CREDES Explained: How to Report a Delphi Study for Publication

Your Delphi study is complete. The data looks good, consensus was reached, and you're ready to write it up. Then the journal rejection comes back: "Insufficient methodological detail in the methods section. Please revise according to established reporting guidelines."

This is one of the most common reasons Delphi studies get rejected or sent back for major revision — not because the study was poorly designed, but because it wasn't reported in a way that allows reviewers to evaluate the methodology. CREDES (Guidance on Conducting and REporting DElphi Studies) exists precisely to solve this problem.

This guide explains what the CREDES delphi reporting guideline requires, where researchers most commonly fall short, and how to write a methods section that holds up to peer review.

What Is CREDES?

CREDES stands for Guidance on Conducting and REporting DElphi Studies. It's a set of methodological recommendations developed through a systematic review of Delphi studies, published in Palliative Medicine in 2017 and maintained on the EQUATOR Network, the central registry for health research reporting guidelines.

CREDES was originally developed for palliative care research, but it has since been adopted across healthcare, education, social sciences, and policy research as the leading standard for Delphi reporting. If you're submitting to a journal that follows EQUATOR guidelines — which includes most high-impact health and social science journals — your methods section should align with CREDES.

The checklist addresses two things: how your study was conducted, which affects validity, and how it's reported, which affects reproducibility. Both matter.

The CREDES Checklist: What Each Item Requires

CREDES is organized around five reporting domains. Here's what reviewers will look for in each.

Justification for Using Delphi

Reviewers want to see that you chose Delphi intentionally, not as a default. Your methods section should explicitly state why Delphi was appropriate for your research question, why alternatives such as a literature review, focus groups, or voting methods were insufficient, and which variant of Delphi you used — traditional sequential, modified Delphi, or Real-Time Delphi — and why.

Weak justification: "A Delphi study was conducted to gather expert opinion."

Strong justification: "Given the absence of high-quality evidence on the topic, a Delphi study was used to systematically aggregate structured expert judgement. A Real-Time Delphi design was selected to reduce the study timeline and minimize panel fatigue across three rounds."

Expert Panel Selection and Composition

This is the section where most rejections originate. Reviewers need to be able to evaluate whether your panelists were genuinely qualified and whether the panel was diverse enough to be representative. Report:

Many researchers report Round 1 participation but not subsequent rounds. CREDES requires all rounds to be reported separately.

Questionnaire Design and Modifications

Describe how the initial item list was generated — literature review, qualitative interviews, or expert suggestion — and how items were operationalized into questionnaire form.

If you modified the questionnaire between rounds, such as adding items, splitting items, or removing items that reached consensus, this must be documented explicitly with the rationale for each change. Changes that aren't documented look like researcher bias.

Also report the response scale used (Likert, VAS, ranking), any demographic variables included, and whether a pilot test was conducted.

Consensus Definition and Analysis

This is the most technically specific section and the one most often under-reported. You must define your consensus criteria a priori — before the study began, not after looking at the data. Report:

A common mistake: defining consensus as "70% agreement" without specifying what the agreement scale was or how borderline items were handled.

Round-by-Round Results Reporting

CREDES requires results to be reported for each round separately, not just the final state. Reviewers need to see how consensus evolved — which items moved, which diverged further, and which were dropped. For each round, report the response rate, the distribution of responses per item (mean, median, IQR, percentage in each response category), which items reached consensus and were removed from subsequent rounds, and a summary of qualitative feedback and how it informed revisions, if collected.

Common CREDES Reporting Failures

Research on Delphi study reporting quality consistently finds the same gaps:

If your study does these things correctly and you report them clearly, you're already ahead of the majority of published Delphi research.

Writing Your Methods Section: A Template Structure

Structure your Delphi methods section in this order:

Keep it precise rather than narrative. Reviewers scan methods sections — clear headings and explicit criteria are easier to audit than prose.

How Durvey Supports CREDES Compliance

Durvey was built with academic publishing in mind. The platform automatically calculates and exports the consensus metrics CREDES requires: IQR per item per round, percentage agreement, convergence rate, and response distributions. The export format maps directly to what your methods and results sections need.

For guideline groups and consensus statements there is a dedicated overview of how this works in a clinical Delphi study. Durvey also supports the survey design best practices that CREDES recommends — standardized response scales, demographic capture, and anonymized feedback display between rounds.