What is the COREQ checklist?

The COREQ checklist is a 32-item reporting guideline for qualitative research that collects data through in-depth interviews or focus groups, published in 2007 by Allison Tong, Peter Sainsbury and Jonathan Craig. COREQ groups its items into three domains: research team and reflexivity (items 1–8), study design (items 9–23), and analysis and findings (items 24–32).

COREQ stands for Consolidated Criteria for Reporting Qualitative Research. The statement appeared as Tong A, Sainsbury P, Craig J, “Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups”, International Journal for Quality in Health Care 2007;19(6):349–357, and is hosted by the EQUATOR Network. This page covers what each of the 32 items requires, how the checklist was assembled and why that assembly has been challenged, which items authors most often leave unreported and by how much, how a reader detects non-compliance in a manuscript, what reviewers write when COREQ has been mishandled, and where COREQ’s structure sits badly with the methods it is applied to.

What COREQ specifies, and what it does not

The COREQ checklist specifies what an interview or focus group study must state in its manuscript. Tong, Sainsbury and Craig wrote that the checklist was intended to “promote complete and transparent reporting among researchers and indirectly improve the rigor, comprehensiveness and credibility of interview and focus-group studies” — a claim about reporting, with rigour named as an indirect consequence rather than the object of measurement.

Three consequences follow, and each is routinely ignored in practice.

COREQ is not a quality appraisal instrument. A study can answer all 32 items completely while describing a poorly conceived sampling strategy and an unconvincing analysis. Reviews nonetheless use COREQ as a scoring tool: de Jong and colleagues, in a meta-review of 1,695 qualitative reviews published in BMC Medical Research Methodology in 2021, found 17% of reviews published after September 2007 used COREQ to appraise the reporting quality of their included studies, a use the original checklist was not built for.

COREQ contains no item on research ethics. There is no item requiring a statement of ethics committee approval, informed consent, participant compensation or data protection, and no item requiring a data availability statement. Journals that treat a completed COREQ form as covering ethics and sharing are covering neither; those obligations sit in ethics and data-sharing reporting instead.

COREQ applies to interviews and focus groups. Participant observation, ethnographic fieldwork, document analysis and qualitative evidence synthesis fall outside its scope, and several items become unanswerable when it is forced onto them.

The 32 items in full

The COREQ item names below are the checklist’s own short labels; the requirement column is condensed. Domain scores run 0–8 for domain 1, 0–15 for domain 2 and 0–9 for domain 3, which is the scoring convention used by reviews that appraise studies with it.

Items 1 to 8 are the domain most authors treat as optional and most reviewers treat as diagnostic. Nothing in that domain concerns the participants; every item concerns the person who asked the questions, on the premise that in interview research the researcher is an instrument whose properties belong in the methods section the way an assay’s properties would.

How COREQ was assembled, and the replication that failed

Tong, Sainsbury and Craig built COREQ by consolidation rather than by consensus. They searched Cochrane and Campbell protocols, Medline, CINAHL, systematic reviews of qualitative studies, the author and reviewer guidelines of major medical journals, and the reference lists of relevant publications; compiled 76 items from 22 existing checklists; then removed duplicates and items that were ambiguous, too broadly defined, or impractical to assess, arriving at 32.

That method is the checklist’s most contested feature, and the challenge is specific rather than rhetorical. Buus and Perron attempted to replicate the four development phases and published the attempt in the International Journal of Nursing Studies in 2020 (102:103452). They reported that they “were rarely able to replicate them completely because of uncertainty about the actual procedures”, and concluded that COREQ’s items “were not credible because the coding processes were not trustworthy and because they decontextualized original checklist items and significantly distorted their meaning in COREQ’s final list of items.”

Two things are true at once here, and a manuscript is better for the author knowing both. COREQ was not developed by the Delphi-style consensus process now expected of reporting guidelines, and a serious replication attempt failed. COREQ is also endorsed by EQUATOR, required by a large number of health journals, and associated with measurable improvement in what qualitative papers actually report. Treat it as a floor for transparency, not as a definition of quality — and where an item does not fit your method, say why in the manuscript rather than leaving the reader to infer an omission.

