Reporting guidelines: which one applies, and how to use it properly
A checklist submitted with the manuscript is not the same as a study reported completely. Used at the end, guidelines catch omissions; used at the start, they prevent them.
Reporting guidelines specify the minimum a reader needs in order to judge a study. They are not quality scales: following CONSORT does not make a trial good, it makes a trial legible.
Which one applies
- Randomised trials: CONSORT, with extensions for cluster, non-inferiority, pilot and other designs.
- Systematic reviews and meta-analyses: PRISMA 2020, with PRISMA-S for the search and PRISMA-P for the protocol.
- Observational studies: STROBE, with RECORD for those using routinely collected health data.
- Diagnostic and prognostic accuracy: STARD.
- Prediction model development or validation: TRIPOD.
- Case reports: CARE. Qualitative research: COREQ or SRQR. Economic evaluations: CHEERS.
The EQUATOR Network maintains the catalogue, and checking it takes a minute that many manuscripts never spend.
Use it before you need it
The usual pattern is to complete the checklist at submission because the journal asks for one. At that point every gap is a problem to be written around, and some cannot be fixed at all — a trial that did not report its randomisation method at the time cannot reconstruct it credibly two years later.
Read the guideline while writing the protocol. Each item names a decision the study has to make. A protocol built against STROBE has already decided how it will define exposure, handle missing data and address confounding, because the checklist asked.
A completed checklist is not complete reporting
Page numbers written next to items are evidence that someone filled in a form. Reviewers and editors increasingly check, and the gap between "item addressed" and "item present on that page" is where a submission loses credibility fastest.
Fill it in honestly, including the items you did not do. "Not applicable" with a reason is a legitimate entry; a page number pointing at a paragraph that does not contain the item is not.
Where they help most
The items people skip are the ones that determine whether a result can be believed: how the sample size was arrived at, how missing data were handled, which analyses were pre-specified and which were not, how confounders were selected, and what the limitations actually are.
Those are also the items reviewers ask about. Reporting them properly the first time removes a revision round.
Frequently asked questions
Which reporting guideline should I use?
CONSORT for randomised trials, PRISMA for systematic reviews, STROBE for observational studies with RECORD for routinely collected data, STARD for diagnostic accuracy, TRIPOD for prediction models. The EQUATOR Network maintains the full catalogue.
When should you use a reporting guideline?
At the protocol stage, not at submission. Each item names a decision the study has to make, so a protocol built against the checklist has already decided how it will define exposure, handle missing data and address confounding.
Is a completed checklist the same as complete reporting?
No. Page numbers next to items show that a form was filled in. Reviewers increasingly check, and a page number pointing at a paragraph that does not contain the item is where a submission loses credibility fastest.