How Ellum decides status, grades and what we publish
Methodology version 1 · Effective October 7, 2026
Summary
- Status comes from primary regulatory sources and carries the date it was recorded as current.
- Evidence grades aren't shown on Ellum yet. When they are, each one will follow the published rubric below.
- Ellum's pages are built from our regulatory records. No physician has reviewed them yet.
- Ellum's pages show no doses. They never tell you what dose to take.
- Commercial plans never change a status or what a page concludes.
Status rules
Every peptide has one regulatory status, shown as one of the labels below. Each status is set from primary sources: FDA records, the Federal Register, DailyMed labels, ClinicalTrials.gov and the WADA Prohibited List. Where we have recorded the source, the page lists it under References.
Every status carries an as-of date: the date it was recorded as current in its source. The same status, with the same date, appears everywhere Ellum states it, so the Index and the peptide's page never disagree. If the date is old, the status may have changed since.
- FDA-approved (Rx)
- Not FDA-approved for this use
- Not FDA-approved · advisory vote, no FDA decision
- Investigational
- Not FDA-approved
- FDA-approved, discontinued
- Approved outside the US
- Supplement
- Cosmetic
Evidence grades
Ellum doesn't show evidence grades on any page yet. When it does, every grade will follow this rubric (version 1.0.0, dated 2026-10-04) and show the version it was graded under. A change to the rubric gets a new version number.
Each paper gets a grade from its design:
- A: a randomized controlled trial in humans, or a systematic review of such trials, with at least 100 people and a clinical endpoint (how people feel, function or fare).
- B: any other human randomized trial or systematic review, or a non-randomized controlled human trial with at least 30 people and a clinical endpoint.
- C: other controlled human trials, and observational human studies (cohort, case-control).
- D: human case reports and case series, every animal study and every lab (in-vitro) study.
A paper drops one grade (never below D) for a declared conflict of interest tied to the result, or a journal's expression of concern. Papers in journals flagged as predatory are capped at D. Retracted papers are never counted. Narrative reviews give context and aren't graded.
Each peptide gets a grade from its counted papers. Papers from the same research team or sponsor count as one group. The first rule that matches sets the grade:
- A, strong human evidence: at least 2 grade-A papers from 2 independent groups, or at least 3 pre-registered grade-A trials.
- B, moderate human evidence: at least 1 grade-A paper, or at least 2 papers at B or better from 2 independent groups.
- C, limited human evidence: at least 1 paper at B or better, or at least 2 grade-C papers from 2 independent groups.
- D, early evidence: any counted paper (animal, lab or case reports).
- Insufficient: no counted papers.
If the newest paper at or above a peptide's grade is 20 or more years old, the grade drops one step (never below D).
Dosing policy
Ellum's pages show no doses today. When dosing sections are added, they will report what the record shows and never tell you what to take:
- FDA-approved drugs: the doses on the FDA label, cited to DailyMed, with the label's warnings.
- Everything else: doses studied in humans, trial by trial, each with its study, design, number of people, schedule, length and result. If no human dosing data exists, the section will say so.
We will never publish personal doses, dose calculators or how-to steps for preparing or injecting a product beyond an approved label. No physician has reviewed a dosing section, and each one will say so.
Update cadence
- Status: every status shows the date it was recorded as current in its source.
- One record, every page: a status comes from one record, so when it changes, the Index and the peptide's page change in the same update.
- We don't promise a fixed recheck schedule yet.
Corrections
If you find an error, email hello@ellumclinic.com with the page and what's wrong. We check it against the primary source, fix it, and log it. Every correction is listed on our corrections log with the date, the page and what changed.
Medical review status
Physician review is not in place yet. No doctor has reviewed any page on Ellum. When a physician reviews a page, the page will name them, give their credentials and show the date of the version they reviewed. Until then, no page carries a reviewer's name.
AI use
How Ellum uses AI, and what it never does, is set out in How Ellum uses AI. AI never picks a dose, diagnoses, or signs off a page as reviewed.
Editorial independence
Ellum plans to open a clinic and sell a small range of non-prescription products. Neither has any say in a peptide's status, what a page concludes, or whether a page exists. We take no money from any company whose products we cover. Full policy: Editorial independence.
AI-crawler policy
Ellum allows every crawler, including search engines, AI search and citation bots, and AI training bots. Our robots.txt allows all of them, and /llms.txt lists our main pages for AI tools.
GPTBotOAI-SearchBotChatGPT-UserClaudeBotClaude-SearchBotClaude-UserPerplexityBotPerplexity-UserGoogle-ExtendedApplebot-ExtendedCCBot
Other methodologies
The Peptide Source Index will rate pharmacies and sellers with its own published method. It isn't on Ellum yet.