[Diagnostic hypothesis generation through engaged peer observation: a quantitative descriptive study in a clinical simulation context].

Background: Learning clinical reasoning (CR) requires practice in a variety of educational settings. As part of the clinical simulation sessions at the University of Ottawa's Faculty of Medicine, pre-clerkship students are paired in dyads to increase the number of practical clinical cases before the clerkship. One student plays the role of a Clinical Student (CS) and the other alternates as a Student Observer (SO). This quantitative descriptive study aims to compare the diagnostic hypothesis generation by SOs with that of CSs to support the usefulness of engaged peer observation as a CR learning strategy in clinical simulation settings.

Methods: Following an interview with a simulated patient, CSs and SOs were asked to generate two diagnostic hypotheses in an electronic form. Responses were compiled, categorized, and compared in terms of equivalent diagnostic hypotheses within the same dyad. The difference in frequency distribution of equivalent hypotheses was statistically analyzed using a chi-square calculation.

Results: The percentage of dyads with at least one equivalent diagnostic hypothesis ranged from 83% to 100%, depending on the scenario. The number of equivalent hypotheses between SOs and CSs was statistically significant (p < 0.01).

Conclusion: SOs appear to be able to generate diagnostic hypotheses similar to those of CSs. The results support the use of engaged peer observation as a learning strategy for CR in clinical simulation settings in pre-clerkship medical education.