Factors associated with consumer engagement in supported housing

Background

Many homeless consumers who enroll in supported housing programs—which offer subsidized housing and supportive services—disengage prematurely, before placement in permanent, community-based housing. We sought to identify factors associated with premature exits from supported housing, before housing placement.

Methods

Using administrative data, we obtained a roster of VA Los Angeles’ supported housing enrollees from 2011-2012. We matched consumers who exited the program before achieving housing (“exiters,” n=51) 1:1 on age, gender, marital status, and race/ethnicity with consumers who achieved housing (“stayers,” n=51). We reviewed medical records to compare diagnoses, healthcare utilization, housing histories, vocational history, and criminal-justice involvement between these two groups. We identified exiters’ housing outcomes and used recursive partitioning to highlight variables best-differentiating exiters from stayers.

Results

Several factors were significantly associated with premature program exits: residing in temporary housing on hospital grounds at program enrollment, substance abuse, hepatitis C, chronic pain, justice-involvement, poor adherence to outpatient care, frequent Emergency Room utilization, and medical/surgical admissions. After exiting, >50% of exiters became street homeless or incarcerated. Recursive partitioning highlighted two variables sufficient to capture information from 42 predictors to classify consumers as exiters vs. stayers. Consumers living on hospital grounds at program enrollment were predicted to exit. Among those living elsewhere, those with excellent adherence to outpatient medical/surgical care were predicted to stay.

Conclusions

Our findings highlight several factors associated with premature exits from supported housing programs. Future work is needed to implement and evaluate interventions addressing these factors, adapted to the context of supported housing programs.