Parts of this chapter have been reproduced with permission from Law et al.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: https://creativecommons.org/licenses/by/4.0/. The text below includes minor additions and formatting changes to the original text.
Summary of findings
‘Function First’ aimed to determine the role of primary care in promoting physical activity and physical function for people with long-term conditions and developed programme theories, which describe what works, for whom and in what circumstances. This study is the first realist synthesis in this area, offering a theory-based consideration of the current literature, views and perspectives of the role of primary care in promoting physical activity and physical function for people with long-term conditions.
This realist evidence synthesis with embedded co-design has resulted in five CMO statements and a co-designed set of resources that embody these CMOs and are designed to facilitate improvements in physical activity and physical function for people with long-term conditions in primary care.
The five CMO statements are as follows:
Primary care settings are characterised by competing demands, and improving physical activity and physical function are not prioritised in a busy practice (context). If the practice team culture can be aligned to promote and support the elements of physical literacy (mechanism), then physical activity promotion will become routine and embedded in usual care (outcome).
Physical activity promotion in primary care is inconsistent and unco-ordinated (context). If specific resources are allocated to physical activity promotion (in combination with a practice culture that is supportive) (mechanism), then this will improve opportunities to change behaviour (outcome).
People with long-term conditions have varying levels of physical function and physical activity, varying attitudes to physical activity and differing access to local resources that enable physical activity (context). If physical activity promotion is adapted to individual needs, priorities and preferences, and considers local resource availability (mechanism), then this will facilitate a sustained improvement in physical activity (outcome).
Many primary care practice staff have a lack of knowledge and confidence to promote physical activity (context). If staff develop an improved sense of capability through education and training (mechanism), then they will increase their engagement in physical activity promotion (outcome).
If a programme is credible to patients and professionals (context), then trust and confidence in the programme will develop (mechanism) and more patients and professionals will engage with the programme (outcome).
We also co-designed a prototype multicomponent intervention with a range of stakeholders, embodying the emerging programme theories and developing resources to promote physical activity and physical function for people with long-term conditions. These consisted of resources designed to help nurture physical literacy among practice staff and complement existing schemes promoting physical activity. Suggestions for changing the layout of the practice and other promotional materials are included and would aim to create an environment encouraging physical activity. Although it is important that all members of staff promote physical activity during routine consultations, resources have been developed for a credible professional who can concentrate on physical activity behaviour change with individual people with long-term conditions. We developed prototype materials to facilitate this behaviour change, as well as resources to signpost to local community assets and to address barriers to the uptake of physical activity, such as community transport schemes. People would be directed to local exercise opportunities, clubs and groups following the adaptation of existing electronic social prescribing directories. These resources would need further development and refinement before being ready for use in practice (see Chapter 7).
Findings in the context of previous research and guidance
This realist synthesis of evidence has drawn from a wide variety of evidence sources and has identified many interventions designed for use by primary care professionals to promote physical activity to people with long-term conditions (e.g. LGM; Movement as Medicine for Type 2 diabetes). These have had varying levels of success in implementation. There are also a growing number of initiatives that provide additional education and resources for GPs and primary care staff on physical activity promotion for different long-term conditions (e.g. Motivate2Move, Moving Healthcare Professionals Programme, Moving Medicine). There is also the ‘Active Practice Charter’ developed by the Royal College of General Practitioners, which is designed to encourage practices to become ‘Active Practices’ by adopting ways of working to encourage physical activity in both patients and professionals. As described as part of this synthesis, research within the area of physical activity promotion interventions in primary care is rapidly developing; for example, the number of PubMed-referenced articles about physical activity within primary care has doubled between 2012 and 2014.153
Recent Public Health England guidance246 on the prevention and management of long-term conditions summarises health and social benefits of physical activity, discusses the scale of physical inactivity and the barriers that people with long-term conditions experience. It then sets out available physical activity resources, programmes and campaigns for the public, physical activity training for health-care professionals, digital toolkits for clinicians, and initiatives in general medical practices such as the ‘Active Practice Charter’ and the parkrun practice initiative. There is a call for action for health-care professionals, local authorities and NHS organisations. What ‘Function First’ adds is:
a set of evidence-based theoretical statements of what works, for whom and in what circumstances when promoting physical activity and physical function for people with long-term conditions, and how this could be embedded in routine practice
a prototype multicomponent intervention of how this could be implemented, which involves practice alignment, a set of resources that sit alongside existing interventions and initiatives and makes a credible intervention for people with long-term conditions and their treating clinicians.
