A Medical Content Review Checklist: The One I Actually Use

Most published checklists for reviewing health content are useless, and they are useless in the same way: they list virtues no one disputes. "Ensure accuracy." "Check the sources." "Use plain language." Nobody disagrees, and nobody is helped, because the list never says what a failure actually looks like or how you would notice one. A checklist earns its place only if each line points at a specific mistake it is designed to catch.

So here is the pass I genuinely run, in the order I run it, with the failure each item exists to catch. It is not a generic template scraped from twenty others. It is the working sequence I have arrived at, and it has the texture of something used rather than something composed.

Why order matters

I run these in sequence for a reason. The early items are about whether the piece is true; the later ones are about whether it is safe, current and honestly attributed. You want the cheap, decisive checks first — a single fabricated dose can end a review before currency or disclosure ever become relevant. Working in order also stops the most seductive failure of all: being so charmed by fluent prose that you forget to verify the things the prose is asserting.

1. The claims audit

The first pass ignores style entirely and lists every factual claim the piece makes — every "X causes Y", "most people experience Z", "the evidence shows". Each claim is then marked as supported, narrower-than-stated, or unsupported. Most weak content is not full of lies; it is full of claims that are true in a narrow sense and have been quietly widened.

The failure it catches: a sentence reading "this condition usually resolves on its own", lifted from a study of mild cases, applied to a reader whose presentation is anything but mild.

2. The dosing and numbers check

Every number in the piece — dose, frequency, threshold, percentage, risk figure — is treated as a claim to be verified against a current reference, not a detail to be trusted because it reads plausibly. Numbers carry an authority that prose does not, which is exactly why a wrong one does more damage.

The failure it catches: a "twice daily" where the formulary says once, or a risk stated as a bare percentage with no baseline — "increases risk by 40%" with nothing to multiply it against.

3. The currency check against guidelines

Each clinical claim is checked against present guidance, not against whatever was true when the source was written. In UK content that usually means NICE, the relevant royal college or specialist society, and the latest safety position from the regulator. A claim does not need to be false to fail here; it only needs to be out of date.

The failure it catches: an article naming a first-line treatment that was first-line three guideline revisions ago and is now reserved for second-line use.

4. The safety-netting check

Does the piece tell the reader, in specific terms, what would mean stop and seek help now? Safety-netting is the line a clinician adds by reflex and a writer almost never adds unprompted, because nothing about writing rewards the unsatisfying caveat. Its absence is the single most common serious flaw I find.

The failure it catches: a thorough, calm article about abdominal symptoms that never once names the features that should send someone to be assessed the same day.

5. The population caveats check

Who is this true for, and who has been silently assumed not to exist? Efficacy and safety claims usually come from defined populations — particular ages, particular health states, people not also pregnant, not frail, not on other drugs. Content that reports a finding as universal has performed a quiet sleight of hand.

The failure it catches: "this is safe in pregnancy" stated flatly, when the underlying evidence covers one trimester, or one indication, or says no such thing at all.

6. The sourcing quality check

Not are there sources, but what kind. A piece that cites guidelines and primary literature is in a different class from one that cites other websites citing other websites. I follow a sample of the references back to where they actually lead, because a citation that exists is not the same as a citation that supports the sentence attached to it.

The failure it catches: a confident statistic footnoted to a source that, when you open it, is a press release summarising a study that found something narrower.

7. The conflict and disclosure check

Who produced this, what relevant expertise do they have, and is there a commercial interest shaping what the piece does and does not say? Content about a treatment, written by a party that sells it, can be accurate sentence by sentence and still misleading in what it chooses to emphasise and omit. The check is for the tilt as much as for the disclosure.

The failure it catches: a balanced-reading explainer that mentions every benefit of an intervention and, on inspection, was produced by its provider — with the disclosure absent or buried.

8. The "reviewed by whom" check

The last item turns the lens on the review itself. Does the named reviewer's expertise actually match the content? A generalist sign-off on dense pharmacology is not a review; it is a signature. This item is partly why I keep the list — to make sure I am the right person to be the name on a given piece, and to decline when I am not. (My companion piece on what a reviewer actually does goes deeper into this.)

The failure it catches: my own name on content I am not the right clinician to vouch for — the failure most worth catching, because it is mine.

Practical takeaways

• A useful checklist names the failure each item catches; one that only lists virtues helps nobody.
• Run the cheap, decisive checks first — a single fabricated number can end a review before currency or disclosure matter.
• The most common serious flaw is not a false statement but a missing safety-net and a silently assumed population.
• Sourcing quality is about what kind of source, followed back to where it actually leads — not merely whether citations exist.
• The final, uncomfortable item turns the lens on the reviewer: am I the right clinician to put my name on this?

What this doesn't mean

It doesn't mean a piece that clears every item is guaranteed correct — evidence shifts, and reviewers miss things. Nor does it mean every article needs all eight checks applied at full depth; a short wellbeing explainer carries less risk than a medication piece, and proportion is part of the judgement. The list is a floor for the work, not a substitute for it.

A closing thought

The reason I keep returning to a written list rather than trusting experience is that experience is exactly what makes you skim. The more familiar a topic, the more confidently your eye slides over the sentence you should have stopped on. A checklist is not there to teach the reviewer what they don't know. It is there to stop a competent reviewer from missing what they do.

Further reading and sources

• Patient Information Forum — PIF TICK criteria for trustworthy health information
• NICE — guidance library (the currency benchmark for UK content)
• British National Formulary (BNF) — the reference for verifying UK doses and interactions
• MHRA — Drug Safety Updates
• General Medical Council — Good Medical Practice (honesty and clarity in communication)
• International Committee of Medical Journal Editors (ICMJE) — recommendations on editorial standards and disclosure