Learning From Errors During Simulation-based Ultrasound Training

During the past decades, simulation-based training has become essential for making trainees fit for clinical practice. Traditionally, trainees are instructed to practice with as few errors as possible during simulation-based training. However, recent evidence from the field of psychology suggests that transfer of learning may improve if trainees are encouraged to commit errors during training. The aim of this study is to assess on performances with real patients the effect of error-management instructions compared to error-avoidance instructions during simulation-based virtual reality ultrasound training.

This study is planned as a randomized controlled trial. Participants include medical students with no prior simulation practice at the fifth or sixth year of medical school. All participants receive 3 hours simulation-based ultrasound training focusing on fetal weight estimation. The participants (N=60) are randomized into two groups: Participants in group 1 are instructed to follow the simulator program step-by-step to achieve the highest possible simulator metric scores by making as few errors as possible. Participants in group 2 are instructed to experiment and explore and to deliberately make errors during training. A simulation-based pre- and post-test is administered before and after training.

Participants are scheduled to perform a transfer test seven to ten days after simulation training. The transfer test consists of fetal weight estimation on a real patient. The primary outcome is weight deviation of participant measurements compared to expert findings. Secondary outcomes include performance assessments of pre-, post- and transfer test performances by blinded ultrasound experts using the Objective Structured Assessment of Ultrasound Skills (OSAUS) scale. Using an alpha level of 0.05, an effect size of 0.80, and a power of 0.80, 25 participants are needed in the two study groups. Loss to follow-up of up to 20% of study participants is anticipated, resulting in 30 participants in each study group.