Parts of this chapter have been reproduced from Law et al.1,2 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http://creativecommons.org/licenses/by/4.0/. The text below includes minor additions and formatting changes to the original text.
This project involved a realist synthesis of literature, with input from key stakeholders, public contributors and study management and advisory groups. Co-production was embedded throughout and included a co-design process where intervention ideas were developed. The process was iterative, with data sources informing each other as the synthesis progressed ().
Schematic showing the iterative, integrated flow of information throughout the synthesis. Arrows indicate how each element informed another. The Study Management Group and Project Advisory Group meetings continuously informed the synthesis throughout (more...)
Co-production of the short title and acronym
Members of the study management group, including public research partners, worked together to develop the short title of the study ‘Function First’ and the strapline ‘Be Active, Stay Independent’. We first discussed project keywords and gathered suggestions to develop 11 options. The final version was chosen following a vote and subsequent discussion.
Summary of the overall methods
We used Realist And Meta-narrative Evidence Syntheses: Evolving Standards (RAMESES) methodological guidance and standards to inform our application of realist methods.81
The first stage of the synthesis developed initial theories about how and why primary care interventions aiming to improve physical function and physical activity among patients with long-term conditions work (or may not work), for whom, and in which circumstances. These theories were developed through two theory-building stakeholder workshops and an early scoping search of published and grey literature. This phase informed theory development and literature searches by helping to develop a shared understanding of the key topic areas and stimulate initial ideas.
Following the theory-building workshops, a list of ‘if . . . then’ statements were created, and the search strategy and inclusion and exclusion criteria were developed and refined. The ‘if . . . then’ statements also formed part of the activities included in the first co-design workshop, where stakeholders took part in activities and games to familiarise themselves with the emerging ideas and match different statements.
After an initial sift of titles and abstracts, we refined our inclusion and exclusion criteria a second time and identified relevant systematic reviews. We created an evidence table to capture relevant realist critique from the systematic reviews. This evidence table and the ‘if . . . then’ statements were then used to develop eight emerging ‘theory areas’. To explore these theory areas, we then conducted 10 individual ‘theory-refining’ interviews with stakeholders. These theory areas were explored during the second co-design workshop, with stakeholders considering how particular intervention assets related to the theory areas.
Data were extracted from papers that were identified as rich and relevant to the eight theory areas, including individual qualitative and quantitative studies, guidelines and the grey literature. Through this process, and iterative discussion, we developed initial ‘candidate’ CMO statements. These were further refined and tested through the organisation of extracted information into evidence tables representing the different sources of evidence, including the individual stakeholder interviews. This enabled exploration of confirming or refuting evidence. We also developed a taxonomy of primary care physical activity interventions for people with long-term conditions.
The CMOs formed part of the final co-design and knowledge mobilisation workshops, where participants considered the emerging CMOs and how they were embodied within the co-designed resources. The ‘final’ five CMO statements were then defined, alongside the co-designed ‘Function First’ product.
As this was an iterative process, there were some changes to the original protocol, and these changes are outlined in more detail below.
Changes to the original proposal
shows the original schematic of project progression.
The original schematic of project stages. See Figure 1 for the actual process of project progression.
All elements of the protocol were completed as planned. However, there was more overlap in the stages than originally anticipated because of the large evidence base in the areas of physical activity, physical function and primary care for people with long-term conditions. This meant that:
- a.
We discussed emerging ‘theory areas’ with stakeholders in the theory-refining interviews, rather than conjectured CMO statements.
This was necessary to enable refinement of the key emerging theories and to avoid artificially ‘jumping ahead’ to CMO development. It allowed for a deeper understanding of the areas involved, strengthening the eventual theories developed.
- b.
The CMOs were not finalised before the first co-design workshops began and developed further as we progressed through the workshops.
