Objectives and study:
There is a large interobserver variability in evaluating mucosal lesions of inflammatory bowel disease (IBD), especially in pediatric patients. This multicenter prospective observational study aims to evaluate interobserver agreement (IOA) among pediatric endoscopists in assigning validated IBD endoscopic scores in children.
Fifteen videos of follow-up ileocolonoscopies in children with IBD were recorded and selected as cases. Eleven pediatric endoscopists from different centers blindly evaluated all videos and calculated scores: either Ulcerative Colitis Endoscopic Index of Severity (UCEIS) or Simple Endoscopic Score for Crohn Disease (SES-CD). Scores from all reviewers were compared in order to calculate IOA for general videos and specific sections. Scores from an expert adult reader were used to calculate possible reviewer's characteristics affecting scores’ reliability.
Intraclass correlation was 0.298 (95% confidence interval [CI]: 0.13–0.55) for ulcerative colitis (UC) and 0.266 (0.11–0.52) for Crohn disease (CD). When a disease activity categorization was adopted (remission, mild, moderate, severe activity) Fleiss kappa coefficient was 0.408 (0.29–0.53) for UC and 0.552 (0.43–0.73) for CD. When stratified by item, vascular pattern of UC was the most reliable item IC: 0.624 (0.321–0.854). In multivariable analysis, none of the reviewer's characteristics affected the readers’ errors.
This multicenter study shows low agreement among pediatric endoscopists in evaluating endoscopic scores in children with IBD. By using disease activity categorization, agreement slightly increased, mostly for CD. All readers showed a low-grade concordance with the expert adult gastroenterologist's evaluations. Future-specific training programs should be considered to increase IOA in using IBD endoscopic activity scores.