When Clinical Evidence Is Incomplete

Start with the decision, not the questionnaire

Define the population, setting, clinical uncertainty, and decision the study should inform. Rare diseases, emerging practice, inconsistent care pathways, and unresolved research priorities can create questions for structured expert input. Ask whether consensus is the appropriate method before choosing the platform.

Separate evidence from expert judgment

Provide a clear account of the evidence available to panelists and the gaps it leaves. Delphi findings describe the judgments of the participating panel under the stated conditions. Agreement is not proof of treatment benefit, and a panel does not replace comparative clinical research.

Make rigor visible

• Justify panel expertise, stakeholder representation, and recruitment.
• Define rating dimensions, consensus thresholds, feedback, and stopping rules in advance.
• Preserve independent judgments and document changes to statements.
• Report participation, valid-response denominators, non-ratings, and unresolved items.

Do not assume one panel size or study duration fits every question. The required perspectives, scope, and expected attrition should drive planning.

Treat disagreement as useful evidence

A pooled result may conceal differences between specialties, countries, patients, or practice settings. Inspect distributions and comments, and compare groups only where the design and available counts support interpretation. Persistent dissent can reveal conditions that make a recommendation inappropriate or difficult to implement.

Design outputs for their intended audience

A committee may need a concise decision brief; a publication team may need methods, tables, figures, and supporting records. Define each deliverable in advance. If findings will inform a submission or HTA evidence package, involve the appropriate specialists and confirm requirements with the intended recipient.

ACCORD supports transparent reporting of consensus methods in biomedicine. CREDES was developed for Delphi studies in palliative care; consider its relevance to your study context. These resources guide reporting and do not certify software or guarantee study quality.

Choose software and support to match the study

Surveylet supports Delphi collection, controlled feedback, round snapshots, and exports. Your team can lead the study or request scoped implementation and reporting support. Plan comparisons between rounds when stability forms part of your protocol, including the measures, participant matching, and stopping criteria.

Medical consensus research · Support options · Final reporting