Feasibility of self-reflection as a tool to balance clinical reasoning strategies

Clinicians are believed to use two predominant reasoning strategies: system 1 based pattern recognition, and system 2 based analytical reasoning. Balancing these cognitive reasoning strategies is widely believed to reduce diagnostic error. However, clinicians approach different problems with different reasoning strategies. This study explores whether clinicians have insight into their problem specific reasoning strategy, and whether this insight can be used to balance their reasoning and reduce diagnostic error. In Experiment 1, six medical residents interpreted eight ECGs and self-reported their predominant reasoning strategy using a four point scale (4S). Self-assessed reasoning strategy correlated with objective assessment by two clinical experts using a post hoc talk-aloud protocol (ρ = 0.69, p < 0.0001). Reporting an analytic strategy was also associated with 40% longer interpretation times (p = 0.01). In Experiment 2, twenty-four residents were asked to reinterpret eight ECGs with instructions customized to their 4S. Half of the ECGs were reinterpreted with instructions to use the opposite reasoning strategy to that reported, and half with instructions to use the same reasoning strategy. ECG reinterpretation scores did not differ with potentiating compared to balancing reasoning instructions (F1,188 = 0.22, p = 0.64). However, analytic instructions were associated with improved scores (F1,188 = 15, p < 0.0001). These data suggest that clinicians are able to recognize their reasoning strategies. However, attempting to balance reasoning strategies through customizable instructions did not result in a reduction in diagnostic errors. This suggests important limitations to the widespread belief in balancing reasoning strategies to reduce diagnostic error.