Methods: The PEP outreach team, comprised of clinicians, social workers, and peer outreach workers, and the IRIS research team reviewed existing quality of life measures to identify an instrument to modify. The team selected the EUROHIS QOL-8 as the consensus starting measure. Modifications were made to the measure including the addition of two items, changing the questions time frames, and altering some of the wording of the items to better reflect the lived experience of individuals being surveyed. The final instrument included demographic data and 10 items of the modified quality of life instrument. The PEP outreach team included the modified tool in their assessment of consumers and attempted to interview individuals six weeks following the initial interview. The data was then sent to the IRIS team for analysis. Psychometric assessments of the tool included internal consistency reliability, test-retest reliability, and item response item analysis.
Results: The PEP outreach team conducted 197 initial assessments and 32 six week follow-up assessments. Internal consistency reliability was excellent as estimated using both Cronbach’s alpha (0.84) and Omega coefficient (0.86). Six-week test-retest reliability was poor (0.65) as assessed by interclass correlation using a multilevel model with time nested in individuals. Rating Scale Rasch Based Item Response Models indicated good item fit to the restrictive Rasch model and item difficulty parameters were spread across the quality-of-life latent trait supporting use of the tool to measure change.
Conclusions: The modified measure scored high on internal consistency reliability and had excellent IRT item fit properties. The test-retest reliability was poor. These two findings are consistent with the hypothesis that the measures do measure an individual’s state, but that homeless individuals who had both baseline and six-week interviews had high levels of variability on the reported quality of life domains. We believe that this reflects instability in lived experience on the street. These preliminary data support the use of this QOL tool to measure quality of life and to continue to understand variability and measurement of change for individuals experiencing homelessness and OUD.