Ahmad A, Alnoah Z, Kochman ML, Krevsky B, Peikin SR, Mercogliano G, Bailey M, Boynton R, Reynolds JC. Endoscopic simulator enhances training of colonoscopy in a randomized, prospective, blinded trial. Gastrointestinal Endoscopy. 2003;57(5):S1499


Abstract

Methods used to train fellows have not changed in decades. New endoscopic simulators have the potential to recreate the challenges of performing colonos¬copy using virtual reality technology. Aim: To determine the impact of using an endoscopic simulator for the acquisition of shills to perform colonoscopy. Methods: Nine first-year G1 fellows from 4 programs were enrolled in a single-blind. randomized, prospective, multicenter study of the AccuTouch Endoscopy Simulator. Four fellows were randomized to simulator use during the first month of fellowship. All fellows abstained from performing actual colonoscopies during this time. During the next 2 months, attendings blinded to the randomization evaluated fellow skills during actual colonoscopy. Fellows were scored on time to cecal intubation, distance achieved without attending intervention (1 = rectosigmoid. 2 = transverse colon, 3 = ascending colon, 4 = cerum), number of attending interventions, and overall attending assessment (.scale of 1-10). Their ability to improve with time was also assessed by comparing scores from the early to the late half of the evaluation period. Results: Fellows enrolled in the study obtained a total of 132 evaluations. Fellows used simulator models of 6 flexible sigmoidoscopies and 18 colonoscopies for a mean of 32S minutes. The average attend¬ing assessment of fellow performance was 6.6 in the simulator group vs. 5.2 in the control group (p < 0.001). The distance achieved without attending intervention was 2.7 in the simulator group vs. 2.1 in the controls (p = 0.001). Attending assessment of performance; over time revealed a significant improvement in scores in the simulator group (6.1 vs. 7.0, p = 0.03) while fellows in the control group did not improve (6.? vs. 6.2). The distance achieved without attending intervention was higher in the simulator group during the early evaluation period (2.6 vs. 2.0, p = 0.02). This difference was maintained during the late evaluation period (2.8 vs. 2.0, p = 0.02). No significant differences were seen in the number of attending interventions or time to completion between the two groups. Conclusions: The use of endoscopic simulators enhances the performance of colonoscopies by GI fellows within the first two months of training. Follow-up is in progress to determine if this difference persists