To our knowledge, this realist synthesis is the first to explore the role of primary care in promoting physical activity and physical function for people with long-term conditions. There are a limited number of existing realist evidence syntheses in the area of physical activity promotion for people with long-term conditions. In the primary care setting, a realist review has explored what works, and why, in the identification and referral of adults with comorbid obesity to weight management services. Similar mechanisms to the current study emerged including those operating at an individual level, such as having the confidence to talk about weight management and recognition of the importance and value of weight management. At an interpersonal level, mechanisms included discussing weight management at the right time and without blame. However, most mechanisms were operating at primary care practice level, requiring changes to systems or culture, such as developing consensus around working to promote weight management, not assuming ‘one size fits all’ and improving communication between practices and weight management services. Importantly, this review also identified contextual factors including varying patient and practitioner characteristics and competing priorities.247
In an evidence synthesis, which included a realist approach, different versions of a health-related lifestyle advisor role in improving health were explored. Levels of acceptability were high, with advisors helping to remove barriers to prescribed behaviour and creating facilitative social environments. However, there was only limited evidence suggesting that lifestyle advisors have a positive impact on health knowledge, behaviours and outcomes, and there was some ambiguity about the role and the impact of the lay and peer characteristics of the interventions,248 thus indicating that further development of this role and subsequent research is required. Finally, similar evidence in a realist review found potential benefits of promoting physical activity when health care was located in leisure settings including theories relevant to the creation of a physical environment that re-enforces a physical activity culture and supports behaviour change, improvement of staff and patient experience, increased collaboration and co-ordination between health professionals, and increased awareness of facilities. Theories explaining the challenges of embedding physical activity in the NHS concerned the logistics of service delivery and inconsistency of clinical schedules.249 One example of a strategy designed to create a culture of physical activity across a larger population is ‘Move More’ in Sheffield, which has been commissioned in three locations where medical intervention and physical activity are combined.250
As far as we are aware, this study is the first to co-design intervention resources that embody realist programme theories, particularly in the area of physical activity promotion for the primary care management of people with long-term conditions. There are limited examples of the application of realist methods to facilitate intervention development. One such example is a study developing a complex intervention for improving outcomes in elderly patients following hip fracture. Three programme theories were developed and used to inform intervention development: improving patient engagement by tailoring the intervention to individual needs, reducing fear of falling and improving self-efficacy to exercise and perform activities of daily living, and co-ordination of rehabilitation delivery.251 Studies have also used participatory methods to co-create new ways of working. For example, ‘Movement as Medicine’ included stakeholder work to develop a prototype intervention136 and the Choose to Move programme in Canada has been shown to enhance physical activity, mobility and social connectedness in older adults.252 This study also highlighted the central role of community-based partner organisations when scaling up complex interventions in diverse settings, alongside the need to interact with stakeholders to develop collaborations and engage at multiple levels, in multiple sectors.252
Strengths of this work
This project had a strong and embedded co-production element. People with long-term conditions, primary care health professionals, researchers with relevant expertise and experience and other relevant stakeholders were involved regularly at all stages, both as research participants and as members of the study oversight groups. The Study Management Group and independent Project Advisory Group provided scrutiny and advice, but also academic input into the development of the programme theories and co-design processes. Research participants were involved in co-creation activities designed to facilitate input into the development of programme theories and associated Function First resources. An initial stakeholder analysis strengthened this co-production element by ensuring that we captured insight from relevant groups.
We carried out systematic, comprehensive and transparent literature searches to identify a wide range of evidence that could be used to identify underlying programme theories. We used Covidence software to enable a team of reviewers to contribute to all aspects of the review process involving a large data set of publications. The theory-building work informed the inclusion and exclusion criteria, which were refined in an iterative manner. Implementation of the ‘good and relevant enough’ criteria at the data extraction stage allowed a more inclusive approach within the synthesis. Discussion among review team members, and iteration as the review progressed and programme theories were developed, was essential to ensure rigour. Discussion about the nature and quality of the evidence also helped to inform thinking about the emerging programme theory as the synthesis developed.
We co-designed a set of tangible resources that embodied the programme theories. These resources were designed to be flexible and to augment existing initiatives that are working successfully. They need further development and refinement before they can be used in primary care consultations.
