Development and validation of a model combining bowel ultrasound and inflammatory biomarkers for evaluating disease...

This study evaluated the diagnostic accuracy of bowel ultrasound combined with inflammatory biomarkers for assessing ulcerative colitis disease activity.

The single-centre derivation cohort retrospectively enrolled 60 patients with ulcerative colitis from our institution who underwent colonoscopy and bowel ultrasound. Patients were stratified into active disease and remission groups. Bowel ultrasound parameters and clinical features were analysed. Diagnostic performance was assessed using receiver operating characteristic curve analysis. The resulting model was externally validated from 35 patients at an independent tertiary hospital retrospectively. A nomogram was plotted for clinical application.

In the derivation cohort (31 remission: age 42.6 ± 18.3 years, male 48%; 29 active: age 44.4 ± 14.5 years, male 48%), univariate analysis identified bowel wall thickness (cutoff value 3.95 mm), submucosal index (cutoff value 46.1%), bowel wall flow (cutoff value grade 2), faecal calprotectin (cutoff value ≥ 200 μg/g), and erythrocyte sedimentation rate (cutoff value 26.5 mm/h) as significant predictors of disease activity. Bowel wall thickness and flow were independent risk factors. The combined model demonstrated an area under the curve of 0.908 (95% CI 0.829-0.986). External validation (19 remission and 16 active) provided preliminary evidence supporting the model's discriminative ability, showing an area under the curve of 0.862 (95% CI 0.743-0.981). The nomogram provided individualised risk prediction with acceptable calibration in both cohorts.

An integrated model combining bowel ultrasound and inflammatory biomarkers was developed and externally validated for the assessment of ulcerative colitis activity. The model, presented as a clinically practical nomogram, exhibits acceptable calibration, supporting its potential as a valuable noninvasive tool for patient management.