The integration of CMO development with the co-design workshops was an unanticipated strength of the research process and meant that the refinement of the final CMOs benefited greatly from further stakeholder input, including suggestions for additional literature searches. In addition, the co-design process was not limited by a final set of CMO statements and was more authentic as a result. Chapter 6 discusses this further.
Data sources, analysis and synthesis
The data sources and how they were analysed and synthesised are described in the following sections. The realist synthesis of literature including stakeholder interviews, co-design workshops and knowledge mobilisation event are described in the Chapters 3 and 4.
Early scoping exercise of published and grey literature
To gain familiarity with the literature and aid with the identification of keywords, we carried out a preliminary scoping of the literature to retrieve reports, theses, key articles, systematic reviews and any relevant websites to help inform our formal search strategy. This scoping exercise was informed by previous work in this area,82–84 proposal work-up, early study management, and Project Advisory Group meetings and discussions with patient and public representatives.
Theoretical landscape
As part of the above scoping exercise and to stimulate early thinking about important areas, we identified the overarching theories and frameworks that we determined as likely to inform the realist synthesis. We drew from theories that address a wide social context including theories and models relating to physical function (e.g. International Classification of Functioning, Disability and Health69); environmental factors and individual compensation strategies;85 psychological theories of motivation, behaviour and behaviour change relevant to patients and health professionals (e.g. self-efficacy and self-determination theory,86,87 intention and behaviour,88 health beliefs, planned behaviour);89,90 interventions based around Capability, Opportunity, Motivation – Behaviour (COM-B) principles;91,92 and the self-regulation of illness,93,94 sociological theory (e.g. governmentality,95 habitus,96 social and peer support97,98); implementation theories (e.g. diffusion,99 knowledge to action100); and organisational theories relevant to how interventions fit into different ways of delivering services and pathways.1,101,102
Stakeholder analysis
An initial stakeholder analysis helped us to identify and target the most relevant groups for the different stages of the synthesis.103 It included representation from people with long-term conditions, primary care professionals working in general medical practices, allied health professionals, third-sector organisations, council-funded initiatives, social care, policy-makers and commissioners of services (). The stakeholder analysis was used to ensure that no particularly important groups were missed. The stakeholder analysis took place across two or three study management meetings and one external Project Advisory Group meeting, and involved a process of feedback and iteration. This was consistently monitored, and care was taken to incorporate any missing perspectives. Researchers with knowledge and experience relevant to the syntheses were also identified as stakeholders as the project progressed.
Summary of the stakeholder analysis. LTC, long-term condition; OT, occupational therapy.
Theory-building workshops
Data collection using LEGO® Serious Play®
Creative methods, borrowed from the field of co-design, were employed to structure the two theory-building workshops and elicit the views and experiences of all stakeholder representatives, including a facilitated session using LEGO® Serious Play® (LEGO, Billund, Denmark). This method has been used previously in service improvement work, training of health professionals, and research.104–108 The specific choice to use LEGO Serious Play as opposed to other participatory approaches was because of the tangibility of working with LEGO. LEGO Serious Play enabled a move from ‘research’ activity into ‘co-design’ activity through making tangible outputs. By using these tangible forms, it set some expectations early on in the programme of work about the transition from intangible to tangible (i.e. from evidence to actionable tools). Participants who attended these two workshops are detailed in Report Supplementary Material 1.
Following a series of skills-building activities, each individual created and described LEGO models (physical metaphors) in response to the following questions: ‘What does physical function mean to you?’ and ‘What are your experiences of maintaining physical function?’.
Participants were encouraged to play with the LEGO as they considered the questions, connecting bricks while they thought. This gave the participants’ hands, and part of their mind, an occupation while they pondered the topic in question. What is built by the end of the reflective period then becomes a physical ‘tool’ that can be used to enhance the expression of those thoughts back to the wider group. We also asked participants to summarise their thoughts in three words written on a Post-it Note (Post-it® Brand; 3M, Cynthiana, KY, USA). As shown in , some participants chose to augment this communication further by using illustrations as well as their LEGO model. This demonstrates one of the core values behind the LEGO Serious Play methods; these media give people a greater opportunity for self-expression (using LEGO model plus verbal description) than the sole use of speech or writing.
