‘I wouldn't push that further because I don't want to lose her’: a multiperspective qualitative study of behaviour change for long‐term conditions in primary care

Abstract Background Health outcomes for long‐term conditions (LTCs) can be improved by lifestyle, dietary and condition management‐related behaviour change. Primary care is an important setting for behaviour change work. Practitioners have identified barriers to this work, but there is little evidence examining practices of behaviour change in primary care consultations and how patients and practitioners perceive these practices. Objective To examine how behaviour change is engaged with in primary care consultations for LTCs and investigate how behaviour change is perceived by patients and practitioners. Design Multiperspective, longitudinal qualitative research involving six primary health‐care practices in England. Consultations between patients with LTCs and health‐care practitioners were audio‐recorded. Semi‐structured interviews were completed with patients and practitioners, using stimulated recall. Patients were re‐interviewed 3 months later. Framework analysis was applied to all data. Participants Thirty‐two people with at least one LTC (chronic obstructive pulmonary disease, diabetes, asthma and coronary heart disease) and 10 practitioners. Results Behaviour change talk in consultations was rare and, when it occurred, was characterized by deflection and diffidence on the part of practitioners. Patient motivation tended to be unaddressed. While practitioners positioned behaviour change work as outside their remit, patients felt uncertain about, yet responsible for, this work. Practitioners raised concerns that this work could damage other aspects of care, particularly the patient–practitioner relationship. Conclusion Behaviour change work is often deflected or deferred by practitioners in consultations, who nevertheless vocalize support for its importance in interviews. This discrepancy between practitioners’ accounts and behaviours needs to be addressed within primary health‐care organizations.


Background
Long-term conditions (LTCs) are common in the general population, especially for people over the age of 50, 1 and the number of people living with LTCs is expected to rise. 2,3 LTCs such as chronic obstructive pulmonary disease (COPD), coronary heart disease (CHD), diabetes and asthma currently affect between 1.7 and 6% of the general practice population in England. 4 Effective management of the most prevalent LTCs requires patients to make dietary and lifestyle changes. [5][6][7] Behaviour change, such as smoking cessation and increasing physical activity, can improve the prognosis of people with LTCs. [8][9][10] Finding ways to support people with LTCsand to promote behaviour changeis a key challenge facing health-care services and governments globally. [11][12][13] Primary care is considered the optimum health-care setting to manage people with LTCs, 12,14 and LTC management accounts for a large proportion of the workload. 15 Primary care is community-based, accessible and able to offer continuity of care and of patientpractitioner relationshipskey requirements for managing patient needs over time. 16,17 Several initiatives in the United Kingdom, such as the NHS and Social Care Long-term Conditions Model 1 and the Quality and Outcomes Framework (QOF), 18 aim to improve quality of care and health outcomes for patients through system changes in primary care. As part of these system changes, there is a focus on patient self-management, supported by initiatives such as the Expert Patient pro-gramme. 19 The chronic care model suggests that 70-80% of people can self-manage their conditions with support from health-care services. 20 This support includes information provision, lifestyle and dietary advice, and behaviour change interventions around issues such as smoking and alcohol use.
Despite evidence of, and a developing consensus regarding, effective behaviour change techniques for the use in primary care consultations 21 and interventions, 22 behaviour change work may be absent in practice. 23 Barriers to behaviour change work in primary care have been identified. Practitioners cite lack of confidence and of knowledge about effective techniques as reasons for not engaging in this work. [23][24][25] Also, practitioners mention a lack of time within consultations or describe patient characteristics such as lack of motivation, knowledge or ability as reasons for not including behaviour change work. 23,[25][26][27] Blakeman et al. 28 found that practitioners prioritized the patient-practitioner relationship over engaging in self-management talk in consultations.
The aim of the current research was to investigate broadly the role of primary care consultations in the care of people with LTCs and their influence on patients' health-care use over time. Using audio-recordings of consultations, and interviews with patients and practitioners, guided by stimulated recall, a multiperspective approach offered a way to gain insight into primary care practice alongside the perceptions of patients and practitioners. 29 In this study, we added a longitudinal dimension with patient interviews at two time-points, to investigate whether patients perceived the consultation as influencing health choices over time. This paper specifically examines one aspect of primary care consultations: practices of behaviour change. The aim of this paper was to investigate how practitioners and patients engaged in behaviour change in primary care consultations and how instances of behaviour change work were perceived by patients and practitioners.

