalence ratios and differences . Am J Epidemiol 2005 ;162 :199 –200 . 10.1093/aje/kwi188 15987728 32 Miettinen OS Theoretical epidemiology: principles of occurrence research in medicine . 19 edn John Wiley , 1985 . 33 Peer N , Lombard C , Steyn K Rising alcohol consumption and a high prevalence of problem drinking in black men and women in Cape Town: the CRIBSA study . J Epidemiol Community Health 2014 ;68 :446 –52 . 10.1136/jech-2013-202985 24403519 34 Brittain K , Mellins CA , Phillips T Social support, stigma and antenatal depression among HIV-infected pregnant women in South Africa . AIDS Behav 2017 ;21 :274 –82 . 10.1007/s10461-016-1389-7 35 Lee AC , Katz J , Blencowe H National and regional estimates of term and preterm babies born small for gestational age in 138 low-income and middle-income countries in 2010 . Lancet Glob Health 2013 ;1 :e26 –36 . 10.1016/S2214-109X(13)70006-8 25103583 36 Cooke RW Conventional birth weight standards obscure fetal growth restriction in preterm infants . Arch Dis Child Fetal Neonatal Ed 2007 ;92 :F189 –92 . 10.1136/adc.2005.089698 16547077 37 Malaba T , Phillips T , Petro G Timing of ART initiation in pregnancy and birth outcomes in South African women The Conference on Retroviruses and Opportunistic Infections (CROI) Boston , 2016 . 38 Powis KM , Kitch D , Ogwu A Increased risk of preterm delivery among HIV-infected women randomized to protease versus nucleoside reverse transcriptase inhibitor-based HAART during pregnancy . J Infect Dis 2011 ;204 :506 –14 . 10.1093/infdis/jir307 21791651 39 Mofenson LM Antiretroviral therapy and adverse pregnancy outcome: the elephant in the room? J Infect Dis 2016 ;213 : 1051 –4 . 10.1093/infdis/jiv390 26265779 40 Accortt EE , Cheadle AC , Dunkel Schetter C Prenatal depression and adverse birth outcomes: an updated systematic review . Matern Child Health J 2015 ;19 :1306 –37 . 10.1007/s10995-014-1637-2 25452215 41 Dunkel Schetter C , Tanner L Anxiety, depression and stress in pregnancy: implications for mothers, children, research, and practice . Curr Opin Psychiatry 2012 ;25 :141 –8 . 10.1097/YCO.0b013e3283503680 22262028 42 Sanchez SE , Alva AV , Diez Chang G Risk of spontaneous preterm birth in relation to maternal exposure to intimate partner violence during pregnancy in Peru . Matern Child Health J 2013 ;17 :485 –92 . 10.1007/s10995-012-1012-0 22527763 43 Capaldi DM , Knoble NB , Shortt JW A systematic review of risk factors for intimate partner violence . Partner Abuse 2012 ;3 :231 –80 . 10.1891/1946-6560.3.2.231 22754606 44 Henderson J , Gray R , Brocklehurst P Systematic review of effects of low-moderate prenatal alcohol exposure on pregnancy outcome . BJOG 2007 ;114 :243 –52 . 10.1111/j.1471-0528.2006.01163.x 17233797 45 Pfinder M , Kunst AE , Feldmann R Preterm birth and small for gestational age in relation to alcohol consumption during pregnancy: stronger associations among vulnerable women? Results from two large Western-European studies . BMC Pregnancy Childbirth 2013 ;13 :49 10.1186/1471-2393-13-49 23433310 46 O'Leary CM , Nassar N , Kurinczuk JJ The effect of maternal alcohol consumption on fetal growth and preterm birth . BJOG 2009 ;116 :390 –400 . 10.1111/j.1471-0528.2008.02058.x 19187371 47 Paintner A , Williams AD , Burd L Fetal alcohol spectrum disorders-- implications for child neurology, part 1: prenatal exposure and dosimetry . J Child Neurol 2012 ;27 :258 –63 . 10.1177/0883073811428376 22351188 48 Jones KL The effects of alcohol on fetal development . Birth Defects Res C Embryo Today 2011 ;93 :3 –11 . 10.1002/bdrc.20200 21425437 49 Gårdebjer EM , Cuffe JS Periconceptional alcohol consumption causes fetal growth restriction and increases glycogen accumulation in the late gestation rat placenta . Placenta 2014 ;35 :50 –7 . 10.1016/j.placenta.2013.10.008 24239160 50 Nykjaer C , Alwan NA , Greenwood DC Maternal alcohol intake prior to and during pregnancy and risk of adverse birth outcomes: evidence from a British cohort . J Epidemiol Community Health 2014 ;68 :542 –9 . 10.1136/jech-2013-202934 24616351 51 Chai J , Fink G , Kaaya S Association between intimate partner violence and poor child growth: results from 42 demographic and health surveys . Bull World Health Organ 2016 ;94 :331 –9 . 10.2471/BLT.15.152462 27147763 52 Kinsella MT , Monk C Impact of maternal stress, depression and anxiety on fetal neurobehavioral development . Clin Obstet Gynecol 2009 ;52 :425 –40 . 10.1097/GRF.0b013e3181b52df1 19661759 53 Rahman A , Iqbal Z , Bunn J Impact of maternal depression on infant nutritional status and illness: a cohort study . Arch Gen Psychiatry 2004 ;61 :946 –52 . 10.1001/archpsyc.61.9.946 15351773

PMC005xxxxxx/PMC5372147.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01473010.1136/bmjopen-2016-014730PaediatricsProtocol1506171917121704Impact of the Early Start Denver Model on the cognitive level of children with autism spectrum disorder: study protocol for a randomised controlled trial using a two-stage Zelen design Touzet Sandrine 12Occelli Pauline 12Schröder Carmen 34Manificat Sabine 5Gicquel Ludovic 67Stanciu Razvana 8Schaer Marie 9Oreve Marie-Joelle 10Speranza Mario 1011Denis Angelique 12Zelmar Amelie 12Falissard Bruno 1213Georgieff Nicolas 1415Bahrami Stephane 1116Geoffray Marie-Maude 1415The IDEA Study GroupGrisi Stéphane Marignier Stéphanie Useo Jean-Marc Pourrat Alain Peter Chloé Mengarelli Flavia Gallifet Natacha Rumillat Fanny Tenant Gaelle Sonié Sandrine Zimmerman Marc Malo Valérie Jacob Céline Grosmaître Bahrami Stéphane Queste Camille Raffeneau Florence Gilet Sophie Ammeloot Mélanie Danion-Grilliat Anne Florence Emilie Vecchionacci Valérie Janssen Lucie Dubrovskaya Anna Rohmer Monique Lambs Bénédicte Delvenne Véronique Carlier Sophie Ducenne Lesley Colin Cyrille Bouveret Laeticia 1 Pôle Information Médicale Evaluation Recherche, Hospices Civils de Lyon, Lyon F-69003, France2 Laboratoire Health Services and Performance Research, EA 7425 HESPER, Université de Lyon, Lyon F-69008 France3 Department of Child and Adolescent Psychiatry, Hopitaux universitaires de Strasbourg, Strasbourg F-67000, France4 CNRS UPR 3212—Team 9, Strasbourg University, Strasbourg F-67000, France5 Centre Hospitalier Saint Jean de Dieu, Lyon F-69008, France6 Pôle Universitaire de Psychiatrie de l'Enfant et de l'Adolescent, Centre Hospitalier Spécialisé Henri Laborit, Saint Benoît F-86280, France7 Child and Adolescent Psychiatry Department, Unité de Recherche Clinique, Université de Poitiers, Poitiers F-86000, France8 Hôpital Universitaire des Enfants Reine Fabiola, Université Libre de Bruxelles, Bruxelles 1020, Belgium9 Office Médico-Pédagogique, University of Geneva, Geneva, Switzerland10 Service Universitaire de Psychiatrie de l'Enfant et de l'Adolescent, Centre Hospitalier de Versailles, Le Chesnay F-78150, France11 EA 4047 HANDIReSP, Université de Versailles Saint-Quentin-en-Yvelines, Versailles F-78000, France12 Université Paris-Sud, CESP, INSERM, UVSQ, Université Paris-Saclay, U1178, Maison de Solenn, Paris cedex 14, France13 Department of Public Health, AP-HP, Hôpital Paul Brousse, Villejuif F-94800, France14 Department of child and adolescent psychiatry, Centre Hospitalier le Vinatiers, Bron F-69500, France15 Université de Lyon, Lyon F-69008, France16 CIC 1429, INSERM, AP-HP, Hôpital Raymond-Poincare, Garches F-92380, FranceCorrespondence to Marie-Maude Geoffray; marie-maude.geoffray@ch-le-vinatier.fr2017 27 3 2017 7 3 e01473013 10 2016 17 2 2017 27 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction Early intervention for autism spectrum disorder (ASD) in the European French-speaking countries is heterogeneous and poorly evaluated to date. Early intervention units applying the Early Start Denver Model (ESDM) for toddlers and young children with ASD have been created in France and Belgium to improve this situation. It is essential to evaluate this intervention for the political decision-making process regarding ASD interventions in European French-speaking countries. We will evaluate the effectiveness of 12 hours per week ESDM intervention on the cognitive level of children with ASD, over a 2-year period. Methods and analysis The study will be a multicentre, randomised controlled trial, using a two-stage Zelen design. Children aged 15–36 months, diagnosed with ASD and with a developmental quotient (DQ) of 30 or above on the Mullen Scale of Early Learning (MSEL) will be included. We will use a stratified minimisation randomisation at a ratio 1:2 in favour of the control group. The sample size required is 180 children (120 in the control and 60 in the intervention group). The experimental group will receive 12 hours per week ESDM by trained therapists 10 hours per week in the centre and 2 hours in the toddlers' natural environment (alternatively by the therapist and the parent). The control group will receive care available in the community. The primary outcome will be the change in cognitive level measured with the DQ of the MSEL scored at 2 years. Secondary outcomes will include change in autism symptoms, behavioural adaptation, communicative and productive language level, sensory profile and parents' quality of life. The primary analysis will use the intention-to-treat principle. An economic evaluation will be performed. Dissemination Findings from the study will be disseminated through peer reviewed publications and meetings. Trial registration number NCT02608333 (clinicaltrials.gov); Pre-results. Autism spectrum disorderEarly Start Denver ModelTwo-stage Zelen designBehavior TherapyRandomized Controlled TrialMulticenter Study ==== Body Strengths and limitations of this study The study population will be recruited from the general population of children with autism spectrum disorder (ASD) which strengthens the external validity of the study. The Early Start Denver Model (ESDM) will be conducted for 12 hours per week: 10 hours in an intervention unit and 2 hours in the toddlers' natural environment (home, nursery or preschool) encouraging generalisation of their skills. An innovative two-stage Zelen design will be used to avoid the disappointment of parents whose children are followed in the control group; these parents will not be informed of the ESDM intervention among other children. Blinding of participants will not be feasible due to the nature of the intervention; the evaluators will, however, be blinded to the intervention. An economic add-on evaluation will be performed to put direct and indirect costs against clinical and social outcomes. Introduction Background and rationale Autism spectrum disorder (ASD) is a heterogeneous developmental disorder with impairments in reciprocal social interaction and communication, a restricted repertoire of interests and behaviours and atypical sensory reactivity.1 The clinical presentation of ASD is frequently associated with intellectual disability and other developmental disorders, such as attention deficit hyperactivity disorders, and specific motor and language disorders.1 2 Overall ASD prevalence is about 0.7–1%.3 4 The prognosis involves individuals, families and society.5–7 Early intervention for ASD in European countries is heterogeneous and poorly evaluated to date.8 In French-speaking European countries, most children and their families have access to a public consultation centre specialised in autism. Families and children have regular consultations with a public psychologist or child and adolescent psychiatrist, but meeting schedules vary greatly (weekly, monthly or once a quarter). Moreover, children can also have speech and language therapy and/or occupational therapy and/or individual or group psychotherapy. Occupational therapy and individual or group psychotherapy are frequently based on psychoanalytic or psychodynamic traditions.9 Additional interventions may include behavioural therapies or the principles of Treatment and Education of Autistic and Related Communication Handicapped Children (TEAACH).10 Group psychotherapy is a relationship-based intervention in small groups (three to four children for two adults), generally conducted in sessions of 1 hr 30 min, twice a week. Of all interventions, speech and language therapies are those most frequently reported.8 These therapies can vary in frequency (30 min per session and up to four times per week). All these different treatments can be delivered in public or private centres. Most are refunded by social national funds and mandatory health insurance. With regards to schooling, children have access to regular preschool with an optional special needs assistant or they may attend special preschools for children with more severe disabilities. In France, the majority of children with ASD below the age of four have <4 hours interventions per week and go to preschool with a special needs assistant. Educational professionals are connected with healthcare professionals. An early comprehensive, intensive, behavioural intervention may improve the children's developmental trajectory.11 12 Among the various early interventions, the Early Start Denver Model (ESDM) has the advantage of having been developed for young non-verbal children with ASD.13 The ESDM is a comprehensive developmental and behavioural intervention aiming to promote optimal social interactions between the child and their environment to enable the child to learn from their environment. It integrates applied behaviour analysis with developmental and relationship-based approaches. The ESDM involves a therapist who individually interacts with the child; parents or other childcare professionals (eg, nursery/preschool) can be taught ESDM techniques to use them in daily life and thereby improving generalisation of the children's skills. ESDM delivered at home 20 hours per week has demonstrated significant gains in the cognitive level of toddlers.14 The ESDM methodology is precisely detailed in a manual, including a fidelity rating scale, and rigorous training.13 Since 2011, early intervention units applying the ESDM for toddlers and young children with ASD have been created in France and Belgium to improve early intervention in French-speaking countries. In these intervention units, 12 hours per week ESDM intervention is provided through public funding. The effectiveness of 12 hours per week ESDM has yet to be demonstrated.12 15 It is, therefore, essential to evaluate this in a population of young children with ASD in order to provide data for the political decision-making process regarding ASD interventions. Objectives The main objective is to evaluate effectiveness of 12 hours per week ESDM on the cognitive level in children aged 15–36 months with ASD. Cognitive level will be assessed through the developmental quotient (DQ) of the Mullen Scale of Early Learning (MSEL). The secondary objective is to measure the effectiveness on autism symptoms, behavioural adaptation, communicative and productive language levels and sensory profiles. Parents' quality of life will be also assessed. An economic evaluation will be performed to put direct and indirect costs against clinical and socioeconomic outcomes. Methods/design Study design The study will be a multicentre, randomised controlled trial, using a two-stage modified Zelen design (figure 1).16 17 The experimental group will consist of toddlers with ASD receiving 12 hours per week ESDM and the control group will consist of toddlers with ASD receiving care available in the community. Figure 1 Flow diagram of study protocol. In the two-stage Zelen design, all parents of eligible children consent to participate in a longitudinal observational study, and in a second step, children are randomly allocated to the intervention group. Core trial information is presented in table 1 (WHO Trial Registration Data Set). Table 1 WHO trial registration data set Data category Information Primary registry and trial identifying number clinicaltrials.gov NCT02608333 Date of registration in primary registry 11 May 2015 Secondary identifying numbers French institutional review board (2015–013B) sponsor (69HCL15_0278) Source of monetary or material support French Ministry of Health (Programme de Recherche sur la Performance du Système de soins) Fondation de France (member of the Network of European Foundations for Innovative Cooperation) Primary sponsor Centre Hospitalier Le Vinatier, Bron, France Secondary sponsor Not applicable Contact for public queries Marie-Maude Geoffray (PI), email: marie-maude.geoffray @ch-le-vinatier.fr Sandrine Touzet, email: sandrine.touzet@chu-lyon.fr Amelie Zelmar, email: amelie.zelmar@chu-lyon.fr Contact for scientific queries Marie-Maude Geoffray (PI), email: marie-maude.geoffray @ch-le-vinatier.fr Sandrine Touzet (scientific contact), email: sandrine.touzet@chu-lyon.fr Public title Impact of the Early Start Denver Model (ESDM) on the cognitive level of children with autism spectrum disorder. Scientific title Impact of the ESDM on the cognitive level of children with autism spectrum disorder: study protocol for a randomised controlled trial using a two-stage Zelen design. IDEA study. Countries of recruitment France and Belgium Health condition or problem studied Autism spectrum disorder (ASD) Interventions Intervention: ESDM for toddlers and young children with ASD Description: ESDM delivered by trained therapists 10 hours per week at the intervention unit and 2 hours per week in the toddlers' natural environment (home, nursery, preschool). Control: care available in the community, i. e. consultations with a psychologist or child and adolescent psychiatrist, speech and language therapy, occupational therapy, individual or group psychotherapy. Key inclusion and exclusion criteria Age: between 15 and 36 months Sex: male or female Inclusion criteria: diagnosis of ASD (Diagnosis and Statistical Manual of Mental Disorders, Fifth Edition, and Autism Diagnosis Observation Schedule), DQ of 30 or above at the MSEL, family living within 40 min of an ESDM unit. Exclusion criteria: serious neurological or physical condition, diagnosis of Rett syndrome, family unavailable for a regular follow-up and intervention. Study type Interventional Allocation: randomised 1:2; parallel assignment; blinding: assessor blind; two-stage modified Zelen design Date of first enrolment NA Target sample size 180 Recruitment Status NA Primary outcome Change in developmental quotient (DQ) measured using the MSEL, scored at 24 months Key secondary outcomes Child development, autism symptoms, quality of life of parents, healthcare resource use. Setting The study will be conducted in France (five centres) and in Belgium (one centre). The centres are located in university or general hospitals. All centres have an ASD diagnostic unit and a separate ESDM