Alcohol-associated liver disease (ALD) is a leading cause of preventable liver-related morbidity and mortality, with hospitalizations for ALD increasing significantly in recent years. Many patients with ALD and active alcohol use disorder (AUD) are discharged from the hospital without receiving evidence-based behavioral interventions, despite hospitalization representing a teachable moment when patients may be more motivated to change. Psychosocial support in this setting is often limited by short inpatient stays, provider shortages, and workflow barriers. Novel digital health approaches are needed to increase access to structured, personalized behavioral support during hospitalization.
This project aims to assess the feasibility, usability, and clinical relevance of SOBER-VR, a virtual reality (VR) and artificial intelligence (AI)-powered digital therapeutic that delivers motivational interviewing (MI) to hospitalized patients with ALD and active alcohol use. SOBER-VR uses immersive VR to enhance emotional salience and a large language model (LLM)-driven conversational agent to simulate reflective, MI-consistent interactions. The intervention is brief (10-15 minutes), self-guided, and delivered using sanitized, standalone VR headsets with real-time support from trained staff.
Using the NIH ORBIT Model for behavioral intervention development, this Phase Ib single-arm feasibility study will recruit 30 hospitalized patients with ALD and active alcohol use. Participants will use SOBER-VR during their inpatient stay and complete structured assessments post-discharge. The study will evaluate feasibility (recruitment, engagement, and retention), acceptability, usability, and safety of SOBER-VR. Preliminary signals of clinical impact, such as readiness to change, attitude to AUD treatment, among others, will also be explored.