EQUATOR Network: Choosing a Reporting Guideline

The EQUATOR Network is an international library of reporting guidelines for health research, not a guideline in its own right. Its name stands for Enhancing the QUAlity and Transparency Of health Research, and its practical function is to answer one question fast: given the study I actually ran, which checklist am I expected to report against, and is there a more specific version than the one I had in mind. It indexes several hundred guidelines along with their extensions, and it tracks which version is current.

Knowing the library exists matters more than it sounds, because the commonest reporting failure is not ignoring a guideline. It is following the wrong one, or following a superseded version, or following a core checklist when a far more specific extension was available. All three are avoidable in about ten minutes.

Why a guideline follows your design, not your field

Reporting guidelines are organised by what you did rather than what you studied. A randomised trial in cardiology and a randomised trial in physiotherapy report against the same checklist, while a trial and a cohort study in the same clinic report against different ones. Authors searching by disease area therefore tend to conclude that nothing applies to them, when in fact something does.

The operative question is the design. Did you allocate an intervention, or observe what happened anyway? Did you collect the data yourself, or use records gathered for another purpose? Are you reporting one patient, a sample, or a synthesis of other people's studies? Each of those turns leads to a different guideline, and the choice is rarely genuinely ambiguous once the design is stated plainly.

Which guideline applies to which design

The mapping below covers the guidelines most researchers need. Each links to our own walkthrough where we have one, and the official wording and files live with the guideline and on EQUATOR.

Two of those rows deserve a note. A protocol guideline and a report guideline are not alternatives, they are used in sequence, which is why trials appear twice and systematic reviews appear twice. And where a coding or agreement statistic is expected, as it is in the qualitative guidelines and in review screening, the coefficient has to match the coding design: our guide to choosing between Cohen's kappa and Krippendorff's alpha covers which applies when.

One reporting guideline in wide use is deliberately absent from the table. ENTREQ, for qualitative evidence synthesis, is covered inside our overview of qualitative evidence synthesis methods rather than as a standalone page.

Extensions usually beat the core checklist

Most of the major guidelines have grown a family of extensions for specific contexts, and an extension is both more useful to you and better received by reviewers. There are extensions covering cluster randomised trials, pilot and feasibility trials, non-inferiority trials, patient-reported outcomes, harms reporting, routinely collected health data, artificial intelligence and prediction models, network meta-analysis, individual participant data, and abstracts as a distinct format.

The practical rule is to search the library for your design, then read the list of extensions attached to the guideline before downloading anything. If an extension fits, cite the extension and say so in the methods. Reporting a cluster randomised trial against the core trial checklist, when a cluster extension exists, invites exactly the criticism you were trying to avoid.

Version matters, and several changed recently

Guidelines are revised, and reviewers compare your paper against whichever version they know. Three recent changes are worth holding in mind. CONSORT was updated to CONSORT 2025, replacing the 2010 statement and moving to 30 items. SPIRIT, the protocol counterpart, was updated to SPIRIT 2025 with 34 items. CHEERS was updated in 2022 and now carries 28 items rather than the 24 of the 2013 version.

Name the version in your methods. Writing that the study is reported in accordance with CONSORT 2025 takes four extra characters and removes an entire class of reviewer query.

A reporting guideline is not an appraisal tool

This is the distinction that causes the most methodological trouble, and it is worth being blunt about. A reporting guideline asks whether a study is adequately described. An appraisal instrument asks whether it was well conducted. The two are independent: a badly conducted study can be reported completely, and an excellent study can be reported poorly.

So a checklist score is not a quality score, and using one as the other is a real error rather than a stylistic preference. When you are appraising studies inside a review, the tools you want are different instruments entirely: AMSTAR 2 for appraising systematic reviews, ROBINS-I for non-randomised studies of interventions, QUADAS-2 for diagnostic accuracy studies, the QUIPS tool for prognostic factor studies, and the Newcastle-Ottawa Scale for cohort and case-control studies. The certainty of the body of evidence is then a third question again, handled by the GRADE framework.

How to use the library in practice

Three steps, in order. State your design in one sentence, without reference to your topic. Search the library for that design and read the extension list before choosing. Then complete the checklist against your final draft, recording page and section numbers rather than ticks, and cite the guideline and version in your methods.

The completed checklist is usually submitted as a supplementary file. It is also the fastest self-review available: an item you cannot point to a page for is, almost always, an item a reviewer is about to ask about.