EDITORIAL POLICY & GOVERNANCE· VERSION 1.0·2026

What publishes here, and what does not

Most medical reference sites ask you to trust them. This one asks you to check. Every claim on this site carries a source, every source carries an identifier, and every identifier has been verified against the issuing registry before the claim was allowed to appear. This page documents the rules that produce that — including what we do when we get something wrong, which we have, and which is recorded rather than quietly repaired.

The publication gate

Nothing reaches a reader by being written. It reaches a reader by passing every one of these, in order. Content that fails any step remains visibly marked as a draft.

Identifier verification. A paper must carry a DOI confirmed against Crossref's registry — not transcribed from a citation, not inferred from a pattern, but retrieved. Documents without DOIs — guidelines, statutes, official reports — must carry a canonical URL on the issuing body's own domain.
Metadata agreement. The stored title, authors, year, and journal must match the registry record. Disagreement is treated as a defect in our record, not the registry's.
Grounded content. Summaries are written only from the retrieved abstract or the primary document. An entry with no retrievable abstract carries no summary rather than an invented one.
Physician review. A qualified reviewer reads the entry and signs it. Machine-drafted content is labelled as such and cannot publish on its own.
Publication. Only then. And the gate is enforced in the database, not by convention — the publish operation itself tests these conditions and refuses rows that fail.

That last point is the one we would ask a skeptic to test. The rules are not a statement of intent; they are executable conditions attached to the act of publishing. Entries have been held back by them, including entries we wanted to publish.

Evidence grading

Grades describe the strength of the evidence behind a source, not its importance. A consensus document can matter enormously and still be consensus.

AMultiple events or populations, adequate denominators, findings reproduced independently.

BSingle well-conducted study with clear denominators, or consistent findings across small series.

CSingle event, descriptive series, or data without a stated denominator. Useful, not generalizable.

CONSENSUSExpert or organizational position not derived from primary data.

REGULATORYA binding legal requirement. Graded separately because law is not evidence — a statutory ratio may have no empirical basis and still be mandatory.

Editorial structure

Roles are defined below. Seats marked unseated are not yet filled, and we say so rather than leaving the impression of a board that does not exist. No content publishes without a named reviewer, whatever the state of the roster.

Editor-in-Chief: Owns the publication gate and final sign-off on every published entry, synthesis, and case. Accountable for corrections. Seated — Brian Adkins, MD · University of Kentucky.
Domain editors: One per evidence domain — prediction, heat, crowd dynamics, toxicology, cardiac, mass-casualty, medical direction and law, concerts, endurance, pediatric, motorsports, communications. Responsible for the state-of-the-evidence synthesis in their domain and for flagging when new literature changes it. unseated
Regulatory editor: Maintains the 51-jurisdiction atlas: monitors amendment and repeal, re-verifies against official sources, and dates every jurisdiction's last review. unseated
Methods reviewer: Audits the library for integrity defects — wrong identifiers, mismatched metadata, duplicate or unsupported records — independently of the people who wrote the entries. unseated
Curation: Literature surveillance, structured extraction, drafting from primary sources, and maintenance of the verification pipeline. Machine-assisted; every output labelled and gated. Operating.

Corrections policy

A reference work that has never published a correction is either very new or not looking. We log ours publicly, with the date, what was wrong, how it was found, and what changed. Corrections are not edits — the record of the error stays.

Critical: A citation that does not identify a real, retrievable work; an entry conflating two or more sources; a duplicate record. Action: the entry is withdrawn from public view immediately, not repaired in place, and the withdrawal is logged.
Major: Wrong author, year, journal, or identifier, where the underlying work is real and identifiable. Action: corrected, dated, and logged.
Minor: Formatting, truncated author lists, grade classification. Action: corrected; logged in aggregate.
Unresolved: We can see a record is wrong but cannot determine the right value from an authoritative source. Action: the entry is flagged as unresolved on its own page. We would rather show a reader an acknowledged gap than a confident error.

Read the corrections log →

Disclosure & independence

This site accepts no advertising, no sponsorship, and no payment for placement. No entry appears, and no recommendation is written, in exchange for anything. Where a reviewer has a relationship relevant to a topic — employment, consultancy, event medical direction, expert testimony — that relationship is disclosed on the content it touches.

Machine assistance is disclosed rather than hidden. A substantial share of the drafting, extraction, and verification here is machine-assisted. Every such output is labelled, none publishes without human review, and no summary is written from anything other than a retrieved source document. We think that combination is more trustworthy than an unlabelled human-only claim, and we would rather be judged on the gate than on the byline.

Versioning & how to cite

Content here changes as the literature and the law change — four US jurisdictions altered their event-medicine regulations inside a single recent twelve-month window. Every substantive page therefore carries a version and a review date, and we ask that citations include the date accessed.

Mass Gathering Medicine Institute. [Page title]. Version [n], [year]. Available at: [URL]. Accessed [date].

Syntheses and the regulatory atlas are versioned independently of the site. A superseded version is retained and reachable, so that a citation made last year still resolves to what it said last year.

What this site is not

It is not legal advice; the regulatory atlas is a research reference and statutes change. It is not a clinical protocol; nothing here overrides medical direction or local standing orders. It is not accredited for continuing education credit. And it is not complete — the evidence base in this field has substantial gaps, several of which we document explicitly rather than paper over, because a clinician who knows the limits of the evidence makes better decisions than one who has been handed false certainty.