CADeT-CS

Deep Learning Computer-aided Polyp Detection Reduces Adenoma Miss Rate: A United States Multi-center Randomized Tandem Colonoscopy Study (CADeT-CS Trial)

Patient / Population Intervention / Exposure Comparison Outcome

In 232 patients undergoing screening or surveillance colonoscopy at four US academic centres, CADe-assisted colonoscopy performed first reduced the adenoma miss rate at tandem colonoscopy compared with high-definition white-light colonoscopy performed first.

N
232
Design
Multicentre single-blind randomised tandem-colonoscopy trial, USA
Endpoint
Adenoma miss rate (AMR)
Relevance
2
ResultAMR 20.12% vs 31.25% (OR 1.80, 95% CI 1.08–3.02; P=0.0247) in 223 analysed patients; sessile serrated lesion miss rate 7.14% vs 42.11% (P=0.0482); first-pass adenomas per colonoscopy 1.19 vs 0.90 (P=0.0323). First-pass ADR 50.44% vs 43.64% (P=0.3091).
Glissen Brown JR, et al. Clin Gastroenterol Hepatol. 2022;20(7):1499-1507.e4. 10.1016/j.cgh.2021.09.009
Discussion & critique

First US evidence for CADe and the first tandem design in the cluster, showing that the device finds adenomas human eyes pass over rather than merely more hyperplastic polyps. Small (232 randomised, 223 analysed), endoscopists unblinded by necessity, a miss-rate surrogate, and a non-significant first-pass ADR difference; the serrated-lesion result rests on 14 and 19 lesions.