Association Between Adenoma Detection Rate and Prevalent Colorectal Cancer Detection Rate in a National Colonoscopy Registry

Abstract

Background & Aims: Although the adenoma detection rate (ADR) is associated with postcolonoscopy colorectal cancer risk, it is unknown to what extent this reflects missed colorectal cancer (CRC) vs missed precancerous lesions. The association between physician ADR and prevalent CRC detection during colonoscopy were evaluated. Methods: The study used the cross-sectional 2019–2022 GI Quality Improvement Consortium data for more than 1.73 million colonoscopies performed by 3567 endoscopists for screening or abnormal fecal test follow-up from 683 US endoscopy units. Endoscopist ADR and sessile serrated lesion detection rate (SSLDR) were determined based on screening examinations. Results: CRC was detected in 0.3% of screening and 1.5% of follow-up colonoscopies. From lowest to highest endoscopist ADR quintile, CRC detection increased from 26.6 (95% CI, 24.4–27.9) to 33.1 (95% CI, 29.7–33.7), and from 107.8 (95% CI, 96.2–129.4) to 164.7 (95% CI, 140.8–188.6) per 10,000 screening and abnormal fecal test follow-up colonoscopies, respectively. In multivariable models with lowest ADR quintile as reference, the odds ratios of CRC detection in the highest ADR quintile were 1.27 (95% CI, 1.14–1.41) for screening and 1.50 (95% CI, 1.16–1.93) for abnormal fecal test follow-up colonoscopies. Compared with high-ADR/high-SSLDR endoscopists, the ORs of CRC detection were lower for low-ADR endoscopists irrespective of SSLDR (high-SSLDR, 0.87; 95% CI, 0.80–0.96; low-SSLDR 0.92; 95% CI, 0.85–0.98), but similar for high-ADR/low-SSLDR endoscopists. Conclusions: ADR reflects prevalent CRC detection as well as detection and removal of CRC precursors. Our findings suggest that postcolonoscopy CRC is not uncommonly due to missed CRC, especially among endoscopists with low ADR.