Can we teach reflective reasoning in general-practice training through example-based learning and learning by doing?

Purpose: Flaws in physicians’ reasoning frequently result in diagnostic errors. The method of deliberate reflection was developed to stimulate physicians to deliberately reflect upon cases, which has shown to improve diagnostic performance in complex cases. In the current randomised controlled trial, we investigated whether deliberate reflection can be taught to general-practice residents. Additionally, we investigated whether engaging in deliberate reflection or studying deliberate-reflection models would be more effective. Methods: The study consisted of one learning session and two test sessions. Forty-four general-practice residents were randomly assigned to one of three study conditions in the learning session: (1) control without reflecting (n = 14); (2) engaging in deliberate reflection (n = 11); or (3) studying deliberate-reflection models (n = 19). To assess learning, they diagnosed new cases in both a same-day test and a delayed test one week later. In the delayed test, participants were additionally asked to elaborate on their decisions. We analysed diagnostic accuracy and whether their reasoning contained key elements of deliberate reflection. Results: We found no significant differences between the study conditions in diagnostic accuracy on the same-day test, p =.649, or on diagnostic accuracy, p =.747, and reflective reasoning, p =.647, on the delayed test. Discussion: Against expectations, deliberate reflection did not increase future reflective reasoning. Future studies are needed to investigate whether residents either did not sufficiently learn the procedure, did not adopt it when diagnosing cases without instructions to reflect, or whether the reflective-reasoning process as itself cannot be taught.