unit. They receive a French-speaking population including socioeconomically disadvantaged groups. Participants Inclusion criteria Children will be included if they meet the following criteria: (1) diagnosis of ASD based on the Diagnosis and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) criteria; (2) diagnosis of ASD on the Autism Diagnosis Observation Schedule (ADOS-2)18 and Autism Diagnosis Interview (ADI-R) for toddlers (we will include children above or equal to 13, which is the research cut-off);19 20 (3) aged between 15 and 36 months; (4) DQ of 30 or above at the MSEL and (5) family living within 40 min of an ESDM unit. Exclusion criteria Exclusion criteria will be as follows: (1) serious neurological or physical condition, such as epilepsy requiring medication or severe sensory impairment presented by the child that would interfere with the intervention; (2) diagnosis of Rett syndrome and (3) family unavailable for a regular follow-up and intervention. Intervention conducted in the experimental group: 12 hours per week ESDM Eligibility criteria for ESDM therapists Therapists will be speech language therapists, occupational therapists, clinical nurses and psychologists—all specialised in autism. They will all have received formal ESDM training. To ensure treatment fidelity, all therapists will have to have a score above 80% on the ESDM fidelity scale, verified once a year by an official instructor.21 ESDM description The intervention will be provided 10 hours per week at the intervention unit and 2 hours per week in the toddlers' natural environment (home, nursery and preschool). Therapists will apply the ESDM principles outlined in the manual.13 According to the manual, 20–25 behavioural and developmental objectives will be set every 12 weeks based on observations made by therapists and parents. ESDM intervention will be modified according to the child's development as detailed in the ESDM manual. ESDM is an individualised therapy with one therapist per child. However, according to the children's objectives, two children and two therapists can work together on social interaction. For 2 hours per week in the toddlers' natural environment, the majority of the time therapy will be delivered at home, alternatively by the therapist and the parent under the supervision of the therapist. Parents will be coached in ESDM by therapists. Therapists will give them information on their child's functioning and ESDM techniques, and guide them with positive feedback to practice ESDM with their child. Moreover, parents will be encouraged to use ESDM in daily activities and for periods of special ESDM time (around 30 min per day in addition to the 2 hours weekly therapy). Moreover, the therapist will convey information about autism and ESDM techniques and share the child's developmental objectives with childcare professionals of the children's nurseries and preschools. At least once a month, parents will have an additional consultation with the referent child psychologist or psychiatrist in order to assist the family further in the understanding of ASD, coping and to reinforce support for parents and siblings. Four family workshops (2 hours each) per year will be proposed to parents by professionals in order to share general information about autism, rights and social aids and to solve problems related to challenging behaviour, food selectivity and disorders, together with other parents of children with ASD. Strategies to improve adherence to ESDM interventions At least once a month, the referral child psychologist or psychiatrist will contact the parents in order to accompany the family and reinforce the motivation of the parents so that their child attends each ESDM session. Interventions that are permitted during the trial Parents will be free to seek other care available in the community. Control group Parents with toddlers allocated to the control group will be informed and referred to care available in the community. Psychiatrists and psychologists of ASD diagnostic units will guide parents toward interventions such as weekly speech and language therapy, occupational therapy as well as individual or group psychotherapy. All different treatment approaches will be quantified (type, frequency, intensity and duration) and taken into account in comparative analyses. Allocation sequence generation and randomisation All eligible toddlers, whose parents have provided consent for them to participate, will be randomised using stratified minimisation at a ratio 1:2 in favour of the control group (one toddler assigned to the intervention group for two toddlers assigned to the control group). A random element will be used, so that children will be assigned to the treatment arm which will minimise the imbalance with 90% probability. The minimisation factors will be the centre, the children's age, the DQ score on the MSEL and the severity of the ADOS score at inclusion. Randomisation will be performed using a computer program generated in the statistical software SAS by the centre for clinical research of the university teaching hospital of Lyon (Hospices Civils de Lyon, France). The algorithm will be held and controlled centrally by an independent statistician. Details of stratified minimisation will be provided in a separate document unavailable to those who enrol toddlers. Once informed of group allocation, trial administrators will inform the lead investigator of the centre, who will plan ESDM intervention delivery if applicable, or simple follow-up. Child assessment All children will be evaluated in an ASD diagnostic unit of the participating centres. The diagnosis of ASD will be confirmed by expert child psychiatrists at the time of inclusion. Children of both groups will benefit from early evaluation and will have the same follow-up evaluations over the 2-year study period in the diagnostic units. Blinding Owing to the nature of the intervention, parents of the children or ESDM therapists cannot be blind to the allocation group. ESDM therapists will not be involved in the diagnosis. The MSEL and other instruments will be administered by an assessor blind to the allocation group. At assessment points, children, parents and schools will be reminded that the assessor is blind to treatment allocation and that anything related to this should not be discussed with them. Blinding of outcome assessors and data analysts will be fully achieved. Outcome measures and tools All outcomes will be assessed at baseline, 12 and 24 months of follow-up (figure 2). Additionally, demographic and socioeconomic data will be obtained at the same times points for descriptive purposes. Figure 2 Schedule of enrolment, interventions and assessments (SPIRIT template). The primary outcome will be the change in DQ measured using the MSEL, scored from baseline to 24 months. Secondary outcome measures will focus on child development and autism symptoms, quality of life of parents and healthcare resource use. To assess child cognitive level - The MSEL is a direct observation tool measuring cognitive levels for children from birth to 68 months.22 It is a reliable, validated test widely used to evaluate the paediatric ASD population.23 24 As most of the children are likely to score too low to use the Early Learning Composite Score of the MSEL manual, we will calculate the DQ score (ie, developmental age equivalents divided by chronological age). As the Early Learning Composite Score, the DQ score will be obtained from four subscales (fine motor, visual reception, expressive language and receptive language). DQ is usually used for children with intellectual disabilities or developmental delays.23 25–27 The MSEL has been translated and back-translated in French for the purpose of the study. To assess child autism symptoms and change in autism symptoms - The ADOS-2 is a semistructured standardised observation tool, which measures core autism symptoms, that is to say reciprocal sociocommunicative interactions, repetitive and restrictive behaviours.18 A score in social affect and restrictive repetitive behaviours is measured, and an overall autism severity score will be calculated.28 - The Brief Observation of Social Communication Change (BOSCC) instrument measures social changes in children with ASD.29 Social changes are observed using standardised activities protocol between therapist and child. The instrument is scored from videos. To assess child behavioural adaptation - The Vineland Adaptive Behaviour Scales second version (VABS-2) measures personal and social skills needed for everyday living.30 It assesses socialisation, communication, motor and daily living skills, based on parent interviews. Standard scores for each of the four studied domains are provided and a composite standard score is derived from the four domains. To assess child communicative and productive language level - The Communication and Symbolic Behaviour Scales Developmental Profile (CSBS-DP) is a self-administered parent report about social communication, expressive speech/language and symbolic functioning. It provides an overall score.31 - The Dyadic Communication Measure for Autism (DCMA)32–34 is a direct observation instrument of the communication between a parent and a child with autism. It rates parental and child mutual shared attention, child communication (initiation and response) and parental communication style (synchronous/asynchronous). Coding is done based on a video of the parent and child playing together. - The ‘development of expressive language’ is a standardised French scale (Development du Language de Production en Français—DLPF) measuring the development of expressive language in French, based on a self-administered parent report.35 It was derived from The MacArthur-Bates Communicative Development Inventories.36 The level four of the DLPF will be administered at each assessment. The score obtained is the number of words and length of sentences. Assessment sensory reactivity - The Dunn's Sensory Profile for toddlers from 7 to 36 months determines how children process sensory information in everyday situations.37 The six sensory system scores (auditory, visual, touch, movement, body position and oral expression) and the four sensory pattern scores (seeking, avoiding, sensitivity and registration) will be used. It is a self-administered parent report. To assess the parent burden and quality of life - The CareQuol-7D is a validated instrument to measure the impact on the carer's quality of life, that is, the parents.38 It is a self-administered parent report that provides an overall score. Intervention process measures Irrespective of group assignment, parents may look for other care available in the community (additional speech language therapy, occupational therapy, interventions by educators or psychologists providing behavioural analysis, etc). In both groups, type, frequency, intensity and duration of interventions will be collected. Moreover, in the experimental group, adherence to the ESDM will be monitored with the fidelity scale of the published manual.13 Economic evaluation An economic evaluation will be performed alongside the clinical trial to put direct and indirect costs related to interventions against clinical and social outcomes. The evaluation will be conducted from a general societal perspective, encompassing perspectives of the national health and social care systems, as well as patients. A cost-consequences framework will be used to accommodate all relevant perspectives and outcomes. Time horizon will be of two years for the primary analysis, in line with study duration. Healthcare, social and personal resources devoted to childcare (day care and medical) will be measured throughout the study.39 Patient day care organisation, school attendance and medical and social care interventions during the preceding year will be recorded during each study visit (baseline, 12 months and 24 months). National unit costs for care and interventions will be applied to estimate direct costs. Parental employment changes related to ASD will be recorded and valued using the human capital approach. Analysis of socioeconomic data will be adapted to trial results and to the nature and diversity of toddlers' care in the control group, which are unknown to date, and will therefore be exploratory in nature. The cost-consequences framework will allow concomitant publication of all relevant costs and outcomes, avoiding overemphasis on any of them.40 Participant enrolment Toddlers will be enrolled by psychiatrists of the centres. The number of toddlers usually in care in the participating centres should be sufficient for the recruitment required. Nevertheless, in order to facilitate inclusions, healthcare professionals, usually meeting with children with ASD (paediatricians, speech language therapists and occupational therapists), will be contacted by postal mail to inform them of the study, or informed by specific meetings. Informed consent according to the two-stage Zelen design When toddlers meet the criteria for participation, parents will be informed by an expert child psychiatrist about the 24-month observational study, including an early evaluation and two evaluations over the follow-up period. They will be asked to sign written informed consent. Children will be subsequently randomised into intervention or control groups. Only the parents of children randomised to the intervention group will then be asked if they are willing to participate in an additional study involving a 12 hours per week ESDM intervention over the 24-month period. If they agree, they will be asked to sign a second written informed consent for the ESDM intervention. The parents of the children allocated to the control group will not be told about this further study or about the intervention group, but their child will be followed up as agreed initially (baseline evaluation and two yearly follow-up evaluations). Parents with children allocated to the ESDM intervention group may refuse to consent to the intervention, as it is demanding. In which case, children will remain in the study for assessment, as agreed initially. We believe this ‘crossover’ will have an acceptable probability (<10%).41–43 Sample size The primary outcome is the DQ measured using the MSEL at 24 months after randomisation. Given the limited number of 12 hours per week ESDM interventions available at the participating centres, an unequal randomisation with a ratio 1:2 (experimental: control groups, respectively) will be used. At 24 months after the randomisation, a mean difference of at least 15 points in the DQ (SD=25) is expected between the groups. This hypothesis was based on the results of two studies. One of theses studies aimed to evaluate an ESDM intervention where children received 15–20 hours group-based and 1 hour with one therapist for one child. The effect size found was of 10 points for the DQ.44 The other study was a pilot study conducted in the principal investigator's (PI) centre, where children received ESDM 12 hours per week with one therapist for one child, over 9 months, which found an improvement of 11 DQ points at the end of the study. Assuming that the 12 hours per week ESDM with one therapist for one child will be more intense than the group-based intervention, and carried out over a longer period than the pilot study, we expect a greater improvement in DQ. However, we do not expect the same improvement the second year of intervention because it is reported that children gain most of the improvement during the first year.14 To detect such a difference with a two-sided 5% significance and a power of 90%, we calculate that we will need to recruit 132 children (44 in the intervention and 88 in the control group). The required sample size has to be inflated to accommodate for the Zelen design and its dilution bias, and for attrition and dropout during the study period. Assuming a 10% non-consent rate for children randomised to the intervention group increases the sample size by 1/(1–0.10)²=20%,45 and allowing a dropout rate of 10%, according to the previous randomised controlled trial,14 will increase the total sample size to 180 children (120 in the control and 60 in the intervention group — see table 2). Table 2 The steps involved in calculating the sample size of the study Step Hypothesis Total number of patients to be included Total number of patients to be included in the intervention group Total number of patients to be included in the control group #1 To detect a difference of at least 15 points in the DQ (SD=25) with a two-sided 5% significance and a power of 90% with a ratio 1:2 (intervention:control) 132 44 88 #2 A 10% cross-over rate in the control group involve a 20% dilution rate 132+27=159 44+9=53 88+18=106 #3 A 10% drop-out rate over the 24-month follow-up 159+21=180 53+7=60 106+14=120 After the 1st step, a total of 132 patients (44:88) are required to detect a difference between groups under assumptions. After the 2nd step, a total of 159 patients (53:106) are required to correct the 20% dilution bias due to the cross-over effect. To maintain the 1:2 ratio, we rounded the sample size up to the next whole number (27 patients rather than 26). After randomisation, all patients will be then followed up during 24 months. After the 3rd step, a total of 180 patients (60:120) are to be included to account for ∼10% drop-out rate during the 24-month follow-up. Again, we rounded the sample size up to the next whole number to keep the 1:2 ratio between groups. DQ, developmental quotient. Data collection and management The study data will be collected on a secure electronic case report form (eCRF) that will be available at each centre through an internet portal. No personal identifying information will be mentioned on the eCRF. Each subject included in the study will be assigned a unique identification number that will consist of the identification number of the investigational centre, the initials of the patient and the chronological inclusion number of the patient. Multiple external validation checks will be applied: examination of the source documents and crosschecking with the data recorded in the eCRF as to its accuracy, the presence of missing data and the consistency of data. The eCRF will only include the data necessary for the analysis to be reported in a scientific publication. All study data will be stored securely in the University Hospital of Lyon. All electronic data will be secured on a password-protected laptop. Paper-based study documents will be stored in a secure filing cabinet at each centre. All electronic documents containing names or personal identifying information, necessary for the follow-up of the