Evaluation of a Supportive Housing Program for Adults Who Are Unhoused and Have Complex Health Needs

The study design is a cohort study with a propensity score-matched control group and difference-in-difference analysis to evaluate intervention effectiveness. A hybrid type 1 effectiveness-implementation study framework is used that incorporates mixed methods to determine clinical effectiveness and explore implementation and participant well-being.

Using a quasi-experimental design approach, participants are not randomly assigned to the intervention group. Participants are selected based on their utilization of health services, satisfying the other inclusion criteria, and accepting to participate in the Initiative. Selection favored those individuals with the most health service utilization. All intervention participants will be invited to participate in the evaluation study. A control group will be created by selecting individuals from administrative hospital records that are propensity score matched to the individuals in the treatment group. Four control individuals will be selected for every one treatment individual (1:4 pair). Patient data from multiple time points, before and after intervention implementation, will be used to strengthen the analysis and to compare immediate and longer-term effects to the treatment group. The propensity-score matching procedure is described below.

The hybrid type 1 effectiveness-implementation study framework was used to guide the selection of study aims and outcomes to focus primarily on clinical effectiveness, while also exploring implementation-related factors. This approach helps to combine questions concerning the effectiveness of the intervention with questions about how best to implement it. The clinical, service, and implementation outcomes, according to Proctor's Outcomes for Implementation Research, will be analyzed in conjunction to provide a complete understanding of the program mechanisms of change.

Mixed methods will be used to combine quantitative and qualitative data in a convergent approach. The quantitative and qualitative data will be collected in parallel, primarily, to merge the information and compare the results for triangulation of the findings. Triangulating data from multiple methods and sources will allow us to comprehensively answer the evaluation questions, minimize bias, and increase the credibility of the results. Data collection sources are described below.

All participants that were selected and agreed to participate in the Social Medicine Housing Initiative will be invited to participate in the study. The tenants of Social Medicine Housing Initiative must consent to participate in the study. The participants will be housed at the newly constructed housing building in Toronto or will be housed in units that become available throughout the city. There are no exclusion criteria that would exclude participants who are receiving housing from participating in the research study. If participants are able to independently provide consent, or do not want to provide consent for the study, they will be excluded.

To qualify to participate in the Social Medicine Housing Initiative, patients must have had 2 or more inpatient admissions or 6 or more visits to the emergency department in the past 6 months. They must also be currently without housing, which is defined as having spent at least one night in a shelter or having received assistance from city-funded street outreach providers in the past 90 days. Alternatively, the patient can participate in Social Medicine Housing Initiative if they identify as Indigenous. Historically marginalized groups, including people with disabilities, women, gender-diverse individuals, and Indigenous people, are given priority for housing.

University Health Network program coordinators will identify potential participants of the Social Medicine Housing Initiative through the health system data of the University Health Network hospitals. Once identified, the potential participant will be asked by a member of the community organization that will manage the housing (Fred Victor Organization) if they would like to move into the housing and participate in the Social Medicine Housing Initiative. After the participants confirm their involvement in the Social Medicine Housing Initiative, coordinators at the University Health Network will speak with them to invite them to participate in the study and to conduct the informed consent process. University Health Network coordinators will explain the purpose, process, risks and benefits of the study. Participants will then confirm their enrollment in the study by providing written informed consent.

All tenant study participants will be asked to participate in mixed-methods surveys and in-depth interviews to assess their well-being, their perception of the program, and their personal journey. All interviews will last between 45 and 60 minutes and will be audio-recorded so they can be transcribed for analysis. All surveys will be conducted in the housing facility or another private location that is comfortable for the participant. The surveys will be administered through the REDcap application for managing online surveys and databases(8). The survey will be conducted by a member of the University Health Network research team using a smartphone or tablet. University Health Network will hire research and interview specialists with experience in mixed-methods surveys and experience working with vulnerable populations.

Administrative health data will be obtained through an Applied Health Research Questions Research Project from the Institute of Clinical Evaluative Sciences of Canada. ICES is a non-profit, independent research instate funding by the Ontario Ministry of Health and the Ministry of Long-Term Care. The Institute of Clinical Evaluative Sciences of Canada has legal status under Ontario's health information privacy law which allows it to collect and analyze healthcare and demographic data, without consent, for health system evaluation and improvement. Using a data sharing agreement, an The Institute of Clinical Evaluative Sciences of Canada researcher will provide relevant information on health care resource utilization for study participants. The administrative health data housed at the Institute of Clinical Evaluative Sciences of Canada will be used to conduct the propensity score matching and to obtain the data to assess the outcomes described in Aim 1, Aim 2, and Aim 4. The Institute of Clinical Evaluative Sciences of Canada data will be used to measure the following indicators: Emergency department visits, Days spent in the hospital, Inpatient Utilization for Substance Use Disorder, Inpatient Utilization for Severe Mental Illness, Mortality, Primary Care Visits, Specialist Visits.