A strength of co-design is that it involves giving new ideas tangible form and testing how these will work in the real world. It is ‘solution’ focused, looking to introduce something into the world that will create change. It considers the ‘solution’ at various scales, from the specific to the general, from the individual to the mass market. It considers technical as well as aesthetic matters. Co-design is democratic and pragmatic, ensuring that the issues relating to desirability, feasibility and viability have been considered by the involved stakeholders both throughout the generation of ideas and the embodiment of those ideas into tangible forms. It provides a solution within constraints; for example, if cost is the limiting factor, then the solution will be tailored to fit the budget.
The use of material culture such as LEGO Serious Play and the physical prototypes to share ideas and knowledge in co-design processes is a vital component for democratising the process by removing the exclusivity of technical language. It is also a ‘designerly’ skill and inherent to ‘designerly’ ways of thinking. This makes it a notable feature that distinguishes ‘designerly’ co-design (led by and facilitated by designers such as in this study) from ‘design-like’ co-design (led by and facilitated by non-designers).253 The visual and physical objects made become boundary objects for sharing ideas and knowledge. We hypothesise that this relationship with material culture has implications between the final physical manifestation (the intervention) and knowledge mobilisation or use of the intervention. Material objects can mediate and afford people agency, guide a person through a series of actions or words, and give them a tool with which to perform a specific action. For example, when a person unpacks a new mobile telephone, the package, its contents and the way it is packed should lead the consumer through a series of intuitive steps. This agency that the consumer is afforded is mediated by the design of the box, its contents and the way that it is arranged.
The use of qualitative interviews to help refine the programme theories is an established part of realist method and was extremely valuable in this study. As described by Mukumbang et al.,254 these interviews helped to ‘reinforce and maintain theoretical awareness and contribute to trustworthiness’. They were also helpful in other ways: guiding further literature reviewing, enabling more in-depth exploration with stakeholders we had not yet been able to hear from, contributing to a greater understanding of context and acting as a sense-checking point for the developing ideas.
Overall, the realist approach offers a theory-driven explanation that considers theoretical depth, breadth and transferability. Guided by realist principles, it is based on the assumption that reality is dynamic. Theorising was not purely deductive (theory-testing) or inductive (theory-building), thus providing more scope for deeper theorising and greater ontological depth, and potentially wider applicability of the resulting theories.255 The key methodological strength of a realist approach is the requirement to pay attention to context. The analytical stance assumes that individuals, including patients and primary care staff, and the systems in which they live and work, bring different perspectives to the promotion of physical activity and physical function. Consequently, different patterns of outcomes may be observed, which can be specific to time and place, and, in this sense, the successful promotion of physical activity in primary care can be described as emergent. Any structured attempts to improve activity therefore meet the definition of a complex intervention and require particular methods for development and evaluation. Although the purpose of some systematic review approaches is to manage and control for context in the specification of an overall effect size, a realist approach has enabled us to focus our analysis on the interactions between context and physical activity interventions. We have done this by specifying conjectured, contingent hypotheses about what might work and how physical activity programmes generate engagement and success.
Finally, this study was conducted in a highly iterative, at times almost cyclical manner. Originally intended as a largely linear and sequential process, it emerged during the process of conducting the study that the iterative nature of the interaction between elements of the study added unanticipated strength and richness. It facilitated a greater integration of the different sources of data and enhanced the depth and breadth of the findings. This demonstrated that it is possible to deliberately overlap stages in a realist synthesis with embedded co-design, enabling the process to be more emergent. This became a much more natural evolvement of knowledge and helped to improve our collective understanding of the evidence into a shared body of knowledge, which was then embodied into the ‘product’. In fact, the striking ease with which the two methodological approaches (realist review and co-design) related to each other led us to reflect on their similarities and differences. ‘What works, for whom in specific circumstances’ connects the two as a fundamental way of expressing realist synthesis, yet also an important aspect of design. Although similar, realist is more cognitive and theoretical, whereas design is more practical, tangible, based on making tools, artefacts or materials. However, employing a design–evaluation cycle is hypothesised to improve the design and implementation of complex innovation by using programme theories to develop design propositions, which are evaluated through realistic evaluation, resulting in further refinement of programme theories.256
Challenges, limitations and alternative approaches
A realist approach generates evidence-based hypothetical ‘recommendations’ that are developed at a specific time and place, by specific stakeholders. They are context specific and, as with all methods based on qualitative findings, they may not apply in a different set of circumstances.
The literature searching identified a large volume of literature due to the broad research area. The aim was to be comprehensive to capture a diverse range of evidence. This meant that we had to manage large numbers of publications and had to adapt our reviewing methods to deal with this large volume. The Covidence software was helpful in managing the data and for screening abstracts. Screening guidelines and inclusion/exclusion criteria went through multiple stages of refinement by the project group who all understood and agreed on the final criteria. This meant that all members of the project group could take part in the screening of abstracts.
