High-end intestinal ultrasound versus mid-end systems benchmarked against tandem ileocolonoscopy in inflammatory bowel disease (HUMID): a paired prospective, validating confirmatory study

Le résumé fourni par la source

Background: Intestinal ultrasound (IUS) is redefining inflammatory bowel disease (IBD) monitoring, but global adoption remains limited by the cost of high-end systems. This study evaluated whether widely available mid-range equipment provides comparable diagnostic accuracy without additional cost. Methods: We conducted a prospective, cross-sectional, paired diagnostic accuracy study at a single IBD centre in India from September 2024 to October 2025. We included patients aged 18-75 years with confirmed ulcerative colitis (UC) or Crohn's disease (CD). Trained operators performed same-day blinded assessments using a mid-end (Siemens ACUSON S2000) and a high-end system (Samsung RS80 EVO). Blinded ileocolonoscopy with central review served as the reference. Endoscopic remission was defined as Ulcerative Colitis Endoscopic Index of Severity (UCEIS) ≤ 1 for UC and Simple Endoscopic Score for Crohn's Disease (SES-CD) ≤ 2 for CD. Sonographic activity was assessed using the Milan Ultrasound Criteria (MUC) for UC and the International Bowel Ultrasound Simplified Activity Score (IBUS-SAS) for CD. Diagnostic accuracy, segment-wise performance, and reclassification impact were evaluated. Paired comparisons used McNemar testing for proportions, receiver operating characteristic (ROC) analysis with area-under-the-curve (AUC) comparison, and agreement analyses using Bland-Altman and concordance methods. This study is registered with ClinicalTrials.gov, NCT06938295. Findings: 450 patients underwent paired IUS and ileocolonoscopy (239 UC, 211 CD; median age 36 years; 32% female). In UC, sensitivity was 94.2% with the mid-end and 97.1% with the high-end system (Δ +2.9%; 95% CI -3.8 to 9.6), with identical specificity (54.5%). In CD, sensitivity was 85.1% versus 87.6% (Δ +2.5%; 95% CI -8.2 to 13.0) and specificity 62.0% versus 60.0% (95% CI -28.0 to 24.2); accuracies differed by <3%. Segment-wise performance showed no significant differences across any bowel regions. ROC analyses showed good-excellent discrimination [AUC: 0.75-0.93 (UC); 0.83-0.94 (CD) in various segments], with overlapping curves and strong score correlation. Reclassification analysis showed a net gain of +6 patients (diagnostic odds ratio 1.8, 95% CI: 0.9-3.5) for high end machine. Bland-Altman and concordance analyses demonstrated minimal bias and high agreement. Decision-curve analysis confirmed equivalent net clinical benefit. Interpretation: High-end IUS conferred no measurable diagnostic or clinical advantage over mid-range equipment when examinations were performed by trained operators using standardised scoring systems, supporting mid-range systems as scalable and cost-effective tools for IBD monitoring. Funding: No external funding.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.