study, will be stored separately from other study data and protected by a code number. Access to these files will be limited to research staff involved in the study. The statistician for the final analysis will receive checked and validated data from the eCRF with no personal identifying information. There are no current plans for granting public access to the full protocol, participant-level data set or statistical code. However, if researchers wish to access the data set (eg, for conduct of secondary analysis or meta-analysis) the project management committee will facilitate this. Statistical considerations Owing to the Zelen design and its dilution bias, the analysis will be undertaken using the intention-to-treat principle. All randomised patients will be analysed according to their allocation group, regardless of whether they will start or complete the intervention protocol. Patients who will be randomised to the intervention but who refuse their allocated treatment and who opt for usual care (control group) will be analysed as though they had received the intervention. The analysis by intention-to-treat will maintain the baseline comparability of the groups. There is no interim analysis planned. Demographic and clinical characteristics, as well as baseline data, will be presented to assess the baseline comparability of the two groups. The comparability will be verified on the main characteristics using the Student's t-test or, when appropriate, the Wilcoxon-Mann-Whitney rank test for continuous variables and the χ² test or the exact Fisher test if the conditions of application of the χ² test were not fulfilled for the qualitative variables. Concerning the analysis of the primary outcome, descriptive statistics will be presented for each group as the mean change (SD, 95% CIs) in DQ score from baseline to 24 months after randomisation. The longitudinal association between ESDM intervention and DQ scores over the 24-month study period will be estimated using a linear mixed-effects regression model.46 Mixed models have the advantage of dealing with missing values because they use all of the available data from a patient over the study period. The DQ score at each follow-up visit will be defined as the dependent variable. The first model will include the following as independent variables: study group, time. The group will be coded as a dichotomous variable. Time will be treated as a continuous variable to account for the variability of the point in time at which the measurements are obtained. The intercept and slope for time will be specified as random effects to account for the initial difference and change difference over time between individuals. An interaction effect between intervention and time will be investigated to test if the slope of change is different between intervention and control groups. A repeated statement will be added to model the best covariance structure of the within-patient error, according to the Akaike's information criterion. This basic model will be adjusted for all minimisation variables (centre age at diagnosis, DQ score at baseline and autism severity score on the ADOS-2 scale). Adjusted models will additionally control for other patient characteristics, family characteristics (social class based on occupation and education level) and the total number of ESDM hours toddlers will have received over the study period. Additional covariates can be included if they are significant at 0.05 level according to the likelihood ratio test or if they alter the coefficient of the intervention effect >10%, when removed from the analysis. All models will be performed with the SAS MIXED procedure. Model assumptions will be verified according to residual analysis. If most of the assumptions are not met, other alternatives such as transformation for DQ score or generalised linear mixed model will be examined. For the secondary outcomes, change in other scores from baseline to 24 months after randomisation will be analysed in the same way as the primary outcome. All tests will be two sided and carried out at the 5% level of significance. Statistical analyses will be performed using SAS 9.3 (SAS Institute, Cary, North Carolina, USA) by the centre for clinical research of the teaching hospital of Lyon (Hospices Civils de Lyon, France). Organisation The Trial Steering Committee (TSC) will be responsible for overseeing the progress of the trial and will meet at regular intervals. The TSC includes the PI, the investigators of the centres and the trial coordinators. The TSC has developed the study protocol and is responsible for data collection, management, publications and the final data set. The committee is responsible for finding solutions to unforeseen questions/problems that may arise in the course of the study. According to French law, the study does not require a formal data monitoring committee as it is a trial with known minimal risks. Dissemination The study team will be committed to full disclosure of the results of the trial. The results of the study will be disseminated at several national and international meetings, and as articles published in national and international peer-reviewed journals. The study will be implemented and reported in line with the CONSORT statement. Each paper or abstract will be submitted to the appropriate subcommittee for review of its appropriateness and scientific merit prior to submission. The study team will adhere to defined authorship criteria as per the International Committee of Medical Journal Editors. Discussion The 12 hours per week ESDM intervention We expect that our strategy of 10 hours of ESDM at the intervention unit and 2 hours in the toddlers' natural environment (home, nursery or preschool) will encourage use of the ESDM by parents at home and in natural environments of daily life and thus reinforce the intervention. The study will be the first randomised multicentre controlled trial investigating the effectiveness of a behavioural and developmental intervention such as ESDM in a French-speaking European population. The present study will be a randomised clinical trial with a large sample size of very young children with ASD receiving a therapist-ESDM semi-intensive intervention during 2 years. The inclusion criteria are wide enough to take into account the heterogeneity of ASD at this age and will not exclude children with genetic disorders associated with ASD (with the exception of Rett syndrome) as in other studies.32 47 48 Moreover, this study will address needs of non-verbal toddlers for whom few studies are available.12 32 48–50 The two-stage Zelen design Contrary to the classic randomisation procedure for clinical trials, the Zelen design proposes that randomisation is performed before patients give their consent to participate. More specifically, in the two-stage Zelen design, all parents of eligible children consent to participate in a longitudinal observational study, and second, children are randomly allocated to the intervention group. Only the parents of these children are informed about the intervention, and give their written consent for it. This design presents some significant advantages. First, it should avoid a strong feeling of disappointment among parents, and reduce the risk of study withdrawal if their child is not randomised to the intervention group, as 12 hours per week ESDM is highly desired by families.43 Thus, it may promote participation and limit withdrawal of consent.42 Furthermore, the use of the Zelen design will minimise the bias typically encountered in traditionally designed randomised clinical trials where participant expectations may influence study outcomes, that is, parents' expectation for their child and their participation in the behavioural intervention. An effectiveness study As the aim of this study is to evaluate the 12 hours per week ESDM strategy, compared with the care available in the community within a broad, heterogeneous and representative population, we will conduct an effectiveness study.51 The control group of this study will receive heterogeneous interventions available in the community with different intensity and practices.8 In the intervention group, we will notify to what extent the participants are compliant with the ESDM intervention. The outcome measures Because the needs of children with ASD are complex, a diversity of outcome measurement tools is used to collect evidence about the child's progress.52 Their cognitive abilities play a central role in the manifestation of core and associated symptoms in ASD.23 The MSEL is commonly used as a measure of cognitive and language skills in ASD. The MSEL covers the range of age and skills differences (eg, verbal or non-verbal children). It has been used in previous studies and allows comparison with other papers.14 The authors would like to thank Sally Rogers, Elizabeth Fulton, Cynthia Zermut and Marie Rocha for their training in ESDM. They are grateful to Catherine Lord and Rebecca Grzadzinski for use of the BOSCC. Particular acknowledgements are addressed to Jonathan Green for his expert advice and to Catherine Aldred for her training in DCMA methodology. They thank all teams of the different study sites for their ongoing support of this project. Thanks also to Paul Belhouchat, a member of the Centre for Autism Resources of Lyon (Rhône-Alpes) who provided references. The authors are grateful to Philip Robinson for his help with manuscript preparation and Isabelle Bluche, Translator-Editor, Translands Traductions, for editing this paper Collaborators: Lyon—CH du Vinatier: MMG, NG, Stéphane Grisi, Stéphanie Marignier, Jean-Marc Useo, Alain Pourrat, Chloé Peter, Flavia Mengarelli, Natacha Gallifet, Fanny Rumillat, Gaelle Tenant; Lyon—CH Saint Jean de Dieu: SM, Sandrine Sonié, Marc Zimmerman, Valérie Malo, Céline Grosmaître Jacob; Versailles—Paris: Stéphane Bahrami, BF, MJO, MS, Camille Queste; Poitiers: LG, Florence Raffeneau, Sophie Gilet, Mélanie Ammeloot; Strasbourg: Anne Danion-Grilliat, Emilie Florence, CS, Valérie Vecchionacci, Lucie Janssen, Anna Dubrovskaya, Monique Rohmer, Bénédicte Lambs; Bruxelles: Véronique Delvenne, RS, Sophie Carlier, Lesley Ducenne; Genève: MS; Lyon—HCL: Cyrille Colin, AD, PO, ST, AZ, Laeticia Bouveret. Contributors: ST, MMG, SB and PO conceived and designed the project, and MM-G is leading the coordinator of the trial. ST, MMG and PO drafted the protocol and procured the project funding. MMG is responsible for study implementation, staff training and supervision. SB and MMG designed and are leading the economic evaluation. ST and AD contributed to the sample size calculation, the randomisation procedure and the statistical plan, and are responsible for data management, randomisation and statistical analysis. BF gave advice to the randomisation procedure. MMG, SM, CS, MSp, RS, NG, MJO and MSc are responsible for recruitment and evaluation of children. All authors critically reviewed and approved the final version of the manuscript. Funding: This study is supported by a grant from the Programme de Recherche sur la Performance du Système de soins (PREPS 14-0533) from the French Ministry of Health (Ministère chargé de la Santé, Direction de l'Hospitalisation et de l'Organisation des Soins) and a grant from the Fondation de France, a member of the Network of European Foundations for Innovative Cooperation. Disclaimer: The funders and sponsor have no role in study design, data collection, management, data analysis and interpretation of data, in the writing of the report or in the decision to submit the manuscript for publication. Competing interests: None declared. Patient consent: Obtained. Ethics approval: The ethical issues raised by Zelen design were discussed by the Ethics Committee of the university teaching hospital of Lyon; the study appeared to protect the children and the Zelen design was an ethically acceptable methodological alternative for this behavioural trial, and was therefore given approval. Ethics approval was obtained from the French Institutional Review Board (reference No 2015-013 B—“Comité de Protection de Personnes”) and the French Data Protection Agency (reference DR-2015-319—“Commission Nationale Informatique et Liberté”). Ethics approval was also obtained from the Belgium Institutional Review Board (reference P2016/554, reference CCB B406201630678—“Comité d'Ethique Erasme-ULB”). Amendments will be reported to relevant regulatory parties. Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: The trial statisticians will have access to the data set for the analysis of trial outcomes. The PI will have access to the data and will take full responsibility for the analysis and publication of the results. Once the main analyses have been undertaken, data will be available to principal and other investigators subject to approval of data analysis plans by the steering committee. ==== Refs References 1 American Psychiatric Association . Diagnostic and statistical manual of mental disorders . 5th edn Arlington, VA : American Psychiatric Publishing , 2013 . 2 Charman T , Pickles A , Simonoff E IQ in children with autism spectrum disorders: data from the Special Needs and Autism Project (SNAP) . Psychol Med 2011 ;41 :619 –27 . 10.1017/S0033291710000991 21272389 3 Yeargin-Allsopp M The prevalence and characteristics of autism spectrum disorders in the ALSPAC cohort . Dev Med Child Neurol 2008 ;50 :646 10.1111/j.1469-8749.2008.03059.x 18754911 4 Fombonne E The prevalence of autism . JAMA 2003 ;289 :87 –9 . 10.1001/jama.289.1.87 12503982 5 Howlin P Living with impairment: the effects on children of having an autistic sibling . Child Care Health Dev 1988 ;14 :395 –408 . 10.1111/j.1365-2214.1988.tb00591.x 3067904 6 Hayes SA , Watson SL The impact of parenting stress: a meta-analysis of studies comparing the experience of parenting stress in parents of children with and without autism spectrum disorder . J Autism Dev Disord 2013 ;43 :629 –42 . 10.1007/s10803-012-1604-y 22790429 7 Ganz ML The lifetime distribution of the incremental societal costs of autism . Arch Pediatr Adolesc Med 2007 ;161 :343 –9 . 10.1001/archpedi.161.4.343 17404130 8 Salomone E , Beranová Š , Bonnet-Brilhault F Use of early intervention for young children with autism spectrum disorder across Europe . Autism 2016 ;20 :233 –49 . 10.1177/1362361315577218 25916866 9 Evans B How autism became autism: The radical transformation of a central concept of child development in Britain . Hist Human Sci 2013 ;26 :3 –31 . 10.1177/0952695113484320 10 Schopler E , Mesibov GB , Hearsey K Structured teaching in the TEACCH system,1995 . In: Schopler E , Mesibov GB eds. Learning and cognition in autism . New York : Kluwer Academic/Plenum , 243– 267 . 11 Warren Z , McPheeters ML , Sathe N A systematic review of early intensive intervention for autism spectrum disorders . Pediatrics 2011 ;127 :e1303 –11 . 10.1542/peds.2011-0426 21464190 12 Maglione MA , Gans D , Das L Technical expert panel . Nonmedical interventions for children with ASD: recommended guidelines and further research needs . Pediatrics 2012 ;130 (Suppl 2 ):S169 –78 . 10.1542/peds.2012-0900O 23118248 13 Rogers S , Dawson G Early Start Denver Model for young children with Autism: promoting language, learning, and engagement . Traduit par Rogé. Edition Dunod , Paris : Guilford Press , 2013 . 14 Dawson G , Rogers S , Munson J Randomized, controlled trial of an intervention for toddlers with autism: the Early Start Denver Model . Pediatrics 2010 ;125 :e17 –23 . 10.1542/peds.2009-0958 19948568 15 Fava L , Strauss K Response to early intensive behavioral intervention for autism--an umbrella approach to issues critical to treatment individualization . Int J Dev Neurosci 2014 ;39 :49 –58 . 10.1016/j.ijdevneu.2014.05.004 24866707 16 Torgerson DJ , Roland M What is Zelen's design? . BMJ 1998 ;316 :606 10.1136/bmj.316.7131.606 9518917 17 Campbell R , Peters T , Grant C Adapting the randomized consent (Zelen) design for trials of behavioural interventions for chronic disease: feasibility study . J Health Serv Res Policy 2005 ;10 :220 –5 . 10.1258/135581905774414150 16259688 18 Lord C , Rutter M , DiLavore PC Autism diagnostic observation schedule . Los Angeles, CA : Western Psychological Services , 2012 . 19 Lord C , Rutter M , Le Couteur A Autism diagnostic interview-revised: a revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders . J Autism Dev Disord 1994 ;24 :659 –85 . 10.1007/BF02172145 7814313 20 Kim SH , Thurm A , Shumway S Multisite study of new autism diagnostic interview-revised (ADI-R) algorithms for toddlers and young preschoolers . J Autism Dev Disord 2013 ;43 :1527 –38 . 10.1007/s10803-012-1696-4 23114567 21 McConachie H , Fletcher-Watson S , Working group 4, COST action ‘enhancing the scientific study of early autism’ . Building capacity for rigorous controlled trials in autism: the importance of measuring treatment adherence . Child Care Health Dev 2015 ;41 :169 –77 . 10.1111/cch.12185 25163398 22 Mullen EM Mullen scales of early learning. AGS Edition. Circle Pines, MN : American Guidance Service , 1995 . 23 Bishop SL , Guthrie W , Coffing M Convergent validity of the mullen scales of early learning and the differential ability scales in children with autism spectrum disorders . Am J Intellect Dev Disabil 2011 ;116 :331 –43 . 10.1352/1944-7558-116.5.331 21905802 24 Swineford LB , Guthrie W , Thurm A Convergent and divergent validity of the Mullen Scales of early learning in young children with and without autism spectrum disorder . Psychol Assess 2015 ;27 :1364 –78 . 10.1037/pas0000116 25894712 25 Lord C , Risi S , DiLavore PS Autism from 2 to 9 years of age . Arch Gen Psychiatry 2006 ;63 :694 –701 . 10.1001/archpsyc.63.6.694 16754843 26 Pickles A , Anderson DK , Lord C Heterogeneity and plasticity in the development of language: a 17-year follow-up of children referred early for possible autism . J Child Psychol Psychiatry 2014 ;55 :1354 –62 . 10.1111/jcpp.12269 24889883 27 Rogers SJ , Estes A , Lord C Effects of a brief Early Start Denver model (ESDM)-based parent intervention on toddlers at risk for autism spectrum disorders: a randomized controlled trial . J Am Acad Child Adolesc Psychiatry 2012 ;51 :1052 –65 . 10.1016/j.jaac.2012.08.003 23021480 28 Gotham K , Risi S , Pickles A The Autism Diagnostic Observation Schedule: revised algorithms for improved diagnostic validity . J Autism Dev Disord 2007 ;37 :613 –27 . 10.1007/s10803-006-0280-1 17180459 29 Grzadzinski R , Carr T , Colombi C Measuring changes in social communication behaviors: preliminary development of the Brief Observation of Social Communication Change (BOSCC) . J Autism Dev Disord 2016 ;46 :2464 –79 . 