Breaking down automaticity: Case ambiguity and the shift to reflective approaches in clinical reasoning

Context: Two modes of case processing have been shown to underlie diagnostic judgements: analytical and non-analytical reasoning. An optimal form of clinical reasoning is suggested to combine both modes. Conditions leading doctors to shift from the usual mode of non-analytical reasoning to reflective reasoning have not been identified. This paper reports a study aimed at exploring these conditions by investigating the effects of ambiguity of clinical cases on clinical reasoning. Methods: Participants were 16 internal medicine residents in the Brazilian state of Ceará. They were asked to diagnose 20 clinical cases and recall case information. The independent variable was the degree of ambiguity of clinical cases, with 2 levels: straightforward (i.e. non-ambiguous) and ambiguous. Dependent variables were processing time, diagnostic accuracy and proposition per category recalled. Data were analysed using a repeated measures design. Results: Participants processed straightforward cases faster and more accurately than ambiguous ones. The proportion of text propositions recalled was significantly lower (t[15] = 2.29, P = 0.037) in ambiguous cases, and an interaction effect between case version and proposition category was also found (F[5, 75] = 4.52, P = 0.001, d = 0.232, observed power = 0.962). Furthermore, participants recalled significantly more literal propositions from the ambiguous cases than from the straightforward cases (t[15] = 2.28, P = 0.037). Conclusions: Ambiguity of clinical cases was shown to lead residents to switch from automatic to reflective reasoning, as indicated by longer processing time, and more literal propositions recalled in ambiguous cases. © 2007 Blackwell Publishing Ltd.