Science and health story verification checklist

What to take away

• Name the exact paper, version, protocol, registration, and regulatory record.
• Check the research question, design, participant flow, outcome changes, and analysis before writing conclusions.
• Reproduce central numbers and report absolute effects, uncertainty, and harms.
• Separate research publication, trial registration, and product approval.
• Review every headline and visual against the narrowest material limit.

Use this checklist for health studies, medical products, epidemiology, laboratory findings, and scientific images. Add reviewer names, dates, results, and evidence locations. A checked box should mean the underlying document was inspected, not that someone recognized the publisher.

Study identity and version

• Title, authors, institution, journal or repository, DOI, and date recorded
• Conference abstract, preprint, accepted manuscript, final paper, and later versions distinguished
• Correction, expression of concern, and retraction status checked
• Paper, supplement, protocol, registration, and analysis plan saved
• Institutional release and author publicity kept separate from the paper
• Version used by the story identified in the reporting file

Research question and design

• Population, intervention or exposure, comparator, outcome, and period written plainly
• Laboratory, animal, observational, experimental, diagnostic, qualitative, or synthesis design named
• Design can answer the question asked in the story
• Recruitment and selection process examined
• Random assignment, concealment, masking, and control condition recorded where relevant
• Potential confounding, selection, reverse causation, and measurement error considered

Protocol and outcome audit

• Primary and secondary outcomes copied from the earliest plan
• Outcome definitions, instruments, and time points match the report
• Eligibility, enrollment goal, stopping rule, and analysis population compared
• Amendments dated and explained
• Subgroup and exploratory analyses labeled
• Unreported planned outcomes queried

Participant and sample flow

• Screened, eligible, assigned, treated, followed, and analyzed counts reconciled
• Exclusions and missing data explained
• Attrition compared across groups
• Sample size and power tied to the main question
• Study population matches the people named in the story
• Cell, animal, or model findings are not described as demonstrated human outcomes

Statistical result

• Main calculation reproduced from reported inputs
• Events and denominators retained
• Absolute and relative effects both assessed
• Confidence interval or other uncertainty carried into the story
• Statistical significance separated from clinical importance
• Multiple comparisons and subgroup claims treated cautiously
• Model assumptions and sensitivity analyses reviewed

Body of evidence

• Earlier studies and current evidence synthesis found
• Consistency, replication, and contrary results described
• Study type, quality, sample size, and relevance assessed across the body
• One new paper is not presented as replacing all earlier evidence without support
• Mechanistic plausibility is not substituted for demonstrated patient benefit

FDA's framework for the scientific evaluation of health claims assesses study type, methodological quality, sample size, relevance, replication, consistency, and the totality of the evidence. The guidance addresses food-label health claims, but those review dimensions are also useful checks against building a news conclusion from one favorable result.

Risk, benefit, and harm language

• Baseline risk and absolute change reported when available
• Follow-up duration stated
• Serious and common adverse events reviewed
• Withdrawals due to harms checked
• "Safe" replaced with a bounded description unless the wider record supports it
• No-result and uncertain-result language used accurately
• Patient outcome distinguished from biomarker or surrogate outcome

Funding and control

• Funders, sponsors, employers, patents, consulting, and relevant holdings disclosed
• Study design, data custody, analysis, manuscript control, and publication rights identified
• Author and sponsor comments checked against source documents
• Conflicts treated as risk information, not proof that a finding is false
• Independent expert has no material undisclosed tie to the work

Product and regulator status

• Product identity, manufacturer, formulation, dose, route, and jurisdiction matched
• Intended use and population matched to the regulator record
• Approved, authorized, cleared, registered, listed, and investigational used precisely
• Trial registration not described as approval or endorsement
• Off-label study or use identified accurately
• Regulatory date is current for publication

The FDA's page on product approval explains that premarket requirements differ by product category: new drugs and some devices require review, while cosmetics and dietary supplements generally do not receive the same prior approval. Verify the category before attaching an FDA verb to a product.

Images, charts, and data displays

• Figure traced to the paper and original data when available
• Cropping, contrast, splicing, duplication, and labels examined in context
• Chart denominators, units, uncertainty, and baselines shown
• Micrograph, model output, and illustrative stock image labeled correctly
• Finding an anomaly is not reported as misconduct without investigation
• Alt text and caption carry the same qualification as the story

Publication review

• Headline, deck, body, pull quote, caption, metadata, and social copy agree
• Causal verbs match the design
• Population and period remain visible
• Material limits appear near the finding
• Medical advice and evidence reporting remain separate
• Correction plan covers all formats and distribution surfaces

Common questions

Must every item be checked for a short news brief?

The depth can vary, but study identity, design, central result, uncertainty, conflicts, current status, and claim wording cannot be skipped safely.

Does FDA registration mean FDA approval?

No. The meaning depends on the product category and program. Use the specific agency record and exact regulatory term.

Is an image discrepancy proof of fabrication?

No. Preserve the image, compare originals, ask the authors and journal, and distinguish a visible anomaly from a finding about cause or intent.

What if the study conflicts with a systematic review?

Explain the difference in population, intervention, outcome, design, and date. Do not select the newer result solely because it is new.