10.1007/s10803-016-2782-9 27062034 30 Sparrow S , Cicchetti DV , Balla DA Vineland adaptive behavior scales . 2nd edn Vineland™-II : Pearson , 2005 . 31 Wetherby AM , Allen L , Cleary J Validity and reliability of the communication and symbolic behavior scales developmental profile with very young children . J Speech Lang Hear Res 2002 ;45 :1202 –18 . 10.1044/1092-4388(2002/097) 12546488 32 Green J , Charman T , McConachie H Parent-mediated communication-focused treatment in children with autism (PACT): a randomised controlled trial . Lancet 2010 ;375 :2152 –60 . 10.1016/S0140-6736(10)60587-9 20494434 33 Aldred C , Green J , Emsley R Brief report: mediation of treatment effect in a communication intervention for pre-school children with autism . J Autism Dev Disord 2012 ;42 :447 –54 . 10.1007/s10803-011-1248-3 21512834 34 Pickles A , Harris V , Green J Treatment mechanism in the MRC preschool autism communication trial: implications for study design and parent-focussed therapy for children . J Child Psychol Psychiatry 2015 ;56 :162 –70 . 10.1111/jcpp.12291 25039961 35 Bassano D , Labrell F , Champaud C Le DLPF : Un nouvel outil pour l’évaluation du développement du langage de production en français. Enfance, n°2. Presses universitaires de France 2005 http://www.umr7023.cnrs.fr/o-Le-DLPF-un-nouvel-outil-pour-l.html 36 Fenson L , Dale PS , Reznick JS MacArthur communicative development inventory: users guide and technical manual . San Diego, CA : Singular , 1993 . 37 Dunn W Sensory profile 2 user's manual . San Antonio, TX : Psychological Corporation , 2014 . 38 Hoefman R , Payakachat N , van Exel J Caring for a child with autism spectrum disorder and parents’ quality of life: application of the CarerQol . J Autism Dev Disord 2014 ;44 :1933 –45 .24577786 39 Byford S , Cary M , Barrett B Cost-effectiveness analysis of a communication-focused therapy for pre-school children with autism: results from a randomised controlled trial . BMC Psychiatry 2015 ;21 :316 10.1186/s12888-015-0700-x 40 Glick HA , Doshi JA , Sonnad SS Economic evaluation in clinical trials . 2nd edn Oxford : University Press , 2015 . 41 Schellings R , Kessels AG , ter Riet G The Zelen design May be the best choice for a heroin-provision experiment . J Clin Epidemiol 1999 ;52 :503 –7 . 10.1016/S0895-4356(99)00009-8 10408988 42 Homer CSE Using the Zelen design in randomized controlled trials: debates and controversies . J Adv Nurs 2002 ;38 :200 –7 . 10.1046/j.1365-2648.2002.02164.x 11940133 43 Adamson J , Cockayne S , Puffer S Review of randomised trials using the post-randomised consent (Zelen's) design . Contemp Clin Trials 2006 ;27 :305 –19 . 10.1016/j.cct.2005.11.003 16455306 44 Eapen V , Crnčec R , Walter A Clinical outcomes of an early intervention program for preschool children with autism spectrum disorder in a community group setting . BMC Pediatr 2013 ;13 :3 10.1186/1471-2431-13-3 23294523 45 Zelen M Randomized consent designs for clinical trials: an update . Stat Med 1990 ;9 :645 –56 . 10.1002/sim.4780090611 2218168 46 Brown H , Prescott R Applied mixed models in medicine . 2nd edn Oxford : Wiley , 2006 . 47 Bearss K , Johnson C , Smith T Effect of parent training vs parent education on behavioral problems in children with autism spectrum disorder: a randomized clinical trial . JAMA 2015 ;313 :1524 –33 . 10.1001/jama.2015.3150 25898050 48 Pickles A , Le Couteur A , Leadbitter K Parent-mediated social communication therapy for young children with autism (PACT): long-term follow-up of a randomised controlled trial . Lancet 2016 ;388 :2501 –9 . 10.1016/S0140-6736(16)31229-6 27793431 49 Kasari C , Lawton K , Shih W Caregiver-mediated intervention for low-resourced preschoolers with autism: an RCT . Pediatrics 2014 ;134 :e72 –9 . 10.1542/peds.2013-3229 24958585 50 Wetherby AM , Guthrie W , Woods J Parent-implemented social intervention for toddlers with autism: an RCT . Pediatrics 2014 ;134 :1084 –93 . 10.1542/peds.2014-0757 25367544 51 Glasgow RE , Lichtenstein E , Marcus AC Why don't we see more translation of health promotion research to practice? Rethinking the efficacy-to-effectiveness transition . Am J Public Health 2003 ;93 :1261 –7 . 10.2105/AJPH.93.8.1261 12893608 52 McConachie H , Parr JR , Glod M Systematic review of tools to measure outcomes for young children with autism spectrum disorder . Health Technol Assess 2015 ;19 :1 –506 . 10.3310/hta19410

PMC005xxxxxx/PMC5372148.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01350810.1136/bmjopen-2016-013508Rehabilitation MedicineProtocol150617271737Single-stage osseointegrated reconstruction and rehabilitation of lower limb amputees: the Osseointegration Group of Australia Accelerated Protocol-2 (OGAAP-2) for a prospective cohort study Al Muderis Munjed 123Lu William 4Tetsworth Kevin 56Bosley Belinda 3Li Jiao Jiao 4 1 The Australian School of Advanced Medicine, Macquarie University, North Ryde, New South Wales, Australia 2 School of Medicine, University of Notre Dame, Auburn, New South Wales, Australia 3 Norwest Private Hospital, Bella Vista, New South Wales, Australia 4 Biomaterials and Tissue Engineering Research Unit, School of AMME, University of Sydney, Sydney, New South Wales, Australia 5 Royal Brisbane Hospital, Herston, Queensland, Australia 6 School of Medicine, University of Queensland, Brisbane, Queensland, AustraliaCorrespondence to Dr Jiao Jiao Li; jiaojiao.li@sydney.edu.au2017 22 3 2017 7 3 e01350817 7 2016 17 2 2017 20 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Introduction Lower limb amputations have detrimental influences on the quality of life, function and body image of the affected patients. Following amputation, prolonged rehabilitation is required for patients to be fitted with traditional socket prostheses, and many patients experience symptomatic socket–residuum interface problems which lead to reduced prosthetic use and quality of life. Osseointegration has recently emerged as a novel approach for the reconstruction of amputated limbs, which overcomes many of the socket-related problems by directly attaching the prosthesis to the skeletal residuum. To date, the vast majority of osseointegration procedures worldwide have been performed in 2 stages, which require at least 4 months and up to 18 months for the completion of reconstruction and rehabilitation from the time of the initial surgery. The current prospective cohort study evaluates the safety and efficacy of a single-stage osseointegration procedure performed under the Osseointegration Group of Australia Accelerated Protocol-2 (OGAAP-2), which dramatically reduces the time of recovery to ∼3–6 weeks. Methods and analysis The inclusion criteria for osseointegrated reconstruction under the OGAAP-2 procedure are age over 18 years, unilateral transfemoral amputation and experiencing problems or difficulties in using socket prostheses. All patients receive osseointegrated implants which are press-fitted into the residual bone. Functional and quality-of-life outcome measures are recorded preoperatively and at defined postoperative follow-up intervals up to 2 years. Postoperative adverse events are also recorded. The preoperative and postoperative values are compared for each outcome measure, and the benefits and harms of the single-stage OGAAP-2 procedure will be compared with the results obtained using a previously employed 2-stage procedure. Ethics and dissemination This study has received ethics approval from the University of Notre Dame, Sydney, Australia (014153S). The study outcomes will be disseminated by publications in peer-reviewed academic journals and presentations at relevant clinical and orthopaedic conferences. OsseointegrationLower limb amputeesSingle-stage surgery ==== Body Strengths and limitations of this study This study is the first study to describe, as well as report on the safety and efficacy of, a single-stage procedure for the osseointegrated reconstruction of amputated limbs. This study may therefore have significant influence on the standard of treatment for patients with lower limb amputations undergoing osseointegration surgery, and reverse the concept that a two-stage procedure is required. This study has a relatively large sample size of 105 patients, which resembles one of the largest patient cohorts among studies published to date reporting on the outcomes of osseointegrated reconstruction of amputated limbs. This study does not directly compare the outcomes of using osseointegrated prostheses to the outcomes of using socket prostheses as the traditional method of treating patients with lower limb amputations. This study has a relatively short follow-up period of 2 years, which does not allow the examination of longer term outcomes and risk of adverse events. Introduction Lower limb amputations almost inevitably result in major changes to the quality of life, function and body image of the affected patients,1–4 and are still associated with considerable morbidity and mortality.5 For patients with bilateral above-knee amputations, over 90% are eventually confined to a wheelchair due to the difficulty of mobilising with prostheses attached to both lower limbs.6 The effect of amputation is greater in younger and more active patients who must adapt to a drastic reduction in mobility and the associated psychological consequences,4 which also often restricts their ability to seek or continue employment.7 The return-to-work rate for lower limb amputees is between 22% and 67%, with many having to change their occupation or work only part-time.8 Data pooled from study populations in the UK, Europe, USA and parts of the Asia-Pacific indicate that the global incidence of all forms of lower extremity amputation ranges from 5.8 to 31 per 105 in the total population.9 Following amputation, prolonged rehabilitation is often required for patients to be fitted with traditional suspended socket prostheses. The fitting generally becomes more difficult with amputation at higher levels, and often results in characteristic symptoms of local pain, skin ulceration and discomfort.2 10 11 In addition, patients with a short stump, skin grafts, scarring or heterotopic bone formation may be entirely unable to use socket prostheses, or choose not to use them due to the problems associated with socket fitting. A survey of 97 patients with transfemoral amputation in Sweden reported a very high prevalence of problems related to the use of socket prostheses, including: 72% with symptomatic heat and sweating of the stump, 62% with sores or skin irritation from the socket, 61% with interference to mobility and 51% with pain in the stump when standing or walking.2 In addition, these patients consistently reported a significantly diminished quality of life when compared with able-bodied participants. The typical socket prosthesis used for transfemoral amputations has also been shown to hinder the range of motion of the involved hip, which contributes further to difficulties in ambulation.12 The problems associated with socket prostheses have remained largely unsolved, despite extensive and continuing research into socket design and manufacturing.13 Consequently, at least one-third of all amputees still experience symptomatic socket–residuum interface problems, leading to reduced prosthetic use and markedly reduced quality of life.10 14 Osseointegration has emerged over the past two decades as a dramatically different approach for the reconstruction of limb loss due to amputation, which can overcome many of the issues associated with traditional socket prostheses.15–17 The surgical procedure involves direct attachment of the prosthesis to the skeletal residuum, in a process analogous to that used in uncemented hip arthroplasty. A titanium intramedullary implant is press-fitted into the residual bone of the amputated limb, which becomes rapidly incorporated into the bone over the few months following surgery. This implant is continuous with an abutment that penetrates the skin through a small permanent opening, which is used for attachment of the prosthetic limb. Osseointegrated prostheses can therefore completely eliminate the problematic socket–residuum interface due to the intimate structural and functional connection between the intramedullary implant and host bone. The Food and Drug Administration (FDA) recently approved osseointegration surgery for amputees in the USA, although for humanitarian indications only.18 Nevertheless, this technology has been employed internationally for over 20 years and increasingly in Europe, the UK and Australia over the past decade. Osseointegration surgery for the reconstruction of amputated limbs has mainly been performed using two types of implants. Cannulated screw-fixation implants, originally developed in Sweden, achieve skeletal integration as a result of bone on-growth. Treatment requires two operations, where an initial procedure is performed to insert the intramedullary implant, and a second procedure is performed 6–9 months later to create the percutaneous skin opening for allowing abutment attachment and prosthesis fitting.19 Alternatively, press-fit macroporous surface structure implants allow skeletal integration by bone penetration and ingrowth, which also require a two-stage operation with a 6–8-week interval between stages.20 Several prospective case series describing these two techniques have been published, which reported on the major clinical benefits of the osseointegration approach, including improvements in the patients' quality of life,21 prosthetic use,21 22 body image,22 hip range of motion,23 sitting comfort,12 donning and doffing,21 osseoperception24 and walking ability,16 25 as well as very acceptable levels of risk with respect to implant stability26 and infection rates.17 27 Recently, a comprehensive surgical and rehabilitation protocol (OGAAP-1) has been developed for procedures involving osseointegrated implants which are currently used in Australia and the Netherlands, with press-fit designs that encourage bone ingrowth.28 This protocol emphasises an integrated approach for the management of patients with lower limb amputations, and revolves around two-stage osseointegration surgery with a 4–6-week interval between stages. Preliminary results have demonstrated significant improvements in quality of life and functional outcome measures,28 while maintaining very reasonable complication rates.29 Until very recently, the vast majority of osseointegration procedures worldwide have been performed in two stages. From the time of the initial surgery, these procedures require up to 12–18 months for the completion of reconstruction and rehabilitation with screw-fixation implants, and at least 4–5 months even under the accelerated OGAAP-1 protocol.28 Since April 2014, however, a single-stage osseointegration procedure has been routinely performed by the Osseointegration Group of Australia (OGA) under the Osseointegration Group of Australia Accelerated Protocol-2 (OGAAP-2). This protocol reduces the overall time required for the definitive osseointegrated reconstruction and rehabilitation of lower limb amputees to ∼3–6 weeks, which is substantially shorter compared with any currently available two-stage procedure. This study aims to assess the safety and efficacy of the single-stage OGAAP-2 procedure in a sufficiently large group of patients. The main hypothesis is that the single-stage OGAAP-2 procedure will result in more rapid progression to unrestricted full weight-bearing with an osseointegrated reconstruction, within an even shorter time frame compared with the two-stage OGAAP-1 procedure, without an increased risk of associated adverse events. Study objectives The overall objective of this study is to assess the safety and efficacy of the single-stage OGAAP-2 procedure for the osseointegrated reconstruction of lower limb amputations, and compare the benefits and harms with the outcomes obtained using the OGAAP-1 procedure within 2 years of follow-up. Specifically, this involves: Assessing the objective functional outcomes after osseointegrated reconstruction of amputated limbs using the single-stage OGAAP-2 procedure, with the 6 Minute Walk Test (6MWT), Timed Up and Go (TUG) and K-levels, compared with preoperative data and also data obtained using the previously employed two-stage OGAAP-1 procedure. Assessing the subjective patient-reported quality-of-life outcomes after osseointegrated reconstruction of amputated limbs using the single-stage OGAAP-2 procedure, with the Questionnaire for persons with a Trans-Femoral Amputation (Q-TFA) and the Short Form Health Survey 36 (SF-36), compared with preoperative data and also data obtained using the previously employed two-stage OGAAP-1 procedure. Examining the prevalence of adverse events after osseointegrated reconstruction of amputated limbs using the single-stage OGAAP-2 procedure, including infection, revision surgery, fractures and implant failures, compared with data obtained using the previously employed two-stage OGAAP-1 procedure. Methods and analysis Overview of study design The current prospective cohort study is designed to assess the safety and efficacy of the single-stage OGAAP-2 procedure for the osseointegrated reconstruction of lower limb amputations within 2 years of follow-up. The OGAAP-2 procedure is a comprehensive programme for the management of patients with lower limb amputation using single-stage osseointegration surgery (figure 1). This procedure is exclusively performed by the OGA in Australia, and has been routinely performed on eligible patients since March 2014. Patients are evaluated by validated outcome measures preoperatively and postoperatively. Preliminary data have been obtained from an initial pilot study with 10 patients, which have been used to provide the sample size estimate for the current study. The outcomes of this study will be compared with those obtained using the previously employed OGAAP-1 procedure at the same follow-up time points. Figure 1 Overview of the single-stage Osseointegration Group of Australia Accelerated Protocol-2 clinical programme and outcomes evaluation for the osseointegrated reconstruction of lower limb amputations. Patient selection Eligibility criteria The inclusion criteria are age over 18 years, unilateral transfemoral amputation and experiencing socket-related problems or difficulties in using socket prostheses (including patients who are ambulatory with assistive devices or non-ambulatory with short stumps and non-reconstructable limb pathology). The exclusion criteria are smoking, psychological instability, pregnancy, limb exposure to radiation, ongoing chemotherapy, immunosuppression, diabetes, peripheral vascular disease and non-compliance during preoperative