Should we prioritize knowledge gain or process improvement when teaching diagnostic reasoning? Insights from how medical students engage with deliberate reflection.

INTRODUCTION: The complexity of diagnostic reasoning ensures there can be no singular educational intervention to perfect its teaching. The need to flexibly adapt to needs of trainees and learning environments demands deeper understanding of available strategies. Deliberate reflection offers one technique that is central to debate in this area, but the literature remains insufficiently clear about why it can be effective and, hence, what purposes it can serve. By better understanding how deliberate reflection works we can better design learning exercises that guide students toward clinical expertise and better practice.

METHOD: Ten pre-clerkship medical students with experience applying deliberate reflection completed a mixed-methods think aloud study. They were asked to verbalize everything that passed through their minds while analyzing clinical vignettes, before and during instruction to apply a deliberate reflection grid. Two cases of varying difficulty were presented. Inductive thematic coding was applied to transcripts with quantification of codes to contextualize when different types of ideas were expressed.

RESULTS: Participants' statements suggest that, while using the deliberate reflection grid, medical students shifted away from hypothesis generation toward more explicit hypothesis testing; further, the grid encouraged and supported ongoing reasoning in moments of uncertainty. Each shift altered perceptions of diagnostic likelihood and drew participants' attention to clinical features they did not fully understand.

DISCUSSION: Contrary to debate about whether one should teach diagnostic reasoning by emphasizing knowledge gain versus shifting knowledge application strategies, the patterns indicate deliberate reflection to function as a process-oriented scaffold that simultaneously directs people toward additional learning (enabling knowledge gain) and supports retrieval of prior knowledge (de-biasing through procedural direction). Thus, rather than focusing exclusively on correcting cognitive biases or increasing clinical knowledge, educational interventions may be most effective when they guide learners through deliberate and comparative reasoning processes for engaging and building their knowledge base.