Main findings and comparisons with previous research
Our findings suggest that practitioners and patients perceived that engagement in self-management among patients with multimorbidity is contingent on the three inter-linked factors of capacity, responsibility and motivation. The availability and expression of these factors was moderated by the presence of high or low deprivation and/or the presence of mental and physical multimorbidity.
Structural capacity explained how patients’ ability to engage in self-management practices was partly based on their access to social, economic, and material resources. Physical and emotional capacity were also key ingredients in enabling patients to carry out multiple tasks and seize opportunities to use social resources to support pathways to self-management. Economic hardship reduced structural and emotional capacity, while interpretive capacity, in the form of know-how, increased opportunities for some patients to adopt self-management behaviours that had synergistic effects on multiple problems. As such, capacity was a necessary pre-cursor to patient engagement in self-management, but it was not sufficient.
Similarly, Shippee et al. proposed a cumulative complexity model to describe how clinical and social factors accumulate over time and interact to complicate patient care, including self-care, among patients with multimorbidity [
27]. In this cumulative complexity model, the ability of patients to self-manage hinges on a balance between patient `workload’ of demands and patient capacity to address increasingly complicated demands. As with our findings, patient capacity here refers to the availability of financial and social resources as well as health literacy, but also includes level of physical functioning. Workload is used to denote the extent to which patients expend time and energy on health care and self-management, and on coping with life in general. In this sense Shippee et al usefully distinguish between resources that make up capacity and the use of resources that might drain capacity, further illuminating the notion that patient capacity has dynamic properties and can vary between and within patients depending on individual characteristics and circumstances. Our findings add to this model by building in an understanding that the level of certain forms of capacity is partly dependent on patient motivation. However, neither our model or that of Shippee et al has been subject to formal empirical testing yet.
Practitioners offered qualified support for the notion that all patients with multimorbidity, regardless of level of symptom or illness burden, should share responsibility to self-manage, but only if patients had the interpretative capacity to weigh up the risks and benefits of self-management practices. There is strong evidence that lower levels of health literacy are associated with poorer health status, but the notion that health literacy influences self-management through motivational and volitional determinants is contested [
28]. Our practitioner data might support the argument that health literacy precedes responsibility for adopting self-management practices, but this relationship would need to be confirmed in a mediation-moderation model [
29].
Patients tended to agree with practitioners’ narratives that they shared some responsibility for self-managing their illnesses. Patients’ narratives about responsibility to self-mange did however deviate to some extent along deprivation lines. Patients living in the most deprived areas tended to hold their doctor as responsible for managing and monitoring their health, whereas patients living in the least deprived areas favoured the pursuit of more individually oriented approaches to living with and managing their illnesses. This is not to assume however that patients who had greater aspirations to self-manage necessarily elevated individual over social responsibility. We have previously shown in a separate analysis that attitudes to self-management among this sample of patients did not vary between frequent or non-frequent attenders in primary care, or that self-management was more likely to be adopted by particular types of patients with specific combinations of illnesses [
30]. This is in keeping with the previous findings of Townsend et al who suggested that use of healthcare among people with multimorbidity is underscored by individual characteristics of patients and their social conditions rather than specific forms of morbidity [
31]. Hurd and Clarke make the compelling argument that patients’ engagement in self-management is in part driven by a desire to regain control over their bodies [
32], which appeals to the notion that chronic illness, and indeed multimorbidity, present opportunities for patients to invoke a moral discourse and reinstate agency and self-determination through self-management [
33].
Sinnott et al found that opportunities to engage patients with multimorbidity in patient centred approaches founded on shared decision making were limited in patients who lacked motivation [
34]. Likewise we found that even where patients had capacity and responsibility, full engagement in self-management was unlikely to be successful in the absence of self-motivation. Expectations and health beliefs figured prominently in the way patients articulated motivation to engage in the full array of self-management practices, especially supported forms of self-management. Attendance at support groups was low among patients with multimorbidity because most patients did not expect that such support groups could positively transform their health or their ability to self-care. Socio-economic deprivation also shaped patients’ motivation. Patients from deprived areas were often socialised to perceive that their poor health was the norm, which limited their ambitions to improve health by self-management. By contrast patients from less deprived areas had greater motivation to self-manage because they had more opportunities to model the healthy lifestyle behaviours of peers and family. Finally, both patients and practitioners agreed that motivation was prone to being negatively affected by the presence of depression, which in turn could reduce responsibility and emotional capacity to self-manage.
The finding that
motivation to engage in self-management may be reduced in the presence of socioeconomic deprivation is supported by a US study by Clark et al which found that patients with multimorbidity from deprived areas had lower expectations for health and successful ageing than patients from more affluent areas [
35]. The concept of `successful ageing’ is premised on the idea that health promotion and disease prevention, principally through the adoption of health protective behaviours, can delay the onset of the harmful effects of ageing [
36]. As with the patients in our study, Clark et al found that socio-economically vulnerable patients tended to focus on keeping within self-imposed limits, avoided “doing too much” and aimed to achieve basic goals such as rest and alleviation of pain rather than investing time and energy in more complex lifestyle changes. By contrast, patients with private health insurance shared “life goals” such as seeing their grandchildren grow up, working, socialising and enjoying their hobbies and saw health promotion as key to successful ageing. This difference in perspective can partly be explained by what is known as a future time perspective [
37] which is a learned cognitive style of information processing that favours a focus on the future and is partly conditioned by socio-economic context [
38]. In our sample, patients from deprived areas were socialised to see ill health as an expected outcome and focused on day-to-day existence rather than changing their behaviour to improve their health in the future. Additionally we also found that patients were less likely to engage in supported forms of self-management owing to a lack of expectation that such groups would benefit their health. This explanation maps onto the health belief model which suggests that individuals will act in a specific way if they believe or expect that they will benefit from those actions [
39],[
40]. Or, alternatively, it might be for some patients multimorbidity is not perceived as a significant additional burden and as such they do not feel the need to seek out further support [
31],[
41].
