On the record
Every reference work gets things wrong. What separates the ones worth trusting is whether a reader can find out. When we publish an error it goes on this record, dated, with the wording we published, what was wrong with it and what replaced it, and it stays here after the fix.
Corrections to date
None yet. When a published page turns out to be wrong, the correction will be listed here with the wording we published, what was wrong with it and what the page says now.
What counts as a correction
Pages change all the time, for one of three reasons, and only the first is a correction. Keeping the three apart matters: fold routine updates into the record and a real error has somewhere to hide.
- A correction. We published something inaccurate, such as a wrong attribution, a misstated result, or a claim the evidence failed to support. It goes on the record above.
- An evidence update. The literature moved, with a new trial or a changed grade, and the page moved with it. What it said was accurate when written, so the change lives in the registry’s dated snapshots instead.
- Catalogue growth. We added coverage of something new. A gap we fill was simply a gap, and what we have yet to catalogue has its own page.
How a correction is made
A fix travels the same road as any other change, however embarrassing the error. A change to a graded claim is re-checked against the primary source, passes the same publication gates as a brand-new claim, and lands in a hash-chained snapshot that every page on the site reads. You can check the snapshot the page you are reading was built from.
Sometimes the error sits upstream, in a paper we cite. Each study record carries an integrity status, and the registry is checked against retraction and erratum notices. When a retraction or expression of concern lands against a cited study, a person decides what happens next, because removing evidence deserves at least as much scrutiny as adding it.
Who is responsible
Every release has a person behind it. Someone authorises each one, and the audit chain records who did so next to the snapshot it produced, so each published state traces back to whoever signed it. Because the chain is hash-linked, altering any entry after the fact would break every later link in it.
Peregrine Biopharma funds the platform and holds no editorial authority. Grades and conclusions follow the evidence, and that independence is written into the funding agreement. A public record of our own mistakes depends on it.
Spot something wrong? Tell us, whether it is a fact or the grade we put on a claim. For anything that concerns harm, use Safety and reporting.