Boston low-threshold shelters (LTS) expand access to housing for chronically unsheltered people with substance use disorder, findings from a comparison analysis

Citation

Yang, S., Braverman Bronstein, A., Desmarais, J., Miah, S., Nair, S., Garcia, K., Clark, M., & Ojikutu, B., Boston low-threshold shelters (LTS) expand access to housing for chronically unsheltered people with substance use disorder, findings from a comparison analysis. Presentation at BMC Grayken Center Together for Hope Conference, Boston, MA.

Authors

Shanyin Yang
Ariela Braverman Bronstein
Jeffrey Desmarais

Background

There is an over-representation of Black and Latinx individuals experiencing substance use disorder (SUD) among the homeless population in Boston, MA. In addition, there has been an increase in overdose deaths in recent years, particularly among Black individuals. Low threshold housing with comprehensive services have emerged as an effective and evidence-based strategy to support individuals living with substance use disorders who are experiencing unsheltered homelessness. As a response to the increasing number of individuals who experience both unsheltered homelessness and SUD, in January 2022, the city of Boston established six low-threshold sites (LTS). The Boston Public Health Commission launched the Boston Pathways project to evaluate this policy with a focus on racial equity. BPHC engaged the Institute for Community Health (ICH) to conduct the equity evaluation.

Objective

LTS offer low-barrier transitional housing and support services, focused on stabilization and permanent housing placement and stabilization. Compared to traditional shelters, LTS have reserved beds and lockers, with no sobriety requirements or curfews. As part of a multi-year mixed-methods program evaluation, this study compares housing outcomes between the LTS and Traditional emergency shelter guests.

Methods

Variables

  • Outcome: Permanent housing placement
  • Exposure: Group: LTS or control
  • Covariates: Race and ethnicity, age group, gender, disabilities other than SUD, income or benefits from any source

Data Sources

  • Homeless Management Information System
  • LTS group: Most recent shelter stay for LTS guests between January 1, 2022, and December 31, 2024
  • Control group: Individuals residing in traditional shelters between January 1, 2018, and December 31, 2021*, restricted to a) experiencing chronic homelessness**, and b) self-reported SUD during most recent stay

*A different time-period was used for the comparison group to simulate a counterfactual scenario where the LTS did not exist; **Chronic homelessness = Self-report in the lasts stay of being homeless for at least 12 months over the last 3 years OR bed-stay count in shelters higher or equal to 365 over the last 3 years prior to their last stay.

Analysis

  • Descriptive : Assessed difference by intervention group in permanent housing placements and demographic variables using frequencies and percentages comparing through chi square tests.
  • Inferential: Association between intervention group and the risk of permanent housing placement was assessed through a competing risk model. Permanent housing was defined using destination after shelter exit; guests still in the shelter at the end of the study period were right censored, and all other destinations were treated as competing risks. Models adjusted for race and ethnicity, gender, age group, income and benefits from any source, having a disability other than SUD, and shelter type (congregate vs. non-congregate). Errors were clustered by shelter.

Results

Summary of Findings

Guests in LTS between January 2022 and December 2024 were placed in permanent housing at 2.2 times the rate as guests in traditional shelters between January 2018 and December 2021, after adjusting for race and ethnicity, gender, age, income and benefits from any source, and type if shelter (SHR 2.21; 95% CI 1.27,3.85) .

Conclusion

These findings, while having limitations, suggest that the LTS is associated with higher rates of permanent housing placement compared to traditional shelters, specifically for chronically unsheltered individuals with SUD. Overall, the results emphasize the importance of policymaking that supports harm reduction and Housing First approaches and low-barrier access in facilitating successful housing transitions.