Developing a checklist for reporting research using simulated patient methodology (CRiSP): a consensus study.

OBJECTIVES

Simulated patients are increasingly used to measure outcomes in health services but reporting is suboptimal. This study aims to create a checklist for the reporting of simulated patient (SP) methodology.

METHODS

This was a Delphi study. The authors of health service research studies using SP methodology were invited to participate. Round 1 questionnaire assessed the applicability of the TIDieR (Template for Intervention Description and Replication) reporting checklist for SP methodology and asked for rewording of/additional items. Responses were thematically analysed to generate Round 2 items in which participants rated each item for importance (seven-point Likert scale) and median, mode and IQR were calculated. In Round 3, participants were invited to rescore their Round 2 responses. Consensus was defined as an IQR ≤ 1 (Extremely important) and median ≤ 2 (Very important). All consensus items were considered for inclusion in the checklist. Similarly, worded items were rationalised and items not specific to SP methodology or other existing checklists were excluded.

KEY FINDINGS

Twenty-nine authors participated in Round 1 and a further seven for Rounds 2 and 3. Twenty-six responses were analysed for Round 1, 30 for Round 2 and 28 for Round 3. There was consensus on 29 of 54 items in Round 2 and 45 of 63 items in Round 3. The final checklist comprised 28 items.

CONCLUSIONS

A new reporting checklist to guide the reporting of studies, using simulated patients, complementary to CONSORT or STROBE, has been developed and will now be tested for usability.