To identify programme theories as soon as possible at the start of the review, we carried out a broad literature search early in the project. This search aimed to capture descriptions of all interventions that had been designed to improve physical function in primary care for people with long-term conditions. This gave us a large library of relevant papers that we could select from but presented a large number of data to sift. An alternative option would have been to build up the initial programme theory areas based on preliminary scoping and a review of reviews, and then to design a purposive search designed for each of the identified theory areas.
The grey literature search included a review of websites and publications of relevant UK organisations. However, only a small number (five items of grey literature and five guidelines) were included in the final data set. We recognise the importance of these other sources of evidence for avoiding publication bias, to provide useful contextual information and to understand interventions and programmes that are already used in practice. We faced a challenge in capturing undocumented knowledge. Although we did not extensively survey social media communication channels, the review benefited from the strength of the stakeholder group and attendees at the workshops who were able to identify additional unpublished evidence.
Despite having identified a large number of published data on physical activity interventions for people with long-term conditions, we found that in most of these publications there was a lack of detailed description (thick description) of the organisational context, delivery planning, patient characteristics and the circumstances that led to the success of the intervention. We also found fewer published reports of negative results, or descriptions of complications that had led to interventions not being as successful as planned. However, we aimed to identify, and present in this report, the clearest evidence describing what lies beneath the success of an intervention.
As a result, coverage of what ‘does not work’ may appear to be less well covered than ‘what works’. Some of this is due to framing (i.e. we tended to frame our statements and programme theories more positively than negatively). However, it is important to note that although more difficult to find, we did make a point of exploring ‘what does not work’ in the literature. For example, ‘one size fits all’ approaches, and interventions introduced without consideration of existing schemes and pathways, appeared to work less well.
The LEGO Serious Play method, facilitated by experienced researchers, successfully enabled all participants to reflect and build models to help them explain their thoughts regarding the questions posed in the ‘theory-building’ workshops. However, the following two challenges are standard to the practice of co-design. First, there were some challenges associated with instilling confidence in participants to build models on an abstract topic. However, facilitators led participants through a series of skills-building ‘warm-up’ type activities that helped participants feel comfortable using the methods. Second, one participant explained how the activity was an emotional process because the activity and questions posed were reflective in nature and likened it to ‘therapy’. This was not a new finding and is common to many methods that encourage reflection, deeper thought and explanation. We reiterated at future workshops that although there would be scheduled breaks, it was also completely acceptable for anyone, to take a break at any time if they felt that they needed a physical or emotional break. There were two ‘new’ challenges that arose during the LEGO workshops. First, one of the workshop participants was partially sighted but was able to contribute fully to the workshops; we spent more time with them, explaining what was in front of them and using larger LEGO blocks. Second, it proved more challenging to encourage professional participants to reflect on their professional experiences of working with people with long-term conditions, rather than their own personal experiences of physical activity and physical function.
Co-design is often criticised for being too specific, focusing on the needs of the participants in the process, resulting in personal rather than generalisable solutions. Design strategies, such as the use of personas, is a mechanism for addressing this as it asks co-design participants to design for other people (the personas) that are more widely representative. This weakness was mitigated further by introducing the evidence through the medium of card games at the start of the co-design process. This shared the wider, generalisable evidence with the co-design participants. Thereafter, the visual story of the whole project and the list of evolving CMOs was displayed on the wall for every co-design workshop, ensuring that the participants maintained this wider context and understanding of the evidence throughout the co-design activities.
There are many initiatives promoting physical activity. There are challenges when introducing yet another physical activity programme in a saturated environment. From the outset, we aimed to complement and not compete with these campaigns. We acknowledged and addressed this by involving representation from relevant bodies in our activities and including a specific search for existing initiatives and campaigns. We believe that many of these initiatives could address some of our programme theories, for example the promotion of physical literacy among all practice staff, and facilitating behaviour change within the consultation. However, these initiatives do not address all of our five programme theories and are individually insufficient to promote physical activity and preserve physical function in all of the people with long-term conditions managed in primary care. We have designed a multicomponent intervention, in which several of the components could be interchanged with materials that have already been developed.
We also acknowledge that despite using a stakeholder analysis, our range of stakeholders included people with similar socioeconomic status, ethnicity and attitudes and, therefore, the findings may be limited in their application. However, although indirect, these varying considerations were represented in the literature and in workshop activities that involved considering issues from the perspective of a range of different people.
