Randomized Clinical Trial of Virtual Reality Simulation Training for Transvaginal Gynecologic Ultrasound Skills

Abstract

OBJECTIVES: To compare the impact of virtual reality simulation training and theoretical teaching on the ability of inexperienced trainees to produce adequate virtual transvaginal ultrasound images. METHODS: We conducted a randomized controlled trial with parallel groups. Participants included inexperienced residents starting a training program in Paris. The intervention consisted of 40 minutes of virtual reality simulation training using a haptic transvaginal simulator versus 40 minutes of conventional teaching including a conference with slides and videos and answers to the students' questions. The outcome was a 19-point image quality score calculated from a set of 4 images (sagittal and coronal views of the uterus and left and right ovaries) produced by trainees immediately after the intervention, using the same simulator on which a new virtual patient had been uploaded. Experts assessed the outcome on stored images, presented in a random order, 2 months after the trial was completed. They were blinded to group assignment. The hypothesis was an improved outcome in the intervention group. Randomization was 1 to 1. RESULTS: The mean score was significantly greater in the simulation group (n = 16; mean score, 12; SEM, 0.8) than the control group (n = 18; mean score, 9; SEM, 1.0; P= .0302). CONCLUSIONS: The quality of virtual vaginal images produced by inexperienced trainees was greater immediately after a single virtual reality simulation training session than after a single theoretical teaching session.