Who decides, on what basis, and what happens when we are wrong.
This page answers one question an evidence site has to answer out loud: who decides, on what basis, and what happens when they get it wrong.
Who writes and who reviews
EBL is written by Arnold Sandlow, D.C., who is responsible for every grade on the site. His credential and license status appear on every page, and a build check confirms that: no page may carry the byline without the credential, and the license claim on the site must match the license that is actually active.
There is currently no independent clinical reviewer. That is the most important limitation on this page and it is stated rather than implied. Every automated check on this site verifies that a citation exists, is not retracted, is attached to the claim it belongs to, and does not exceed what its endpoint The outcome a trial measures to decide whether the treatment worked, such as heart attacks, blood pressure or a symptom score. allows. None of them verifies the interpretation. Where a judgment has been made by a language model reading adversarially rather than by a person, the record says so in those words.
How a grade is decided
The rules are published in full on the methodology page, and they are enforced by the build rather than by memory. Animal and mechanistic data never enter a grade. Observational Research that watches what people already do rather than assigning a treatment. It can show that two things travel together, not that one causes the other. evidence provides context and never independently earns a letter. The endpoint type sets the ceiling: a surrogate biomarker A measurable sign in the body, such as a blood test value, used to track a condition or the effect of a treatment. or a self-reported scale cannot reach Grade A no matter how large the trial.
A grade is assigned or moved only after a documented search for the human randomized evidence on the exact intervention and use, and the search is stored with the record: the query, the database, the date it was run, how many papers came back and how many tested the graded claim. A missing citation is not a failed trial. When a record is found to cite less than the literature holds, the record is marked for evidence review and the letter stays where it is until the search is done; when the search finds the trial the letter was missing, the citation is added, and when it finds that no trial of the exact claim exists, the absence is stored the same way and re-run on every deploy. Since October 2026 the build refuses a moved letter that carries no search record, and it counts, on every deploy, the graded records that still have none.
Every graded record carries a recorded judgment on whether the cited papers actually support the claim at the stated strength. The current coverage, and what remains open, is published on the evidence integrity page and regenerated on every deploy.
Money, and where it is not
EBL earns nothing on any item it grades, in any account, at any tier. That is not a policy statement, it is a build check: a deploy fails if any commission is claimed on anything carrying a letter. Active commercial relationships are named individually on the disclosure page rather than described as a category.
Grades are set from published trials and posted before any commercial relationship exists. There is no commission a letter could be worth moving for, because on graded items there is no commission at all.
When a grade changes
A letter cannot move silently. Every change is dated, published, and carries its reasoning, including the changes that lowered a grade after a closer reading. The log is public and the build refuses to deploy a moved letter that has no entry.
New trials, retractions and published corrections are checked against every cited paper on every deploy. A grade resting on a withdrawn paper fails the build.
Corrections
If something here is wrong, say so. Send it through the contact page with the page and, if you have it, the paper. Claims that turn out to be wrong are corrected and the correction is published rather than quietly edited.
Claims this site has examined and rejected are recorded, and a build check searches the published site on every deploy to make sure none of them has crept back. That register exists because a corrected claim reappearing is a real failure mode, not a theoretical one.
What this page does not claim
No editorial board. No peer review. No institutional affiliation. One author, an automated verification chain, and a recorded judgment trail that anyone can check and anyone can dispute. Saying that plainly is worth more than an advisory list that does not do any work.
It is not a quality measure. A longer reference list is not better evidence, and a page can cite forty papers and still rest on none of them. The grade on any entry here comes from what its trials actually tested and in whom, never from how many were attached to it. Counting references measures breadth, and breadth is the least interesting property evidence has.
It is not fixed. The figure moves whenever a paper is added, retired or superseded, which is why every statement of it here carries the date it was generated. Two documents written a week apart may honestly carry different numbers, and saying so is more accurate than forcing them to agree.
It is not a claim that every paper was read the same way. What each citation was checked for, and what remains unverified, is set out separately in the citation audit, which lists every defect found in this site's own citations rather than only the ones found in other people's.