Grand rounds in methodology: when are realist reviews useful, and what does a ‘good’ realist review look like?

Abstract

Research in the quality and safety field often necessitates an approach that supports the development of an in-depth understanding of how a complex phenomenon occurs, or how an intervention works. Realist review is an increasingly popular form of evidence synthesis that provides a theory-driven, interpretive approach to secondary research. Realist reviews offer quality and safety researchers the opportunity to draw on diverse types of evidence to develop explanatory theory about how, when and for whom interventions ‘work’ or outcomes occur. The approach is flexible, iterative and practical, typically drawing on the experience of policymakers, practitioners and patients throughout the review. With the increasing use of realist reviews, some common misconceptions about the approach have become evident in the literature. This paper introduces what is involved when planning and conducting a realist review, and where the approach can offer most value, as well as outlining common challenges that researchers may face when adopting the approach, and recommended solutions. Our aim is to support researchers who are considering conducting a realist review to understand the key principles and concepts involved, and how they can go about producing high-quality work.

Screening and identification of fetal alcohol spectrum disorder in criminal legal settings: A realist review

“…We used a flexible and adapted realist review approach that considered a wide variety of data sources and focused on contextual factors impacting screening. Indeed, the realist review approach is intended to be flexible, purposeful, contextual, iterative and theoretical and is guided by the question ‘what works for whom in what circumstances and in what respects?’ (Duddy & Wong, 2022; Pawson et al., 2005, p. 32). As such, this approach was used to understand how, when, why, for whom, and by whom FASD screening tools and/or strategies are used, as well as what specific indicators may aid in identifying people with FASD in criminal legal contexts.…”

“…As such, this approach was used to understand how, when, why, for whom, and by whom FASD screening tools and/or strategies are used, as well as what specific indicators may aid in identifying people with FASD in criminal legal contexts. As our research focused on multiple tools as well as an array of possible indicators, our approach did not use a context–mechanism–outcome configuration to describe results (Duddy & Wong, 2022). However, our approach does offer a discussion of important contextual factors that could have impacted the validity of screening instruments (e.g.…”

“…In the current review, we aimed to synthesise the evidence regarding FASD screening in criminal legal contexts and to develop a conceptual framework for understanding which screening tools, items and/or indicators and characteristics, may be effective in accurately identifying individuals with FASD. A modified realist review approach (Duddy & Wong, 2022; Rycroft‐Malone et al., 2012) was used to delineate the characteristics of FASD screening tools, items, and/or indicators, that may prove to be effective in identifying FASD in criminal legal contexts. In addition, we considered potential implementation factors, with a focus on the populations and settings where screening may take place, points in the adjudicative process where screening may be most needed, and the professionals likely to conduct screening.…”

Improving the mental health and mental health support available to adolescents in out-of-home care via Adolescent-Focused Low-Intensity Life Story Work: a realist review

“…Realist approaches are grounded in the assumption that the same intervention will not work everywhere and for everyone 30. A realist review is a theory-driven interpretive type of literature review.…”

Health visiting in the UK in light of the COVID-19 pandemic experience (RReHOPE): a realist review protocol

“…Mechanisms may be found at different ‘levels’ (eg, organisation or individual level), but to make the claim that something is functioning as a mechanism, it is important to consider for which outcome this claim is being made. Importantly, mechanisms are not merely an intervention, an activity,or an intervention component 27 28. Hence, realist reviews will produce knowledge claims about causes of outcomes in certain contexts using the context–mechanism–outcome configuration (CMOC) heuristic, as this links the three main concepts of a realist analysis.…”