Virtual patient simulations for medical education: increasing clinical reasoning skills through deliberate practice

Virtual Patient Simulations (VPS) are web-based exercises involving simulated patients in virtual environments. This study investigates the utility of VPS for increasing medical student clinical reasoning skills, collaboration, and engagement. Many studies indicate that VPS provide medical students with essential practice in clinical decision making before they encounter real life patients. The utility of a recursive, inductive VPS for increasing clinical decision-making skills, collaboration, or engagement is unknown. Following a design-based methodology, VPS were implemented in two phases with two different cohorts of first year medical students: spring and fall of 2013. Participants were 108 medical students and six of their clinical faculty tutors. Students collaborated in teams of three to complete a series of virtual patient cases, submitting a ballpark diagnosis at the conclusion of each session. Student participants subsequently completed an electronic, 28-item Exit Survey. Finally, students participated in a randomized controlled trial comparing traditional (tutor-led) and VPS case instruction methods. This sequence of activities rendered quantitative and qualitative data that were triangulated during data analysis to increase the validity of findings. After practicing through four VPS cases, student triad teams selected accurate ballpark diagnosis 92 percent of the time. Pre-post test results revealed that PPT was significantly more effective than VPS after 20 minutes of instruction. PPT instruction resulted in significantly higher learning gains, but both modalities supported significant learning gains in clinical reasoning. Students collaborated well and held rich clinical discussions; the central phenomenon that emerged was "synthesizing evidence inductively to make clinical decisions." Using an inductive process, student teams collaborated to analyze patient data, and in nearly all instances successfully solved the case, while remaining cognitively engaged. This is the first design-based study regarding virtual patient simulation, reporting iterative phases of implementation and design improvement, culminating in local theories (petite generalizations) about VPS design. A thick, rich description of environment, process, and findings may benefit other researchers and institutions in designing and implementing effective VPS.