On the Shoulders of Others: A mixed-methods evaluation of a theory-informed peer health promotion group for homeless men in recovery

Individuals engaged in substance use recovery have higher rates of co-morbid health conditions, and are more likely to face significant challenges in appropriately managing their health. The concurrent experience of homelessness can present additional barriers for health-promoting behavior. The purpose of this study is to examine how a peer-driven, goal-oriented, and theory-informed health education program can influence positive change in homeless men, and achieve the long-term goals of (1) enhanced self-management of health, (2) enhanced health literacy, and (3) enhanced comfort in engaging with the medical system.

We describe the development of this program in three stages: planning, implementation, and evaluation. This program, entitled "On the Shoulders of Others", took place at a faith-based recovery and transitional shelter for homeless men.

Planning. In collaboration with a group of shelter residents, we identified the SAMHSA Whole Health Action Management (WHAM) peer support group model as a basis for designing program structure. In addition, the Transtheoretical Model (TTM) served as the conceptual foundation upon which we developed the health topics addressed. Through the Hanlon Method for Prioritizing Health Problems, we then conducted a mixed-methods assessment of health issues of concern among shelter-based healthcare providers, shelter staff, and residents to adapt and operationalize these concepts.

Implementation. We implemented the biweekly, 2-hour support group over the course of 12 months, reaching 8-14 individuals at each session. We covered health concerns in both biological and psychosocial domains, and included topics such as the Science of Stress and The Power of Human Connections.

Evaluation. We conducted a participatory evaluation mapping onto our theory of change, involving a convergent parallel mixed-methods design to understand both process and outcomes. Data was captured at each session through participant observation, pre- and post-tests, goal-tracking sheets, and semi-structured interviews. Qualitative data was transcribed and analyzed by independent investigators through an iterative consensus-building process, and quantitative data was tracked over the course of program implementation. All findings were validated against a panel of program participants to corroborate and attenuate interpretations from the investigator's perspective.

This program reached 114 unique individuals. Process measures (e.g. health goal attainment, provider visits) offer promising evidence for this approach in supporting recovering individuals towards managing their health. Outcome measures further indicate program potential: 72% of participants demonstrated improvements in health literacy, 24% more effectively adhered to medication or health behavior regimes, and 81% reported improved trust with their healthcare provider. Implications for scale-up and adaptation for similar settings are discussed.