Evaluation of Washington State’s Community Prevention and Wellness Initiative: Exploring the Influence of Community Readiness on Program Outcomes

Introduction: Community Prevention and Wellness Initiative (CPWI) is a community coalition model targeting adolescent substance use and related risk factors. We evaluated its effectiveness among three cohorts of CPWI communities by comparing whether CPWI 10^th graders differed significantly from non-CPWI 10^th graders in 2016 compared to baseline. Results from multilevel modeling on propensity score adjusted data suggest that CPWI had varying levels of positive effects on the three cohorts. For instance, in 2008, Cohort 1 communities were at significantly higher level of risk for alcohol and marijuana use, and risk factors in peer-individual, family, school, and community domains. By 2016, these communities were successful in closing the risk gap for alcohol use, and risk factors in three domains. By 2016, Cohort 3 communities closed some gaps in two domains, but they did not close any gaps in substance use, and in other two domains. Given the variability in outcomes across cohorts, we plan to re-evaluate the effectiveness of CPWI by accounting for variability in community readiness.

Method: As community readiness data is not available for WA communities, we will conduct a three-round Delphi study to gather expert consensus on variables that could be considered as proxies for community readiness. In the 1^st round, we will interview 10 to 15 prevention researchers and practitioners to identify proxies for readiness. We will code transcripts to identify unique proxy variables. The 2^nd and 3^rd Delphi round will consist of questionnaires asking experts to rate their agreement with whether the listed variables could be considered as proxies for community readiness. We will analyze data from 2^nd and 3^rd rounds for expert consensus using median scores and interquartile deviations. Experts will have three weeks to complete each round of questionnaire on Qualtrics. Once the final list of proxies is created, we will identify a subset of proxies with publicly available data for WA communities. Then, we will include these variables in our propensity score model to account for community readiness. Study timeline: Data collection for 1^st round will begin in December, 2^nd in February, and 3^rd in March. Statistical analyses will be completed by April.

Results: We will generate a list of expert-agreed proxies for community readiness. We will also calculate CPWI program effects controlling for readiness.

Conclusion: Inclusion of community readiness data in statistical analysis could provide more insight into the effectiveness of CPWI in reducing substance use outcomes. Results of our Delphi study will contribute to prevention science by providing a list of proxy variables for community readiness which can be used by researchers and practitioners in their work.