Corpus

Standards

Editorial Policy

Corpus Atlas publishes health content, which means being wrong has consequences. This page sets out the rules every guide is written to — what gets sourced, how confidently a finding may be stated, which authorities take precedence, how often pages are re-checked, and what happens when something turns out to be wrong.

Editorial principles

Six rules govern everything published here. They are applied at the point of writing, not retrofitted at review.

  • Every specific claim is sourced

    Numbers, mechanisms and clinical thresholds are traced to a named primary source — a peer-reviewed study, a systematic review, or a guideline from a recognised body such as NICE, the NHS, the WHO, the AHA or the NIH. General physiological background that appears in any standard textbook is not individually cited.

  • Evidence strength is stated, not implied

    Interventions carry an explicit evidence grade — Strong, Moderate, Emerging or Limited — so a well-replicated finding is never presented with the same confidence as a promising single trial.

  • Education, never diagnosis

    Corpus Atlas explains how the body works and what the evidence supports. It does not diagnose, does not prescribe, and does not replace a clinician who can examine you. Red-flag symptoms are called out explicitly and routed to professional care.

  • Uncertainty is shown, not hidden

    Where the evidence is genuinely contested, guides say so and describe the disagreement, rather than picking the more confident-sounding side.

  • Commercial content is separated and disclosed

    Product comparisons appear only after the educational content, never in place of it, always include honest trade-offs, and never change what the evidence section says.

  • Corrections are made openly

    When a page is found to be wrong it is corrected and its review date updated. Substantive corrections are noted on the page rather than quietly edited away.

How evidence strength is graded

Every intervention discussed on the site carries one of four grades. The grade describes the state of the evidence, not how promising the idea sounds.

Strong
Supported by multiple well-conducted randomised trials, systematic reviews or meta-analyses, and reflected in major clinical guidelines.
Moderate
Supported by consistent observational evidence or a smaller number of randomised trials; directionally reliable but with meaningful residual uncertainty.
Emerging
Early trials or strong mechanistic rationale, not yet replicated at scale. Plausible but not yet something to act on with confidence.
Limited
Weak, conflicting or largely preclinical evidence. Included for completeness, and explicitly flagged as not yet actionable.

Preferred sources

These are the authorities treated as definitive when guidance conflicts, in order of precedence.

  1. 01
    Cochrane Library

    Systematic reviews — the highest tier of synthesised evidence.

  2. 02
    NICE Guidance

    UK clinical guidelines and evidence summaries.

  3. 03
    NHS Health A to Z

    Patient-facing clinical reference for UK practice.

  4. 04
    World Health Organization

    Global public-health guidance and definitions.

  5. 05
    PubMed / MEDLINE

    Primary peer-reviewed literature.

  6. 06
    National Institutes of Health

    Office of Dietary Supplements fact sheets and NIH research summaries.

  7. 07
    American Heart Association

    Cardiovascular guidelines and scientific statements.

How often pages are re-checked

Different content ages at different rates, so the review interval differs by section. Any page can be pulled forward in the queue when guidance changes before its interval is up.

Conditions and symptoms
Reviewed every 12 months, or sooner when major guidance changes.
Nutrients and supplements
Reviewed every 12 months against current intake reference values.
Body part and system guides
Reviewed every 18 months; anatomy changes slowly, guidance does not.
Exercises and technique guides
Reviewed every 24 months.
Calculators and tools
Formulae verified against their source publication every 12 months.

What a review actually involves is described on the medical review process page.

Corrections

When a page here is wrong, the fix is a published change rather than a quiet edit. Corrections are made openly: the page is corrected, its review date is updated, and substantive corrections are noted on the page itself rather than removed from the record.

  • A substantive correction — one that changes what a reader would do or believe — is noted on the page, so the change is visible rather than silently absorbed.
  • Typographical fixes, broken links and clarifications that do not change meaning are made without a note.
  • Any corrected page has its review date updated, so the freshness signal reflects the real check.
  • Reports of factual errors go to the front of the editorial queue, ahead of new content.

To report an error, use the details on the contact page. Include the page URL and the specific claim you believe is wrong.

Common questions

Does every sentence on Corpus Atlas carry a citation?
No, and deliberately so. Numbers, mechanisms and clinical thresholds are traced to a named primary source — a peer-reviewed study, a systematic review, or guidance from a body such as NICE, the NHS, the WHO, the AHA or the NIH. General physiological background that appears in any standard textbook is not individually cited, because a citation on every clause makes the genuinely load-bearing ones harder to spot.
What happens when two guidelines disagree?
Corpus Atlas works down a fixed order of preferred sources, starting with Cochrane systematic reviews and NICE guidance, then the NHS, the WHO, PubMed-indexed primary literature, the NIH and the American Heart Association. Where the disagreement is real rather than a matter of emphasis, the guide describes the disagreement instead of picking the more confident-sounding side.
How do I report a factual error?
Email the editorial team with the page URL and the specific claim you believe is wrong, ideally with a source. Corrections are prioritised over everything else in the queue, and substantive ones are noted on the corrected page.

Related

See authors and reviewers for who stands behind each guide, and the medical disclaimer for the limits of what this site can tell you.