How to Write Up Qualitative Research Using COREQ and SRQR

Reporting qualitative research means letting readers retrace how you reached your themes: who collected the data, who took part and why, when collection stopped, how codes became themes, and which quotations support them. Use the Consolidated criteria for reporting qualitative research (COREQ) for interviews and focus groups, and the broader Standards for Reporting Qualitative Research (SRQR) for other designs.

COREQ, published in 2007, has 32 items in three domains: research team and reflexivity, study design, and data analysis and reporting. SRQR, published in 2014, has 21 items and was built to fit a range of qualitative paradigms, approaches and methods.

Where the two checklists sit among other study designs is set out in our reporting guideline finder for surgical and clinical authors.

Choose COREQ or SRQR by how the data were collected

COREQ was developed for studies that use in-depth interviews or focus groups. It was not designed for ethnography, observation or document analysis.

SRQR is a general standard, so it can also be used for an interview or focus-group study. It covers the whole article, from title and abstract to funding, while COREQ covers the research team, the methods and the findings. SRQR asks you to identify your qualitative approach, recommends naming the research paradigm behind it, and asks which techniques you used to enhance trustworthiness.

If your qualitative material is free-text answers inside a questionnaire, the sampling frame and response rate still need reporting, as described in reporting a questionnaire survey study. Our author guidelines ask authors to match the manuscript to the relevant EQUATOR checklist.

Say who asked the questions before saying who answered

Reflexivity is the researchers' account of how their position, assumptions and relationships may have shaped the data and its interpretation. COREQ's first domain asks who conducted the interviews or focus groups, with their credentials, occupation, gender, experience and training. It also asks whether a relationship existed with participants before the study, what participants knew about the interviewer, and what the interviewer brought to the topic.

Clinical settings make this concrete. A patient interviewed by the surgeon who operated on them may say different things from one interviewed by an independent researcher. Neither arrangement is wrong, but the reader needs to know which it was.

Write reflexivity as method, not confession: two or three sentences in the Methods, plus a line in the limitations if the relationship could have shaped the answers.

Justify the sample by the question, not by a borrowed number

Purposive sampling means choosing participants because they can speak to the research question, for example patients at different stages of treatment. Say which characteristics you sampled for and why. If you used convenience or snowball sampling instead, call it that. COREQ also asks how participants were approached and how many declined or dropped out, with reasons.

Data saturation is the point at which further data add no new codes or themes relevant to the question. It has several definitions, so state the one you used and how you judged it. A sentence such as "the last [n] interviews produced no new codes" can be checked; "saturation was reached" on its own cannot.

Not every qualitative approach uses saturation as a stopping rule. If yours did not, say what fixed the sample instead, such as the size of a small, defined population. Never claim saturation you did not assess.

Report the interviews so another team could repeat them

Describe the interview guide, whether it was piloted and whether it changed as analysis progressed, and supply it as a supplementary file. State whether anyone was interviewed twice, and whether field notes were taken. Say who transcribed the recordings, whether transcription was verbatim, and whether transcripts were returned to participants for comment or correction.

Interviews involve human participants, so they need whatever approval your institution requires, as explained in ethics approval and informed consent. If the consent form covered audio recording and publication of anonymised quotations, say so.

Full transcripts are hard to anonymise, because a story can identify a person without a name. Our data availability policy allows data to be "Restricted for ethical or legal reasons" if the statement gives the reasons and how requests are handled.

Show how codes became themes

Name the analytic approach and cite a methodological source for it. Thematic analysis, framework analysis, grounded theory, interpretative phenomenological analysis and content analysis are different methods with different steps. "Thematic analysis was used" says little until you name whose version you followed.

Then report the mechanics COREQ asks for: how many people coded the data, a description of the coding tree, whether themes were set in advance or derived from the data, and which software, if any, managed the data. Software manages the codes, but the themes are the researchers' interpretation; if you used an automated or AI-assisted coding feature, say which and how.

Say what you did to enhance trustworthiness, and whether participants gave feedback on the findings. SRQR gives member checking, an audit trail and triangulation as examples. They are options, not requirements, so report the ones you used and why.

Present each theme as a finding, with identified quotations

A theme name should state a finding, not a topic. "Communication" is a topic; "Patients filled gaps in explanation with advice from relatives" is a finding. Explain each theme in your own words, then support it with quotations.

COREQ asks whether quotations were presented and whether each one was identified, for example by participant number. Use a label that carries only what the reader needs, and remove details that could identify someone in a small unit. COREQ also asks whether data and findings are consistent and whether minor themes and diverse cases are described, so report the participants who did not fit.

Avoid percentages, because a purposive sample is not chosen to represent a population. If a count helps, give it as a number of participants.

The theme below is invented to illustrate the structure; it is not real data.

Theme 2: Patients filled gaps in explanation with advice from relatives
Participants who felt the consent discussion had been rushed described asking family members to interpret what they had been told. Three of the twelve participants said the family explanation, not the clinic's, shaped their decision.
"My daughter looked it up and told me what the surgeon meant." (P07, patient, interview)
Two participants, both with a nursing background, did not describe this pattern; they read the written information themselves.

Keep the IMRaD structure, with two changes

The IMRaD format (Introduction, Methods, Results and Discussion) still structures a qualitative paper, with two differences. The introduction usually ends on an open research question or aim rather than a hypothesis to test. The results section, often headed Findings, carries interpretation, because a theme is already an analytic claim; SRQR asks for evidence, such as quotations or field notes, to substantiate the analytic findings.

The discussion relates the themes to earlier work, considers whether they may transfer to other settings, and states the limitations.

A fill-in template for reporting qualitative research

Copy the rows that apply into your Methods and replace every bracket. The third column names the COREQ domain each row answers.

If a row does not apply, say the step was not taken and why; a blank leaves the reviewer to guess. If you spot an error in this guide, tell us how to correct it.