S1252 Validation of a Novel Colonoscopy Simulator for Distinguishing Novice, Intermediate, and Expert Endoscopists

Introduction: Simulation-based training and assessment has the potential to reduce risks to patient safety and discomfort. Previous studies demonstrated construct validity in both virtual and physical colonoscopy simulators by being able to distinguish between subjects based on their level of endoscopic experience. We aim to assess the internal structure validity of a novel colonoscopy simulator in being able to differentiate expertise levels and to assess how well the simulator’s tasks measure different aspects of colonoscopy skills. Methods: This is a multi-center study involving 1 United States tertiary academic hospital and 1 Canadian community teaching hospital. Novices (<50 colonoscopies), intermediates (50-500 colonoscopies), and experts (>500 colonoscopies or attending physicians) were recruited. Participants completed 1 easy and 1 advanced level on the colonoscopy simulator model. Metrics recorded on the simulator include elongation of sigmoid colon, compression of colon, reaching sigmoid junction, reaching appendix, and time to completion. Video recordings of participants were de-identified and rated by expert endoscopists using a modified version of the Joint Advisory Group Direct Observation of Procedural Skills (JAG DOPS) criteria. The primary outcome was comparison of colonoscopy simulator-generated metrics and modified JAGDOPS scores between participant groups. Results: Eleven novices, 9 intermediates and 11 experts were recruited for the study. There was a statistically significant difference in all colonoscopy simulator metrics between intermediates and experts when compared to novices (P < 0.001). Expert and intermediates had significantly higher mean modified JAGDOPS scores compared to novices (P < 0.001) at both easy and advanced level. Mean insertion time was significantly lower and withdrawal time was longer in intermediate and expert groups compared to novices (P < 0.001). Conclusion: This next generation colonoscopy simulator model has construct validity in that it distinguishes the performance level of novices from intermediates and experts. This may suggest a benefit to implementing simulation-based education in the early stages of gastroenterology training. Moreover, simulation-based training with this colonoscopy model prior to initiating real-world practice has the potential to improve patient safety and training outcomes.