This work was grounded in the commitment to centering the voices of youth with lived expertise as well as responding to priorities defined by community partners. The overall study drew on key principles of community-based participatory action research (CBPAR) axiology, and the qualitative component was framed with a critical social theoretical lens [20-23] (Textbox 1).

This pilot study used a convergent mixed methods (quantitative and qualitative data collected concurrently and the findings combined), 2-arm parallel RCT (participants randomly assigned to either the intervention or control group), open-label (participants and research team aware of random assignment) design with 1:1 allocation (equal number of participants in each study arm) embedded within a CBPAR framework. The study was conducted collaboratively with 4 community partners who work with youth who are experiencing or have experienced homelessness: (1) Covenant House Toronto (Toronto, Ontario), (2) Living Rock (Hamilton, Ontario), (3) The RAFT (St. Catharines, Ontario), and (4) StepStones for Youth (Toronto, Ontario).

The study was prospectively registered on ClinicalTrials.gov (NCT05781503) and received Unity Health Toronto Research Ethics Board approval (#22-230). All protocol amendments were reported to both these organizations. The only protocol deviation was submitted in October 2023, which allowed community partners to send the rent subsidies directly to youth for those at risk of losing their housing due to refusal of third-party payments from landlords, or fear losing their housing due to disclosing a history of homelessness.

This pilot feasibility and acceptability study was designed with the intention of generating data and hypotheses to inform a full-scale study; therefore, no formal sample size calculation was performed. Instead, we chose a pragmatic sample size of 40 participants, based on the financial resources available to provide rent subsidies and coaching over a 12-month period.

Eligible youth ages 16-24 years who had left homelessness (defined by the Canadian Observatory on Homelessness [24] as unsheltered, emergency sheltered, provisionally accommodated, and at risk of homelessness) within the past 12 months and living or planning to live in market rent housing were identified by our community partners. This age mandate was chosen because this is the age group served by our community partners. We chose to target the first year of exiting homelessness because our collective experience has shown that this is a particularly precarious time for youth in terms of mental health challenges and the risk of returning to homelessness. Foster care was included in our definition of homelessness given the time-limited and provisionally accommodated nature of these living arrangements, which put youth at high risk of homelessness.

In addition to the above conditions, recruited and enrolled participants were required to meet the following criteria (Textbox 2).

Study participants were recruited from the cities in which our community partners are located: Toronto, Ontario (Greater Toronto Area population 6.7 million), Hamilton, Ontario (population 785,000), and St. Catharines, Ontario (St. Catharines-Niagara population 416,000). The initial introduction to the study was facilitated by our community partners. Frontline staff (eg, case workers, housing workers) working with youth who have experienced homelessness were responsible for identifying youth who may be interested in the study and met the inclusion and exclusion criteria. Staff provided potential participants with information about the study and the study consent form to review. Potential participants were directed to contact the study lead research coordinator by email or phone, who then conducted eligibility screening over the phone. Once participants were deemed eligible for the study, they were connected to a member of the research team. To build rapport with community partners and study participants, one dedicated research team member was assigned to each partner site.

Each research team member reached out to eligible youth for an initial Zoom call. Free and informed verbal consent was first obtained after carefully reviewing the information and consent form with the participant. During the same Zoom call, electronic consent was also obtained once the research team member emailed participants the secure baseline data collection web link and a unique login and password. The baseline data collection questionnaire was programmed so that participants were unable to proceed until they selected “Yes” for consent. The research team member stayed on the call while the participants completed the baseline questionnaire.

Once participants completed the baseline questionnaire the research team member opened a sequentially ordered randomization file from a sequentially ordered electronic randomization folder to obtain the participant’s group assignment. The participant’s group allocation was noted, and participants were informed immediately if they had been allocated to the intervention (rent subsidies and co-designed leadership guide+coach) or control (rent subsidies alone) group (Figure 1 [25]). Randomization was stratified by site using block randomization (random block sizes of 2 and 4). In keeping with typical community-based RCTs with psychosocial interventions, blinding of participants, community partners, and research staff to allocation group was not pragmatic due to the nature of the intervention [26].