Methods
This study was part of an NIHR Programme Grant, CHOICE, 30 and took place in North-West England. Ethics approval was received from North West 8 Research Ethics Committee -GM East, 10/H1013/74.

Design
This was a longitudinal qualitative study of primary care consultations for people with LTCs. A multiperspective approach was taken, combining consultation recordings and interview data from patients and practitioners. 31

Data collection
Primary care practices were invited to take part in the study by letter, email and/or phone. Using publicly available QOF data, practices with high prevalence rates in at least one of the four LTCs compared with the median for England were identified and approached, with the aim of achieving a sample with a diverse socioeconomic and geographic spread within the region of recruitment. Practice recruitment ceased when a sufficiently large and diverse sample of practices had agreed to participate. Researchers attended practices on agreed dates. We sampled for two types of consultations: disease reviews where the patient was invited to attend by the practice, and patient-initiated appointments booked on or before the day of attendance. Relevant consultations were those that involved patients with one or more of four conditions: CHD, asthma, diabetes and COPD.
Consultations were identified in two ways: either the consulting practitioner (general practitioner or practice nurse) was holding a chronic disease clinic for one or more of the included LTCs when the researcher attended and therefore all expected patients were eligible to participate; or the practitioner reviewed a general appointment list and selected all patients attending with one or more of the included conditions. On arrival at general practices, patients met reception staff who provided participant information sheets. A researcher then approached every patient to ask about participation.
We adopted a purposive sampling approach, aiming for maximum variation across conditions, age and gender of patients, and type of health-care practitioner.
Patients and practitioners were interviewed separately after the recorded consultations, using a semi-structured interview guide (Tables 1 and  2) and stimulated recall; 32 short snippets of the consultation were played back during the interview to prompt thoughts and reflections. Snippets were identified using a consultation topic guide (see Table 3), and two of the researchers discussed which parts to focus on during stimulated recall, based on the on-going analytical work. Practitioners were interviewed once about all consultations they recorded; patients were interviewed following their consultation and then invited to be followed up for a period of 3 months to gain insight into patients' choices around health-care use over time. Follow-up with patients involved regular telephone contact and the completion of health-care use diaries. Patients were then asked for a final interview at the end of 3 months. We adopted a two-tiered approach to consent: patients gave initial written consent on the day to audio-recording of their consultation and then were given up to 7 days to consider further participation and written consent to retention of the recordings. Where patients gave informed consent, their data were retained and analysed; however, full cases were prioritized in the multiperspective analysis presented below. Two patients' data were excluded from   the following analysis as their consultations dealt solely with an acute issue and included no discussion of LTCs.

Analysis
All audio-recordings (consultations and interviews) were transcribed verbatim and anonymized. Data were analysed using an integrative framework approach, adapting Ritchie and Spencer's approach 33 for the incorporation of multiple perspectives on the same event (consultation). Specifically, the research team analysed transcripts inductively, comparing data within case and between case. 34 The framework for the analysis was initially based on our analysis of the consultations, guided by predetermined topics of interest (Table 3). These topics were then tracked through the analysis of the first patient interview and practitioner interview for each case and summarized within case. These summaries shaped the telephone follow-up calls and the follow-up interview guide. Themes arising during analysis and data collection were discussed within the team and incorporated into further recruitment and topic guides and into the framework for analysis.
Cases were re-analysed as the framework developed over time. The research team was multidisciplinary, incorporating expertise in primary care, psychology and social anthropology. Data collection ended when the team were satisfied that analytical saturation was reached, with no new themes developing from analysis. For the analysis, behaviour change topics included the following: commonly targeted health behaviours and specific self-management behaviours known to affect LTCs. These were regarded as 'behaviour change topics' only if brought up in the context of reviewing patient behaviour related to one or more of the relevant LTCs. Mood management did not arise in this context; however, it arose in one consultation in relation to bereavement. 29