screening and evaluation. Patient screening and recruitment Prospective patients are referred by rehabilitation physicians, specialists, general practitioners (GPs) or prosthetists, or arise from direct enquiries. All prospective patients are asked to complete a secure online enquiry form found on the OGA website. If the patient has consented to communication, team contact occurs in the form of a phone call to confirm the patient's medical and social demographic history, prosthetic history, general expectations and compliance, as well as to discuss the exclusion criteria. Pain and psychological questionnaires are then provided to the patient, and used as tools for assessing the patient's psychological well-being and identifying any chronic pain issues. Patients who satisfy the inclusion criteria are invited to attend a specialised osseointegration clinic, during which preoperative clinical, radiological and psychological examinations are performed, and baseline values of outcome measures are recorded. The clinic also gives an opportunity for prospective patients to receive specialist team counselling, as well as to engage in peer-to-peer interaction regarding the surgical procedure and recovery. After an individual consultation with the specialist team, suitable patients are enrolled for osseointegration surgery using the OGAAP-2 procedure, and informed consent for data collection is obtained. The first definitive patient undergoing osseointegrated reconstruction using the OGAAP-2 procedure was enrolled on 9 April 2014. Enrolment is ongoing at the time of publication and is expected to be complete by December 2016. Study intervention Preoperative management Surgical planning is conducted by using the patient's radiographs, CT scans and dual-energy X-ray absorptiometry (DXA) scans to determine the anatomy of the skeletal residuum, as well as to allow patient-specific selection of the implant type and size, and external prosthetic components. A preoperative physical training programme is recommended for all patients. This consists of muscle strengthening and visualisation, and core strengthening exercises (including abdominals and upper body) for wheelchair-bound patients, and pregait training aimed at increasing the range of movement (with particular emphasis on the residuum hip flexors and adductors) for prosthetic users. Osseointegration surgery The surgical procedure involves permanent insertion of the Osseointegrated Prosthetic Limb (OPL; Permedica s.p.a; Milan, Italy) (figure 2). This osseointegration implant system is specifically designed for press-fit fixation and consists of two components. The intramedullary stem component has proximal splines to facilitate initial rotational stability, and a distal flare to provide initial axial stability. It also has a macroporous surface structure resembling cancellous bone to facilitate bone penetration. These design features provide rigid initial stability, as well as permanent implant anchorage as a result of bone ingrowth. The dual-cone adaptor component connects the intramedullary stem to the external prosthesis and has a highly polished surface coated with titanium-niobium dioxide to minimise soft tissue friction and adhesions. The proximal aspect of the dual-cone adaptor that connects to the intramedullary stem is provided with a safety pin, which fails under excessive torsional forces to prevent periprosthetic fracture or implant breakage. Figure 2 The Osseointegrated Prosthetic Limb implant system (Permedica s.p.a; Milan, Italy) used for the osseointegrated reconstruction of trans-femoral amputations under the Osseointegration Group of Australia Accelerated Protocol-2 procedure. The implant consists of an intramedullary stem component and a dual-cone adaptor component. The procedure is performed in a single stage under the OGAAP-2 protocol (figure 3). Prophylactic intravenous antibiotics using 2 g of cephazolin is administered prior to the procedure, in accordance with standard arthroplasty antibiotic prophylaxis protocols.30 The first part of the operation involves soft tissue preparation and implantation of the intramedullary stem, which comprises a major portion of the surgery. This includes guillotine amputation of the stump without shortening of the bony residuum, and reorganisation of the residual muscle groups around the bone end using absorbable purse string sutures. Haemostasis is achieved, and neuromas are identified and terminalised by shortening the nerve endings and embedding them into the surrounding fatty tissue to avoid adhesions or tethering to muscles. Additional soft tissue optimisation is then performed with refashioning of the stump and excision of excess subcutaneous fat to achieve a minimised soft tissue envelope. The final stage of soft tissue preparation involves identifying the area of soft tissue opposing the distal end of the bony residuum, and removing the fat beneath a circular portion of skin in this area. The dermis is then sutured around the periosteum, and the wound is closed using metal staples with minimal internal sutures. Under fluoroscopy, a k-wire is inserted through the de-fatted skin at the tip of the bony residuum, followed by application of a circular skin corer to create a percutaneous opening. Figure 3 Steps involved in single-stage osseointegration surgery under the Osseointegration Group of Australia Accelerated Protocol-2. (A) Identification and terminalisation of neuromas during soft tissue preparation. (B) Completed guillotine amputation of the stump. (C) Reaming of the medullary canal and preparation of the distal femur to accommodate the flange of the intramedullary stem component of the osseointegration implant. (D) Insertion of the intramedullary stem by impacting with a mallet. (E) Creation of the stoma site and insertion of the dual-cone adaptor component of the osseointegration implant. (F) Stoma site postoperation, showing the inserted dual-cone adaptor and remaining parts of the abutment. After haemostasis is achieved, the medullary canal is prepared under fluoroscopic guidance using sequential flexible reamers followed by implant-specific broaches. Broaching is undertaken to the same size as the planned implant size. However, in the presence of osteoporosis as indicated by examination of the DXA scan, broaching is undertaken to one size under that of the planned implant size. The distal portion of the femur is prepared using specific rasps to match the distal flange of the implant. The intramedullary stem is then inserted into the medullary canal of the residual femur by impacting with a mallet to achieve a mechanically stable press-fit fixation. The second part of the operation involves attachment of the transcutaneous dual-cone adaptor to the intramedullary stem. The appropriate size is selected by inserting the adaptor sizing guide through the skin opening. The selected dual-cone adaptor is placed into the Morse taper sleeve on the distal end of the intramedullary stem and locked by impaction, then further secured with an internal locking screw. The distal taper sleeve, bushing and distal abutment screw are then attached to the distal aspect of the dual-cone adaptor. The bushing contains a two-pin external fail-safe mechanism, which limits the risk of periprosthetic fracture. Once completed, local anaesthetic is infiltrated proximal to the terminalised nerves, as well as surrounding the stoma and incision site. Postoperative management Pain management involves the administration of intravenous and epidural pain medications for the first 3 days following surgery, and oral analgesics thereafter. Wound care involves daily dressing changes with dry ribbon gauze and application of antiseptic solution for 10 days postoperation. Patients are then advised to wash the implant–skin interface two times per day with warm tap water and soap, and to pat dry the skin opening with disposable paper towels. Patients are generally discharged 5–7 days after surgery. Staples are typically removed after 3–4 weeks. Rehabilitation Postoperative rehabilitation proceeds in three phases. Phase I is initiated 3 days after surgery and consists of the patient applying a static axial load of 20 kg for 20 min two times per day. The load is increased by 5 kg/day until either 50 kg or 50% of the patient's body weight is reached. Phase II is initiated when the patient has reached the recommended axial loading level. The patient is fitted with a rehabilitation prosthesis, which includes a light leg with a stable locked knee. The patient continues to perform core strengthening and balance exercises, and gait exercises aided by parallel bars. Phase III is initiated when the patient is safely mobilising using the rehabilitation prosthesis. The patient is then fitted with the definitive prosthesis and proceeds to daily weight-bearing. Postoperative rehabilitation should be concluded within 3–6 weeks following osseointegration surgery. Ongoing physiotherapy After rehabilitation, the patient performs daily weight-bearing using the definitive prosthesis, first on two crutches for 6 weeks, then on a single crutch in the opposite hand for another 6 weeks and unaided thereafter. Meanwhile, further gait training is recommended that focuses on fall prevention and management, balance, walking and ascending and descending slopes. Study outcomes and follow-up Study outcomes The outcome measures of this study include: (1) objective functional outcomes measured using the 6MWT, TUG and K-levels, (2) subjective patient-reported quality-of-life outcomes measured using the Q-TFA and SF-36 and (3) prevalence of adverse events including infection, revision surgery, fractures and implant failures after osseointegrated reconstruction of amputated limbs using the single-stage OGAAP-2 procedure. A description of each of these validated outcome measures is provided below. In addition, perioperative parameters including total surgical time (hours), intraoperative bleeding (significant (>300 cc) or not significant), postoperative pain (numerical rating scale), total time in hospital (days), time to definitive prosthetic fitting (days) and time to full weight-bearing (days) will be recorded and compared with results obtained using the previously employed OGAAP-1 procedure. The 6MWT involves documenting, in metres, the maximum distance covered by the patient in 6 min by continuously walking along a 25 m long level enclosed corridor.31 The TUG involves documenting the time, in seconds, required for the patient to rise from a standard armchair, walk to a line on the floor 3 m away, turn, walk back to the chair and sit down again.32 K-levels are defined by the US Medicare system to categorise the ability of lower limb amputees to ambulate and navigate the environment, and uses a five-level functional classification system (K0–K4, where K0 is defined as inability to ambulate and a prosthesis does not enhance quality of life or mobility, and K4 is defined as ability for prosthetic ambulation that exceeds basic ambulation skills as exemplified by a child, active adult or athlete).33 The results of these functional outcome measures obtained after osseointegrated reconstruction using the OGAAP-2 procedure will be compared with preoperative baseline values, as well as with the results obtained using the previously employed OGAAP-1 procedure at the same follow-up time points. The Q-TFA is a self-report measure developed for transfemoral amputees using a socket or osseointegrated prosthesis to reflect prosthetic use, prosthetic mobility, problems and global health, each in a separate score (0–100).34 The SF-36 is a short-form health survey designed for use in clinical practice and research, and includes one multi item scale that assesses the eight health concepts of physical functioning, social functioning, role limitations due to physical problems, bodily pain, general mental health, role limitations due to emotional problems, vitality and general health perceptions.35 The results of these quality-of-life outcome measures obtained after osseointegrated reconstruction using the OGAAP-2 procedure will be compared with preoperative baseline values, as well as with the results obtained using the previously employed OGAAP-1 procedure at the same follow-up time points. The monitoring and recording of adverse events involves identifying infections related to the osseointegrated implant on the basis of clinical and radiographic findings, and grading them into five levels of severity: 0, no infection; 1, low-grade soft tissue infection; 2, high-grade soft tissue infection; 3, deep bone infection or 4, septic implant failure.29 Management of infections is recorded as (A) oral antibiotics, (B) parenteral antibiotics, or (C) surgical debridement and/or explantation.29 Other adverse events are recorded and categorised into: stoma hypergranulation, redundant soft tissue, proximal femoral fracture, inadequate osseointegration with replacement of implant, breakage of intramedullary component and breakage of dual-cone component safety pin.29 The prevalence of adverse events after osseointegrated reconstruction using the OGAAP-2 procedure will be compared with the results obtained using the previously employed OGAAP-1 procedure at the same follow-up time points. Study follow-up All patients are followed for a minimum period of 2 years, with routine clinical monitoring conducted and outcomes assessed at set intervals of 6 weeks, 3, 6 and 12 months after surgery and annually thereafter. Radiographic outcomes are assessed at all follow-up intervals scheduled. 6MWT, TUG, Q-TFA and SF-36 outcomes are assessed at 6 and 12 months, and annually thereafter. DXA scans are performed at the 12-month follow-up, and annually thereafter. All adverse events are managed when identified as appropriate for the clinical conditions encountered, and recorded at the time of occurrence. Mechanisms of reporting adverse events include entries in the patient's diary booklet, reports from the patient's GP and any contact made directly to the clinical team. The minimum follow-up period of 2 years is chosen as our previous experience indicates that peak gains in function and quality of life as a result of osseointegrated reconstruction of amputated limbs are obtained within 12 months postoperation,28 and changes in related outcomes can be expected to plateau thereafter. Our previous experience also shows that 2 years of follow-up will provide sufficient time for the monitoring of any major adverse events (such as infection) following surgery.29 Given the short time period required for definitive osseointegrated reconstruction and rehabilitation under the OGAAP-2 procedure, the most significant changes in measures of benefits and harms are expected to be observed within 6 months of follow-up. Protecting against sources of bias Owing to the nature of this study with one treatment group (patients undergoing single-stage osseointegration surgery under the OGAAP-2) and no simultaneous control group (previously obtained data from patients undergoing osseointegration surgery under the OGAAP-1 will be used for comparison at the same follow-up time points), the assessors of outcome measures cannot be blinded. However, bias will be unlikely for the functional and quality-of-life outcome measures since these do not require the assessor's subjective judgement. The results of functional outcome measures (6MWT, TUG, K-levels) depend completely on the patients' physical performance, while the results of quality-of-life outcome measures (Q-TFA, SF-36) are derived completely from surveys of patient responses. To avoid bias, the assessors collecting these results will not be involved in the data analysis aspect of the study. For the reporting of adverse events, patients are advised to see their GP for suspected infection, and cases of infection are typically managed by the GP unless surgical intervention is required. The GPs are not part of the core clinical and research team conducting this study, and will therefore provide an unbiased diagnosis of any infection cases related to osseointegration surgery. Information on cases of infection is relayed back to the OGA team, and an infection grading is assigned only after the case has been resolved. The level of severity of the infection (0–4) is determined by a group of surgeons and infection specialists to minimise bias, while the method of management (A–C) is assigned depending on the type of intervention and does not require assessor judgement. Adverse events other than infection are reported as the number of cases occurring and are not subjected to bias. Bias relating to surgeon expertise and protocol adherence is eliminated since all operations will be performed by a single surgeon, who is the developer of the OGAAP-2 protocol and initiator of this study. Statistical plan and data analysis Sample size estimate Owing to the absence of previous literature on osseointegrated reconstruction using the OGAAP-2 procedure, data from an initial pilot study with 10 patients have been used to provide the most accurate sample size estimate possible. An a priori sample size calculation for analysis of variance (ANOVA) has been performed to determine the number of cases required to demonstrate a statistically significant (p=0.05) difference between preoperative and postoperative data for the functional and quality-of-life outcome measures. The sample size was calculated using the difference between the means for four of the main validated outcome measures (6MWT, TUG, Q-TFA and SF-36) based on the pilot study data (table 1), assuming that α=0.5 and β=0.10 with a power of 0.90 (1−β). Based on these parameters, the sample size calculations indicate that a minimum of 84 participants is necessary to provide a statistical power of 90%. To account for dropouts related to death and loss to follow-up (conservative estimate of 25%), a total of 105 patients will be recruited for the current study. Table 1 Sample size calculations comparing preoperative and postoperative pilot study data for the functional and quality-of-life outcome measures, obtained for osseointegrated reconstruction under the Osseointegration Group of Australia Accelerated Protocol-2 procedure Pilot study outcome measures Preoperative mean Postoperative mean SD Estimated N 6MWT 165 392 178 13 TUG 10.5 6.2 8.6 84 Q-TFA (global score) 45.0 71.3 21.6 15 SF-36 (physical component score) 39.3 46.4 12.4 64 Data analysis Continuous outcome variables will be summarised by calculating the mean and SE, and a Kolmorogov-Smirnov test will be used to determine the normality of the data. Categorical outcome variables will be presented as frequencies and percentages. ANOVA will