Understanding the process of a research utilisation intervention in policy agencies: What, how and why?

Abstract

Background: There are increasing calls to strengthen the use of research in health policymaking, but little is known about how intervention strategies may be received by policymakers, or how different contexts are likely to affect engagement and uptake. This research seeks to understand, and provide transferable information about, how research utilisation interventions function in different policy settings, and how they can be made more fit for purpose. Approach: This thesis focuses on Supporting Policy In health with Research: an Intervention Trial (SPIRIT). SPIRIT is a multi-component research utilisation intervention that was implemented in six health policy agencies in Sydney. Taking a process-orientated perspective, my mixed methods research examines facets of the intervention design and implementation, and attempts to both describe and explain how a range of participants perceived and interacted with SPIRIT, and with what effects. Work produced: Six studies are included. Five published papers: (i) a realist scoping review of interventions aimed at building policymakers’ capacity to use research in their work; (ii) a participant observation study of concept development in our team; (iii) an account of theory-focused fidelity assessment; (iv) an analysis of the views, behaviours and impacts of the intervention’s internal facilitators; and (v) a realist evaluation of how participants experienced SPIRIT and the causal pathways through which intervention strategies appear to generate process effects. A further chapter explores the wider context and considerable challenges of increasing research use in policymaking. Key contributions: This thesis delivers new conceptual and methodological approaches for understanding how and why complex interventions function as they do. First, the findings describe how SPIRIT was implemented and perceived, and offer provisional explanations for the marked variation in engagement between the six intervention sites. Using an in-depth realist process evaluation approach, I was able to identify and test possible causal mechanisms, and to make empirically grounded recommendations for program improvement (e.g. improved strategies for identifying and supporting internal facilitators). Second, the thesis makes methodological contributions. It advances the use of realist process evaluation, which is still rarely used, unpacks the role of process effects in a realist scoping review of research utilisation interventions, and presents a novel approach for managing fidelity assessment within flexible interventions. The inclusion of pragmatic tools and worked examples makes this work concrete. Lastly, the research contributes to theory in the field. It furthers our understanding of the dynamic and highly situated connections between policymakers’ diverse information needs and practices, and different kinds of research utilisation intervention and implementation strategies. It brings this learning together in provisional transferable propositions and provides access to the substantial empirical and conceptual work that sits behind them. Together, these contributions offer guidance for the design, implementation and evaluation of future intervention studies in this field and beyond. The thesis concludes by highlighting areas where further understanding is required if the ambitions of research-informed policymaking are to be advanced.