Sample characteristics
Thirty-nine practices were approached and six participated in the study. Reasons given for non-participation were as follows: involvement in other research and being too busy to participate, particularly because of QOFrelated activity. Across 6 primary care practices, 10 practitioners took part, including 5 general practitioners (GPs) and 5 practice nurses (PNs). Of 65 patients approached in practice waiting rooms, 34 were recruited into the study and agreed to audio-recording of their consultations. Of those who declined participation, many did not give a reason, but some said that they were too ill or too busy to participate or that their consultation was not about their LTC. Patients were mainly White British (82.4%), 65% male, with ages ranging between 34 and 87. Most patients had more than one LTC (73.5%; 29.4% had at least two of COPD, CHD, asthma and diabetes) (see Table 4 for participant characteristics).  Twenty-nine consultations were audiorecorded by practitioners, with the number recorded per practitioner varying between 1 and 8 (mean = 3). On 5 occasions, consultations were not recorded due to technical errors; in these instances, patients were still invited to take part in an interview about the consultation.
There were 18 patients from whom consultation data, patient first interview, patient second interview and practitioner interview were all availablethese constituted full cases. Of the remaining 16 patients, 4 declined participation in the follow-up interview; 7 declined interviews but gave consent to retain their consultation; and 5 took part in either one or two interviews, but lacked a recorded consultation. Retention during follow-up was high (81.5%), with 22 patients completing two interviews. Figure 1 details patients' recruitment and retention.
In total, 84 transcripts were included in the analysis, generated from 32 patients and 10 practitioners, across six primary care practices. Table 5 summarizes the behaviour change talk present in 27 consultations. Smoking, medication use and diet were the three most common behaviour change topics brought up in consultations as targets for behaviour change. Specific self-management programmes or strategies, as means to address behaviour change, were only mentioned in four consultations; these included pulmonary or coronary rehabilitation programmes (n = 2) and rescue packs of antibiotics and steroids for COPD (n = 2).

Findings
Two interlinked themes arose from the multiperspective analysis around behaviour change. The first theme, 'Engaging in behaviour change talk', drawing on all three sets of data, was that behaviour change talk was characterized by deflection and diffidence on the part of practitioners. Patient cues about motivation were often not followed up, and patients tended to remain unmotivated, uncertain or overwhelmed by the idea of behaviour change. The second theme, 'Perceptions of behaviour change and the role of the practitioner', drawing on the practitioners' accounts, was that practitioners often positioned behaviour change work as beyond their scope or ability to achieve, and potentially damaging to other aspects of their work, notably the patient-practitioner relationship. Despite offering vocal support for behaviour change and self-management, practitioners frequently minimized their ability and responsibility to focus on this work in practice.

Engaging in behaviour change talk
There were four primary behaviour change techniques employed by practitioners in consultations: advising change without discussion, giving or gathering information, tentative suggestions for patient to change and deferring an issue for discussion at a later point or with another practitioner. The use of these techniques fitted into one of two broad practitioner styles: diffident and deflecting. The diffident style was characterized by advising or suggesting change without discussion; sometimes, information-gathering occurred before a withdrawal from discussion. The deflecting style was characterized by information-gathering, followed by deferral of discussion. Both styles occasionally involved information-giving, but this was characteristically one-sided; that is, the practitioner gave information, but did not encourage discussion about this information. In only two instances, practitioners engaged with a motivational approach to behaviour change talk, where there was exploration of the patient's motivation, barriers and facilitators for change.

