Development of a Clinical and Biomarker‐Based Score to Predict Microscopic Colitis in Patients With Chronic Watery Diarrhoea

BACKGROUND: Colonoscopy with stepwise colonic biopsies is the only method for confirming microscopic colitis in patients with chronic watery diarrhoea. Several clinical scores (Cotter, Kane and Atieh) have been proposed to guide biopsy decisions; however, their diagnostic accuracy remains limited. AIMS: To develop a derivation-stage clinical and biomarker-based score incorporating faecal calprotectin to predict microscopic colitis in patients with chronic watery diarrhoea and to assess its diagnostic accuracy compared with previously reported scores. METHODS: A prospective observational study was conducted at two centres, including adults with chronic watery diarrhoea. Clinical data, faecal calprotectin and colonoscopy with biopsies were recorded. Multivariate logistic regression was used to identify independent predictors of microscopic colitis. Score weights were derived from regression β coefficients. Diagnostic performance was assessed using ROC curve analysis, calibration metrics and bootstrap internal validation with 1000 resamples. RESULTS: , and faecal calprotectin > 81 μg/g. It showed excellent discrimination with an AUC of 0.90 (optimism-corrected AUC 0.88), and no false-negative cases were observed at the sensitivity-driven cut-off. Previously published scores showed lower discrimination in this cohort, with AUCs ranging from 0.63 to 0.72. CONCLUSION: A score integrating simple clinical variables and faecal calprotectin showed promising diagnostic accuracy for microscopic colitis in this derivation cohort and may support future risk-stratification strategies in patients with chronic watery diarrhoea.