Background
Physical activity is comparable to smoking and diet in terms of influence on chronic disease outcomes [
1]. Hence, alongside population-wide physical activity promotion, interest is growing in interventions targeted towards populations whose conditions would benefit from physical activity [
2]. One such approach, exercise referral schemes (ERS), has proliferated rapidly in recent years [
3]. To date, randomised controlled trials and observational studies have provided limited evidence of long-term effects on physical activity [
4,
5]. However, despite substantial heterogeneity in their design [
6], limited attention has been paid to understanding how the delivery of ERS might be improved [
7,
8]. Attending not only to whether ERS 'work', but also to 'what works, for whom and under what circumstances' [
9] is crucial in attempting to improve their impacts.
The often limited long-term impacts of ERS have been attributed in part to limited uptake and adherence, with current evidence indicating that 1 in 3 referred patients do not attend a first appointment, while completion rates range from 12-52% [
4,
10]. Furthermore, recent observational studies have indicated that adherence is often patterned by patient characteristics. A number of studies for example indicate that mental health patients are less likely to complete ERS [
7,
11,
12], that women are more likely to enter though less likely to complete [
7,
11,
13] and that older patients are more likely to complete [
11,
14]. Socioeconomic patterning has varied across schemes. For example, uptake in one [
15] was lower in deprived areas and among patients with lower education [
16], while in another, participants were of lower SES than the population average, although those from the most deprived areas were least likely to participate [
13]. Another reported higher referral volumes in deprived areas in London, with uptake and adherence equal between groups [
14].
In attempting to improve the effectiveness of ERS, it is crucial to go beyond describing patterning in uptake and adherence, and towards understanding the reasons for this limited and often socially patterned reach. Gidlow and colleagues [
8] have highlighted the need for qualitative research in order to explore these issues. Indeed, recent years have seen increasing numbers of largely qualitative studies exploring processes through which ERS might influence adherence and behavioural change. Perhaps the most common theoretical model guiding this literature is self-determination theory (SDT; [
17]). SDT emphasises the need for interventions such as ERS to support three basic psychological needs; autonomy, competence and relatedness. Through supporting these needs, SDT argues that externally regulated changes, such as taking up exercise on instruction of a health professional, may over time become internally regulated. That is, performed for intrinsic enjoyment or because of associations with personally valued goals.
Studies have therefore focused upon the roles of components such as professional supervision and guidance in supporting these needs. Markland and Tobin [
18] for example reported higher identified motivation (i.e. acting due to a sense of the behaviour as personally important) when patients perceived that professionals were supportive of autonomy and competence. An empathic, patient-centred approach was highlighted in one qualitative study as a welcome contrast to more authoritarian dealings with health professionals [
19]. In another [
20], the authors concluded that older women integrated exercise into their identities through development of a sense of mastery, as they were supported by the exercise professional. Indeed, access to an instructor who is both knowledgeable and provides effective interpersonal support has emerged as a key determinant of adherence in almost all qualitative studies [
20‐
23]. Many early trials forming the basis for assessments of the efficacy of ERS however offered limited or unspecified professional support beyond consultations to discuss an exercise programme (e.g. [
15,
24,
25]). Indeed, some qualitative studies have attributed poor adherence to limited continuity in professional support provision [
26] or a perceived need for more interpersonal support than provided [
27].
In addition to professional support, roles of collective exercise with patients in a similar position are emphasised throughout this evidence base as crucial in facilitating adherence [
20‐
23,
26,
28,
29]. In Markland and Tobin's aforementioned study for example, patients reported less external regulation when reporting higher levels of social assimilation into the exercise environment. Higher perceptions of personal relatedness (i.e. perceptions of supportive interpersonal relationships within the exercise environment) were associated with more internal regulation. Furthermore, Stathi and colleagues [
28] conducted longitudinal interviews with older patients in a 12 month exercise programme, concluding that whilst support for knowledge, self-efficacy and competence was crucial in facilitating adherence, intrapersonal changes were more likely to lead to long-term behavioural change where accompanied by formation of long-term interpersonal networks. Structures for fostering social support networks have however been absent or not described within many ERS, with many simply offering discounted access to mainstream services. A minority of trialled schemes have focused on provision of group-based exercise opportunities [
30,
31], whilst few have explored how social networks are maintained in order to support ongoing activity. Indeed, qualitative studies identify disappointment amongst some patients at limited opportunities for social interaction within ERS, a desire for patient-only classes [
23,
29], or a sense of being 'dropped' after a short-term programme [
31].
