About
Sources & Methodology
The primary sources we rely on, and how we verify what we publish.
Our Default Position
When a Phase 3 trial reads out, the most reliable initial source is the manufacturer's press release on the day of announcement; the most authoritative source is the subsequent peer-reviewed publication in a major journal; and the most operationally useful source for design and endpoint detail is ClinicalTrials.gov.
We work in that order. Press releases are noted as such, with publication-pending status, and updated to cite the peer-reviewed paper once it lands. We avoid letting press-release framing creep into how we describe trial results — for example, distinguishing between "topline announcement" and "fully published data."
Primary Sources We Use
- Eli Lilly investor relations — Manufacturer press releases for TRIUMPH, TRANSCEND, and SYNERGY readouts; quarterly earnings transcripts with pipeline updates
- ClinicalTrials.gov — Trial protocols, enrollment status, NCT identifiers, primary and secondary endpoints, completion dates
- U.S. Food and Drug Administration (FDA) — Regulatory guidance documents, warning letters relevant to GLP-1 class compounding, FDA review documents when published
- European Medicines Agency (EMA) — European regulatory pathway, pediatric investigation plans, MAA timelines
- New England Journal of Medicine (NEJM) — Phase 2 obesity trial (Jastreboff et al., 2023) and subsequent retatrutide peer-reviewed publications
- The Lancet — Phase 2 type 2 diabetes trial (Rosenstock et al., 2023) and Lancet Diabetes & Endocrinology coverage
- Nature Medicine — Phase 2 liver fat substudy (Sanyal et al., 2024)
- Diabetes, Obesity and Metabolism — TRIUMPH program design and dose-escalation methodology (Giblin et al., 2026)
- ADA Scientific Sessions — American Diabetes Association annual conference — frequent venue for first presentation of retatrutide and class data
- EASD Annual Meeting — European Association for the Study of Diabetes annual meeting — companion European venue for endocrine readouts
- ObesityWeek — Obesity Society annual conference — primary venue for obesity-specific clinical data
We also reference Reuters, Bloomberg, STAT News, BioSpace, and Endpoints News as secondary journalism sources for context and timing, but never as the sole source for a numerical clinical claim.
How We Handle Conflicting Sources
- When the press release and a peer-reviewed paper differ (this happens occasionally when on-treatment estimands replace topline numbers, or sensitivity analyses change reported figures), we update to the peer-reviewed paper and note the change in the article's Last Updated history.
- When two independent peer-reviewed sources disagree, we present both and explain the methodological reason for the difference (estimand choice, population, duration).
- When journalism contradicts a primary source, we trust the primary source and ask the journalist (or their editor) for clarification.
- When social media or industry forums contradict published data, we treat the published data as authoritative and the social claim as unverified.
Statistical and Methodological Conventions
- Estimands. Phase 3 obesity trials report both an efficacy estimand (treatment effect when actually taken) and a treatment-regimen estimand (effect including discontinuations). We label which estimand we are quoting and avoid mixing them in the same table.
- Responder rates. When reporting "% of participants achieving ≥X% weight loss," we use the same estimand as the mean weight loss reported alongside it, unless explicitly noting otherwise.
- Cross-trial comparisons. We flag every cross-trial comparison explicitly. We do not present cross-trial numerical differences as if they were from a head-to-head randomized trial.
- Subgroup and per-protocol analyses. These are reported as supplementary unless they were pre-specified primary endpoints.
What We Try Hard Not To Do
- Quote a peer-reviewed paper without reading the paper.
- Mix on-treatment and intent-to-treat numbers in the same comparison.
- Use "TRIUMPH" or "TRANSCEND" as if the program were a single trial.
- Cite a press release as if it were a peer-reviewed paper.
- Cite ourselves as a primary source.
- Recommend a specific compounder, supplier, telehealth provider, or pharmacy.
- Round 28.3% to "30%" or 11.3% to "around 10%" in numerical claims.
Reader Verification
Every numerical claim on the site can be traced to a named reference entry. Reference entries live in /references and include authors, journal, year, DOI or URL where available, and the article(s) the reference is cited in. If a claim on the site is not linked to a verifiable source, please flag it via info@retatrutide.med and we will either add the citation or correct the claim.
Related: Editorial Standards · Medical Reviewer · About