Diffident style
This style was reflected in practitioners advising or suggesting change, and then  withdrawing from further discussion. On some occasions, the practitioner enquired into the detail of a behaviour, before shutting down this line of enquiry to move on to another, often biomedical matter. This withdrawal tended to occur at the first sign of reluctance or resistance from the patient. For instance, in the following extract with a female patient with asthma, the nurse asked the patient about stopping smoking: In this segment, while the nurse started by asking about motivation ('any thoughts?'), she closed down the discussion at the first sign of hesitancy. The patient offered one of her triggers for continuing smoking ('something [annoys] you'), which is met by the nurse disengaging and deferring further discussion to a later time ('whenever you feel that you're at that stage'). In interview, the PN indicated that she was trying to encourage the patient 'just to think about quitting', although it was evident that she had little confidence in her ability to do more: PN1 A lot of people, don't want to change. Yes don't want to change. A lot of people will say that whatever normal, well it's the only  Number of responses can be greater than number of consultations as, in some consultations, more than one strategy was employed, for example information gathered, then deferral to another practitioner. † In these instances, patients either indicated that they did not engage in this behaviour, or the practitioner indicated that level of behaviour was satisfactory and did not need addressing. ‡ Medication use is defined as talk around how and when to use currently prescribed medications, rather than talk around new prescriptions. The effect of this style of communication is that the patients were left with a sense of personal responsibility for change, while uncertain about how to change and unclear about what help was available. When asked about how review consultations could help with her health, P3 predicated any hope for better health on her giving up smoking: 'while you're smoking, [the asthma]'s never going to be any better is it? So maybe when you've given up smoking it might get possibly better'. At the same time, she saw herself as the main barrier to change: 'The only thing [PN1] can't give me is the willpower [to stop smoking] (. . .) you're only going to do it when you think you're ready, or feel ready to do it'. Throughout the follow-up period, the patient continued to smoke but wanted to stop for financial reasons. At follow-up interview, she had decided to start Champix (a smoking cessation tablet available on prescription), as another nurse had informed her of its 'good success rate'. In this case, the practice nurse's initial pessimism about the patient's desire to change and her own ability to enable change meant that the patient's motivations and options were unexplored until she saw another practitioner.

Deflecting style
This style sometimes overlapped with a diffident approach in that tentative language and negative phrasing were used as well as active deflection. Mostly, in this style of communication, practitioners gathered information about a patient's behaviour and then either deferred discussion to a later point or suggested that another practitioner was better equipped to deal with the issue. For instance, one nurse considered introducing a new self-management strategy (rescue packs for COPD) following a discussion about recent exacerbations (in consultation with P4, a male patient with COPD: 'I'm just wondering whether in the future we give you a small supply of antibiotics, then you can start on your own, but you'd still need to come and see us afterwards', PN2). When the patient asked for more information, the nurse decided unilaterally to defer the discussion to the next review: 'we'll go through that next time I think, I think we've done too much today', PN2. This deflection seemed to serve the purpose of time management within the consultation, with the nurse deciding to prioritize other aspects of care when it appeared that the issue would require more discussion. It also contained an implicit judgement that the nurse should decide when a patient is ready to be introduced to new self-management techniques.
Another form of deflection involved information-gathering followed by implicit deferral to another practitioner. One GP used this technique twice: in the first instance, with a patient with diabetes who was struggling with his weight; in the second instance, with a patient who had recently started smoking again. On both occasions, despite cues from the patients that they were struggling with behaviour change (in consultation, P21 on diet, a male patient with diabetes: 'I'd like to be better', 'I eat too much', 'I've got to try and manage it') or despondent at the thought of change The GP described his tactic as 'just trying to gauge his level of motivation to want to stop which I don't think was particularly high and he wasn't taking the bait really', GP3. At the same time, the GP also indicated in his interview that 'if they don't want to do it, no matter what a doctor says, they're not going to do it', GP3, suggesting that he had little faith in his ability to influence patient motivation to change.
The deflecting style was characterized by the practitioner controlling the direction of the consultation and determining what topics could be discussed at that time or with them as practitioners. While the nurse and GP offered different reasons for changing direction (the nurse felt too much information had been covered; the GP felt another practitioner was better suited for the discussion), the overall effect is similar to that of the diffident style: the patient is left with a sense of personal responsibility for change, and the difficulties the patient faces with behaviour change are unexplored: Practitioners defined their role as supportive but passive and saw their responsibility as fulfilled by their offer of future support for patients.

