Predictors of housing instability and stability among housing first participants: A 24-month study

This study examined the characteristics of people who experienced housing instability and stability after 24 months of being enrolled in Housing First (HF). A companion study addresses the same objective using qualitative methods. A sequential logistic regression was conducted to determine predictors of unstable and stable housing at 24 months of enrollment in HF in a randomized trial. We applied the Gelberg-Andersen [The BehavioralModel for Vulnerable Populations: Application to medical care use and outcomes forhomeless people. Health Services Research, 34(6), 1273-1302]. Behavioral Model for Vulnerable Populations to identify and group predictor variables. Thirty-one percent of the HF participants (N = 302/977) met the study criteria for housing instability (i.e. stably housed for less than 90% of last six months). Residence in Winnipeg, longer accumulated lifetime homelessness, and higher levels of substance use predicted unstable housing. Residence in Toronto and Montreal, older age, being in an ethno-racial minority group (other than Indigenous), higher income, higher perceived housing quality, and having a family physician predicted stable housing. The findings of the study have implications for strengthening HF supports to better address the needs of HF participants who may be at risk of housing instability.