Syntes: Learning and deskilling effects of artificial intelligence in colonoscopy among endoscopists with different levels of experience: a pragmatic, prospective trial (2026)

Källa: https://doi.org/10.1055/a-2858-7084 (Endoscopy, 2026; PMC13552228)

Verified content

• [full text] Three Norwegian clinics, November 2021 – June 2025; 13 endoscopists (7 inexperienced, 6 experienced), 5,013 colonoscopies; each endoscopist went through phase 1 (no AI), phase 2 (with computer-aided detection, CADe) and phase 3 (CADe removed), in fixed order, as their own controls.
• [full text] Inexperienced: polyp detection (PDR-5) 31.9% before, 39.5% with CADe (OR 1.43, 95% CI 1.11–1.84), 36.3% after removal (OR vs before 1.03, 95% CI 0.79–1.34). Experienced: no significant change between phases. No significant phase 1 vs phase 3 change overall or in either group.
• [full text] Mean time from first colonoscopy: phase 2 reached at 17.7 vs 7.6 months and phase 3 at 23.8 vs 11.8 months (inexperienced vs experienced).

Limits

• [full text] Some inexperienced endoscopists did few phase-3 colonoscopies.
• [assessment] Not randomised, fixed phase order, 13 endoscopists; endpoint is polyps ≥5 mm, not adenoma detection as in Budzyń et al. 2025.

Added 2026-10-16 by the project (week run 2026-10-16). [abstract] = from the published abstract; [full text] = read in the full text; [assessment] = the project's own assessment.