Further to Figure 1, the participant excluded post randomization was found to be ineligible before the intervention was initiated. We will not be using data from the ineligible participant. A new youth was then enrolled and randomized. In addition, the participant who withdrew during follow-up did so in order to meet the criteria for a different (longer-term) rent subsidy program.

Youth in both arms (n=40) were provided monthly rent subsidies (CAD $700/month Hamilton and St. Catharines; CAD $800/month Toronto) for 12 months, which was generally paid directly to landlords and facilitated by our community partners (refer to Trial Design re: protocol deviation; at the time of recruitment, the conversion rate was CAD $1=US $0.7357). Youth who participated in TYOH 1.0 suggested the TYOH 2.0 research team should encourage participants to delegate a portion of the money they would have spent on rent toward a savings account, which they would be able to access at the end of the study. This idea received unanimous support from our community partners and was suggested to all participants at enrolment and suggested periodically during the study. Youth randomized to the intervention group (n=20) were also provided a co-designed leadership guide and assigned a study coach. The control group (n=20) was offered the co-designed leadership guide at the end of the study.

The strengths-based leadership guide (Finding Home: A Guide for Youth in Transition) was co-designed with 12 youth who had experienced homelessness, including youth who participated in TYOH 1.0. NST drafted the original version of the guide and cofacilitated a full-day co-design workshop with research team member MD. MD is an experienced peer engagement specialist who draws on her lived expertise of homelessness to collaborate with researchers and youth who have experienced homelessness. Significant enhancements were made to the guide based on feedback from workshop participants.

The Leadership Guide contained 12 chapters with the overarching aim of enhancing identity capital along with providing strategies to achieve participant-identified goals (Table 1). Each chapter contained 4 activities (eg, a self-reflection exercise or listening to a podcast). It was suggested youth complete one chapter every month; however, they were also encouraged to go at their own pace. Copies were provided in print and electronic formats. While the guide was designed for individual study, it was also meant to serve as a reference point for individual and group discussions with the study coaches.

The role of the coach was to speak with youth about how to draw on internal resources to help orient them toward their preferred future. This role was different from case management or mentorship in that a specific coaching methodology (Brief Solution-Focused Coaching) [27] was used to walk alongside youth as they developed strategies to reach self-defined goals (Table 2). Each coach was instructed to devote 2.5 days/week to the intervention and was given a caseload of 9-11 youth. The coaches were instructed to meet individually with each youth on their caseload every two weeks and collectively with the rest of the youth in their group every month.

The 2 study coaches were selected by a panel that included 2 leaders from our community partner organizations, 2 research team members, and 3 youth from TYOH 1.0 who work as paid experts on the TYOH 2.0 Community Partner and Researcher Advisory Board. Each of the coaches had their own coaching practices outside the homelessness sector, which they maintained during the study. The coaches met monthly with our youth advisors and team members to provide updates as to how the coaching sessions were progressing.

The primary outcomes of this pilot study were feasibility and acceptability (Table 3). Secondary outcomes included proxy indicators of socioeconomic inclusion based on our clinical expertise and research with this population: (1) education, employment, and training; (2) housing security; and (3) identity capital (Table 3). The relative timing of interventions and assessments is described in Figure 2.

The decision to proceed to an adequately powered definitive trial will be based on feasibility and acceptability (Table 3). The research team is particularly interested in quantitative metrics pertaining to coaching session attendance (eg, number attended out of a possible 24 sessions), leadership guide engagement, and intervention dropout rates, as well as qualitative data from the focus groups. We predetermined that we would not proceed to a definitive trial if we found intervention attendance or engagement was less than 50% (ie, average coaching attendance of less than 12 sessions over one year), more than 30% of participants (n=6) dropped out of the intervention, or qualitative feedback from intervention participants during focus group sessions was overwhelmingly negative.

Questionnaires were completed via a secure weblink, which was emailed to participants along with a unique login and password, at baseline and 4-, 8-, and 12-months post randomization. Those unable to access the weblink were offered the option of using computers at MAP Centre for Urban Health Solutions or at their respective community partner agency.