Which items authors actually omit

The most useful evidence on COREQ compliance comes from de Jong and colleagues’ 2021 meta-review, which extracted 2,775 COREQ appraisals of individual qualitative studies from 139 reviews and compared studies published before 2007 with those published from 2009 onwards. Mean total score rose from 15.51 to 17.74 out of 32 (p < 0.001), and 13 of the 32 items improved after Bonferroni correction. Sixteen of the 32 items still averaged below 0.5 — reported in fewer than half of appraised studies.

The items that fail most often cluster in one place. In the post-2008 sample, transcripts returned to participants was reported in about 15% of studies, participant knowledge of the interviewer in 16%, relationship established before the study in 18%, participant checking in 19%, interviewer characteristics in 23%, presence of non-participants in 26%, experience and training in qualitative methods in 26%, repeat interviews in 28%, field notes in 34% and data saturation in 35%.

Against that, the items nearly everyone satisfies are the ones a manuscript would report anyway: sample size in 96% of studies, quotations presented in 92%, description of the sample in 90%.

Two patterns are worth naming. First, the reflexivity domain is where compliance is worst: mean domain 1 score was 2.86 out of 8 after 2008, against 8.51 out of 15 for study design and 6.20 out of 9 for analysis and findings. Second, more than half the reviews in that meta-review had adapted COREQ rather than applying it as published — dropping items or merging it with CASP, QualSyst or MMAT — with median completeness of 25 of the 32 items. A “COREQ score” quoted in a review is frequently not a score on the published checklist.

How COREQ non-compliance is detected in a manuscript

COREQ non-compliance is detected by reading the methods section for named agents and named decisions, not by counting words. Six checks find most of it.

Search for the agent of every verb in the methods. “Semi-structured interviews were conducted” leaves item 1 unanswerable. If no sentence in the manuscript says which author conducted them, items 1 through 8 are all dead at once, because each depends on knowing who was in the room.

Search the whole manuscript for the string “saturation”. Where it appears exactly once, in a sentence of the form “data saturation was reached”, item 22 is asserted rather than reported. A compliant answer states when saturation was assessed, against what, and whether further interviews were conducted to confirm it.

Search for “transcri”. Manuscripts overwhelmingly contain “transcribed verbatim” and nothing about item 23. Transcription and returning transcripts to participants are different items; so are returning transcripts (item 23) and returning findings (item 28), which are conflated constantly.

Count the participant identifiers attached to quotations. If the findings section presents twelve quotations and only three distinguishable identifiers, items 29 and 31 are formally satisfied and substantively hollow, and a reviewer will notice before the checklist does.

Compare the recruitment numbers. Item 13 asks for refusals and dropouts. A paper that reports 24 interviews and never states how many people were approached has an unreported denominator, which is the qualitative equivalent of a missing flow diagram.

Read the submitted checklist against the pages it cites. The 2026 analysis by Buus, Ong, Einboden, Juel and Perron in Qualitative Health Research examined eight articles alongside their self-reported COREQ responses and found authors supplying “misleading, inconsistent, or excessive information, prioritizing checklist completion over substantive engagement with quality principles”. Unfitted “N/A” responses against items that plainly apply, and checklist entries that contradict the manuscript they cite, are two of the patterns that analysis documents.

PerfectPaper’s qualitative and mixed-methods rigour agent runs these checks against the manuscript text rather than against the author’s own checklist entries, which is the distinction that matters: a completed form reports what the author believes is on the page.

What reviewers say when COREQ is mishandled

Reviewer comments on COREQ compliance are recognisable and repetitive. The following phrasings, or close variants, appear constantly on qualitative submissions to health journals.

“The authors state that data saturation was achieved but provide no account of how this was determined.” “It is not clear who conducted the interviews or what relationship they had with participants.” “No information is given about the researchers’ background, training or prior assumptions; reflexivity is not addressed.” “The completed COREQ checklist indicates that item 23 was addressed, but I could not locate the corresponding text on page 8.” “The methods do not name a methodological approach; ‘thematic analysis’ is asserted without reference to a specific formulation.” “Coding was performed by a single author with no description of how interpretive consistency was addressed.” “Please state how many individuals were approached and how many declined.”

