First Real-World Evidence of an AI-Enhanced Digital Collaborative Care Model to Improve IBS Symptoms.

Background And Aims: Health systems struggle to deliver guideline-recommended multidisciplinary care to patients with irritable bowel syndrome (IBS). Digital collaborative care models (DCCMs) that integrate technology with experienced providers offer a promising solution for improving IBS management. We aimed to evaluate whether a novel DCCM improved clinical outcomes in IBS.

Methods: A prospective, longitudinal uncontrolled single-arm study design was used to assess the Ayble Health program. Participants were recruited online via social media, clinic, or employer. Data were prospectively collected from 202 participants (78% female; 78% white) with active IBS symptoms at baseline (≥ 75 on the IBS symptom severity scale (IBS-SSS)) and completed at least one follow-up symptom survey. All participants engaged in at least one care pathway: (1) a multidisciplinary care team, (2) a nutrition program with a personalized elimination diet, and (3) a brain-gut behavioral therapy (BGBT) program. Each pathway was supported by AI algorithms trained on a large, multimodal GI dataset to identify and communicate key trends in patient-reported outcomes, further personalizing care plans.

Results: Of the 202 participants, 197 (98%) participated in the nutrition pathway, 152 (75%) the BGBT pathway, and 156 (77%) the care team pathway. The majority of participants (62%) enrolled in all three pathways. Participants experienced a 140-point decrease in IBS-SSS, on average, with 86% experiencing a ≥ 50-point reduction.

Conclusion: The novel DCCM successfully delivered evidence-based care to participants with active IBS symptoms, with clinically significant, sustained symptom relief. Randomized clinical trials are recommended to assess cost and treatment efficacy compared to standard of care approaches.