A New Risk Prediction Model for Detecting Endoscopic Activity of Ulcerative Colitis.

Background/Aims:: Ulcerative colitis (UC) is an incurable, relapsing-remitting inflammatory disease that increases steadily. Mucosal healing has become the primary therapeutic objective for UC. Nevertheless, endoscopic assessments are invasive, expensive, time-consuming, and inconvenient. Therefore, it is crucial to develop a noninvasive predictive model to monitor endoscopic activity in patients with UC. Methods:: Clinical data of 198 adult patients with UC were collected from January 2016 to August 2022 at Huadong Hospital, China. Results:: Patients with UC were randomly divided into the training cohort (70%, n=138) and the validation cohort (30%, n=60). The receiver operating characteristic curve value for the training group was 0.858 (95% confidence interval [CI], 0.781 to 0.936), whereas it was 0.845 (95% CI, 0.731 to 0.960) for the validation group. The calibration curve employed the Hosmer-Lemeshow test (p>0.05) to demonstrate the consistency between the predicted and the actual probabilities in the nomogram of these two groups. The decision curve analysis validated that the nomogram had clinical usefulness. Conclusions:: The nomogram, which incorporated activated partial thromboplastin time, fecal occult blood test, β2-globulin level, and fibrinogen degradation products, served as a prospective tool for evaluating UC activity in clinical practices.