The last three are the ones that decide papers. Items 9, 24 and 13 are cheap to satisfy and expensive to leave open, because each one licences a broader objection about whether the analysis can be evaluated at all — the shape of methods objections generally. A reviewer who cannot tell how themes were derived cannot endorse them, and will say so in the recommendation rather than in the comments.

COREQ, SRQR and ENTREQ: which one applies

COREQ applies to primary studies using in-depth interviews or focus groups. Two adjacent guidelines cover what COREQ does not, and choosing the wrong one is a visible error at journal submission.

SRQR — Standards for Reporting Qualitative Research, O’Brien, Harris, Beckman, Reed and Cook, Academic Medicine 2014;89(9):1245–1251 — is a 21-item standard written to accommodate the full range of qualitative paradigms, approaches and methods rather than one data collection mode. Where a study uses observation, documents or a design that COREQ’s interviewer-centred items do not describe, SRQR is the better fit, and several journals accept either.

ENTREQ — Enhancing Transparency in Reporting the Synthesis of Qualitative Research, published in 2012 by members of the same group that produced COREQ — is a 21-item checklist across five domains for syntheses of qualitative studies, not for primary studies. Reviews that appraise their included studies with COREQ and report themselves against ENTREQ are doing the correct thing with both.

A practical note on selection: many journals name a guideline in their instructions to authors, and many do not. Zhao, Zeng, Hu and Ma surveyed 29 Chinese nursing journals in a 2026 PLOS One study and found 2 (6.9%) mentioned COREQ or SRQR in their instructions for authors, with 11 of 24 surveyed editors (45.8%) aware of COREQ. Absence of a requirement in the instructions is not evidence that reviewers will not apply one, which is worth weighing during journal selection.

Where COREQ fights the method

Several COREQ items encode methodological positions that substantial parts of the qualitative community reject, and a manuscript is stronger for engaging with this than for silently complying.

Item 22, data saturation. Saturation originates in grounded theory, where it describes theoretical categories rather than a count of new codes. Applied to reflexive thematic analysis, where meaning is generated through interpretation rather than discovered at a stopping point, the concept does not transfer, and researchers working in that tradition argue explicitly that a saturation claim is incoherent there. Reporting “saturation was not sought, because the analytic approach does not use the concept, and the sample size rationale was X” answers item 22 correctly. Blank or “N/A” does not.

Item 23, transcripts returned, and item 28, participant checking. Both encode member checking as a marker of rigour. Its status is contested: returning transcripts frequently yields little, and returning findings can conflate participant agreement with analytic validity, since participants are not positioned to assess a cross-case interpretation. Neither item requires you to have done it; both require you to state whether you did.

Item 25, description of the coding tree. A coding tree presumes a hierarchical, structured coding approach. Approaches that treat coding as an evolving interpretive process do not produce one, and reporting that is an answer rather than a failure.

Overcompliance. The 2026 Qualitative Health Research analysis names the failure mode directly, describing an “overobedience” in which authors strive to meet criteria “even when they were irrelevant or unsuitable”. A checklist response that claims an item you did not do is worse than one that explains why the item does not apply — and it is the kind of internal inconsistency an overclaim check is built to surface before a reviewer does.

Filling in the checklist so it survives review

Complete the COREQ form against the submitted manuscript, item by item, with a page and paragraph reference for each, and do it after the final revision rather than before, because page references drift with every edit.

Three habits separate a form that helps from one that irritates. Give a location, not a section name: “p. 6, para 2” is checkable, “Methods” is not. Where an item does not apply, write the reason in one clause rather than “N/A”; the reason is itself reportable content and several journals now publish the completed checklist as supplementary material where readers will see it. Where an item is satisfied by a supplementary file — an interview guide against item 17, for instance — say which file.

The larger habit is to draft the methods section from the checklist rather than to audit it afterwards. Items 1 through 8 in particular are almost impossible to add convincingly at revision, because the reflexive account they ask for is a description of decisions taken during fieldwork; written retrospectively to satisfy a form, it reads exactly as retrospective, and reviewers say so. Working the domain into the first draft is a matter of academic writing practice rather than compliance.