Quantitative analysis was performed using the intention-to-treat principle; that is, all participants were included and analyzed in the groups they were originally randomized. Baseline characteristics of the intervention and control groups were summarized using descriptive statistics (ie, mean, SD, median, and IQR for continuous variables, and frequencies and proportions for categorical variables).

There was no interim quantitative analysis planned to guide a decision to stop the study early for this pilot study [25]. We decided before the study began to either stop the intervention early or make adjustments to the intervention if feedback from participants during the focus group sessions was overwhelmingly negative and/or the majority of youth stopped participating in the intervention.

Focus groups with those in the intervention group were conducted by NST and MD in Toronto, Hamilton, and St. Catharines at 4, 8, and 12 months post randomization. Focus group questions primarily centered around intervention acceptability but also explored the impact of the intervention on identity capital and socioeconomic inclusion (eg, connection to broader social networks). Analysis began during and after the first data generation session, meaning the questions asked evolved over time based on our preliminary interpretations of the data [32-34]. Keeping in mind that data generation context impacts participant responses [34-36], focus groups were held in a community setting to help minimize researcher-participant power imbalance and increase participant comfort. Participants unable to attend in-person focus groups were offered the option of virtual attendance.

All focus groups were audio recorded using a password-protected device. Audio recordings were transcribed verbatim by a member of the research team, and the transcripts were uploaded to the web-based application Dedoose [37] for storage and retrieval. One member of the research team served as an observer and note taker at each focus group session to document nonverbal communication (eg, eagerness or disinterest) as well as preliminary analytic insights based on listening to the discussion. In addition, each focus group facilitator documented field notes as soon as possible after the meeting to capture their own observations and reflections on the sessions [33-35]. Food was also provided at the focus groups.

The team used reflexive thematic analysis to help make sense of the data [38]. Reflexive thematic analysis is comprised of 6 iterative phases (data familiarization, coding, initial theme generation [patterns anchored by a shared idea or concept], theme development and review, theme refining, and writing up). Crucially, this form of analysis also requires engagement with theory (eg, critical social theory), reflexivity (awareness of social location on analysis), and interpretation (findings not seen as self-evident) [38]. Before each monthly qualitative data analysis session, 2 team members read the most recent focus group transcript multiple times, coded (“tagged”) data relevant to intervention acceptability and our assumption that identity capital was a mediating factor in socioeconomic inclusion (and looked for data that might disprove that assumption), and compared codes to previous transcripts [32,33,38].

At the monthly analysis sessions led by NST, codes were discussed (and revised or deleted as needed) and organized in a code book, clustered into initial themes, and eventually refined and synthesized into key themes. Analysis was primarily inductive (moving from data to conceptualizing); however, deductive reasoning (moving from conceptualizing to data) was used when we wanted to understand new data through the lens of our emerging conceptual framework. The analysis team also formulated new focus group questions based on our preliminary analytic insights. Focus group participants were asked for their perspectives on the emerging interpretations during subsequent focus groups, and these perspectives played a key role in helping shape the data analysis.

As in TYOH 1.0, the aim of our analysis was to go beyond superficial understandings of intervention acceptability. Instead, we engaged in “value-adding” qualitative analysis—interpreting, contextualizing, and adopting a critical posture (eg, attention to power)—with the intention of operationalizing key conceptual insights in a conceptual framework [39]. As we did this, our team embraced the “creative presence of the researcher,” drawing on our own knowledge, insights, and experiences (including the experience of homelessness) to help make sense of the data [39].

A genuine partnership with the community—including youth with lived expertise—has been woven into this study from inception to dissemination. As noted previously, we have been working with most of the community partners on this study for several years, and the leadership guide being used in the intervention was co-designed with youth who have experienced homelessness. Three participants from TYOH 1.0 worked as paid advisors on the TYOH 2.0 Community Partner and Researcher Advisory Board. The Board met approximately every 3 months from June 2022 to October 2024. Our community partners and youth advisors were actively involved in planning TYOH 2.0, including discussions around research questions, outcome measures, recruitment, the burden of the intervention, and the time required to participate.

