Combining Realist approaches and Normalization Process Theory to understand implementation: a systematic review

BACKGROUND: Realist approaches and Normalization Process Theory (NPT) have both gained significant traction in implementation research over the past 10 years. The aim of this study was therefore to explore how the approaches are combined to understand problems of implementation, to determine the degree of complementarity of the two approaches and to provide practical approaches for using them together. METHODS: Systematic review of research studies combining Realist and NPT approaches. Realist methodology is concerned with understanding and explaining causation, that is, how and why policies, programmes and interventions achieve their effects. NPT is a theory of implementation that explains how practices become normalised. Databases searched (January 2020) were ASSIA, CINAHL, Health Research Premium Collection via Proquest (Family Health Database, Health & Medical Collection, Health Management Database, MEDLINE, Nursing & Allied Health Database, Psychology Database, Public Health Database) and PsycARTICLES. Studies were included if the author(s) stated they used both approaches: a scientific Realist perspective applying the principles of Pawson and Tilley’s Realist Evaluation or Pawson’s Realist Synthesis and Normalization Process Theory either solely or in addition to other theories. Two authors screened records; discrepancies were reviewed by a third screener. Data was extracted by three members of the team and a narrative synthesis was undertaken. RESULTS: Of 245 total records identified, 223 unique records were screened and 39 full-text papers were reviewed, identifying twelve papers for inclusion in the review. These papers represented eight different studies. Extent and methods of integration of the approaches varied. In most studies (6/8), Realist approaches were the main driver. NPT was mostly used to enhance the explanatory power of Realist analyses, informing development of elements of Contexts, Mechanisms and Outcomes (a common heuristic in realist work). Authors’ reflections on the integration of NPT and Realist approaches were limited. CONCLUSIONS: Using Realist and NPT approaches in combination can add explanatory power for understanding the implementation of interventions and programmes. Attention to detailed reporting on methods and analytical process when combining approaches, and appraisal of theoretical and practical utility is advised for advancing knowledge of applying these approaches in research. SYSTEMATIC REVIEW REGISTRATION: Not registered.