Managing diabetes in people with dementia: a realist review

Interview procedures

Interviews were conducted either face to face or via telephone by one of three researchers (FB, PRJ and BR). Participants were given a copy of the study information sheet, which provided contact details of the research team, and a consent form, which they were asked to read and sign. Interviews were conducted using an interview schedule and were audio-recorded and transcribed. Ethics approval was obtained from the University of Hertfordshire Health and Human Sciences Ethics Committee with delegated authority (CSK/SF/UH/00106). The interview schedules were designed to explore:

• participants’ experiences of (a) working with people living with dementia and diabetes, (b) living with dementia and diabetes or (c) acting as an informal carer to someone with dementia and diabetes

• current problems and challenges facing people and families who have to manage dementia and diabetes, and what needs to be in place to address the effects of dementia

• what good diabetes care looks like for people living with dementia and what is needed to achieve it

• changes required to improve the management of diabetes in people living with dementia.

Programme theory

If health and social care delivery systems propagate and reinforce positive attitudes towards people living with dementia and diabetes and their families (C), through tailored SM support, then this fosters a belief in staff that people living with dementia and diabetes have the potential to be involved in SM and the right to access diabetes-related services (even when the trajectory is one of deterioration), (M) prompting treatment confidence in people living with dementia and diabetes (M), which leads to engagement in SM practices by people living with dementia and diabetes, their family carers and health-care professionals (HCPs) (O).

Stigma and barriers to care for people living with dementia

There is evidence that social stigma surrounds both dementia and diabetes not only in the general population but also among HCPs. 144 Stigma associated with dementia has an impact on the person living with dementia and may lead to them having poorer access to care than people with similar comorbidities but without dementia. 39 Stigma is also felt by the family carer, for example through increased social isolation. People with diabetes may also be stigmatised and blamed for their condition because of lifestyle choices and/or obesity. 144 Recent work by members of the team around community engagement and dementia-friendly health care has argued that without a foundation of awareness about what it is like to live with dementia, related initiatives will not succeed. A reliance, for example, on single initiatives such as dementia champions is insufficient. 145,146

People with dementia and comorbidity face many problems in accessing health care. This includes a lack of services designed around the needs of people living with dementia, poor communication between services, a lack of training on dementia care for health and social care staff and a reliance on others (such as family carers) to recognise a need for services and seek help. 39 A recent mixed-methods study, which included qualitative interviews and focus groups with clinicians, highlighted deficiencies in access to care and continuity of care for people with dementia and comorbidity. Access to care was affected by clinicians’ previous experiences and their attitudes towards risk. For example, there were contrasting opinions about the appropriateness of taking someone with dementia off insulin. 39,142 Stakeholders also highlighted the need to ensure that people living with dementia received equitable care:

. . . you shouldn’t be sort of swayed one way or the other, just because someone has dementia . . . I think certainly when they first start on their journey I think it’s really important that we do everything we can [of cross-disciplinary training to facilitate appropriate care].

Diab 1 (stakeholder interviews)

Involving people with dementia in self-management

We found 10 studies37,59,62,106,109,110,115,121,132,136 looking at SM interventions for people living with dementia or cognitive impairment, only one of which included people with dementia and diabetes. 59 This controlled study evaluated the impact of personalised education sessions on people with diabetes and cognitive impairment. 59 In this US-based study, certified diabetes educators delivered personalised education sessions over the internet to older adults with type 2 diabetes and cognitive impairment (mild cognitive impairment or early-stage dementia). They found a significant increase in self-efficacy but no difference in glycated haemoglobin (HbA_1c) at 6-month follow-up. The authors say that the face-to-face nature of the contact (via Skype) was beneficial for establishing good rapport with participants. Mountain37 suggests that the lack of studies on SM for people living with dementia is a result of negative perceptions of the abilities of people with early dementia, which ‘leaves them on the periphery of any benefits that can be derived from the current focus of policies that support people with long term conditions’. 37

Studies on SM for people with dementia are mostly qualitative or small pilot or feasibility studies, and most do not report measurable health outcomes. They also tend to focus on couples or people who have a partner, which excludes many who might potentially benefit. 147 This literature does, however, provide valuable insights into what needs to be in place to support independence and appropriate outcomes in people living with dementia, much of which might form the basis for programmes for people with dementia and diabetes. Potentially important components of such programmes include the need to focus on the abilities and strengths of people with dementia rather than their disabilities,109,110 on emotions rather than problems,109 on building confidence62,121 and on respecting participant autonomy and enhancing empowerment. 106 Social interaction may also be important. 106,110,143 A small feasibility study121 suggested that SM interventions may foster independence and reciprocity, and promote social and clinical support. A larger RCT involving 136 people living with dementia–spouse dyads found that a group-based intervention to promote SM in people newly diagnosed with dementia led to improvements in health-related quality of life in spouses and in cognition in people living with dementia, although quality-of-life improvements in spouses were not maintained at longer-term (9-month) follow-up. 106 Further details of these studies can be seen in Appendix 7 and an overview of some key potential aspects of context, mechanism and outcomes can be seen in Figure 6.

