Residential eviction proceedings and overdose mortality in Rhode Island

• In Event: The Intersection of Housing Insecurity and Substance Use Disorder—defining the crisis and identifying policy solutions

Abstract

Introduction/Background: Housing insecurity is a well-established risk factor for overdose, although less is known about the particular effects of eviction on drug use-related outcomes. Prior ecological studies have documented county- and census tract-level associations between residential eviction and overdose mortality, although research has not previously quantified this relationship at a finer geographic resolution nor longitudinally.  

Purpose/Research Question: We sought to estimate the address-level causal effects of residential eviction proceedings on risk of overdose death. We hypothesized that multiple stages of the eviction process (i.e., filing, hearing, and enforcement) would each contribute to elevated overdose mortality.  Data: We constructed an address-level dataset linking several public health surveillance and administrative data sources in Rhode Island (RI) from 2017–2023, including overdose deaths from the RI Office of the State Medical Examiner, eviction records from the RI Judiciary, and registered rental properties from the RI Department of Health. Covariates included address-level housing subsidies, municipality-level drug-related arrests and Low-Income Home Energy Assistance Program recipients, census block group-level American Community Survey demographic characteristics, and time-varying state policy measures such as eviction moratoria and emergency rental assistance. 

Research Design and Methods: Using a target trial emulation with a clone-censor-weight approach, we compared address-level fatal overdose risk of residents corresponding to the following treatment strategies: 1) Received an eviction filing; 2) Received an eviction filing and appeared in a court hearing; 3) Received an eviction filing, appeared in a court hearing, and received an executed writ of eviction; and 4) Did not undergo eviction proceedings. Inverse probability weighted marginal structural models accounted for confounding by time-varying covariates and artificial censoring. We implemented Kaplan-Meier methods to estimate weighted 12-month risks of overdose death for each treatment strategy.  

Results/Findings: Forthcoming results will present the adjusted cumulative incidence of overdose death for each treatment strategy over 12 months of follow-up. 

Conclusion/Implications: Given discrete opportunities for policy intervention at each stage of the eviction process, findings from this work have the potential to elucidate novel avenues for overdose prevention that address residential eviction as a structural driver of overdose mortality

Author

• Alexandra Skinner, Brown University
  Presenting Author