P276 Risk matrix model for prediction of one-year surgery in Crohn’s disease

Abstract

Having a simple score able to predict the risk of surgery, combining clinical, endoscopic and imaging features, could tailor the management of patients with Crohn’s disease (CD). Aim: to prospectively evaluate the one-year risk factors for surgery in CD and to build a risk matrix model for predicting the 1-year probability of surgery. Enrolled CD patients underwent prospectively clinical, laboratory, endoscopy and bowel sonography (BS) examinations within one week. Firstly, the optimal cut-off values for Simple Endoscopic Score for CD (SES-CD), bowel wall thickness (BWT) at BS and disease extension at BS in predicting surgery were identified by ROC curves; then binary logistic regression and Cox’s regression were conducted. Finally, the probabilities of surgery were computed for selected baseline levels of covariates and results were arranged in a prediction matrix model. Among 100 CD patients included in the study (males 58%, mean age 36.2 ± 11.5), 30% underwent surgery within one year. SES-CD>9 (OR 15.3; p < 0.001), BWT >7 mm (OR 15.8; p < 0.001), disease extension >33 cm (OR 8.23; p < 0.001) and stricturing/penetrating behaviour (OR 4.3; p < 0.001) were the only independent factors for one-year surgery both at binary logistic and Cox’s regression. Finally the matrix model combined these risk factors, and the probability of surgery ranged from 0.48% to 87.5% (16 combinations) (Figure 1). The risk matrix model, combining clinical, endoscopic and ultrasonographic findings, can accurately predict the one-year risk of surgery in CD. This tool could be of relevance into clinical practice, as it could be the basis for a tailored management of CD patients.