be used to determine differences between preoperative and postoperative values for each continuous outcome measure. Fisher's exact test will be used to examine the significance of the contingency of preoperative and postoperative K-levels. The Bonferroni corrections will be performed to adjust for multiple comparisons. Multiple two-way contingency tables will be created, and a two-tailed Pearson χ² test with Yates corrections will be used for comparisons as indicated. All statistical analyses will be performed using Systat SPSS V.22.0 (IBM Corp., Armonk, New York, USA), where p<0.05 will be considered significant for all comparisons. Dissemination Ethical considerations All patients included in this study will sign a consent form that provides sufficient information about the study for patients to make an informed decision about their participation. All patient data and personal information will be deidentified and maintained on a password-enabled secure laptop computer. Dissemination Outcomes of the current study will be disseminated by publications in peer-reviewed academic journals and presentations at relevant clinical and orthopaedic conferences. To minimise delay, the preparation of study findings for dissemination will start as soon as data collection for the 2-year follow-up is complete for all enrolled patients. A manuscript containing the major findings of this study will be submitted for publication regardless of the presence or absence of significant outcomes. Discussion The findings of this study will make an important contribution to the small body of literature that is currently available regarding the reconstruction and rehabilitation of lower limb amputees using osseointegrated prostheses. The single-stage OGAAP-2 procedure enables the patient to achieve rapid progression to definitive osseointegrated reconstruction and full weight-bearing after surgery, within a substantially shorter period of time compared with any other currently available two-stage procedure. By allowing single-stage surgery, the new implant design used under the OGAAP-2 procedure eliminates the minimum 6-week delay between surgical stages that is required for two-stage procedures, and also facilitates an accelerated postoperative rehabilitation programme that results in fast progression to unrestricted and unassisted full weight-bearing. The OGAAP-2 procedure therefore dramatically reduces the overall treatment time to 3–6 weeks, which is comparable to that of uncemented total hip arthroplasty procedures.36 The concept of osseointegrated reconstruction originated in the dental implant field. Osseointegrated dental implants were introduced in the 1960s and were initially performed in two stages, where a 3–6-month interval between stages allowed uneventful wound healing prior to further intervention.37 Nevertheless, the immediate loading of dental implants is now routinely permitted, with multiple studies demonstrating the feasibility and predictability of the single-stage approach.38–40 A major research question answered by the current study is whether the same concept of single-stage operation for dental implants can be applied to the osseointegrated reconstruction of amputated lower limbs. If this study demonstrates that the single-stage OGAAP-2 procedure leads to significant improvements in postoperative functional and quality-of-life outcomes compared with preoperative levels without an increased prevalence of adverse events, it will potentially reverse the concept that a two-stage procedure is necessary to achieve the safe and effective osseointegrated reconstruction of amputated limbs. This may have a substantial impact in shaping a new standard of treatment for patients with lower limb amputations, which allows for rapid recovery and the earliest possible return to daily activities. A limitation of this study is that direct comparisons cannot be made with the outcomes of using socket prostheses, as the cohort of patients who are considered for osseointegration surgery are either experiencing problems with their socket prosthesis or are wheelchair-bound. Nevertheless, the current study compares preoperative and postoperative values of functional and quality-of-life outcome measures in patients undergoing osseointegrated reconstruction under the OGAAP-2 procedure, which gives an indication of the benefits of osseointegrated prostheses as a solution for patients who have previously been excluded from other types of prosthetic reconstruction. Another limitation is the relatively short follow-up period of 2 years. Although this time frame is considered sufficient for assessing the important benefits and harms of the OGAAP-2 procedure, it does preclude the examination of long-term outcomes and the long-term risk of adverse events. Larger prospective studies with longer follow-up times will be necessary to fully evaluate the long-term safety and efficacy of osseointegrated reconstruction of lower limb amputations under the OGAAP-2 procedure. The OGA team will continue to collect routine clinical data from the patients of this study after the 2-year follow-up, such that long-term retrospective analyses can be performed at a later date. Contributors: MAM is responsible for study design, patient care and surgical procedure, and manuscript preparation. WL is responsible for data collection and manuscript revision. KT is responsible for study design, statistical planning and manuscript revision. BB is responsible for patient care, data collection and manuscript revision. JJL is responsible for data collection and manuscript preparation and revision. Funding: This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Competing interests: MAM receives royalties for design contributions for the Osseointegrated Prosthetic Limb (OPL; Permedica s.p.a; Milan, Italy) implant system. Ethics approval: University of Notre Dame, Sydney, Australia. Provenance and peer review: Not commissioned; externally peer reviewed. ==== Refs References 1 Ephraim PL , MacKenzie EJ , Wegener ST Environmental barriers experienced by amputees: the craig hospital inventory of environmental factors-short form . Arch Phys Med Rehabil 2006 ;87 :328 –33 . 10.1016/j.apmr.2005.11.010 16500165 2 Hagberg K , Brånemark R Consequences of non-vascular trans-femoral amputation: a survey of quality of life, prosthetic use and problems . Prosthet Orthot Int 2001 ;25 :186 –94 . 10.1080/03093640108726601 11860092 3 Livingston DH , Keenan D , Kim D Extent of disability following traumatic extremity amputation . J Trauma Acute Care Surg 1994 ;37 :495 –9 . 10.1097/00005373-199409000-00027 4 Senra H , Oliveira RA , Leal I Beyond the body image: a qualitative study on how adults experience lower limb amputation . Clin Rehabil 2012 ;26 :180 –91 . 10.1177/0269215511410731 21908476 5 Ploeg AJ , Lardenoye JW , Vrancken Peeters MPFM Contemporary series of morbidity and mortality after lower limb amputation . Eur J Vasc Endovasc Surg 2005 ;29 :633 –7 . 10.1016/j.ejvs.2005.02.014 15878543 6 Ziegler-Graham K , MacKenzie EJ , Ephraim PL Estimating the prevalence of limb loss in the United States: 2005 to 2050 . Arch Phys Med Rehabil 2008 ;89 :422 –9 . 10.1016/j.apmr.2007.11.005 18295618 7 Burger H , Marincek C The life style of young persons after lower limb amputation caused by injury . Prosthet Orthot Int 1997 ;21 :35 –9 . 10.3109/03093649709164528 9141124 8 Burger H , Marinček Č Return to work after lower limb amputation . Disabil Rehabil 2007 ;29 :1323 –9 . 10.1080/09638280701320797 17729080 9 Moxey PW , Gogalniceanu P , Hinchliffe RJ Lower extremity amputations—a review of global variability in incidence . Diabet Med 2011 ;28 :1144 –53 . 10.1111/j.1464-5491.2011.03279.x 21388445 10 Dillingham TR , Pezzin LE , MacKenzie EJ Use and satisfaction with prosthetic devices among persons with trauma-related amputations: a long-term outcome study . Am J Phys Med Rehabil 2001 ;80 :563 –71 . 10.1097/00002060-200108000-00003 11475475 11 Meulenbelt HEJ , Geertzen JHB , Jonkman MF Skin problems of the stump in lower limb amputees: 1. A clinical study . Acta Derm Venereol 2011 ;91 :173 –7 . 10.2340/00015555-1040 21290085 12 Hagberg K , Häggström E , Uden M Socket versus bone-anchored trans-femoral prostheses: hip range of motion and sitting comfort . Prosthet Orthot Int 2005 ;29 :153 –63 . 10.1080/03093640500238014 16281724 13 Collins DM , Karmarkar A , Relich R Review of research on prosthetic devices for lower extremity amputation . Crit Rev Biomed Eng 2006 ;34 :379 –438 . 10.1615/CritRevBiomedEng.v34.i5.20 17206920 14 Lyon CC , Kulkarni J , Zimerson E Skin disorders in amputees . J Am Acad Dermatol 2000 ;42 :501 –7 . 10.1016/S0190-9622(00)90227-5 10688725 15 Brånemark R , Brånemark P , Rydevik B Osseointegration in skeletal reconstruction and rehabilitation: a review . J Rehabil Res Dev 2001 ;38 :175 .11392650 16 Van de Meent H , Hopman MT , Frölke JP Walking ability and quality of life in subjects with transfemoral amputation: a comparison of osseointegration with socket prostheses . Arch Phys Med Rehabil 2013 ;94 :2174 –8 . 10.1016/j.apmr.2013.05.020 23774380 17 Brånemark R , Berlin Ö , Hagberg K A novel osseointegrated percutaneous prosthetic system for the treatment of patients with transfemoral amputation: a prospective study of 51 patients . Bone Joint J 2014 ;96-B :106 –13 . 10.1302/0301-620X.96B1.31905 24395320 18 U.S. Food and Drug Administration . FDA authorizes use of prosthesis for rehabilitation of above-the-knee amputations. http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/UCM455103 16 July 2015. 19 Hagberg K , Brånemark R One hundred patients treated with osseointegrated transfemoral amputation prostheses—rehabilitation perspective . J Rehabil Res Dev 2009 ;46 :331 –44 . 10.1682/JRRD.2008.06.0080 19675986 20 Aschoff HH , Kennon RE , Keggi JM Transcutaneous, distal femoral, intramedullary attachment for above-the-knee prostheses: an endo-exo device . J Bone Joint Surg Am 2010 ;92 (Suppl 2 ):180 –6 . 10.2106/JBJS.J.00806 21123601 21 Hagberg K , Brånemark R , Gunterberg B Osseointegrated trans-femoral amputation prostheses: prospective results of general and condition-specific quality of life in 18 patients at 2-year follow-up . Prosthet Orthot Int 2008 ;32 :29 –41 . 10.1080/03093640701553922 18330803 22 Lundberg M , Hagberg K , Bullington J My prosthesis as a part of me: a qualitative analysis of living with an osseointegrated prosthetic limb . Prosthet Orthot Int 2011 ;35 :207 –14 . 10.1177/0309364611409795 21697203 23 Tranberg R , Zügner R , Kärrholm J Improvements in hip- and pelvic motion for patients with osseointegrated trans-femoral prostheses . Gait Posture 2011 ;33 :165 –8 . 10.1016/j.gaitpost.2010.11.004 21130654 24 Hagberg K , Häggström E , Jönsson S Osseoperception and osseointegrated prosthetic limbs In: Gallagher Desmond D , MacLachlan M , eds. Psychoprosthetics . London : Springer London , 2008 :131 –40 . 25 Frossard L , Hagberg K , Häggström E Functional outcome of transfemoral amputees fitted with an osseointegrated fixation: temporal gait characteristics . J Prosthet Orthot 2010 ;22 :11 –20 . 10.1097/JPO.0b013e3181ccc53d 26 Nebergall A , Bragdon C , Antonellis A Stable fixation of an osseointegated implant system for above-the-knee amputees . Acta Orthop 2012 ;83 :121 –8 . 10.3109/17453674.2012.678799 22489885 27 Tillander J , Hagberg K , Hagberg L Osseointegrated titanium implants for limb prostheses attachments: infectious complications . Clin Orthop Relat Res 2010 ;468 :2781 –8 . 10.1007/s11999-010-1370-0 20473597 28 Muderis MA , Tetsworth K , Khemka A The Osseointegration Group of Australia Accelerated Protocol (OGAAP-1) for two-stage osseointegrated reconstruction of amputated limbs . Bone Joint J 2016 ;98-B :952 –60 . 10.1302/0301-620X.98B7.37547 27365474 29 Al Muderis M , Khemka A , Lord SJ Safety of osseointegrated implants for transfemoral amputees: a two-center prospective cohort study . J Bone Joint Surg Am 2016 ;98 :900 –9 . 10.2106/JBJS.15.00808 27252434 30 Meehan J , Jamali AA , Nguyen H Prophylactic antibiotics in hip and knee arthroplasty . J Bone Joint Surg Am 2009 ;91 :2480 –90 . 10.2106/JBJS.H.01219 19797586 31 Lipkin DP , Scriven AJ , Crake T Six minute walking test for assessing exercise capacity in chronic heart failure . Br Med J (Clin Res Ed) 1986 ;292 :653 –5 . 10.1136/bmj.292.6521.653 32 Schoppen T , Boonstra A , Groothoff JW The timed ‘up and go’ test: reliability and validity in persons with unilateral lower limb amputation . Arch Phys Med Rehabil 1999 ;80 :825 –8 . 10.1016/S0003-9993(99)90234-4 10414769 33 Health Care Financing Administration (HCFA) Common Procedure Coding System HCPCS 2001 . Washington, DC, USA : US Government Printing Office , 2001 . 34 Hagberg K , Rickard B , Hägg O Questionnaire for persons with a transfemoral amputation (Q-TFA): initial validity and reliability of a new outcome measure . J Rehabil Res Dev 2004 ;41 :695 –706 . 10.1682/JRRD.2003.11.0167 15558399 35 Ware JE , Sherbourne CD The MOS 36-Item Short-Form Health Survey (SF-36): I. Conceptual framework and item selection . Med Care 1992 ;30 :473 –83 .1593914 36 Bitar AA , Kaplan RJ , Stitik TP Rehabilitation of orthopedic and rheumatologic disorders. 3. Total hip arthroplasty rehabilitation . Arch Phys Med Rehabil 2005 ;86 (Suppl 1 ):56 –60 . 10.1016/j.apmr.2004.12.015 37 Adell R , Lekholm U , Rockler B A 15-year study of osseointegrated implants in the treatment of the edentulous jaw . Int J Oral Surg 1981 ;10 :387 –416 . 10.1016/S0300-9785(81)80077-4 6809663 38 Brånemark P-I , Engstrand P , Öhrnell L-O Brånemark Novum®: a new treatment concept for rehabilitation of the edentulous mandible. Preliminary results from a prospective clinical follow-up study . Clin Implant Dent Relat Res 1999 ;1 :2 –16 . 10.1111/j.1708-8208.1999.tb00086.x 11359307 39 Gatti C , Haefliger W , Chiapasco M Implant-retained mandibular overdentures with immediate loading: a prospective study of ITI implants . Int J Oral Maxillofac Implants 2000 ;15 :383 –8 .10874803 40 Spiekermann H , Jansen VK , Richter E-J A 10-year follow-up study of IMZ and TPS implants in the edentulous mandible using bar-retained overdentures . Int J Oral Maxillofac Implants 1995 ;10 :231 –43 .7744443

PMC005xxxxxx/PMC5372149.txt

==== Front BMJ OpenBMJ OpenbmjopenbmjopenBMJ Open2044-6055BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR bmjopen-2016-01266510.1136/bmjopen-2016-012665Genetics and GenomicsResearch150616971724Venous thromboembolism in adults screened for sickle cell trait: a population-based cohort study with nested case–control analysis Little Iain Vinogradova Yana Orton Elizabeth Kai Joe Qureshi Nadeem Division of Primary Care, University of Nottingham, Nottingham, UKCorrespondence to Dr Nadeem Qureshi; nadeem.qureshi@nottingham.ac.uk2017 29 3 2017 7 3 e01266516 5 2016 31 1 2017 3 2 2017 Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://www.bmj.com/company/products-services/rights-and-licensing/2017This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/Objective To determine whether sickle cell carriers (‘sickle cell trait’) have an increased risk of venous thromboembolism (VTE). Design Cohort study with nested case–control analysis. Setting General population with data from 609 UK general practices in the Clinical Practice Research Datalink (CPRD). Participants All individuals registered with a CPRD general practice between 1998 and 2013, with a medical record of screening for sickle cell between 18 and 75 years of age. Main outcomes measures Incidence of VTE per 10 000 person-years (PY) among sickle cell carriers and non-carriers; and adjusted OR for VTE among sickle cell carriers compared with non-carriers. Results We included 30 424 individuals screened for sickle cell, with a follow-up time of 179 503 PY, identifying 55 VTEs in 6758 sickle cell carriers and 125 VTEs in 23 666 non-carriers. VTE incidence among sickle cell carriers (14.9/10 000 PY; 95% CI 11.4 to 19.4) was significantly higher than non-carriers (8.8/10 000 PY; 95% CI 7.4 to 10.4). Restricting analysis to confirmed non-carriers was non-significant, but performed on a small sample. In the case–control analysis (180 cases matched to 1775 controls by age and gender), sickle cell carriers remained at increased risk of VTE after adjusting for body mass index, pregnancy, smoking status and ethnicity (OR 1.78, 95% CI 1.18 to 2.69, p=0.006), with the greatest risk for pulmonary embolism (PE) (OR 2.27, 95% CI 1.17 to 4.39, p=0.011). Conclusions Although absolute numbers are small, in a general population screened for sickle cell, carriers have a higher incidence and risk of VTE, particularly PE, than non-carriers. Clinicians should be aware of this elevated risk in the clinical care of sickle cell carriers, or when discussing carrier screening, and explicitly attend to modifiable risk factors for VTE in these individuals. More complete primary care coding of carrier status could improve analysis. sickle cell trait;Venous thromboembolismcohort study ==== Body Strengths and limitations of this study The study used a large primary care database representative of the UK population. The database was interrogated over a 25-year time period, and included a large number of individuals screened for sickle cell. The findings from the main cohort study were consolidated by a nested case–control study. In the primary care database, there is limited recording of non-carrier status and ethnicity. No external source of data was available to verify carrier status and outcomes. Introduction Sickle cell disease (SCD) is a group of recessively inherited haemoglobin disorders causing anaemia, vascular occlusion and increased venous thromboembolism (VTE).1 People with sickle cell trait (SCT) are carriers who inherit one sickle haemoglobin gene and one normal haemoglobin gene. There are an estimated 300 million individuals worldwide with SCT, with the highest frequency of carriage in people of African or Middle Eastern origin, with increased rates among those of Mediterranean descent.2 3 There are at least 240 000 people with SCT in England, and an estimated 4 million individuals in the USA.4 5 SCT has important implications for genetic reproductive risk to offspring but is otherwise considered relatively harmless to the individual's health, manifesting no complications of SCD itself other than, rarely, in situations of severe dehydration or hypoxia such as at high altitude.6 However, concerns about other morbidity and mortality associated with being a sickle cell carrier have long been debated. These include reports of sudden exercise-related death in military recruits and college sportsmen.7–10 The US National Collegiate Athletic Association's mandatory