Mediators of measurement-based care implementation in community mental health settings: results from a mixed-methods evaluation

BACKGROUND: Tailored implementation approaches are touted as superior to standardized ones with the reasoning that tailored approaches afford opportunities to select strategies to resolve determinants of the local context. However, results from implementation trials on this topic are equivocal. Therefore, it is important to explore relevant contextual factors that function as determinants to evaluate if they are improved by tailoring and subsequently associated with changes in implementation outcomes (i.e., via statistical mediation) to better understand how tailoring achieves (or does not achieve) its effects. The present study examined the association between a tailored and standardized implementation approach, contextual factors that might mediate change, and a target implementation outcome in an initiative to implement measurement-based care (specifically the clinical integration of the Patient Health Questionnaire [PHQ-9] for depression) in a community mental health organization. METHODS: Using a cluster randomized control design, twelve community-based mental health clinics were assigned to a tailored or standardized implementation group. Clinicians completed a self-report battery assessing contextual factors that served as candidate mediators informed by the Framework for Dissemination at three time points: baseline, 5 months after active implementation support, and 10 months after sustainment monitoring. A subset of clinicians also participated in focus groups at 5 months. The routine use of the PHQ-9 (implementation outcome) was monitored during the 10-month sustainment period. Multi-level mediation analyses assessed the association between the implementation group and contextual factors and the association between contextual factors and PHQ-9 completion. Quantitative results were then elaborated by analyzing qualitative data from exemplar sites. RESULTS: Although tailored clinics outperformed standard clinics in terms of PHQ-9 completion at the end of active implementation, these group differences disappeared post sustainment monitoring. Perhaps related to this, no significant mediators emerged from our quantitative analyses. Exploratory qualitative analyses of focus group content emphasized the importance of support from colleagues, supervisors, and leadership when implementing clinical innovations in practice. CONCLUSIONS: Although rates of PHQ-9 completion improved across the study, their sustained levels were roughly equivalent across groups and low overall. No mediators were established using quantitative methods; however, several partial quantitative pathways, as well as themes from the qualitative data, reveal fruitful areas for future research. TRIAL REGISTRATION: Standardized versus tailored implementation of measurement-based care for depression. ClinicalTrials.gov NCT02266134, first posted on October 16, 2014