Validation of a Causal Model of Implementation

点击翻译按钮获取中文摘要

PROJECT SUMMARY Advancing causal implementation theory is critical for designing tailored implementation strategies to facilitate clinician behavior change that target specific mechanisms associated with evidence-based practice (EBP) use. Such strategies may be more successful and more efficient than general implementation strategies. The overall goal of this study is to test the generalizability of a conceptual model that posits the causal relationship among variables from organizational and social psychology to predict clinician evidence-based practice (EBP) use. Broadly, the model proposes that organizational factors like climate and culture influence attitudes, norms and self-efficacy, while other organizational factors like workload, resources and organizational friction moderate the pathway between intentions and the use of an EBP. Our recently completed NIMH-funded R21 demonstrated that this model accounted for up to 75% of variance in implementation of three EBPs in community settings. While promising, we were limited by a small sample, and a focus on special education teachers' use of autism interventions. This R01 will test the generalizability of the causal model in a much larger sample from a new group of practitioners, for a different and more widely-used EBP, cognitive- behavioral therapy (CBT). Successful completion would further validate the model, advancing our understanding of the causal pathways in EBP implementation. We will leverage ongoing CBT implementation efforts to recruit approximately 300 mental health clinicians across 40 organizations in two large public health systems (Philadelphia and Texas). After completing training in CBT, clinicians will complete measures of all constructs delineated in the model. Clinicians also will be observed via audio recording delivering CBT with a client on their caseload on two occasions. Following each observation, data on theorized moderators of the intention to behavior gap will be collected via survey. A subset of clinicians who report high intentions and demonstrate low EBP use will be purposively recruited to complete brief semi-structured interviews further assessing reasons for the intention to behavior gap. Our primary dependent variables and implementation outcomes of interest are clinician intentions to use CBT and direct observation of clinician use of CBT. However, as CBT comprises many discrete components that vary in complexity, each discrete component of CBT use will be measured separately. Data will be analyzed via multilevel modeling to test the extent to which intentions and determinants of intention predict each discrete CBT component (Aim 1) and the extent to which organizational and other contextual factors highlighted in the implementation literature predict factors related to intention formation and moderate the association between intentions and CBT use (Aim 2). Results will inform the development of implementation strategies that target modifiable factors explaining substantial variance in intention and in implementation that can be applied broadly across EBPs.