The fact that both our study and Clark et al found that for more deprived patients self-management revolved around taking medicines suggests that medication management is, for some patients, a central focus around which they can be responsible for their illnesses [
42]. This highlights that practices broadly defined as `self-management’ (such as medication taking and lifestyle) can be construed by patients in very different ways.
Strengths and limitations
A particular strength of this study was the inclusion of both patient and practitioner views as previous qualitative studies have tended to focus only on one group [
43],[
44]. This allowed triangulation of emergent themes and the use of deviant cases to illuminate difference and similarity between patient and practitioner accounts. However, although qualitative work does not seek to make probabilistic generalisations it does strive to make logical generalisations to support the development of explanatory theory about how other comparable populations experience similar classes of phenomena [
45]. Our efforts to achieve this level of generalizability may be weakened by the fact that the practitioners in our study were mainly GPs who were partly recruited using a convenience sample, potentially leading to selection bias.
We anticipated that socio-economic deprivation would be an important factor in determining patients- and practitioner attitudes about self-management. We therefore used the IMD to measure the level of deprivation for each area where patients lived and where practitioners worked. While the IMD score can be used to identify the most and least deprived areas in England and to compare the levels of deprivation between areas it cannot identify individuals who are deprived and individuals who are not. As with other area level measures the IMD is thus possibly prone to ecological bias and it is recommended that both individual and area-based measures of socio-economic status are used to capture important differences in social and economic status [
46]. When researching patients with multimorbidity collecting individual level data on home ownership may be particularly relevant as these patients are typically older and/or retired adults for whom the importance of other indicators of socio-economic status is much reduced [
47].
Additionally, it is known that the burden of multimorbidity varies by ethnicity, with South Asian patients more likely to experience cardiovascular multimorbidity [
48]. While self-management of risk factors for cardiovascular multimorbidity improves with increased multimorbidity there are important differences in outcome by ethnicity which remained to be explained [
49]. Less is known about ethnic differences in self-management of other combinations of multimorbidity. We were however unable to assess the relevance of our findings among an ethnically diverse population. It will be important to explore whether the three factors identified in this study as important drivers of self-management practices in multimorbidity generalise to black and ethnic minority groups.
Implications for future research and clinical practice
Despite recent interest in multimorbidity there is still a sparse evidence base on effective interventions [
50]. In the UK and US the chronic disease self-management programme (CDSMP) is the dominant model that underpins educational and organisational interventions to enhance patient self-management in patients with single long term conditions. In this model the capacity and indeed moral responsibility for self-management is believed to rest with the individual patient, leading to an emphasis on psychological interventions to change behaviour by increasing self-efficacy (i.e. confidence to produce a desired outcome). As well as divorcing responsibility for health from health service providers, interventions that focus on individual capacity and responsibility for self-care have been criticised for not taking into account the complexity and plurality of experience of illness and how these experiences shape self-management practices [
51]. There is preliminary evidence that the effects of the CDSMP may be at least as good in those with multimorbidity (if not better) [
18], but levels of engagement remain poor, suggesting that a more socially informed model might increase engagement and demonstrate these benefits more widely.
Taking a lead from whole person [
16] and network [
52] perspectives, this may require a co-ordinated approach not just from primary care, but also from wider networks of local resources that can encourage patients with multimorbidity to engage in sustainable healthy behaviour change [
53]. For example, signposting to existing community groups which help to reduce social isolation - optimising
capacity; and by funding local peer-led healthy lifestyle initiatives - thus optimising
responsibility through positive role modelling. Previous research has shown that general practitioners are reluctant to discuss self-management with patients with one long term condition for fear of alienating patients and disrupting the consultation [
54]; this also equally applies to consultations with patients with multimorbidity [
30]. However, training of primary care practitioners in theoretically informed brief behaviour change techniques [
55], such as motivational interviewing, might enable them to optimise patient
motivation and enhance engagement in self-management.
Key to the success of such initiatives might be to engage patients with multimorbidity in the co-design of interventions. Experience based co-design is an evidence based approach to involving service users in the development phase of intervention design to enhance their experience of treatment and care [
56]. EBCD focuses on drawing out issues that are often hidden from view but are critical to patient experience. Understanding how capacity, responsibility, and motivation combine to support engagement in self-management may offer a way forward to crafting interventions that can benefit both practitioners and patients. This may be especially true in relation to understanding why some patients had greater interpretative capacity to identify opportunities to extract maximum benefit from self-management practices for all their conditions. Here, harnessing patient `know-how’ about optimal methods of self-management in the face of multimorbidity would potentially benefit other patients with less interpretative capacity.
Managing mental and physical multimorbidity is particularly challenging [
57] and the findings from our study alluded to the idea that prioritising the treatment of depression in patients with multimorbidity might increase their motivation to self-manage. However, in the context of primary care, it is difficult for practitioners to routinely prioritise the management depression in people with long term conditions [
21]. Effective management of mental and physical multimorbidity might therefore only be possible in the context of system level interventions such as collaborative care [
58], but the effectiveness of these approaches have not been fully tested in multimorbid populations. Alternatively, there is growing interest in mindfulness interventions that offer patients opportunities to engage in synergistic approaches that address both mental and physical ill health [
59], and further research about the effectiveness of these approaches in multimorbidity is warranted