Another Real-World U.S. Trial of Computer-Aided Detection Fails To Impress
Douglas K. Rex, MD, MASGE, reviewing Nehme F, et al. Gastrointest Endosc 2023 Feb 17.
Randomized controlled trials (RCTs) have generally shown consistent improvements in adenoma detection rates (ADRs) with the use of computer-aided detection (CADe) during colonoscopy, but a recent real-world assessment of polyp detection gains in a single U.S. center was negative. The current study is a second such report, also negative.
At a single tertiary referral center in the U.S., 23 endoscopists utilized a CADe program for colonoscopy, with results compared with those of historical controls. CADe was activated in only 52.1% of overall cases and 55% of screening and surveillance cases, and it was activated more often by high versus low ADR for endoscopists. Overall adenomas per colonoscopy (APC) did not improve after the introduction of CADe (1.08 vs 1.04). When only screening and surveillance cases in which CADe was activated were considered, APC was still not better (1.27 vs 1.17; P=.45). ADR also did not improve.
According to survey responses from endoscopists, before implementation, 62.5% thought CADe would improve ADR, but after implementation, only 23.5% believed it would improve ADR. Among the concerns with CADe during colonoscopy were that it caused too many false positives (82%), was distracting (58%), or prolonged procedure time (47%).
Note to readers: At the time we reviewed this paper, its publisher noted that it was not in final form and that subsequent changes might be made.
Nehme F, Coronel E, Barringer DA, et al. Performance and attitudes toward real-time computer-aided polyp detection during colonoscopy in a large tertiary referral center in the United States. Gastrointest Endosc 2023 Feb 17. (Epub ahead of print) (https://doi.org/10.1016/j.gie.2023.02.016)