Su1050 ENDOSCOPIST PROCEDURAL VOLUME AND COLONOSCOPY OUTCOMES: A SYSTEMATIC REVIEW AND META-ANALYSIS

Abstract

In addition to monitoring adverse events (AEs) and post-colonoscopy colorectal cancers (PCCRC), several indicators are used to assess the overall quality of colonoscopy performance, including adenoma detection rate (ADR) and cecal intubation rate (CIR). It is unclear whether there is an association between an endoscopist's annual colonoscopy volumes and ADR, CIR, AEs or PCCRC. As such, we performed a systematic review and meta-analysis to address this question. A comprehensive electronic search was performed through March of 2019 for any studies assessing the potential association between annual colonoscopy volume and quality indicators, including ADR, CIR, AEs or PCCRC. Pooled odds ratios (OR) were calculated using DerSimonian and Laird random effects models. Subgroup and sensitivity analyses were performed to assess for potential methodological or clinical factors associated with outcomes. This included assessing associations among screening procedures separately. Out of an initial 9,235 studies, 27 were included in our systematic review, representing 11,276,244 colonoscopies. There was no association between procedural volume and ADR (OR 1.00, 95% confidence intervals, CI, 0.98 to 1.02 per additional 100 annual total colonoscopy procedures performed by an endoscopist). CIR was likely improved with each additional 100 annual colonoscopy procedures (OR 1.17, 95% CI 1.08 to 1.28), though the 95% prediction interval (0.96 to 1.43) included the null value of 1.00. There was a trend toward decreased overall adverse events per additional 100 annual procedures that did not meet significance (OR 0.95, 95% CI 0.90 to 1.00), although there was a decreased incidence of colonic perforations with increasing colonoscopy volume. Figure 1 - Forest plots demonstrating the odds of A) detecting an adenoma, B) intubating the cecum, and C) incurring an overall or specific adverse event, per additional 100 annual procedures, for total and screening procedures. The performance of higher annual colonoscopy volumes correlates with higher CIR, but not with ADR or PCCRC. Trends toward lower AE rates were also associated with higher volumes. Data are lacking on endoscopists performing fewer than 100 annual colonoscopies. Future studies should focus on extremes in the performance range to more clearly elucidate associations between volume and colonoscopy outcomes.