Comparing different retrieval practice strategies using virtual patients: A stratified randomized trial

This study evaluated the effectiveness of three retrieval practice strategies: re-solving virtual patient (VP) cases, answering multiple-choice questions (MCQs), and answering short answer questions (SAQs), on long-term memory retention of medical students using a VP simulation platform. Eighty fifth-year medical students participated in a stratified randomized trial conducted in three phases. In the initial learning phase, participants completed a 14-item baseline test (seven MCQs and seven SAQs) to assess prior knowledge and enable stratified randomization. They then engaged with two clinical cases using the Paciente 360® VP platform. One week later, participants were randomly assigned to one of three retrieval practice conditions: re-solving the original VP cases, answering 24 related MCQs, or answering 24 related SAQs. Six weeks after the intervention, participants completed a 40-item retention test (20 MCQs and 20 SAQs), which included both previously encountered and novel questions to assess long-term retention and transfer of learning. Participants in the SAQ condition demonstrated a statistically significant improvement in performance over time, while those in the re-solving the virtual case and MCQ conditions maintained their knowledge levels. No significant differences were observed between performance on repeated versus novel questions or between MCQs and SAQs. Retrieval practice using VP simulations supports knowledge retention, with SAQs yielding the greatest improvement from baseline. Comparable performance on repeated and novel questions suggests that retrieval practice may also promote transfer of learning to new clinical scenarios.