COLO-DETECT
2024
Polyp detection with colonoscopy assisted by the GI Genius artificial intelligence endoscopy module compared with standard colonoscopy in routine colonoscopy practice
In 2,032 adults undergoing screening or symptomatic colonoscopy, computer-aided detection increased mean adenomas per procedure compared with standard colonoscopy.
Result
Primary endpoint: Mean adenomas per procedure
Adenomas per procedure 1.56 vs 1.21 (adjusted mean difference 0.36, 95% CI 0.14–0.57; adjusted IRR 1.30, 95% CI 1.15–1.47; P<0.0001). ADR 56.6% vs 48.4% (difference 8.3%, 95% CI 3.9–12.7; aOR 1.47, 95% CI 1.21–1.78). Adverse events 25 vs 19, none attributed to the device.
- Subtopic
- Adenoma detection
- Design
- Multicentre open-label pragmatic RCT, UK
- Relevance
- Important: one of several pillars.
- Source
- Seager A, et al. Lancet Gastroenterol Hepatol. 2024;9(10):911-923.doi 10.1016/S2468-1253(24)00161-4
Discussion & critique
The largest pragmatic CADe trial in routine practice and the UK evidence base for GI Genius, extending the Repici result to unselected NHS lists with screening and symptomatic indications. Open-label with unmasked colonoscopists, a polyp-count surrogate rather than cancer outcomes, imputation for missing outcome data, and Medtronic funding.