Journal endorsement and the extension in progress

Journal endorsement of COREQ is broad but uneven, and no complete register of endorsing journals exists. EQUATOR lists COREQ among its reporting guidelines; a large number of health, nursing and health services journals name it in their instructions to authors; and de Jong and colleagues noted COREQ had accumulated over 5,600 Web of Science citations by their 2021 meta-review. Against that, the same body of work finds endorsement in instructions to authors is frequently absent even in fields where qualitative research is common.

One extension is in development. Fehring and colleagues published a protocol in JMIR Research Protocols in 2025 (14:e78682) for COREQ+LLM, an extension covering the reporting of large language model use in qualitative research — how models were used in design, processing, analysis or interpretation, and how their probabilistic behaviour, dependence on training data and susceptibility to hallucination were accounted for. The protocol sets out a four-phase development: a scoping review of peer-reviewed literature from January 2020 to April 2025, then a Delphi consensus process among international experts to agree candidate items, then refinement and publication of the extension.

Phase one has since been published as a scoping review (Kempny and colleagues, BMC Medical Research Methodology 2026;26:137). It searched five databases, retrieved 5,049 records (4,201 after duplicate removal), and included 75 studies that used a language model at a substantive stage of qualitative research. Reporting of that use was thin. Seventy-five per cent of the included studies reported no model parameter settings at all; 45% did not state whether the model was reached through an API, a web interface or a local deployment; 13% gave no prompting detail. Coding assistance (43 studies) and theme identification (41) were the commonest applications, and reported agreement between model and human coders ranged from 36% to 99% — a spread wide enough that the number means nothing without the method behind it.

COREQ+LLM itself has not been released, so no manuscript can yet be reported against it. The underlying obligation does not wait for the checklist, though: if a model touched the coding or the derivation of themes, the methods section has to name the model and version, state at which stage it was used, give the prompting approach, and describe how a human verified the output. Items 24 to 27 already reach part of this — a model that codes is a coder, and a model used through a platform is software — so a study that leaves it out is under-reporting against the current checklist, not merely against a future one.

Related

Qualitative and mixed-methods rigour agent · Research methods concepts · Why papers get rejected · How to write a response to reviewers · AI peer review for epidemiology

Review my manuscript

Frequently asked questions

What does COREQ stand for?

COREQ stands for Consolidated Criteria for Reporting Qualitative Research. The checklist contains 32 items across three domains and was published in 2007 by Allison Tong, Peter Sainsbury and Jonathan Craig in the International Journal for Quality in Health Care. The name reflects its method: items were consolidated from 22 existing qualitative checklists.

What is COREQ in qualitative research?

COREQ is the reporting checklist most health journals name for interview and focus group studies. COREQ specifies what the manuscript must state about the research team, the study design, and the analysis — 32 items in total. COREQ governs what is reported, not how the study was designed or whether its findings are sound.

How many items does the COREQ checklist have?

The COREQ checklist has 32 items. Eight cover the research team and reflexivity, fifteen cover study design, and nine cover analysis and findings, which is why domain scores are conventionally reported out of 8, 15 and 9 respectively. Reviews sometimes report adapted versions with fewer items.

What are the three domains of COREQ?

The three COREQ domains are research team and reflexivity (items 1–8), study design (items 9–23), and analysis and findings (items 24–32). The first covers who conducted the interviews and their relationship to participants, the second covers sampling and data collection, and the third covers coding, themes and the evidence presented for them.

Which studies does the COREQ checklist apply to?

COREQ applies to primary qualitative studies collecting data by in-depth interview or focus group. Ethnography, participant observation and document analysis fall outside it; SRQR, a 21-item standard published in 2014, accommodates those designs. Syntheses of qualitative studies are reported against ENTREQ instead.

Is COREQ a quality assessment tool?

No. COREQ is a reporting guideline, and a study can answer all 32 items while describing weak methods. Systematic reviews nonetheless use it as an appraisal score — 17% of qualitative reviews published after 2007 did so, in a meta-review of 1,695 reviews — which is a use the original checklist was not designed for.

Who developed the COREQ checklist and when?

Allison Tong, Peter Sainsbury and Jonathan Craig developed COREQ and published it in 2007, compiling 76 items from 22 existing checklists and reducing them to 32. A 2020 replication by Buus and Perron in the International Journal of Nursing Studies could not reproduce that consolidation and questioned the credibility of the resulting items.