We examined the relationships between SUDs and housing outcomes across racial/ethnic subgroups in a cohort of Veterans housed in HUD-VASH in Los Angeles. We identified an overall association between SUDs and negative PSH exits. However, in analyses stratified by race and ethnicity, we found this association varied by race/ethnic group. There was no statistically significant association between SUDs and negative PSH exits for Black, Non-Hispanic White, and Hispanic/Latino residents. Though it did not reach statistical significance, for residents of Hispanic/Latino ethnicity, the effect of presence of SUDs on negative PSH exits was nearly double that of White and Black subgroups. We observed a statistically significant positive association for Other/Mixed race HEVs. Notably, the relationships between SUDs and negative PSH exits were much stronger among Other/Mixed HEVs compared to other racial/ethnic groups, although this group comprises a small subset of HEVs (4%).
Our findings differ from prior studies that broadly examined SUDs as associated with increased rates of premature or unwanted exits from PSH but did not focus on race/ethnic differences (6). In these data, among most racial/ethnic subgroups, the effects of SUDs on negative PSH exits were not significant, which suggests that current strategies to retain residents with SUDs in PSH, (e.g., improving timely access to supportive services, including behavioral health care) may be effective among these subgroups (4). However, despite these efforts however, our analyses highlight potentially important disparities in PSH housing outcomes among Hispanic/Latino and Other/Mixed race PSH residents with SUDs.
Among Hispanic/Latino and Other/Mixed race residents, disparities in health behaviors, including SUD service utilization, may contribute to the increased effect of SUDs on negative PSH exits. In prior literature, Veterans of Hispanic/Latino and Other/Mixed race/ethnicity were found to have SUD prevalence rates nearly two times that of clinically documented SUD (19). Further indicating a gap in VA treatment receipt for SUD among these minoritized groups, White Veterans diagnosed with SUDs were found to be much more likely to receive treatment for SUD diagnoses as compared to Hispanic/Latino Veterans diagnosed with SUDs (27). This trend is also seen among Asian and NHPI populations. Across the general population, outside of Veteran-specific literature, minoritized communities have been shown to severely underutilize SUD treatment. Underutilization among these populations is often attributed to barriers to access including stigma, cost, lack of knowledge, and cultural attitudes (28). We suspect that tailored implementation approaches designed to increase adoption of evidence-based SUD treatments dissemination within VA (e.g., using peers to activate HEVs from racial/ethnic minoritized groups) may address these disparities and increase health equity within the PSH program.
Prior research has found that other potentially relevant factors in examining relationships between SUDs and negative PSH exits include socioeconomic disparities associated with developing SUDs (29), differential stigma associated with specific substance use (30), and other social factors associated with SUDs (e.g., disparate marketing for substances in low-income and minority communities [31]). Racial/ethnic minority Veterans are also noted to have an increased risk of adverse SUD and psychiatric treatment outcomes (e.g., involuntary hospitalizations, shorter treatment duration) compared to their Non-Hispanic White peers (32). In general, researchers have attributed increased risk of SUDs among racial/ethnic minority populations to differential access to health services, social supports, and other healthy coping mechanisms (e.g., professional/clinic services, social service resources, community infrastructure). We note that, in this study, these disparities may be mitigated in part by the VA infrastructure; during the study period, all HUD-VASH residents were eligible for VA healthcare which awarded them equitable potential access to all health services, including SUD treatment.
Strengths and limitations
The primary strength of this study is its ability to examine longitudinal data for a large subset of PSH enrollees in a system that integrates housing and health services. VA administrative and homeless registry data provides robust information related to diagnoses, date of housing move-in, exits from PSH enrollment, and the competing risk of death.
This study also had limitations. First, misclassification of PSH exits (i.e., negative, positive, neutral) may have occurred. Each exit is categorized using standardized reasons for exit which omit granular details about factors contributing to each participant’s PSH exit. Second, while there a large sample size for the entire cohort, when stratifying by race/ethnicity, small proportions in some subgroups (i.e., Asian, AIAN, NHPI, Other, and Mixed) necessitated collapsing of these subgroups into one category (“Other/Mixed”) which comprised 4% of HEVs. Future studies with larger samples sizes and/or utilizing qualitative methods could help provide greater insights into the potential vulnerabilities of racial/ethnic subgroups with smaller populations. Third, these analyses were based on diagnosed and documented SUDs, which may vary by race/ethnicity. In addition, in this study, we combined all diagnoses of substance use disorders within the relevant time frame (two years prior to housing move-in). Future research would benefit from examining differences in housing retention associated with specific substances used. We note specific complexities in data interpretation related to persons who only had cannabis use disorder to classify them as having a SUD; cannabis was legalized in the state of California in 2016, including at the study site (33). Fourth, it is possible that the high rates of comorbid mental health disorders and SUDs among this population overshadowed effects of SUDs on negative PSH exits. Future studies may benefit from assessing the relationships between comorbid mental health and SUD diagnoses on housing outcomes. Last, as a study conducted with one large and urban VA, it is unclear how much our findings extrapolate to a national HUD-VASH sample, or to homeless-experienced consumers who receive PSH services or health services outside the VA.