# The Housing Instability Scale: Determining a Cutoff Score and Its Utility for Contextualizing Health Outcomes in People Who Use Drugs

## TL;DR

This study examines how housing instability affects health outcomes in people who use drugs, identifying a cutoff score for housing instability and its associations with infections and other health issues.

## Contribution

The study introduces a cutoff score for the Housing Instability Scale and demonstrates its utility in understanding health outcomes among people who use drugs.

## Key findings

- Housing instability was significantly associated with infections, blackouts, and seizures among participants.
- Cluster analysis and Gaussian Mixture Modeling suggested different levels of housing instability.
- SSP use, opioid use, and Xanax use were identified as potential predictors of health outcomes.

## Abstract

(1) Background: Housing instability, a key social determinant of health, disproportionately affects people who use drugs (PWUD), increasing their risk for adverse outcomes. This study explores the relationship between housing status and drug-related health outcomes among PWUD in an urban setting in the Southeastern United States (US) and determines the cutoff point for the Housing Instability Scale (HIS). (2) Methods: We conducted a cross-sectional survey from July to November 2024 among adult PWUD. Recruitment was through syringe services programs (SSPs), nightlife venues, and community outreach. HIS was used to assess housing status, while cluster analysis and Gaussian Mixture Modeling (GMM) were used to suggest a potential cutoff. Logistic regression models were employed to examine associations between housing instability and health outcomes. (3) Results: Among 164 participants (mean age = 41.2; 79.9% Black/African American), the average HIS score was 3.23. Cluster analysis suggested a singular cutoff, while GMM suggested four levels of housing instability. Multivariate logistic regression revealed that housing instability was significantly associated with infections (AOR = 1.55, p = 0.064), blackouts (AOR = 1.47, p = 0.0457), and seizures (AOR = 1.28, p = 0.0667). Overdose and wounds showed no significant association. SSP use, opioid use, and Xanax use were also identified as potential predictors.