Computer-aided colonoscopy minimizes adenoma miss rate

01 Nov 2020
Doctor removes polyps during colonoscopy
Although the prevalence for perforation during colonoscopy is small, patients should not neglect the underlying medical risks.

Computer-aided detection (CADe) systems help reduce adenoma miss rate during colonoscopy, a new study has found.

Researchers conducted a prospective study of patients referred for diagnostic or surveillance colonoscopies. Participants were randomly assigned to receive routine (n=185; mean age, 47.19±10.38 years) or CADe-assisted (n=184; mean age, 47.72±10.82 years) colonoscopies. All patients also underwent a second-pass colonoscopy using the opposite method.

Adenoma miss rate (AMR), the number of adenomas detected in the second-pass round divided by the total number of adenomas, was the primary outcome.

The AMR was significantly lower in the CADe procedures than in the routine white light colonoscopies (13.89 percent vs 40.00 percent; p<0.0001), as was the polyp miss rate (12.98 percent vs 45.90 percent; p<0.0001).

The overall adenoma (ADR) and polyp (PDR) detection rates, as well as adenoma (APC) and polyp (PPC) per colonoscopy, did not differ between patients who underwent either CADe or routine colonoscopy during the first-pass round. However, the PDR (p=0.001), APC (p<0.001), and PPC (p<0.001) were all significantly better when CADe colonoscopy was performed for the first-pass.

Similarly, CADe colonoscopy for the second pass yielded significantly better PDR (p<0.001), APC (p=0.001), and PPC (p<0.001) values, as well as ADR (p=0.043).

On the other hand, CADe colonoscopy consistently detected false polyps (n=67), which were features continuously traced by the CADe system but ultimately deemed as not polyps by the endoscopists. There were no missed polyps.

Editor's Recommendations