DEVELOPMENT OF A PREDICTIVE RISK CALCULATOR FOR ADVERSE EVENTS AND EMERGENCY DEPARTMENT UTILIZATION IN ULCERATIVE COLITIS

Abstract BACKGROUND Patients with inflammatory bowel disease (IBD) frequently experience unplanned emergency department (ED) visits, reflecting both disease activity and systemic barriers to care. Current clinical tools do not provide a simple, actionable way to stratify ED risk in IBD, limiting opportunities for timely intervention and resource allocation. Objective We aimed to develop and validate a practical, clinician-facing risk calculator that predicts the 12-month probability of ≥ 1 ED admission in patients with ulcerative colitis (UC). UC was chosen as the initial model given its more uniform disease pattern compared to Crohn’s disease. METHODS Using nationally representative claims and survey data, we identified key demographic, socioeconomic, and clinical predictors of ED use. Variables included: age, sex, insurance type, geographic region and urbanicity, distance to nearest gastroenterologist, cost-related care delays, prior ED visits, and prior IBD medication exposure (steroids, biologics). A multivariable linear regression model was used to generate individualized probability estimates. RESULTS The calculator outputs: (1) a calibrated probability of ≥ 1 ED visit in the next 12 months, (2) a categorical risk band (low < 10%, medium 10–20%, high >20%), (3) the top three contributing drivers, and (4) an interactive 「what-if」 slider for modifiable factors (e.g., cost delays, distance to care). Prototype one-click actions—such as expedited GI referral or social work consultation—were integrated for clinical workflow usability. CONCLUSION We present the first UC-specific ED risk calculator designed for real-time clinical decision support. By identifying high-risk patients and surfacing modifiable drivers, this tool has the potential to reduce unplanned ED visits, improve care coordination, and inform future expansion to Crohn’s disease populations.