Categories
Change Password!
Reset Password!
Computer-aided diagnosis (CADx) has emerged as a promising tool for predicting polyp histology during colonoscopy, with the capacity to minimize redundant excision of nonneoplastic polyps. However, the true clinical benefits and risks linked with CADx remain uncertain and warrant further exploration.
Computer-aided diagnosis does not offer any added benefit or risk when managing small rectosigmoid polyps (≤5mm) during colonoscopy.
Computer-aided diagnosis (CADx) has emerged as a promising tool for predicting polyp histology during colonoscopy, with the capacity to minimize redundant excision of nonneoplastic polyps. However, the true clinical benefits and risks linked with CADx remain uncertain and warrant further exploration. Hence, this systematic review and meta-analysis was carried out to critically investigate the benefits and risks of employing CADx in colonoscopy for small rectosigmoid polyps management.
Comprehensive searches were conducted in Medline, Embase, and Scopus for studies issued up until 22 December 2023. Researchers included histologically validated diagnostic accuracy studies that examined the real-time performance of clinicians in predicting neoplastic changes in small rectosigmoid polyps, both with and without the assistance of CADx during colonoscopy.
Clinical benefits and risks were assessed by comparing the diagnostic accuracy of endoscopists before and after CADx assistance.
Using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework, the certainty of evidence was examined. The primary benefit was measured by the proportion of nonneoplastic polyps correctly predicted not to require removal with CADx. The primary harm was estimated by the proportion of neoplastic polyps incorrectly classified, leading to their retention in situ. For both outcomes, histology served as the reference benchmark.
In total, 10 studies involving 3620 people with 4103 small rectosigmoid polyps were included in the analysis. Studies investigating CADx alone (9 studies; 3237 polyps) illustrated a sensitivity of 87.3% and specificity of 88.9% for predicting neoplastic changes. In investigations comparing histological prediction prior to and after CADx assistance (4 studies; 2503 polyps), no vital differences were found in the percentage of nonneoplastic polyps anticipated not to require removal (55.4% vs. 58.4%; risk ratio [RR] 1.06), nor in the percentage of neoplastic polyps erroneously left in situ (8.2% vs. 7.5%; RR 0.95).
Despite its promising potential, CADx did not show any substantial incremental benefit or harm in the optical diagnosis of small rectosigmoid polyps during colonoscopy, highlighting the need for further research to define its true clinical value.
Annals of Internal Medicine
Computer-Aided Diagnosis for Leaving Colorectal Polyps In Situ: A Systematic Review and Meta-analysis
Cesare Hassan et al.
Comments (0)