SCT screening programme for student athletes has provoked considerable controversy about whether this will improve outcomes, while also leading to discrimination and confusion.4 11 12 Recently, SCT has also been shown to be associated with chronic kidney disease.13 Arguably of greater relevance to all people who are sickle cell carriers and routine clinical practice are reports of increased risk of VTE.2 14–16 However, rather than draw from the general population, these studies have involved selected samples already at higher risk of thromboembolism, such as hospitalised patients,15 17 18 individuals in the peripartum period19 or have focused solely on African-Americans.16 20 For example, in a case–control study of hospitalised African-Americans, the likelihood of VTE was approximately twice as high among those with SCT compared with non-carriers.15 However, more robust evidence on potential morbidity related to SCT in the wider general population is lacking. In the UK, as far back as 1993, policymakers have recommended that primary care should screen patients opportunistically for SCT to identify potential genetic reproductive risk, for example, when they registered with a practice.21 The identification of SCT has greater significance now that several countries are establishing national population screening programmes for sickle cell. For example, since 2003, women in the UK have been routinely offered screening for sickle cell carrier status in pregnancy to better inform reproductive risk assessment, and all newborn babies are offered screening for SCD (which also identifies carriers).22 Similar newborn screening has been offered in New York State since 1975, with universal newborn sickle cell screening throughout the USA since 2006.23 These screening programmes are increasing the number of people identified with SCT in general populations and so achieving a better understanding of possible implications for their health is needed. The aim of this study therefore was to investigate potential associations between SCT and risk of VTE in the general population. Methods Study design and setting We conducted an open cohort study with nested case–control analysis of patient medical data from the UK's Clinical Practice Research Datalink (CPRD). In the UK, the general practitioner holds the lifelong medical record of all patients registered with their practice and 98% of the UK population is registered with a general practitioner.24 The CPRD contains an anonymised copy of these medical records for more than 12 million patients from over 600 general practices across the UK, comprising one of the largest sources of continuous medical and prescribing data in the UK.25 Patient data regarding primary, secondary and tertiary healthcare usage are documented in an electronic health record using the Read code classification system, based on the International Classification of Diseases V.10. It includes prescription data, medical diagnoses, symptoms, test results, treatments and records of secondary care attendance. Patient records in CPRD are representative of the UK population26 27 and of good quality for epidemiological research, including for VTE events.28–30 Study participants Study participants were registered with 609 CPRD practices between February 1988 and May 2013 and had been screened for sickle cell between age 18 and 75 years. Participants entered the study at the latest of 1 January of the year of their 18 birthday, their date of practice registration or the date that the practice attained the recognised data quality standard for CPRD. Participants were followed up until the earliest of the 1 January of the year of their 75 birthday, date of death, date of transfer out from the practice, the last collection date of data from the practice or the occurrence date of the first VTE event. Patients were classified as having SCT if they had a diagnostic code for SCT or recorded laboratory test result confirming carrier status (with or without a medical code entry indicating screening for sickle cell). For the main analysis, patients were classified as not having SCT (non-carriers) if they had a medical code indicating sickle cell screening at any age and had no result recorded in their medical records, or had a medical code or test results indicating non-carrier status. A sensitivity analysis was undertaken that only included patients with confirmed non-carrier status, thus excluding individuals who did not have a diagnostic code or recorded laboratory test result for non-carrier status. Patients were considered to have had a VTE if they had a validated diagnostic medical code for VTE in their record during the observational period (February 1988–May 2013).29 We included only the first occurrence of VTE between the ages of 18 and 75 years while registered in a CPRD practice. We excluded patients with documented VTE prior to cohort entry or with a history of anticoagulant therapy more than 6 weeks prior to the date of VTE as this could indicate a previous VTE. Individuals with a medical diagnosis of SCD were excluded from the study. Code lists for VTE, sickle cell screening, carrier and non-carrier status are in online supplementary material. 10.1136/bmjopen-2016-012665.supp1supplementary data Main outcome measures Our main outcome for the cohort analysis was the incidence rate of VTE per 10 000 person-years (PY) among carriers and non-carriers. In the nested case–control analysis, our main outcome was the adjusted OR for VTE among carriers compared with non-carriers. Cohort analysis The cohort was described by frequencies and percentages of carriers and non-carriers in groups defined by gender, ethnicity and socioeconomic status. The Townsend Deprivation Index of material deprivation was used to assess the socioeconomic profile of the cohort.31 Incidence rates of first VTE were calculated by dividing the number of new cases by PY of follow-up. Incidence rates for subcategories of VTE were also calculated, including deep venous thrombosis (DVT) and pulmonary embolism (PE) and other, rarer VTE, such as retinal vein thrombosis and cerebral venous thrombosis. We used the Mantel-Haenszel method to compare incidence rates between carrier and non-carrier groups. Data management and statistical analyses were performed using Stata V.13. Nested case–control study We undertook a nested case–control study to investigate the association between VTE outcome and carrier status, taking into account potential confounding factors. The case–control study was based on the underlying cohort where matched controls were randomly selected from all remaining participants at risk, including potential future cases.32 Estimates obtained from nested case–control studies and from underlying cohorts have been shown to be similar.33 This design also allowed us to use data for confounding factors ascertained at the closest time before the VTE date, thus simplifying the analysis of time-dependent exposures. The controls were matched by year of birth and gender and allocated an index date, which was the VTE diagnosis date for their matched case. Controls with anticoagulant therapy at any time before the index date were excluded from the analysis. Case–control analysis We used conditional logistic regression to estimate unadjusted and adjusted ORs and 95% CIs for VTE in carriers compared with non-carriers. For the multivariable analysis, we matched for age and sex, and adjusted for a priori confounders: ethnicity, smoking status, body mass index (BMI) and pregnancy. Validation of VTE events recorded in primary care has demonstrated that the vast majority are supported by relevant hospital investigations or death certification.29 However, to account for any potential misclassification of cases and controls by relying on diagnostic recording of VTE in primary care records, we undertook a subgroup analysis using a more stringent definition of VTE diagnosis, in which VTE occurrence was supported by one of the following: a documented prescription for warfarin or low-molecular weight heparin within 3 months of the date of diagnosis; or evidence of attendance at a clinic for treatment with anticoagulants within 3 months of diagnosis; or death within 1 month of diagnosis.34 35 We included ethnicity, pregnancy, BMI and smoking status as covariates. We selected these because within the UK, minority ethnic groups are more likely to be offered sickle cell testing, risk of first VTE in pregnant women is at least sixfold,36 37 obesity is associated with a 2–3 times increased risk of VTE38 39 and smoking with a 1.5 times increased risk.40 41 BMI and smoking are routinely recorded in UK primary care electronic health records.42 43 Smoking status and BMI data were extracted using the closest record prior to the index date. Smoking status was categorised into: current smokers, ex-smokers and non-smokers. BMI was defined as weight (in kilograms) divided by the square of height (in metres), and categorised into <25, 25–29 and ≥30. Ethnicity was extracted from primary care records or, where not available, from linked hospital electronic health records (HES) and categorised as: White; Asian (comprising ethnic groups from the Indian subcontinent and China); Black (indicative of African ancestry); Other (including mixed ethnicity). Ethnicity was self-reported by patients attending General Practice or Hospital, and was grouped by investigators. Pregnancy included the first 3 months after delivery and was defined using pregnancy codes and the estimated conception date of delivery minus 280 days or delivery date minus gestational age if recorded.44 Since pregnancy and the postpartum period is associated with increased risk of VTE,45 we undertook a sensitivity analysis removing pregnant women from the analysis. Subgroup analysis of VTE and SCT in Black people was also undertaken to allow direct comparison with the previous literature.15 16 18 To retain the statistical power and obtain unbiased results, all observations were included in the analysis. Chained equations were used to impute missing values for the covariates.46 Ten imputed sets were generated and the imputation model: matching variables (age and gender); exposure (carrier, non-carrier); outcome (case or control); covariates (pregnancy, logarithm of BMI, smoking status and ethnicity); UK region (because the ethnicity profile differs by regions); index year.47 The results from the imputed sets were combined using Rubin's rules.46 To ensure the assumption that the data were missing at random, we ran a sensitivity analysis on patients with complete data. Patient involvement The study was a retrospective, quantitative, observational study using information routinely collected by general practices and a similar linked source of census data. The design and the development of outcome measures were necessarily technical, based in part on earlier publications and medical expertise within the team, but primarily on examination of available data in order to maximise inclusiveness and length of study period. As such, patients were not involved in any of these aspects. However, as well as dissemination through this publication, we are working with sickle cell patient groups to produce lay summaries describing the research and its results. With respect to the use of the CPRD database, the CPRD Group has ethical approval from a National Research Ethics Service Committee (NRES) for all research not including patient involvement and using anonymised data. Results Cohort analysis During the study period, 45 746 individuals had a record of sickle cell screening in their medical records. After exclusions (figure 1), the cohort analysis included 30 424 individuals; 6758 individuals with SCT (carriers) and 23 666 without SCT (non-carriers). Overall, 82.5% of individuals were women, with a higher proportion in the non-carrier group than the carrier group (86.9% vs 67.3%, respectively) (table 1). Ethnicity was recorded for more than 80% of the study population and of those with recorded ethnicity, 53.4% of the carrier group and 15.8% of the non-carrier group were Black. Table 1 Characteristics of the study population and numbers and proportions of carriers and non-carriers for each category Demographic characteristics Whole cohort Carriers Non-carriers Total number of patients 30 424 6758 23 666 Age at study entry  Median (IQ range) 28 (22, 36) 32 (25, 41) 28 (22, 34) Age at SCD test  Median (IQ range) 29 (23, 36) 28 (19, 36) 30 (24, 35) Gender  Male (%) 5322 (17.5) 2213 (32.7) 3109 (13.1)  Female (%) 25 102 (82.5) 4545 (67.3) 20 557 (86.9) Ethnicity  White (%) 8678 (28.5) 484 (7.2) 8194 (34.6)  Black (%) 7343 (24.1) 3608 (53.4) 3735 (15.8)  Asian (%) 3408 (11.2) 228 (3.4) 3180 (13.4)  Other (%) 5377 (17.7) 662 (9.8) 4715 (19.9)  Not recorded (%) 5618 (18.5) 1776 (26.3) 3842 (16.2) Socioeconomic status  Quintile 1 (least deprived) 1784 (8.3) 325 (4.8) 1459 (6.2)  Quintile 2 1998 (9.3) 386 (5.7) 1612 (6.8)  Quintile 3 2417 (11.3) 602 (8.9) 1815 (7.7)  Quintile 4 4125 (19.2) 1179 (17.4) 2946 (12.4)  Quintile 5 (most deprived) 6876 (32.0) 2254 (33.4) 4622 (19.5)  Missing 4278 (19.9) 2012 (29.8) 11 212 (47.4) Figure 1 Flow chart of patients in the cohort study. CPRD, Clinical Practice Research Datalink; VTE, venous thromboembolism. One hundred and eighty VTEs were identified within a total follow-up time of 179 503 PY. The median follow-up time was 3.71 years in the carrier group and 4.68 years in the non-carrier group. The incidence of VTE was 14.9 (95% CI 11.4 to 19.4) per 10 000 PY in carriers, and 8.8 (95% CI 7.4 to 10.4) per 10 000 PY in non-carriers (table 2). Considering VTE subgroups, the difference in incidence rates was greatest for PEs and not significant for DVT. Sensitivity analysis with confirmed non-carriers identified low event rates in this subgroup, leading to wide CIs but showed a consistent pattern of results to the main analysis (see online supplementary table S1). Table 2 Incidence rates per 10 000 person-years (PY) for venous thromboembolism (VTE), deep vein thrombosis (DVT), pulmonary embolism (PE) and other VTEs by carrier status Carrier Non-carriers Incidence rate ratio (95% CI) p Value Number of events Incidence rate per 10 000 PY (95% CI) Number of events Incidence rate per 10 000 PY (95% CI) VTE 55 14.9 (11.4 to 19.4) 125 8.77 (7.35 to 10.4) 1.70 (1.24 to 2.33) 0.001 DVT only 24 6.53 (4.37 to 9.74) 83 5.83 (4.70 to 7.23) 1.12 (0.71 to 1.76) 0.2 PE only 26 7.07 (4.81 to 10.4) 36 2.53 (1.83 to 3.51) 2.79 (1.69 to 4.62) <0.001 Other VTE 5 1.36 (0.57 to 3.28) 6 0.42 (0.19 to 0.94) 3.23 (0.99 to 10.6) 0.041 Nested case–control analysis A total of 180 cases were matched to 1775 controls by gender and year of birth. Cases and controls had similar profiles with the majority of participants being women (table 3). The proportion of obese patients was higher among cases (31.1% compared with 20.6% of controls, p<0.001), and women were more likely to be pregnant in cases (29.4% of female cases) than in controls (16.4% of female controls). Table 3 Characteristics of cases and controls: frequencies and proportions Cases frequency (%) n=180 Controls frequency (%) n=1775 Adjusted ORs (95% CIs)* Sex  Male 42 (23.3) 409 (23.0)  Female 138 (76.7) 1366 (77.0) Age at first VTE/index date  18–34 83 (46.1) 828 (46.6)  35–54 62 (34.4) 615 (34.6)  55–74 35 (19.4) 332 (18.7) Ethnicity  White 59 (32.8) 518 (29.2) Reference  Black 47 (26.1) 431 (24.3) 0.74 (0.47 to 1.17)  Asian 14 (7.8) 176 (9.9) 0.82 (0.41 to 1.63)  Other 32 (17.8) 299 (16.8) 0.89 (0.56 to 1.40)  Not recorded 28 (15.6) 351 (19.8) Smoking status  Non-smokers 104 (57.8) 1058 (59.6) Reference  Ex-smokers 34 (18.9) 275 (15.5) 1.18 (0.76 to 1.85)  Smokers 31 (17.2) 331 (18.6) 1.05 (0.67 to 1.63)  Not recorded 11 (6.1) 111 (6.3) BMI  BMI mean (SD) 28.7 (6.0) 26.5 (5.5)  <25 kg/m2 44 (24.4) 694 (39.1) Reference  25–29 kg/m2 59 (32.8) 479 (27.0) 2.05 (1.30 to 3.23)  30+ kg/m2 56 (31.1) 365 (20.6) 2.69 (1.67 to 4.33)  Not recorded 21 (11.7) 237 (13.4) Pregnant  No 127 (70.6) 1484 (83.6) Reference  Yes 53 (29.4) 291 (16.4) 1.45 (1.25 to 1.69) *ORs refer to multivariate analysis on imputed data investigating the effect of the covariates to VTE risk. Cases and controls were matched by age and sex, so ORs were not calculated for these variables. After adjusting for a priori confounders (see online supplementary table S2), carriers were 1.78 times more likely to have a VTE than non-carriers (OR 1.78, 95% CI 1.18 to 2.69) and the odds of PE in carriers increased even further (OR 2.27, 95% CI 1.17 to 4.39) (table 4). The odds of DVT were increased in carriers, although this was not significant (OR 1.43, 95% CI 0.79 to 2.59). Table 4 Unadjusted and adjusted ORs of thrombotic events in patients with sickle cell trait (SCT), compared with patients without SCT, main and subgroup analyses Cases Controls OR* Adjusted OR† p Value Main analysis VTE  Non-carrier 125 1369 1.00 1.00 –  Carrier 55 406 1.55 (1.09 to 2.21) 1.78 (1.18 to 2.69) 0.006 DVT only  Non-carrier 83 841 1.00 1.00 –  Carrier 24 207 1.18 (0.71 to 1.95) 1.43 (0.79 to 2.59) 0.2 PE  Non-carrier 36 449 1.00 1.00 –  Carrier 26 170 2.09 (1.17 to 3.72) 2.27 (1.17 to 4.39) 0.011 Subgroup analysis VTE definition included anticoagulant therapy, clinic attendance or death  Non-carrier 60 725 1.00 1.00 –  Carrier 36 223 2.20 (1.36 to 3.57) 2.71 (1.52 to 4.83) 0.001 *Matched by age and sex. †Adjusted for age, sex, pregnancy, BMI, smoking status and ethnicity. When we restricted the case definition of VTE to include cases with a diagnostic medical code for VTE, supported by a record of anticoagulant therapy, clinic attendance or death, there was a more than twofold increased odds of VTE in carriers (OR 2.71, 95% CI 1.52 to 4.83). After removal of pregnant women from the cohort, the odds of VTE in carriers compared with non-carriers remained consistent at 1.80 (95% CI 1.15 to 2.83). There were 27 VTE events in the subgroup of 7343 Black patients. The incidence rate of VTE was 12.9 per 10 000 PY; 15.5 per 10 000 PY in carriers and 10.6 per 10 000 PY in non-carriers (incidence rate ratio 1.45, 95% CI 0.82 to 2.59, p=0.2). In the case–control analysis, 47 cases were matched by age and gender to 418 controls. After adjusting for covariates, the VTE risk associated with SCT carrier in this subgroup was in line with the main analysis (OR 1.73, 95% CI 0.91 to 3.32, p=0.096). Discussion In a large general population cohort screened for sickle cell over a 25-year period, we have found SCT was associated with a significantly increased likelihood of VTE, in particular, PE. The findings from the cohort study were consolidated by a nested case–control study.48 Analysis of VTE based