Syddansk Universitetsbibliotek - LibGuides
Syddansk Universitetsbibliotek - LibGuides
The realist review method is theory-based and qualitative and aims to map the mechanisms that make complex social interventions work/not work. The method involves several types of studies and to a large extent gray literature.
The method is structured in 5 iterative steps: 1) Identification of program theories, 2) Search for evidence, 3) Article selection, 4) Data extraction and data organization, 5) Evidence synthesis and conclusion. These steps are reviewed and revisited until no more literature is identified that contributes to new understanding of the intervention. The 5 steps are further described in the METHOD.
The synthesis is a narrative, explanatory analysis and the starting point is program theory. (Social) theory is combined with empirical evidence to summarize and further develop the program theory.
The analysis makes recommendations for policy and decision makers about the intervention.
The realist review method is a theory-based and qualitative method originally developed in criminology. It is now primarily used to analyze interventions in the health and social care sector, where many different and interdependent societal or personal factors interact. Whereas other intervention research often focuses on whether an intervention works on a population and to what extent, the realist approach does not focus on the intervention itself, but instead tries to identify the mechanisms that facilitate an intervention to work and under what conditions [1, 2].
The method is particularly suitable for complex social programs involving human decisions and actions, for example. Questions such as: "What workplace interventions are needed to improve the mental health and life satisfaction of caregivers?"[3], "What contexts and mechanisms improve access to medical care among homeless people?" [4], "Why and how does an individual-oriented chronic disease prevention intervention work in relation to health behaviors in groups at increased risk?"[5], "Participant involvement in ergonomics. What works for whom and why?"[6]. The method is therefore suitable for targeting decision makers. It typically takes two or more people to conduct a realist review.
Recognizing that there are many different factors that influence whether an intervention works as intended, a realist research question includes one or more of these elements: What works for whom, how, why, to what extent and under what circumstances, in what way, and over how long? [5]. You try to identify the contexts, mechanisms and outcomes that can describe that a given intervention works, whether it is directly described or whether it is something that can be interpreted from the context, thus opening up the possibility that the learning from the specific interventions can be transferred to others.
The method is described in 5 iterative steps [7]:
Step 1, identifying a preliminary program theory:
In this stage you make your preliminary hypothesis about what works for whom, how, why, to what degree and under what circumstances, in what way, and over how long. This initial hypothesis is called the Initial Program Theory. The realist method sets up its program theory in a CMO configuration, which describes a context in which mechanisms (M) and context (C) interact to generate particular effects/outcomes (O).
It's rather general at first, but as you read into the topic, it becomes more specific. What you are looking for is not necessarily written exactly as hypotheses in the articles, but has to be "dug out" as an underlying understanding or underlying statement.
Below is an example of an analysis of employee involvement in ergonomic changes in the workplace from a realist review [6]. This CMO configuration/program theory is one of 13 that form the basis for discussion and conclusion in the review:
"Employee involvement in the implementation phase (C) increases the durability of the changes (O) as employees take ownership and consider the changes relevant (M)".
Below is a text quote that supports/has formed the background for this CMO configuration:
Quote from: Rasmussen 2017, p. 497: "Processes, barriers and facilitators for implementation of a
participatory ergonomics program among eldercare workers". Rasmussen, CDN, et al, Applied Ergonomics 58 (2017).
"Common factors for the mentioned studies and the results in this study are the importance of direct involvement of the employees, their level of influence in the implementation process and the available resources (in terms of money, time, materials, etc.)."
Step 2, Searching for evidence:
Next, you do a preliminary literature search using several search methods, such as chain search, systematic block search, gray literature search and finally a search via personal, professional networks - all relevant sources you can think of. Along the way, you may find more background statements / hypotheses / program theories in the literature you come across.
The initial searches help to strengthen and possibly further develop the already identified program theories in suitable articles, and as you gain a more detailed knowledge of the topic, you can describe inclusion and exclusion criteria and eventually end up setting up a final search.
Step 3, Article selection:
In this phase, you find suitable studies that meet the inclusion and exclusion criteria and then describe one or more program theories/CMO configurations that can describe that a given intervention does or does not work, for whom, how, why, to what extent and under what circumstances.
Step 4, Data extraction and organization:
In this phase, you extract data from the included studies, i.e. you find article citations that support or refute your program theories, and in this phase you refine the program theory - through discussion between co-authors - to the final version. This is where iterative work comes in - at all stages of the work, program theories and citations are discussed between those working with the data to adjust and finally reach consensus. In this phase, some of the original program theories are discarded, while others are revised and new ones may be added.
Step 5, Evidence synthesis and conclusion:
In this phase, you test your conclusions and recommendations with stakeholders, and you try to implement solutions in real life.
However: In a realist review, one can make do with several concrete recommendations. For a more detailed description of the method, refer to:
http://www.ramesesproject.org/ and the book "Purposeful Program Theory" af Funnel, et al.
The realistic synthesis is initially based on a critique of quantitative meta-analyses and systematic reviews. Ray Pawson, who has helped to develop the realist evaluation approach, criticizes the quantitative meta-analyses and systematic reviews for, through restrictive exclusion processes, removing essential insights about the context in which an effort is implemented. Systematic reviews and meta-analyses can therefore according to Pawson, is not used to explain why or under what circumstances an intervention works [5,8].
University Library of Southern Denmark Odense | Esbjerg | Kolding | Slagelse | Sønderborg +45 6550 2100 | sdub@bib.sdu.dk
