Research Team

How a page gets published

Five functions, in this sequence. A page that fails any of them does not publish; it goes back.

1. 1 Author

   Reads the literature for the compound or the outcome, decides which claims the page will make, and grades each one against the evidence rubric. Every claim is written with its citation attached; nothing is written first and cited afterwards, because that is the order in which citations get invented. Where the honest answer is that no good evidence exists, the author writes that as the finding rather than padding the page.

2. 2 Citation verifier

   Independently re-fetches every identifier on the page and confirms the record exists. Title and year are compared against the registry automatically; the journal the registry returns is put in front of a person to read rather than machine-compared, for the reason given under how we verify citations. This function does not take the author's word for what a paper says; where a summary and an abstract disagree, the abstract wins and the claim is rewritten. Citations that turn out to support only part of the claim attached to them get dropped, and the claim shrinks to what the evidence covers.

3. 3 Primary-document verifier

   Checks every regulatory, approval, enforcement, and anti-doping statement against the primary document itself, never against a review paper's characterization of one. The primary document is the current Prohibited List, the agency communication, or the approved product labelling. This function exists because of a specific failure on this site, described in why verification is not enough: real, correctly quoted papers restated regulatory facts wrongly, and the errors shipped behind citations that checked out perfectly.

4. 4 Adversarial reviewer

   Reads the page trying to break it rather than trying to approve it: looks for the claim that is stronger than its citation, the grade that flatters the evidence, the sentence that would read as instruction to someone determined to take it that way, and the omission that makes an honest page misleading. Findings are recorded and resolved individually. This function is deliberately not the author.

5. 5 Automated compliance gates

   The checks that do not depend on anyone remembering them: no commercial or acquisition pointers, no preparation or administration instruction, no personalized dosing, anti-doping status stated wherever it applies, no dead internal cross-references, a depth floor per page, closed citation keys, and no outbound link that was not derived from one of the page's own identifiers. They run on every build and block a release rather than warning about it. The full list is on the methodology page.

Only then does the page get its "last reviewed" stamp, the date printed near the top of every content page. That date means this whole sequence ran, not that a file was edited. A page more than 12 months past that date fails our build, so nothing can go stale quietly.

Why there is no doctor's name on this site

Peptide Health Lab publishes no individual medical byline on this property. That is a deliberate choice, not an omission we are working to fill.

A clinician's name on an education site that does not treat you is a borrowed trust signal. A real relationship is one where someone examines you, knows your history, and is accountable for your care. A byline transfers that authority onto a page that offers none of those things. It asks you to trust a credential you cannot inspect, in place of evidence you can.

We would rather be checkable than credentialed. Every claim on this site carries a PubMed identifier or a DOI that you can look up yourself, in a database we do not control, and confirm or contradict without our permission. Every evidence grade shows the study designs it was built from. Every regulatory statement was read out of the primary document. That is an authority claim you can audit, and unlike a name, it does not ask you to take anything on faith.

To be unambiguous about what this does and does not mean: nothing on this site is a professional opinion about you, nothing here has been individually cleared for your situation, and no one behind this property is your care provider. The right person to evaluate whether any of this applies to you is a licensed provider who can actually examine you. Our methodology page documents exactly what our process does and does not establish, including where it has failed.

How we handle corrections

If something here is wrong, we want to know. Email [email protected] with the page and the specific claim, and a link or identifier for what you believe the correct answer is, if you have one.

What happens next:

1. 1. We re-check the claim against the primary document rather than against whatever we cited the first time: the paper itself, the current Prohibited List, the approved labelling.
2. 2. If the claim is wrong, we correct it. If the claim is right but the citation behind it does not support it, we treat that as wrong too, and either find the citation that does or narrow the claim to what the evidence covers.
3. 3. The corrected page goes back through the citation verification pass and the automated gates.
4. 4. The page's last-reviewed date moves to the date of the correction, because the old date would otherwise claim we had checked text we had not.

Regulatory, approval, and anti-doping corrections are the ones we most want to receive. Those are precisely the errors our citation checks cannot catch on their own, and the reason the primary-document function above exists.

What would change this

If Peptide Health Lab ever brings a credentialed reviewer into this process, that person will be named, their credential will be stated, and the scope of what they actually reviewed will be stated with it: which pages, which claims, on which date. A credential attached to specific reviewed work is a real signal. A credential attached to a whole website is decoration.

That is a condition, not a plan and not an announcement. Today, no such review has taken place, and nothing on this site should be read as implying otherwise.