Measuring Early Recovery in
Street Outreach
A study led by PEP staff, Gabby Knighton, Executive Director; Achike Oranye, Division Director; Patricia Carter, Program Administrator with academic partnership from IRIS’s Jay Unick, Professor; Jon Gilgoff, Clinical Research Specialist; Eunsong Park, Graduate Research Assistant
People Encouraging People (PEP) & University of Maryland School of Social Work (UMSSW)
Background
Low-barrier, street outreach services to individuals who are unhoused and use drugs serve as important building blocks towards later stages of one’s stabilization and recovery from opioid use disorder (OUD) and other harmful substances. Progress in developing this recovery capital through these early-stage engagements are not well measured using current instruments or deeply studied through research. To meet this need, this study developed and piloted a quality of life instrument, which the study team adapted from the EUROHIS-QOL-8. The aims of the research were to assess the instrument’s reliability to measure this construct, the feasibility of implementing the tool within PEP’s program model, and change in quality of life over time among program participants.
Methods
The EUROHIS-QOL-8 was modified by PEP’s street outreach team and IRIS staff from UMSSW to better reflect the lived experience of program participants. This included adapting a question related to “the conditions of your living place” to instead assess conditions of the place(s) where individuals spent most of their time during the day, and of the place(s) where they stayed at night. Another question related to having enough money to meet your needs was changed to inquire about level of resources. A question about level of safety was also added, making 10 total questions. The adapted instrument was implemented with 209 individuals, with 21 individuals the next day to assess test-retest reliability, and with 18 individuals two months later to assess change over time. To test the tool’s psychometric properties, the team analyzed internal consistency reliability, test-retest reliability, and for validity, conducted an item response analysis. Paired sample t-test statistical analysis was used to assess change over time.
Results
Internal consistency reliability of the adapted quality of life 10-item measure was excellent as estimated using both Cronbach’s alpha (0.84) and Omega coefficient (0.84). ​Next day test-retest reliability was relatively low (0.65). The rating scale item response model found that items had excellent fit to the model, suggesting that the items measure a single construct. There was no statistically significant change in quality of life assessed for the small group of participants who completed the scale at baseline and two-month follow up.
Conclusion & Implications
This project piloted an adapted quality-of-life measure that demonstrated good internal consistency reliability and construct validity, with low test-retake reliability, and non-significant results related to change over time. Both the lack of change over time and the low test-retest reliability may be attributed to the constantly shifting conditions of being unhoused, as all follow-up surveys were conducted with those still without stable shelter. The challenge of finding many participants two months later to administer follow-up surveys is a common barrier to assessing longitudinal change with this population. As PEP considers ways to integrate findings into their programming, one option is to use this instrument or a shorter form version within their street outreach intake process, then again at later milestones along the continuum of care. This accounts for the difficulty of administering surveys outdoors with time-limited interactions. Overall, study results support the use of this tool to measure quality of life and confirm the need to continue developing effective ways to measure change within earlier and later recovery stages for people experiencing homelessness and OUD.