Shaleen Vasavada, MD1, Theresa Nguyen-Wenker, MD, PhD2, Scott A. Larson, MD, PhD1 1Baylor College of Medicine, Houston, TX; 2Baylor College of Medicine / Michael E. DeBakey Veterans Affairs Medical Center, Houston, TX
Introduction: Endoscopy training for gastroenterology fellows is critical for ensuring the safe and effective performance of procedures. Currently, there is no standardized feedback mechanism; thus, the quality and consistency of feedback provided to fellows are highly variable. To address this, we developed a mobile-friendly web application that standardizes evaluations and feedback with the help of generative artificial intelligence (AI), aiming to enhance endoscopy education and skill development for fellows.
Methods: A comprehensive needs assessment was conducted at Baylor College of Medicine Gastroenterology Fellowship to identify key evaluation criteria and incorporate existing tools developed by the American Society for Gastrointestinal Endoscopy (ASGE), such as the Assessment of Competency in Endoscopy (ACE) and the Mayo Colonoscopy Skills Assessment Tool (MCSAT). Collaborating with experienced endoscopists, we designed a user-friendly evaluation framework for both an esophagogastroduodenoscopy (EGD) and colonoscopy. The platform allows attending physicians to log in, evaluate procedures, and provide real-time feedback. A fellow dashboard was developed to give fellows average composite scores for EGD and colonoscopy in addition to an AI-assisted anonymous feedback and suggestions for improvement.
Results: Initial implementation involved evaluating a cohort of gastroenterology fellows across three training years at one pavilion at our institution. Each evaluation was filled out at bedside on a personal mobile device during or immediately after a procedure. Initial pilot phase accounted for 50 evaluations. Both attending and fellow physicians reported improved satisfaction with the feedback process, citing increased specificity of the individualized feedback and actionable insights.
Discussion: The standardized evaluation framework and timely feedback mechanism facilitated by the web application significantly enhanced the educational experience for fellows. By providing consistent and constructive feedback, the platform enabled fellows to identify deficits of specific skills to improve upon. In the near future, a performance tracking feature will be added for program directors to address any fellowship-wide deficiencies. Future plans include a prospective interventional study to determine if our web application leads to significant improvements in endoscopy skills.
Figure: Screenshot example of a 20 question Colonoscopy evaluation on the website
Disclosures:
Shaleen Vasavada indicated no relevant financial relationships.
Theresa Nguyen-Wenker indicated no relevant financial relationships.
Scott Larson indicated no relevant financial relationships.
Shaleen Vasavada, MD1, Theresa Nguyen-Wenker, MD, PhD2, Scott A. Larson, MD, PhD1. P2374 - A Novel Web Application Utilizing Generative Artificial Intelligence to Enhance Endoscopy Education for Gastroenterology Fellows, ACG 2024 Annual Scientific Meeting Abstracts. Philadelphia, PA: American College of Gastroenterology.
