Practices and perceptions of advanced endoscopic imaging techniques for colorectal neoplasia detection in patients with inflammatory bowel disease: report of a survey among Canadian inflammatory bowel disease specialists

Abstract

Background and Aims

This study surveyed Canadian inflammatory bowel disease (IBD) specialists to understand their perceptions, preferences, and practices regarding image-enhancement modalities (including dye-spray chromoendoscopy [DCE], virtual chromoendoscopy [VCE], and artificial intelligence [AI]-aided endoscopy) for colorectal neoplasia (CRN) detection in IBD, to inform future studies and recommendations.

Methods

Developed collaboratively by IBD experts and methodologists, the survey assessed physicians' attitudes toward colonoscopy image-enhancement modalities for CRN surveillance in IBD, evaluating 5 domains: current practices, familiarity with advanced techniques, perceptions of AI-assisted endoscopy, barriers to adoption, and research priorities. It was distributed to IBD specialists affiliated with the Canadian IBD Research Consortium and conducted per The Strengthening the Reporting of Observational Studies in Epidemiology guidelines.

Results

Of 82 specialists invited, 33 responded (40.2%). Most worked in academic centers (90%); 39.4% managed large IBD practices. VCE was the most used advanced imaging modality (84.8%), followed by DCE (33.3%), high-definition white-light endoscopy (HD-WLE) plus nontargeted biopsies (33.3%), AI-assisted colonoscopy (18.2%), and HD-WLE alone (6.1%). Although 60.6% considered DCE and VCE similarly effective, drawbacks included bowel preparation (VCE: 26.6%; DCE: 24.1%) and poor visualization with inflammation (VCE: 26.6%; DCE: 12%). AI-assisted colonoscopy was seen as promising for CRN detection; 78.8% agreed randomized controlled trials are needed, and 39.4% would adopt AI if it improved CRN detection by 5% over HD-WLE.

Conclusions

This survey highlights a shift toward VCE over DCE, perceived drawbacks of chromoendoscopy, limited availability and uncertainty about AI-assisted colonoscopy, and a desire for research to demonstrate AI's utility for CRN detection in IBD.