To date, evidence for the processes through which ERS support adherence and behavioural change has almost exclusively incorporated patient perspectives. Although professionals are well positioned to offer insights into issues such as which patients exhibit the most and least positive responses to such schemes, challenges and solutions in engaging patients in physical activity, and how practices are refined through experience to encourage uptake and adherence, fewer studies have explored views of implementers. The only studies incorporating views of exercise professionals have interviewed a small number alongside patients or referring professionals [
28,
32]. This paper presents analyses of interviews with exercise professionals within the National Exercise Referral Scheme (NERS) in Wales, exploring professionals' experiences of engaging diverse clinical populations in an ERS and emerging practices to support uptake and adherence. Interviews were conducted as part of the process evaluation of NERS, which was evaluated via a pragmatic randomised controlled trial, economic evaluation and mixed-methods process evaluation [
33].
Discussion
The present study offers a number of insights into how and for whom a national ERS was seen by implementers as facilitating adherence and behavioural change, as well as the emergence of local strategies for meeting patient needs within a national scheme. Firstly, in talking about the referral process, there was debate as to whether this should seek to motivate patients, or start from the assumption that patients will only adhere where change is already internally motivated. Markland and Tobin [
18] comment that motivation is inherently external on scheme entry, given that entry is based on recommendation from an authority figure. However, contrary to this view, professionals commented that rather than acting on instruction, many patients had sought referral from their health professional, with the process of fully internalising exercise motivations perceived as more likely where motivations were already somewhat internalised prior to referral.
Some studies have focused upon assessing motivation for change prior to offering primary care based physical activity schemes [
36,
37]. This is however a controversial approach, in terms of the ethics of offering a scheme based upon a subjective judgment of motivation [
38]. Rather than a fixed trait, motivation is largely a fluctuating product of social interaction [
39], and expressions of ambivalence may reflect the manner in which the advice or offer of referral is presented to the patient. Interpreting resistance as a sign that a patient is unwilling to change may lead to the scheme being withheld from patients who would benefit. It may be that with future integration of evidence-based motivational communication strategies such as motivational interviewing [
34,
39] and effective goal setting processes, as recommended by Department of Health Quality Assurance Frameworks [
6], the scheme may better engage patients who are initially more ambivalent about change.
In addition to the role of baseline motivation, professionals discussed perceived linkages between patients' responses to NERS and the conditions with which they were referred. Professionals commonly described greater barriers to uptake and adherence amongst mental health patients, including limited confidence and additional anxieties assimilating into the exercise environment. Demographic variations were also reported, with older patients and women described as exhibiting additional anxieties on entry to the scheme, though sometimes as finding it easier to assimilate and develop a sense of relatedness to others, due to the fact that most patients within NERS classes were also older females. Hence, the explicitly group-based structures may have gone some way towards offsetting tendencies for lower adherence amongst female participants observed in some ERS [
7], though may have enhanced tendencies for higher adherence in older patients. Many also commented that the scheme was better received in more affluent areas, with lower engagement in poorer areas attributed to factors including lower motivation, limited financial resources and limited buy in among GPs. Findings on socioeconomic patterning in uptake and adherence of ERS are at present equivocal [
13,
14] and objective examination of such patterning is a priority for future analysis.
The roles of the professional in encouraging patients to take up the scheme were described as including development of effective strategies to assuage anxieties about entering the programme, consistent with reports of patients in previous qualitative studies regarding the intimidating nature of leisure centre environments [
23]. Some identified initial contact about joining the scheme or initial consultations as opportunities to reassure patients that they would work alongside patients, and that they would not be pushed to do anything they were uncomfortable with. Hence, consistent with SDT [
17], framing the programme in a manner which was autonomy promotive rather than controlling and did not threaten to undermine patients' sense of competence, was crucial in facilitating uptake.
