Advancing Predictive Modeling of Inflammatory Bowel Disease (IBD) Flares: A Data-Driven Approach Using Lifestyle and Psychosocial Factors from a Remote Monitoring Platform

The inclusion of subjective health and modifiable lifestyle and psychosocial data improved flare prediction in contrast to clinical characteristics alone, which was evidenced in the model performance matrices (accuracy, AUC, sensitivity, specificity, PPV, NPV). In multifactorial disorders such as IBD, lifestyle, and psychological stressors may intensify inflammatory responses, all of which can be controlled by lifestyle choices including diet, exercise, and stress management, for which this model underscores the need.