Housing-based support models for individuals experiencing homelessness, substance use and mental health needs: a rapid review

Abstract Background Individuals experiencing homelessness and substance use and mental health needs face high rates of mortality. Stable housing is a critical determinant of health, yet the evidence on existing housingbased support models and their effectiveness for improving behavioural health outcomes for this population remains unclear. Methods Given rapid policy shifts favouring treatment focused approaches to address the needs of this population, a rapid review was undertaken to synthesize published evidence on the key components and outcomes of housingbased support models for adults experiencing homelessness with substance use with or without co-occurring mental health conditions. Five databases (CINAHL, Embase, Medline, APA PsycInfo, and Web of Science) were searched for peerreviewed studies published between 1 January 2000 and 30 April 2025. After screening 13,393 records, 57 studies were included, comprising 8 randomized controlled trials, 30 quasi-experimental studies, 6 qualitative studies, 5 mixed-methods studies, and 8 other designs. Study quality was assessed using Joanna Briggs Institute tools and data were extracted on intervention characteristics; model features; comparators; and housing, substance use, mental health, and service-related outcomes. Results Four program types were identified: [1] scattered site [2] single site, [3] transitional housing, and [4] outreach support to housing programs. Housing First (HF) models consistently improved housing stability compared with treatment first (TF), linear treatment-contingent model or treatment-as-usual (TAU) conditions. HF effects on substance use were heterogeneous; alcoholrelated outcomes improved more consistently than illicit drug use. Mental health outcomes were similarly mixed, with some studies showing no differences between HF and comparison conditions and others reporting improvements in psychological distress. Strength of evidence was strongest for housing stability and mixed for other outcomes. Conclusions Housing based support models, particularly HF, showed the strongest evidence for improving housing stability among people experiencing homelessness with substance use with or without co-occurring mental health conditions. Evidence for reductions in acute healthcare utilization, including psychiatric hospitalization and overall costs, was promising but less definitive. Existing evidence supports scaling up low barrier HF programs, while integrating expanded voluntary substance use treatment, mental health, harm reduction, and wraparound services within housing. While some TF models reported positive outcomes, the evidence base was more limited and heterogeneous than for HF, highlighting the need for further comparative research. More robust longerterm studies are needed to identify which program components that best support substance use health among this highneeds population. Clinical trial number NA.