Commentary on Somers et al. (2015): Housing First--a platform for recovery?

Abstract

Housing First effectively ends homelessness, yet randomized controlled trials have shown that it does not reduce daily substance use among homeless adults experiencing mental illness. Whether Housing First provides a platform for recovery, however, remains unexamined through experimental designs. In order to do so would require measuring ‘readiness to change’. What do we make of Somers et al.’s finding that there are no differences in daily substance use between homeless adults assigned to Housing First (HF) and those assigned to usual care 1? Some researchers have suggested that such findings indicate that HF is not an optimal intervention for homeless adults with active addictions 2. Others have interpreted these same sorts of results as supporting the use of HF for homeless adults with comorbid mental illness and addiction, because substance use is no worse despite providing an independent living arrangement without mandating treatment 3. What is clear is that the lack of improvement in substance use outcomes is at odds with so many other positive findings associated with HF that the authors mention, including superior quality of life, more days stably housed, fewer criminal convictions and fewer visits to emergency rooms. Debates about the effectiveness of HF for homeless adults with active addictions depend on the outcome of interest. If the goal is ending homelessness, HF works. If the objective is reducing substance use, it appears that HF does not work. Yet confusion may persist if this study is viewed as addressing the claim that HF ‘creates a foundation on which the process of recovery can begin’, and that it may ‘serve as a motivator for consumers to refrain from drug and alcohol abuse’ 4. Sommers et al. cite this claim as a rationale for their study even though it was not addressed by their findings. That is, this study did not examine the distinction between readiness to change (RTC) and actual recovery from substance use 5, 6. This is an important distinction when considering that qualitative research supports the idea that HF provides a platform for recovery in a way that usual care does not 7. If this is shown to be true, which would require including a validated measure of RTC 8 in experimental studies of HF, then adding specific interventions to target substance use in HF could result in a dramatic improvement in outcomes. However, it is precisely because research has not yet tested the hypothesis that HF promotes RTC that there has been, as the authors note, limited discussion about ‘the problem that arises when participant choice is rendered moot because effective addiction treatment is simply unavailable’. To date, we simply do not know the scope of this problem. The truth is that we know little about how adults who are in a position to be assigned to HF recover not only from substance use but also from a life-time cumulative adversity 9, 10. Considering that HF is located disproportionately in communities that experience concentrated disadvantage 11 and are associated with health-risk behaviors, including substance use 12, recovery is even more difficult. In fact, few studies have considered whether and how HF tenants compare to their community neighbors 13. For this reason, it is not clear whether effective treatment options need to be tailored to HF or whether such options simply need to be made available in these communities. None.