Practitioners are limited in what they can do
When it came to discussing how practitioners can encourage or enable behaviour change, practitioners presented themselves as limited in the techniques they can use. They presented information-giving and support as the techniques available to them and suggested that these techniques might influence patient motivation to change. Practitioners did not seem confident in the efficacy of these techniques: GPs saw information-giving as an essential aspect of behaviour change work, but often raised doubts over what information alone could achieve: GP3 I could say, 'Here's an information leaflet or a little booklet', that I did have on the shelf on the room, I know exactly where they are which says, 'Diet and Diabetes', I probably didn't do it because I didn't think it'd make any difference. I just, that kind of sense of despondency in some ways that giving him an information booklet I don't think is what he probably needs to make those changes Some practitioners felt that, if they understood a patient's motivations and beliefs, they would be able to use this as a motivational lever to encourage or perhaps scare the patient into changing their behaviour.

GP3
You're gauging how interested people are and whether this, and just testing out whether this is the Achilles heel in his behaviour -this is the information [that] strikes a chord with him that will actually get him to start to think and challenge his own habits, his own behaviour and whether we can sort of run with that PN5 [I say to patients] 'Do you want to be a grandparent whose children and grandchildren come and visit them, or do you want to be a grandparent who can take the grandchildren out? You're getting to that stage [of illness]', and I kind of try and use some of those kinds of things so they can relate it to real life However, these same practitioners used tentative language when talking about their ability to influence patients ('gauging', 'testing', 'sort of run with that'; 'kind of try and use some of those kinds of things') and still talked about the need for patients to drive behaviour change, suggesting that they were not wholly confident that they could push patients towards a more motivated state. The value of a good patient-practitioner relationship outweighs the value of challenging patient behaviour The final argument that practitioners presented as a counter to their responsibility for behaviour change work centred around the patientpractitioner relationship. While practitioners agreed that behaviour change was important, they argued that the potential cost of challenging patient behaviour outweighed the benefitsit risked damaging the relationship without resulting in behaviour change:

PN5
The thing is you do not need to emphasise to a smoker this is bad for you. You don't need to go on and do this and do thatthey know that. And quite often if you do that they will shut down on you and they won't take it on board

Discussion
This study demonstrates a disjunction between how practitioners enact behaviour change in consultations and their expressed commitment to its importance for long-term condition management. In the recorded consultations, any behaviour change talk evident was brief, diffident and limited, with no evidence of patient collaboration. This finding seemed consistent across consultations involving patients with multimorbidity or single morbidity. Practitioners typically employed two styles of behaviour change talk: a diffident, disengaged style, where they tentatively suggest change but withdraw from in-depth discussion; and a deflecting style which moves behaviour change talk outside of the remit of the current consultation. Both styles effectively delegitimized behaviour change talk within the consultations, leaving patients unsure about the role of practitioners in aiding behaviour change.
In discussing behaviour change in interviews, practitioners emphasized its importance and its centrality in LTC management. They described a role for themselves in providing reminders and information to patients about the importance of behaviour change. However, they also emphasized constraints on their ability (and therefore, their responsibility) to effect behaviour change. Behaviour change work was constrained by a perceived lack of effective techniques to influence patient motivation, by the need to ensure a continued patientpractitioner relationship and, ultimately, by the patients themselves. 23,25,28,35 The relegation of behaviour change work to the responsibility of the patients or other practitioners offered a way for practitioners to disengage with this work in their consultations. In line with previous studies, this study highlights the infrequency of work around self-management and behaviour change within primary care 28 and adds further evidence of the uncertainty and reluctance on the part of practitioners to participate in this work. 23 patient and practitioner attitudes towards behaviour change work. 28,34 Both patients and practitioners in this study viewed behaviour change work as the responsibility of the patient, and this in part could explain the absence of this work in consultations. This absence is likely to have a significant effect on patient engagement with change, as support from services and their social environment are important influences on patient attitudes towards, and ability to, self-manage. Without supportive, collaborative relationships and support, it is likely that significant numbers of patients will continue to feel responsible for, yet unable to effect, behaviour change. [36][37][38] There is a growing body of research demonstrating techniques that can result in behaviour change within primary care in the context of clinical trials, which may not require much time to implement. [39][40][41][42][43] In addition, patients describe health practitioners as an important source of external motivation for self-management. 44 Yet successfully embedding behaviour change in practice, and motivating practitioners to change their consultation behaviour, is still elusive.
The practitioners in this study and in primary care in general are embedded in an increasingly busy and scrutinized context where certain tasks are valued and counted (like meeting biomedical targets) to the extent that care can become shaped around fulfilling these tasks, 45 and others (such as less measurable behaviour change work) are not. 29 It may be that practitioners feel both ill-equipped to achieve behaviour change in patients and pressurized to achieve other targets, and the rhetoric of patient control and responsibility enabled them to legitimately distance themselves from this work. 46 It is already known that training and education alone is insufficient to change practitioner behaviour. 47 Practitioners are likely to only engage in behaviour change work if the structures in primary care promote and value this work. 17,48 This would mean a commitment to behaviour change as an activity beyond what is currently evidenced, and a cultural change in how practitioners view their role, and the role of the patient, in relation to behaviour change. Adapting routine consultations to run on a collaborative model of care may aid work around behaviour change if there is a commensurate shift in attitudes to a 'joint responsibility' ethos for achieving set goals. 49 However, this shift is unlikely to occur unless collaborative working becomes the foundation through which other targets, such as QOF, can be achieved.

