Computer-Aided Polyp Detection Increases Adenoma Detection Tate in a High Adenoma Detecting Group: A Multi-Site Community Practice Experience

Document Type


Publication Date



Institutions: Tower Health, USA; U.S. Digestive Health, USA. Disclosure compliance: I understand. Participant disclosure: Parth Desai: NO financial relationship with a commercial interest; Nicholas Giordano: NO financial relationship with a commercial interest; Thomas Wasser: NO financial relationship with a commercial interest; Dale Whitebloom: NO financial relationship with a commercial interest; Nirav Shah: Speaking and Teaching: Medtronic. Introduction: Real-time computer-aided polyp detection (CADe) systems have demonstrated efficacy in increasing adenoma detection rates (ADR) during colonoscopy in randomized controlled trials. Subsequent real-world studies of CADe systems, from tertiary centers, have demonstrated similar ADR in CADe and non-CADe groups. However there remains a paucity of data on the benefit of CADe systems in community-based practice. This retrospective observational study evaluates the influence of CADe system implementation on ADR in a high ADR-detecting community-based gastroenterology practice. Methods: Colonoscopy data performed by 40 gastroenterologists from two ambulatory surgery centers was extracted from the GIQuIC database. Data was compared from before and after CADe implementation time periods: June 2021 through March 2022 (pre-CADe) and September 2022 through June 2023 (post-CADe). Exclusion criteria included colonoscopies performed on patients under age 40 or history of inflammatory bowel disease and those performed by gastroenterologists who opted out of CADe use. Baseline characteristics including age, gender, race, insurance, and risk assessment were analyzed in pre- and post-CADe groups. ADR, advanced adenoma detection rate, and sessile serrated adenoma detection rates were compared pre- and post-CADe implementation. Chi-squared tests were used for categorical variables and T-tests for continuous variables. Results: A total 37,975 colonoscopies were analyzed (pre-CADe, n=15,298; post-CADe, n=22,677). There was a similar distribution of males in the pre- and post-CADe groups (47.3% vs 46.7%, p=0.215). There was a statistically significant but clinically minute difference in the distribution of white compared with non-white patients between the periods (87.3 vs 85.7, p<0.001). There were more high-neoplasm risk patients in the pre-CADe period (53.9% vs 48.4%). ADR increased from the pre- to post-CADe periods (43.5% vs. 47%, p<0.001) across all endoscopists, an 8% relative rate increase. Advanced neoplasms (defined as size ≥ 10mm, high grade dysplasia or villous component) were similar between the two groups (6.7% vs 6.8%, p=0.535). Serrated adenoma detection was diminutively higher in the pre-CADe group (12.0% vs 11.2%, 0.021). Withdrawal time was similar between the pre- and post-CADe groups (10.57 vs 10.61 minutes, p=0.414). Conclusion: Post-CADe implementation was associated with increased ADR within this multi-site analysis, despite a lower proportion of high neoplasm risk patients compared with the pre-CADe group. Although baseline ADR surpassed the national mean of 39% at 43.5%, the post-CADe implementation ADR improved to 47%. High adenoma detecting community practice groups may increase ADR through use of CADe systems. Further studies are needed to assess how CADe systems affect colon cancer-related mortality and morbidity outcomes. [Formula presented] [Formula presented]

Publication Title

Gastrointestinal Endoscopy




6 Supplement

First Page


Last Page



Digestive Disease Week (DDW), 2024 held 2024-05-18 to 2024-05-21 in Washington, US.

This document is currently not available here.

Open Access