Dataset Viewer
Auto-converted to Parquet Duplicate
id
stringclasses
6 values
label
stringclasses
6 values
recordType
stringclasses
5 values
journal
stringclasses
5 values
published
stringdate
2025-01-01 00:00:00
2026-07-01 00:00:00
pmid
int64
39.6M
42.5M
doi
stringclasses
5 values
testArticle
stringclasses
6 values
participants
float64
0
94
duration
stringclasses
3 values
reportedOutcome
stringclasses
6 values
statedLimit
stringclasses
6 values
interestDeclaration
stringclasses
6 values
gates.compoundedArticle
stringclasses
3 values
gates.humanAdministration
stringclasses
3 values
gates.comparator
stringclasses
3 values
gates.equivalenceTested
stringclasses
2 values
gates.independentOfStake
stringclasses
3 values
gateNotes.independentOfStake
stringclasses
1 value
sourceUrl
stringclasses
6 values
chun-2025
Chun 2025 real-world cohort
retrospective cohort
Diabetes, Obesity & Metabolism
2025-03
39,776,038
10.1111/dom.16162
Compounded semaglutide with cyanocobalamin, once weekly, subcutaneous
94
3 months
4.11 kg (SD 2.77), or 4.57 % of body weight (SD 2.96)
No control group, single commercial site, underlying data available on request only. The paper's title omits the cyanocobalamin its abstract reports.
All authors have received salary support from Restore Hyperwellness, the wellness chain whose customers were studied.
pass
pass
stop
stop
stop
null
https://doi.org/10.1111/dom.16162
heor-self-experiment-2025
HEOR analysis with one participant
self-experiment
International Journal of Pharmaceutical Compounding
2025-01
39,921,911
null
Compounded tirzepatide, weekly
1
approximately 4 weeks
Not extracted here; the record is reported by its own author, who was the sole participant.
One participant, no control, no blinding. Sole affiliation is a one-person consultancy; the journal is published for the compounding industry. The paper states it is the first article focused on compounded tirzepatide.
Author and participant are the same person; published in the compounding industry's own journal.
pass
pass
stop
stop
partial
Coded partial rather than stop, and the difference matters. The sole affiliation is a one-person consultancy and the venue is the compounding industry's own journal, so a tie to one side of the comparison is disclosed and real. What is not established is that the author, a funder or an employer sells a compounded GLP-1...
https://pubmed.ncbi.nlm.nih.gov/39921911/
lilly-b12-impurity-2026
Novel tirzepatide/B12 impurity
laboratory analysis
Expert Opinion on Drug Safety
2026-05
42,010,938
10.1080/14740338.2026.2663185
Ten samples of compounded tirzepatide with B12 from compounding pharmacies, medspas and telehealth providers across seven US states
0
null
Potency down to 43 % of the labelled amount; a previously unknown tirzepatide-B12 adduct at up to 10 % of total polypeptide content, with NMR showing altered tirzepatide structure.
Ten samples. Establishes that the adduct occurs and is widespread among products tested; does not measure how often, and tested no people. Clinical significance unknown, and the authors say so.
This manuscript was funded by Eli Lilly and Company; all five authors are Lilly employees. Lilly manufactures the approved tirzepatide products.
pass
stop
partial
partial
stop
null
https://doi.org/10.1080/14740338.2026.2663185
novo-immunogenicity-2026
Impurities and potential immunogenicity
laboratory analysis
Pharmaceutical Research
2026-07
42,533,250
10.1007/s11095-026-04146-9
Products from 36 commercial suppliers across four continents, including 13 injectable compounded semaglutide products, alongside follow-on and oral products
0
null
Every compounded and follow-on sample showed an impurity profile distinct from the originator, including amino-acid deletions and additions and unidentified impurities. Under light exposure, compounded semaglutide diverged significantly in strength, total impurities and high-molecular-weight protein.
The impact of these impurity profiles on safety and efficacy has not been tested clinically; increased immunogenicity is potential, not demonstrated. Follow-on products are a separate class from US compounded products and their findings must not be merged.
Every author is a Novo Nordisk employee and shareholder, or was at the time of the work. Novo Nordisk supplied the originator comparators and manufactures the approved semaglutide and liraglutide products.
pass
stop
pass
partial
stop
null
https://doi.org/10.1007/s11095-026-04146-9
faers-2026
FAERS disproportionality analysis
pharmacovigilance
Expert Opinion on Drug Safety
2026-03
40,285,721
10.1080/14740338.2025.2499670
81,078 GLP-1 receptor agonist reports 2018-2024, of which 707 involved compounded products
null
2018-2024
Elevated reporting odds ratios for preparation errors (48.92; 12.63-189.6), contamination (19.00; 4.24-85.03), compounding and manufacturing issues (8.51; 5.17-14.0) and hospitalisation (2.35; 1.94-2.83); lower for administration errors (0.29) and dosing errors (0.24).
A reporting odds ratio is a disproportionality measure, not a risk estimate, and does not establish cause. The database is voluntary and unverified, the denominator is unknown, and the window coincides with sustained press attention. The authors state the study was not designed to show cause and effect.
Geisinger and Binghamton University, with no product on either side. The lead author sits on the board of the Partnership for Safe Medicines, which contributed $850 to the work; a co-author was on a Pfizer- and Lilly-supported research team years earlier. Disclosures, not disqualifications.
partial
partial
partial
stop
partial
null
https://doi.org/10.1080/14740338.2025.2499670
ada-statement-2025
American Diabetes Association statement
position statement
Diabetes Care
2025-02
39,620,926
10.2337/dci24-0091
None. A professional-society position, not a study.
0
null
No US Pharmacopeia monograph exists for semaglutide or tirzepatide, so no external standard defines a correct compounded preparation and no required test verifies one. Some compounders add vitamin B12 or B6, and some use salt forms in place of the base used in approved products.
A position statement is not a measurement. It establishes the absent standard, not an outcome in any patient.
Funded from ADA general revenue; the statement declares that no other entity, including industry, provided support. The only industry-free source in this set.
stop
stop
stop
stop
pass
null
https://doi.org/10.2337/dci24-0091