Strengths and limitations
This study consisted of a relatively small sample of practitioners, which reflected difficulties in engaging general practices with the study. However, the sample size was similar to other studies adopting a multiperspective approach to analysis (between 8 and 17 practitioners), 28,50,51 an approach which offers rich opportunities for small samples. 31 The main reason practitioners gave for not engaging was that they had too many other demands on their time, notably QOF requirements; however, this was also a concern of practitioners who did participate. It is possible that practitioners who participated were more interested than others in psychological approaches and that non-participating practitioners might give even less priority to behaviour change than those we studied. Nearly half of the patients declined participation when approached, and it is possible that these were more unwell than those who took part. Nevertheless, our sample included a broad range of LTCs and high levels of multimorbidity. In a study of this kind, where patients and practitioners are aware of being recorded, there is always the possibility that behaviours may have been changed as a result. To minimize this risk, the researchers were not present in the consultations.
Behaviour change was not initially an explicit focus of the study from which the present analysis was drawn, which aimed to explore chronic disease management within routine consultations. However, because data collection and analysis proceeded in parallel, we were able to incorporate discussion of behaviour change into respondent interviews early in the study when it emerged as a significant issue for practitioners and patients. Using stimulated recall in interviews enabled practitioners to reflect on their practice and facilitated recollection of particular patients and decisions over time. It also enabled discussion of behaviour change and its importance in relation to specific instances rather than generalized accounts.
A key strength of this study was that it combined analysis of practitioners' and patients' viewpoints with analysis of their consultations. This multiperspective approach offered a rich and nuanced way to explore practitioners' attitudes and experiences of behaviour change alongside enactments of behaviour change in practice.

Conclusion
Practitioners' accounts endorsed the importance of behaviour change for LTCs; however, within primary care consultations, their approach to behaviour change tended to be diffident and limited, with little direct challenge to, or exploration of, patient motivations and beliefs. Practitioners employed three minimizing strategies within their interview accounts, which had the effect of absolving practitioners of responsibility for, and reinforcing current practice around, behaviour change. These strategies positioned patients as responsible for any failure to pursue or achieve behaviour change. In addition, these strategies reinforced the practitioners' control of the content or agenda of consultations in which behaviour change could be dealt with. This in effect left patients feeling personally responsible yet unconfident and unclear about the role of primary care in assisting behaviour change.
It is likely that change in how practitioners engage with patients will only be achieved through an extensive cultural and attitudinal shift within primary care. Collaborative care planning may be one model that could enable behaviour change work within primary care, but to be adopted successfully, this model would need to be accompanied by a change in the attitudes and culture of primary care.