Editorial policy: how retainfo.co.uk is written

This page is the contract behind every page on the site: where the facts come from, how we show our working, the seven rules we never break, and what happens when we get something wrong or the evidence moves.

Where our facts come from

Not all sources are equal. When two disagree, the one higher up this list wins.

1. Peer-reviewed trial papers and regulators. Papers in journals such as the New England Journal of Medicine, The Lancet and Nature Medicine; ClinicalTrials.gov registry entries; MHRA, FDA, EMA and other government publications; UK legislation; NHS and NICE guidance.
2. Company announcements. Eli Lilly press releases and investor statements. These are usually accurate but incomplete, so we say when a figure is "topline" and not yet peer-reviewed.
3. Reputable journalism and professional bodies. For example ABC News (Australia) investigations, The Pharmaceutical Journal, CNBC and NPR for corporate news.
4. Never. Sellers of peptides or medicines, "peptide wiki" sites, forums, social media posts and marketing blogs. We may describe what people say in those places on the community reports pages, but we never link to them or treat them as fact.

How citations work

Every factual claim carries a numbered footnote. Click it and you go to the reference at the bottom of the page, which links to the original document. Each page lists only the sources used on that page, and every source lives in a single site-wide register with its title, publisher, date and address. A page cannot be published if it cites a source that is not in that register. We also keep one verified fact sheet for the whole site, so the same number appears the same way on every page. If two pages disagree, one is wrong; please tell us.

How we report numbers

Trial results can be counted in more than one way, and the difference matters.

• We give the treatment-regimen figure first (everyone randomised, whether or not they stayed on the drug), then the efficacy figure (people who stayed on treatment as intended), and say which is which. The efficacy figure is always larger and is the one headlines quote. See estimand.
• We always give the placebo figure next to the drug figure, because placebo groups lost weight too.
• We say "average", not "up to", unless we are quoting a maximum and say so.
• Trial figures describe the trial product only, never a vial sold online.

How to read a trial walks through this with worked examples.

The seven rules, in plain English

1. We give no instructions on using retatrutide. No doses, no dose steps, no preparation, no injection technique, no calculators. Trial dose levels appear only as trial-design facts on clinical-trials pages. Why: any use outside a trial is use of an unlicensed product of unknown content, and helping with it would work against the purpose of the site.
2. We never help anyone obtain it. No vendor names, no prices, no comparisons, no paid links, no "how to tell a safe one". Why: under the Human Medicines Regulations 2012 an unlicensed medicine cannot be sold, supplied or advertised for human use in the UK.^[1] The MHRA's seizures show what such products actually contain.^[2]
3. No testimonials, no photographs of results, and no ranking of retatrutide against licensed medicines as a recommendation. Why: individual stories are persuasive and unrepresentative, and a licensed medicine with a known safety profile should never be made to look like the worse choice.
4. The same status banner appears on every page. Why: nobody should land on a deep page and miss the most important fact: this drug is not licensed, and nothing sold as it is the trial product.
5. Every factual claim cites a source. Why: you should not have to trust us. You should be able to check.
6. We always route to real help: GP, NHS 111, 999, the MHRA Yellow Card scheme and NHS weight-management services. Any page that discusses harms carries a red-flag box saying when to seek help. Why: harm reduction means getting people to care, not just informing them.
7. The automated checks must pass. Before a page goes live, a script scans it for the phrases above, links to banned domains, missing sources and missing red-flag boxes. Why: rules that rely on memory get broken.

Review cadence

Every page shows a last-reviewed date. We re-check each page at least every 90 days. We re-check sooner, usually within days, when a trial reports, a regulator makes a decision, an enforcement action is announced or a safety alert is issued. The news section records those events.

Corrections

A correction is triggered by a factual error, a figure that no longer matches its source, a source that has moved or been withdrawn, a broken link, or wording that could be read as use guidance or sourcing help. We acknowledge every request within five working days. Errors of fact are fixed as soon as they are confirmed, and the page gets a dated note describing what changed. Typos and formatting are fixed quietly. Contact details are on the about page.

Conflicts of interest

None. Nobody on the team is paid by, invested in, or employed by any company that makes or sells medicines, peptides or weight-loss products, and no business connected to anyone on the team is linked from the site. The build check that blocks seller domains also blocks every owner-connected domain, so the rule holds even if someone forgets it.

When the evidence changes

Topline figures are sometimes revised when the full paper appears. When that happens we update the number, cite the newer source, and note the change on the page. If a finding is reversed or a safety signal emerges, we say so plainly rather than softening earlier pages.

How we use AI tools

Drafting assistance from AI tools may be used to help structure and write pages. Every figure, quotation and citation is checked against the primary source by a human reviewer before publication, and the reviewer, not the tool, is responsible for the page. No page is published unread.