Once patients attended, guidance on exercising within the limits of their conditions was described as crucial in providing patients with the skills to become autonomous exercisers without risking aggravating health problems. However, consistent with the dual roles of educator and provider of interpersonal support described in previous qualitative studies [
19], such talk was commonly inseparable from discussion of the need to provide interpersonal support for confidence and motivation. Indeed, instruction was sometimes seen as secondary to mentoring roles given the perceived vulnerabilities of the client group.
Although as described above, professionals often identified lower adherence and greater anxieties relating to the exercise environment amongst mental health patients, some commented that where provided with sufficient interpersonal support, mental health patients benefitted substantially from the programme. Hence, the degree of opportunity for social interaction within NERS may have gone some way towards overcoming the often reported tendency for lower adherence among mental health patients [
7,
11]. However, variable reports regarding mental health patients' adherence perhaps reflected variable competence in supporting the needs of these patients. Indeed, some expressed a need for further training in dealing with mental health patients, or described discomfort with 'counselling' roles, with concerns expressed regarding the need to balance provision of sufficient interpersonal support to maintain motivation and confidence against risks of dependence upon unsustainable levels of support.
Consistent with research demonstrating the importance of social assimilation into exercise environments in enhancing patients' motivations for exercise [
19,
23,
28], professionals placed substantial emphasis upon roles of other exercisers in supporting or discouraging adherence. Fitter mainstream exercisers were seen as playing a negative role, through providing unrealistic exemplars, undermining internal motivation through thwarting patients' need to feel competent. Conversely, an empathic social environment where it was normal to struggle with illness, and where patients could compare themselves against realistic role models who had progressed after attending the scheme for a while, was seen as helping patients feel that they could become competent, independent exercisers.
However, consistent with conclusions of a recent qualitative study that intrapersonal change are more likely to produce long-term change where accompanied by emergence of long term social support networks [
28], professionals commonly discussed the contingency of ongoing exercise on the maintenance of networks developed during the scheme. Emerging strategies to lessen dependence on programme structures involved fostering social support networks through filtering completers into maintenance classes or exiting patients in clusters, with emerging networks taking over the role of the programme in motivating adherence to physical activity. Nevertheless, a tendency emerged in many areas for patients to be offered indefinite access to NERS classes, due to perceptions that continued activity was often contingent on ongoing support; a level of provision likely to become unsustainable as the scheme progressed.
Though strengths of the present study include a high response rate, with views of professionals in all areas delivering the scheme during the trial represented, the position of both the researcher and participants need to be considered. As typical of evaluative research, the interaction between an evaluator, linked to a trial of a scheme whose future hinged on positive findings, and professionals whose livelihood depend on its continuation, may have produced an understanding which portrayed the scheme in an excessively positive light. Hence, future research will focus upon the consistency of the trends perceived by professionals with findings from quantitative analysis of adherence and the extent to which psychosocial change processes were triggered by the programme, as well as perspectives of patients on the issues described in this paper.
Conclusions
Whether the referral process should seek to direct resources toward motivated patients, or whether it is feasible to integrate activities such as motivational interviewing in order to enhance effectiveness for more ambivalent patients deserves consideration. Efforts to promote uptake among referred patients should emphasise aspects of the service which promote autonomy and which do not threaten to undermine patients' feelings of competence, thus assuaging anxieties regarding entry to the scheme. ERS professionals' training should emphasise providing professionals both with the skills to fulfil instructional roles, and the interpersonal skills to engage patients in physical activity. This appears particularly critical where dealing with patients likely to face additional anxieties assimilating into exercise environments, such as mental health patients. However, there is also a need to understand how patients' transition from the scheme ought to be best supported to avoid change becoming contingent on the support of the programme. Though some individuals may access ERS largely for the advice of an instructor without the need for wider social support, for many patients, provision of explicit opportunities for social interaction through patient-only group classes offers a means of lessening anxieties and assisting social assimilation into the exercise environment. This context potentially provides realistic role models to support patients' sense of competence, and opportunities to develop social networks supportive of longer term change. The effectiveness of emerging strategies to foster social networks to support the transition to autonomous activity, such as filtering patients into maintenance classes, or exiting patients in clusters, deserves close attention.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
GM led the development and implementation of the process evaluation under the supervision of SM and LM, conducted analyses and drafted the paper. SM managed the overall evaluation. All authors commented on drafts and approved the final manuscript.