Example models built by participants in the theory-building workshops to reflect on and describe their interpretation of (a) what physical function meant to them; (b) how they maintained physical function; and (c) an example of a ‘shared landscape’. (more...)
This process helped to develop a shared understanding of the key topic areas and stimulate initial ideas and thoughts for theory development. It exposed researchers involved in the workshop to first-hand lived experiences of people with long-term conditions and professionals working in primary care. These LEGO models were then incorporated into a shared ‘landscape’ to explore which aspects of these experiences helped or hindered the maintenance of physical function (see ).
As shown in , photographics of the models were captured. Participant descriptions were audio-recorded and then transcribed for analysis and interpretation, and were used to shape emerging theories.
Data analysis workshop
The project team members convened for a face-to-face data analysis workshop to interpret the data gathered from the two LEGO Serious Play workshops. The data from the two workshops comprised anonymised transcripts from the audio-recordings of the workshop dialogue, images of the models, three-word text descriptors and images of the landscapes with annotations. The photograph of the LEGO model and the associated transcribed description were linked so that during the group analysis session it was possible to view the LEGO model and the description from each anonymised participant at the same time. This helped to bring the description in the transcript ‘to life’.
The models themselves served three functions:
Building these models aided reflection as people considered the question about their experiences.
The models were a tangible visual aid to assist people in their explanation of their reflections, or personal theories in their heads. Once their initial explanation was ‘out’, the explanation became refined through careful probing and ‘cross-examination’ by the facilitator until the explanation stood alone without the need of the model to ‘support’ it.
The models acted as a tangible representation for all participants in the workshop of each person’s theories.
The models were built in response to two specific questions; thus, the transcripts and images were divided into the relevant sections for each participant. This enabled consideration of the responses of each individual holistically and in depth. Researchers also returned to the transcription to understand the ‘full picture’ of the dialogue between participants, as it was the ‘connecting’ dialogue between participants’ responses that often illustrated how the threads of explanatory narratives were formed.
The transcripts and image models were divided between the project team members who identified any explanatory statements coming from the participants (see Report Supplementary Material 2). For example, statements that explained why they (or others they knew) had (or had not) done certain things, why some things had or had not happened, why health-care professionals had (or had not) done specific things. To gain familiarity with the process, the copies of the first two examples were shared among project team members who reviewed them together as ‘worked examples’. We discussed the statements that identified and moved towards a shared understanding of what we were all looking for.
In some instances, it was helpful to refer to the images of the individual models, the landscapes and annotations. However, there was common agreement that the annotations had often been stated ‘out loud’ and captured on the audio-recording, and the images of the models were limited in their usefulness. The images occasionally prompted a memory but were considered to be more useful in the moment of creation and immediate explanation (in the participatory theory-building workshops) rather than as a longitudinal record of the event for later analysis. We were seeking the theory in the heads of the stakeholders, and so we were reliant on their explanations. Therefore, the models could be defined as a data extraction tool (i.e. something to elicit the personal theories from stakeholders’ heads into forms that others could engage with). Once this was complete, the model no longer served a purpose as the recorded explanation formed the data. Moreover, the models themselves were not interpreted by the research team; it was the transcripts of participants’ explanations that were used to inform theory-building.
All explanatory statements were highlighted, discussed and compiled into a list of 28 ‘If . . . then . . .’ explanatory statements (see Report Supplementary Material 3). These nascent theories were used to identify early clues to possible ‘contexts’, ‘mechanisms’ and ‘outcomes’. The information from these workshops was then used to refine the literature searches and the inclusion and exclusion criteria. For example, these workshops raised our awareness of the importance of studies detailing the wide range of barriers to and facilitators of physical activity and the implementation of interventions, as well as the range of community-based physical activity opportunities.
