Policy introduced: 8 September 2026. This policy sets the standard for new articles and substantive updates. It does not certify that every older article has completed medical review.
Choosing evidence
Clinical recommendations should be supported by current, relevant guidance from recognised health authorities or professional organisations, and by suitable research evidence. Systematic reviews and well-conducted studies should be considered in context, including their limitations.
A single study, laboratory finding or animal experiment must not be described as proof that a treatment is safe or effective in people. Preprints and other findings that have not completed peer review should be identified as such.
Making claims traceable
Links should lead to the specific guidance, paper or report supporting the nearby claim. Authors must check that the source actually supports the wording, population and outcome described. A reference list alone does not validate every statement in an article.
Where useful, include the source title, organisation or authors, publication or update date and a DOI or stable link. Evidence search dates should reflect research actually completed; they must not be automatically refreshed.
Uncertainty and commercial sources
Explain meaningful uncertainty, conflicting findings and differences between patient groups. Manufacturer and commercial pages may describe their own products, but should not be the sole basis for independent medical claims.
Images and illustrations
Identify illustrations as illustrations and patient photographs as photographs. AI-generated images must not be presented as patient cases, diagnostic evidence or images taken during a clinician’s own practice. Captions and explanatory text must describe the image’s limitations.
Readers can report a missing, broken or misleading citation through our Contact page.