Transfer of Clinical Reasoning Trained With a Serious Game to Comparable Clinical Problems:A Prospective Randomized Study

INTRODUCTION: Serious Games can be used effectively for clinical reasoning training in medical education. Case specificity of learning outcome elicited by Serious Games has not been studied in detail. This study investigated whether learning outcome elicited by repeated exposure to virtual patient cases is transferable to other cases addressing similar problems.

METHODS: In this monocentric, prospective, randomized trial, 69 fifth-year medical students participated in ten 90-minute sessions of using a computer-based emergency ward simulation game. Students were randomized to 3 groups (A, B, AB) and subsequently exposed to up to 46 different virtual patients. Group A was presented with 2 specific cases that were not shown in group B and vice versa. Group AB was exposed to all 4 specific cases. After 6 weeks of using the game as a learning resource, students were presented with 4 cases addressing similar problems. For each case, an aggregate score reflecting clinical reasoning was calculated, and performance was compared across groups.

RESULTS: In the final session, there were no significant between-group differences regarding the sum score reflecting aggregated performance in all 4 cases (A: 66.5 ± 7.2% vs. B: 61.9 ± 12.4% vs. AB: 64.8 ± 11.1%, P = 0.399). An item-by-item analysis revealed that there were no between-group differences regarding correct therapeutic interventions.

CONCLUSIONS: Previous exposure had limited impact on subsequent performance in similar cases. This study suggests that case specificity of learning outcome elicited by the serious game is low and that acquired higher-order cognitive functions may transfer to similar virtual patients.