Integrating data analytics, peer support, and whole health coaching to improve the health outcomes of homeless veterans: Study protocol for an effectiveness-implementation trial

Background: Homelessness is a strong determinant of acute care service utilization (inpatient hospitalization, emergency department visits) among US adults. Data analytics, peer support, and patient-centered approaches can collectively offer high-quality care for homeless patients who frequently utilize acute care ( "super utilizers "). However, few outpatient programs have integrated these components and tested their effectiveness for this patient population.Objective: To test the effectiveness and implementation potential of a novel intervention that integrates data analytics with peers trained in whole health coaching ( "Peer Whole Health ") to reduce use of acute care among homeless adults.Methods: Using a randomized controlled trial design at two US Veterans Health Administration Medical Centers, we plan to enroll 220 veterans in primary care on VHA's Homeless Registry who are flagged on a super-utilizer clinical dashboard. Participants will complete a baseline interview, be randomized to Enhanced Usual Care (EUC; primary care and data analytics) or EUC plus 18 sessions of Peer Whole Health over 6 months, and be reinterviewed at 3, 6, and 9 months. Qualitative interviews with primary care staff and patients will identify facilitators and barriers to more widespread implementation of the intervention.Discussion: The primary hypothesis is that those who receive the intervention will have fewer total days of allcause hospitalization. If confirmed, the findings can provide healthcare systems that serve homeless superutilizers with a high-value approach to care that can be integrated into primary care services and reduce overall costs for these patients.