S1269 Impact of Endoscopist Characteristics on Recurrence After Endoscopic Large Colonic Polyp Resection

Introduction: Although endoscopist experience and skill are believed to be critical to ensuring complete endoscopic mucosal resection (EMR) of large colorectal polyps, very little is known about variation between endoscopists and its effect on completeness of resection. We aimed to examine the association between individual endoscopist characteristics and recurrence after cold and hot EMR. Methods: We used data of a recent randomized trial comparing cold with hot EMR of ≥20 mm non-pedunculated colorectal polyps. Participating endoscopists completed a survey prior to study commencement. Our primary outcome measure was the endoscopists’ variation in the rate of residual or recurrent neoplastic polyps (termed “recurrence”) at first surveillance colonoscopy. Secondary outcomes included association of endoscopist characteristics with recurrence. Multivariable analysis was applied to adjust for polyp complexity (size, location, advanced histology). Results: Among 660 patients 714 large polyps were removed by 29 endoscopists. Corresponding survey data were available for 680 EMRs (95.2%). Recurrence varied broadly from 12.7 to 45.8% for any EMR (P = 0.002), 0 and 50.0% for hot EMR (P < 0.001), and 11.9 and 46.7% for cold EMR (P = 0.129). There was no correlation between hot and cold EMR. In univariable analysis, endoscopists with longer resection times, fewer study EMRs, and prior advanced endoscopy training had a greater risk of recurrence. After adjusting for polyp complexity, the risk of recurrence increased for endoscopists with longer resection times following any EMR (OR 1.02 per minute increment, 95% CI [CI] 1.01-1.03), hot EMR (OR 1.02, CI 1.01-1.04) and cold EMR (OR 1.02, CI 1.00-1.03). The risk for recurrence following cold EMR was lower for endoscopists with a greater number of lifetime cold EMR procedures (OR 0.999 per 1 EMR increment, OR 0.999-1.000, P = 0.016). Years since training, lifetime number of colonoscopies, or annual EMR volume were not associated with recurrence. Conclusion: This first formal analysis of endoscopist characteristics and recurrence after large polyp EMR showed a broad variation between endoscopists, while poor outcomes after hot and cold EMR were not correlated. Although most endoscopist characteristics were not associated with recurrence, a greater recurrence with longer resection times suggests that sufficient experience is required to ensure low recurrence rates, particularly for cold EMR.