Nomogram-Based Outcome Prediction Models in Patients with Inflammatory Bowel Disease

Inflammatory Bowel Disease (IBD) has been proven to affect the quality of life of patients and their families, and the outcomes of IBD are varied.

This study intends to combine clinical, endoscopic, and laboratory information to develop individualized nomogram models for IBD outcome prediction.

Patients with Crohn's disease and ulcerative colitis were retrospectively included and divided into different groups (clinical remission, clinical response, endoscopic remission, and biological remission groups) according to their status. Multi-dimensional information was collected and analyzed with single-variable logistic regression, followed by a subset selection method. Features that survived selection steps were used to develop nomogram models; the performance was evaluated with the area under the curve (AUC).

After single-variable logistic regression, the feature numbers in clinical remission, clinical response, endoscopic remission, and biochemical remission groups were 17, 13, 15, and 11, respectively. After the second stage of feature selection, only 9, 5, 8, and 7 features remained, respectively. The nomogram models were based on these selected features, and the AUCs of the four groups were 0.79, 0.88, 0.78, and 0.89, respectively.

This study provides a new perspective on outcomes in patients with IBD. The nomogram models proposed in this study showed great performance (all AUCs >0.70) and could provide information for managing IBD patients.