solely on medical code recording does include the potential for false-positive diagnoses. However, 84% of VTEs reported in primary care have been validated through hospital records or death records.29 Our study findings were consistent when a more stringent definition of VTE (combining evidence from medical diagnoses, drug therapies and/or clinic attendance) was used. Strengths and limitations of the study A major strength of this study is the investigation and confirmation of a significant association between SCT and VTE in a very much larger, more socially and ethnically diverse, general population than previous studies. The CPRD contain data from patients that are considered representative of the general UK population.49 Our cohort analysis included more than 30 000 individuals screened for sickle cell, with almost 7000 carriers and over 23 000 non-carriers, with an incidence of VTE observed over nearly 180 000 years of person time. We also note the VTE incidence in non-carriers is similar to that reported in the UK general population.48 We recognise that only people who had screening and/or testing documented in their medical record were included, potentially missing others in the population who had been screened for sickle cell. However, the introduction of universal sickle cell and thalassaemia antenatal carrier screening in the UK in 2003 will have improved recording, with universal screening across the entire antenatal population in areas with higher prevalence of SCD (more than 1.5 per 10 000 births).50 This is reflected in the study population, which, as expected, was mostly screened around reproductive age with proportionally more women, while still representative of the broad ethnic and social diversity of the UK general population. The majority of people had a record of screening but no result recorded and these participants were assumed to be non-carriers in our main analysis, as this would be consistent with clinical practice. However, some with no results recorded may have been carriers. The presence of carriers in the unconfirmed non-carriers (no test) group would reduce the effect size of the main analysis. Further, GPs are more likely to document carrier state than non-carrier state introducing recall bias. Restricting participants to only those with documented carrier and non-carrier status resulted in incidence rates and incidence rate ratios that were consistent with the main analysis, although underpowered due to the restricted number of events. We also only used patient data that were considered to be of sufficient quality (in terms of completeness and accuracy) as determined by CPRD (ie, identified as ‘up to standard’ within the database). However, we were only able to take account of confounding factors that are recorded in the electronic health record. Similarly, while free-text information on diagnostic imaging is inconsistent and not accessible to researchers in electronic primary care records, the prevalence of PE in the carrier group could potentially reflect ascertainment bias, if, for example, clinicians had a lower index of suspicion for lung problems in carriers, with more frequent diagnostic imaging. Comparison with recent studies Our findings are consistent with previous North American studies in demonstrating a relationship between SCT and VTE, but confirm this not only in a much larger general population sample but also across socioeconomic groups and ethnicities. Based on our subgroup and adjusted analyses, the difference in VTE incidence rates could not be attributed to difference in the ethnicity of carriers and non-carriers in our study. Our results are also consistent with previous observational studies with smaller or more selected populations.15 16 18 20 Similar to our findings, Austin et al15 reported a higher risk of PE among hospitalised black Americans with SCT, with no increased risk of DVT in carriers compared with non-carriers. An earlier retrospective cohort study of hospitalised black men found PE occurring in 2.2% of carriers compared with 1.5% in non-carriers, though diagnosis of PE was based solely on clinical criteria.18 Recent cohort studies, again of African-Americans only, have reported no increased overall risk of VTE but up to a twofold increased risk of PE.16 20 The possible causes for this increased risk of VTE remain unclear but may, in part, be related to increased coagulation activity found in SCT.51 Ideally, future research should explore the relationship between SCT and VTE in a prospective birth cohort systematically offered neonatal screening for sickle cell with carrier and non-carrier state accurately documented. If primary care records are to be used in future research, the coding of non-carrier status needs to be improved, for example, through direct coding of carrier status when results are transferred electronically to primary care records from pathology laboratories. Implications for clinical practice Currently, individuals identified with SCT are advised that their carrier state is relatively harmless to their health, with caution required only in unusual conditions of severe dehydration or hypoxia such as extreme exertion, anaesthesia or high altitude. Although the absolute numbers of VTE events were small, our results indicate that clinicians should make patients with SCT aware of their increased risk of VTE and stress the particular importance of attending to modifiable risk factors, such as obesity and smoking. Compared with other conventional risk factors, the risk of VTEs with carrier status is similar to pregnancy, obesity and prolonged bed rest.36 The evidence may now be sufficiently robust to consider SCT a significant non-modifiable risk factor in evaluating patients' VTE risk. In the clinical setting, this could facilitate modification of other coexistent risk factors of VTE, such as obesity and immobility. In relation to sickle cell screening, identification of carrier status is perceived as useful to inform individuals about potential reproductive genetic risk—antenatally and among parents of newborns.52 53 The current finding of increased risk of VTE in sickle cell carriers adds a further dimension to the appropriate provision of information to patients in such screening contexts. Given the experience of African-American communities in the recent past, caution is needed to avoid stigmatisation and community anxiety11 54 55 in advancing appropriate care of people with SCT. Previous researchers have observed the Black community, at genetic risk of sickle cell, are also disadvantaged in society. A situation exacerbated by racism.56 Such issues also deserve further consideration. However, we also note our findings are drawn from a large, socially and ethnically diverse general population with SCT. Thus, this study's implications for clinical practice are not confined to a particular ethnic community, but rather apply to all. The authors thank Drs Vence Bonham and Sarah Donohue for their expert advice at the conception of the study. There have been no conference presentations of this study. Contributors: NQ, JK, EO and YV are responsible for study conception and design. All authors are responsible for acquisition, analysis or interpretation of data. IL is responsible for drafting of the manuscript. NQ, JK, EO and YV are responsible for critical revision of the manuscript for important intellectual content. YV and IL are responsible for statistical analysis. NQ and JK obtained funding. All authors are responsible for administrative, technical or material support. NQ and EO are responsible for study supervision. All authors had full access to the data in the study and can take full responsibility for the integrity of the data and the accuracy of the data analysis. Funding: This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Competing interests: None declared. Ethics approval: We obtained ethical approval from the CPRD Independent Scientific Advisory Committee of the UK Medicine and Healthcare products Regulatory Agency (protocol approval no. 13_097RA). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: The authors confirm that, for approved reasons, some access restrictions apply to the data underlying the findings. Data are collected and owned and licensed by The Clinical Practice Research Datalink Group, The Medicines and Healthcare Products Regulatory Agency, 5th Floor, 151 Buckingham Palace Road, Victoria, London SW1W 9SZ, UK. Contact +44 20 3080 6383. See https://www.cprd.com/dataAccess/default.asp#OnlineDataGOLD ==== Refs References 1 Serjeant GR , Serjeant BE Sickle cell disease . 3rd edn . Oxford : Oxford University Press , 2001 . 2 Tsaras G , Owusu-Ansah A , Boateng F Complications associated with sickle cell trait: a brief narrative review . Am J Med 2009 ;122 :507 –12 . 10.1016/j.amjmed.2008.12.020 19393983 3 Modell B , Darlison M Global epidemiology of haemoglobin disorders and derived service indicators . Bull World Health Organ 2008 ;86 :480 –7 . 10.2471/BLT.06.036673 18568278 4 Jordan LB , Smith-Whitley K , Treadwell MJ Screening U.S. college athletes for their sickle cell disease carrier status . Am J Prev Med 2011 ;41 (Suppl 4 ):S406 –12 . 10.1016/j.amepre.2011.09.014 22099365 5 England. PH . NHS Sickle Cell and Thallasaemia programme Fact and Figures. Secondary NHS Sickle Cell and Thallasaemia programme Fact and Figures. http://sct.screening.nhs.uk/statistics 6 Kark JA , Ward FT Exercise and hemoglobin S . Semin Hematol 1994 ;31 :181 –225 .7973777 7 Kark JA , Posey DM , Schumacher HR Sickle-cell trait as a risk factor for sudden death in physical training . N Engl J Med 1987 ;317 :781 –7 . 10.1056/NEJM198709243171301 3627196 8 Maron BJ , Doerer JJ , Haas TS Sudden deaths in young competitive athletes: analysis of 1866 deaths in the United States, 1980–2006 . Circulation 2009 ;119 :1085 –92 . 10.1161/CIRCULATIONAHA.108.804617 19221222 9 Harmon KG , Drezner JA , Klossner D Sickle cell trait associated with a RR of death of 37 times in National Collegiate Athletic Association football athletes: a database with 2 million athlete-years as the denominator . Br J Sports Med 2012 ;46 :325 –30 . 10.1136/bjsports-2011-090896 22442191 10 Jones SR , Binder RA , Donowho EM Jr Sudden death in sickle-cell trait . N Engl J Med 1970 ;282 :323 –5 . 10.1056/NEJM197002052820607 5410817 11 Bonham VL , Dover GJ , Brody LC Screening student athletes for sickle cell trait: a social and clinical experiment . N Engl J Med 2010 ;363 :997 –9 . 10.1056/NEJMp1007639 20825310 12 Carter B , Dyson SM Actor network theory, agency and racism: the case of sickle cell trait and US athletics . Social Theory Health 2015 ;13 :62 –77 . 10.1057/sth.2014.17 13 Naik RP , Derebail VK , Grams ME Association of sickle cell trait with chronic kidney disease and albuminuria in African Americans . JAMA 2014 ;312 :2115 –25 . 10.1001/jama.2014.15063 25393378 14 Key NS , Derebail VK Sickle-cell trait: novel clinical significance . Hematology Am Soc Hematol Educ Program 2010 ;2010 :418 –22 . 10.1182/asheducation-2010.1.418 21239829 15 Austin H , Key NS , Benson JM Sickle cell trait and the risk of venous thromboembolism among blacks . Blood 2007 ;110 :908 –12 . 10.1182/blood-2006-11-057604 17409269 16 Bucknor MD , Goo JS , Coppolino ML The risk of potential thromboembolic, renal and cardiac complications of sickle cell trait . Hemoglobin 2014 ;38 :28 –32 . 10.3109/03630269.2013.832689 24099594 17 Austin H , Lally C , Benson JM Hormonal contraception, sickle cell trait, and risk for venous thromboembolism among African American women . Am J Obstet Gynecol 2009 ;200 :620.e1 –3 . 10.1016/j.ajog.2009.01.038 19306959 18 Heller P , Best WR , Nelson RB Clinical implications of sickle-cell trait and glucose-6-phosphate dehydrogenase deficiency in hospitalized black male patients . N Engl J Med 1979 ;300 :1001 –5 . 10.1056/NEJM197905033001801 431593 19 Pintova S , Cohen HW , Billett HH Sickle cell trait: is there an increased VTE risk in pregnancy and the postpartum? PLoS ONE 2013 ;8 :e64141 10.1371/journal.pone.0064141 23717554 20 Folsom AR , Tang W , Roetker NS Prospective study of sickle cell trait and venous thromboembolism incidence . J Thromb Haemost 2015 ;13 :2 –9 . 10.1111/jth.12787 25393788 21 Johnson AG Report of a working party of the Standing Medical Advisory Committee on sickle cell, thalassaemia and other haemoglobinopathies . London : HMSO , 1993 . 22 Ryan K , Bain BJ , Worthington D Significant haemoglobinopathies: guidelines for screening and diagnosis . Br J Haematol 2010 ;149 :35 –49 . 10.1111/j.1365-2141.2009.08054.x 20067565 23 Benson JM , Therrell BL Jr History and current status of newborn screening for hemoglobinopathies . Semin Perinatol 2010 ;34 :134 –44 . 10.1053/j.semperi.2009.12.006 20207263 24 NHS Digital . Attribution data set GP-registered populations 2010–2012. http://content.digital.nhs.uk/pubs/gpregpop10 25 Walley T , Mantgani A The UK general practice research database . Lancet 1997 ;350 :1097 –9 . 10.1016/S0140-6736(97)04248-7 10213569 26 Lawrenson R , Williams T , Farmer R Clinical information for research; the use of general practice databases . J Public Health Med 1999 ;21 :299 –304 . 10.1093/pubmed/21.3.299 10528957 27 Crooks C The epidemiology of upper gastrointestinal bleeding . University of Nottingham , 2013 . 28 Herrett E , Thomas SL , Schoonen WM Validation and validity of diagnoses in the General Practice Research Database: a systematic review . Br J Clin Pharmacol 2010 ;69 :4 –14 . 10.1111/j.1365-2125.2009.03537.x 20078607 29 Lawrenson R , Todd JC , Leydon GM Validation of the diagnosis of venous thromboembolism in general practice database studies . Br J Clin Pharmacol 2000 ;49 :591 –6 . 10.1046/j.1365-2125.2000.00199.x 10848723 30 Cardwell CR , Abnet CC , Cantwell MM Exposure to oral bisphosphonates and risk of esophageal cancer . JAMA 2010 ;304 :657 –63 . 10.1001/jama.2010.1098 20699457 31 Morris R , Carstairs V Which deprivation? A comparison of selected deprivation indexes . J Public Health Med 1991 ;13 :318 –26 .1764290 32 Etminan M Pharmacoepidemiology II: the nested case-control study—a novel approach in pharmacoepidemiologic research . Pharmacotherapy 2004 ;24 :1105 –9 . 10.1592/phco.24.13.1105.38083 15460170 33 Wacholder S Bias in full cohort and nested case-control studies? Epidemiology 2009 ;20 :339 –40 . 10.1097/EDE.0b013e31819ec966 19289961 34 Grainge MJ , West J , Card TR Venous thromboembolism during active disease and remission in inflammatory bowel disease: a cohort study . Lancet 2010 ;375 :657 –63 . 10.1016/S0140-6736(09)61963-2 20149425 35 Walker AJ , Card TR , West J Incidence of venous thromboembolism in patients with cancer—a cohort study using linked United Kingdom databases . Eur J Cancer 2013 ;49 :1404 –13 . 10.1016/j.ejca.2012.10.021 23146958 36 Anderson FA Jr, Spencer FA Risk factors for venous thromboembolism . Circulation 2003 ;107 (Suppl 23 ):I9 –16 . 10.1161/01.CIR.0000078469.07362.E6 12814980 37 Sultan AA , West J , Tata LJ Risk of first venous thromboembolism in and around pregnancy: a population-based cohort study . Br J Haematol 2012 ;156 :366 –73 . 10.1111/j.1365-2141.2011.08956.x 22145820 38 Cushman M Epidemiology and risk factors for venous thrombosis . Semin Hematol 2007 ;44 :62 –9 . 10.1053/j.seminhematol.2007.02.004 17433897 39 Huerta C , Johansson S , Wallander MA Risk factors and short-term mortality of venous thromboembolism diagnosed in the primary care setting in the United Kingdom . Arch Intern Med 2007 ;167 :935 –43 . 10.1001/archinte.167.9.935 17502535 40 Cheng YJ , Liu ZH , Yao FJ Current and former smoking and risk for venous thromboembolism: a systematic review and meta-analysis . PLoS Med 2013 ;10 ;e1001515 10.1371/journal.pmed.1001515 24068896 41 Pomp ER , Rosendaal FR , Doggen CJM Smoking increases the risk of venous thrombosis and acts synergistically with oral contraceptive use . Am J Hematol 2008 ;83 :97 –102 . 10.1002/ajh.21059 17726684 42 Bhaskaran K , Forbes HJ , Douglas I Representativeness and optimal use of body mass index (BMI) in the UK Clinical Practice Research Datalink (CPRD) . BMJ Open 2013 ;3 :e003389 10.1136/bmjopen-2013-003389 43 Booth HP , Prevost AT , Gulliford MC Validity of smoking prevalence estimates from primary care electronic health records compared with national population survey data for England, 2007 to 2011 . Pharmacoepidemiol Drug Saf 2013 ;22 :1357 –61 . 10.1002/pds.3537 24243711 44 Vinogradova Y , Coupland C , Hippisley-Cox J Use of combined oral contraceptives and risk of venous thromboembolism: nested case-control studies using the QResearch and CPRD databases . BMJ 2015 ;350 :h2135 10.1136/bmj.h2135 26013557 45 Sultan AA , Tata LJ , West J Risk factors for first venous thromboembolism around pregnancy: a population-based cohort study from the United Kingdom . Blood 2013 ;121 :3953 –61 . 10.1182/blood-2012-11-469551 23550034 46 Royston P Multiple imputation of missing values . Stata J 2004 ;4 :227 –41 . 47 Newgard CD , Lewis RJ Missing data: how to best account for what is not known . JAMA 2015 ;314 :940 –1 . 10.1001/jama.2015.10516 26325562 48 Parker C , Coupland C , Hippisley-Cox J Antipsychotic drugs and risk of venous thromboembolism: nested case-control study . BMJ 2010 ;341 :641 . 49 Herrett E , Gallagher AM , Bhaskaran K Data resource profile: Clinical Practice Research Datalink (CPRD) . Int J Epidemiol 2015 ;44 :827 –36 . 10.1093/ije/dyv098 26050254 50 NHS_England . NHS public health functions agreement 2015–16 . Service specification no.18. NHS Sickle Cell and Thalassaemia Screening Programme. 10100 PIPP , 2014 . 51 Westerman MP , Green D , Gilman-Sachs A Coagulation changes in individuals with sickle cell trait . Am J Hematol 2002 ;69 :89 –94 . 10.1002/ajh.10021 11835343 52 Kai J , Ulph F , Cullinan T Communication of carrier status information following universal newborn screening for sickle cell disorders and cystic fibrosis: qualitative study of experience and practice . Health Technol Assess 2009 ;13 :1 –82 . 10.3310/hta13570 53 Locock L , Kai J Parents’ experiences of universal screening for haemoglobin disorders: implications for practice in a new genetics era . Br J Gen Pract 2008 ;58 :161 –8 . 10.3399/bjgp08X277276 18318970 54 Whitten CF Sickle-cell programming—an imperiled promise . N Engl J Med 1973 ;288 :318 –19 . 10.1056/NEJM197302082880612 4682674 55 Dyson S , Boswell G Sickle cell and deaths in custody . London : Whiting and Birch , 2009 . 56 Krieger N , Chen JT , Waterman PD The inverse hazard law: blood pressure, sexual harassment, racial discrimination, workplace abuse and occupational exposures in US low-income black, white and Latino workers . Soc Sci Med 2008 ;67 :1970 –81 . 10.1016/j.socscimed.2008.09.039 18950922