We endeavored to weave ethical considerations into all aspects of the study design, including our decision to use a CBPAR methodology, ensure all participants received portable rent subsidies (not just the intervention group), and offer the co-designed leadership guide to those in the control group after the study was over. The consent form highlighted that participants could withdraw from the entire study or just the identity capital intervention at any time, and this would not impact their monthly rent subsidy. Participants received honoraria via e-transfer within 24 hours of each data collection session (CAD $20 [US $14.28] for completing quantitative questionnaires and CAD $40 [US $28.57] for attending focus groups). Efforts were made to mitigate potential privacy and confidentiality concerns at all stages of the study (Table 4).

We intentionally did not choose deficit-focused questions or scales for the T2-T4 quantitative data collection sessions, and questionnaires were programmed to allow participants to skip questions if they chose. The focus group questions also targeted participant strengths, and participants were reminded at the start of each session that they did not have to answer any questions that made them uncomfortable. NST and members of the research team met monthly with the study coaches. These meetings provided an opportunity to assess or discuss any unintentional harms or risks to intervention participants, which were followed up immediately.

The study was approved by the Unity Health Toronto Research Ethics Board (REB# 22-230). Any protocol amendments were approved by the research ethics board. Free and informed verbal and electronic consent was obtained from all participants before the initiation of intervention and research activities.

We anticipate that this pilot study will provide important new insights regarding the feasibility and acceptability of an intervention using coaching and a co-designed leadership guide to target identity capital for youth transitioning out of homelessness. Although the study is not powered to detect statistically significant differences in the proxy indicators of socioeconomic inclusion outcomes, we hope to see signals in this data—especially in the Multi-Measure Agentic Personal Scale [12]—that the intervention holds promise. As we have found with our previous research [10,11], we anticipate the rigorous qualitative component will strongly influence our decision-making as to the next steps.

Homelessness in Canada is increasing, with 44% of those experiencing homelessness reporting they were first homeless before the age of 25 (with a spike between the ages of 15-19 years of age) [41]. This represents a crucial missed opportunity for homelessness prevention. Interventions like TYOH 2.0 offer promise as a way to prevent young people from returning to homelessness and potentially becoming chronically homeless adults. To be clear: we are not suggesting with this intervention that young people should “bootstrap” themselves out of homelessness. Rather, our aim is to provide young people with some of the tangible (eg, money toward rent) and intangible (eg, fostering identity capital) resources typically provided or nurtured—at least in part—through living at home.

Over 60% of Canadian youth (20-24 years of age) live with at least one parent [42]. This is important to consider when thinking about what to expect from young people attempting to transition out of homelessness. At a minimum, we need to adopt a holistic response to youth homelessness and think beyond housing when considering what supports might foster socioeconomic inclusion. We anticipate using the findings from this study to develop a multicomponent intervention that incorporates aspects of this current intervention (potentially revised) along with components that may need to be added to better facilitate socioeconomic inclusion for youth exiting homelessness.

Working alongside community partners to disseminate findings with the aim of highlighting sociostructural inequities, building community capacity, and improving the lives of the youth we work alongside is fundamental to this work. We anticipate disseminating our findings broadly in a variety of formats such as oral presentations, open-access plain language summaries or reports, and scientific journal papers. We are committed to dissemination beyond academic audiences. For example, for TYOH 1.0, we produced an animation of our study findings along with a documentary film [43]. Whenever possible, we intend to discuss our study findings alongside a lived expert involved in this research.

This study has limitations. All youth were connected to urban-based social service agencies in the province of Ontario; youth living in rural locations and outside of Ontario may not take up the intervention the same way. In addition, all of the quantitative instruments were based on self-reports and thus subject to social desirability bias. We also chose quantitative measures that we believe signal socioeconomic inclusion; it is plausible that these measures do not adequately capture this complex concept. Finally, this pilot study was not powered to detect statistical significance; findings beyond feasibility and acceptability must be interpreted with caution.

Findings from this RCT will help inform the way we conceptualize the types of support that are necessary to sustain successful exits from homelessness. This is important given the limited peer-reviewed evidence base on interventions designed to foster socioeconomic inclusion for youth exiting homelessness. The intervention was co-designed with youth who have experienced homelessness, and their voices will continue to inform the next iteration of this work.