Compounded GLP-1 products: six published sources coded for human evidence, comparator and independence

An editorially selected audit of six published sources bearing on whether compounded GLP-1 products have the same human evidence as approved products. Each source is coded against five explicit gates covering the test article, human administration, comparator, equivalence and independence.

Read the evidence-led article: https://lifesco.re/edge/do-compounded-glp-1-drugs-have-the-same-human-evidence/

Archived version and DOI: https://doi.org/10.5281/zenodo.22017428

What the coding shows

  • Two of the six audited sources administered a compounded product to a person.
  • None tested clinical equivalence or bioequivalence against the approved product.
  • One source passed the audit's independence gate outright.

Scope and responsible use

This is not a systematic review or literature census. It is not medical, dosing, sourcing or purchase advice, a safety verdict, or proof that an untested claim is false.

The complete selection rules, gate definitions, limitations, independence statement and versioning policy travel with the data in PROTOCOL.md. The CSV files are intended for tabular exploration; the JSON files preserve the full structured record and source notes.

Provenance

  • Publisher: lifescore Edge
  • Editorial methodology and correction route: https://lifesco.re/edge/methodology/
  • Dataset version: 1.0.0
  • Evidence reviewed: 2026-08-18
  • Licence: CC BY 4.0, limited to the original selection, scheme, coding and text described in PROTOCOL.md

Citation

@dataset{lifescore_compounded_glp1_evidence_audit_1_0_0,
  author    = {{lifescore}},
  title     = {Compounded GLP-1 products: six published sources coded for human evidence, comparator and independence},
  year      = {2026},
  version   = {1.0.0},
  publisher = {lifescore},
  doi       = {10.5281/zenodo.22017428},
  url       = {https://doi.org/10.5281/zenodo.22017428}
}
Downloads last month
24