FIGURE 6.

Summary of evidence around SM programmes for people living with dementia. CI, cognitive impairment; MCI, mild cognitive impairment.

Fostering confidence

The importance of confidence as a mechanism is underlined in a number of studies. For example, a systematic review130 of self-care interventions for older adults with diabetes (not dementia) found that interventions using concepts of self-efficacy, self-determination and proactive coping were effective in influencing diabetes self-care behaviours, leading to improved health outcomes. For example, proactive coping helped people to anticipate threats and act accordingly. In addition, work by Clare et al. 86 and Claire148 suggests that there is a link between independence, functional ability and self-care behaviour and feelings of confidence or self-efficacy in people living with dementia and their family carers. Clare et al. 86 argue that a loss of confidence can promote disability:

Negative influences can contribute to the development and maintenance of ‘excess’ disability – where the extent of functional disablement is greater than would be predicted by the degree of impairment: an example would be where an individual loses confidence.

Clare et al. 86

Therefore, interventions that promote confidence or self-efficacy may lead to increased engagement in SM activities. This ties in with literature on enablement, in which the focus is on what the person with dementia can do both cognitively and functionally to maintain their quality of life. 148 The process of enablement has been sponsored by the Department of Health in England in their 2010 guidance Nothing Ventured, Nothing Gained,149 which made it clear that HCPs need to be proportionate in their assessment of risk with people living with dementia and that a ‘safety first’ management perspective disempowers people living with dementia. Although we found no literature on enablement for people living with dementia and diabetes, there is some evidence that teaching goal-directed activities can improve the cognitive functioning of people living with dementia. 85,86 It has been suggested that the enablement philosophy can also be applied to identifying and treating accompanying physical illness and providing information and support to caregivers,148 although this has not been tested in people living with dementia and diabetes.

We included several studies that focused on the functional abilities of older people with multimorbidity. A case study138 of co-ordinated care for people with complex chronic conditions supported the idea that care co-ordination programmes should focus on supporting service users and carers to become more functional, independent and resilient. The authors suggested that this is preferable to a purely clinical focus on managing or treating medical symptoms. Two studies58,65 involved occupational therapists (OTs) delivering interventions aimed at improving the functional abilities of older people with multimorbidity. One, which found improvements in activities of daily living, suggested that an important aspect of the intervention was that OTs used motivational interviewing and goal-setting to understand patient priorities and to negotiate solutions in partnership. 58 In the other, a RCT of community OT for people living with dementia and their carers, the authors found that 10 sessions of OT led to improvements in mood, quality of life and health status at the 12-week follow-up. 65,95,96 The intervention involved the use of compensatory strategies to adapt activities of daily living to the disabilities of patients and adapt environments to their cognitive disabilities. Primary caregivers were trained to use effective supervision, problem-solving and coping strategies to sustain both their own and the patient’s autonomy and social participation. A potential mechanism was an increased sense of control.

Although these studies do not focus on SM for dementia and diabetes, and although they focused on people living with mild or early-stage dementia, they may provide transferable learning for the development of SM support for people living with dementia and diabetes, for example the focus on family relationships, maintenance of an active lifestyle, psychological well-being and techniques for coping with memory changes. Furthermore, they promote positive attitudes, with an emphasis on equality and collaboration between participants and facilitators. 106 Working with families is clearly key. However, studies highlight the need to ensure that the voice of the person living with dementia is heard and that their needs are balanced with those of the carer. 115 Qualitative studies on SM support for people living with dementia found that information provision may be aimed at carers, leaving people living with dementia feeling powerless,115 that carers may inadvertently take control away from people living with dementia109 and that people living with dementia can find support inappropriate or stifling. 132 As one stakeholder put it:

. . . an intervention should work at a level that people . . . particularly early stages of dementia . . . can be included . . . so it’s not decisions being made about them . . .

Dem 1

Further RCTs of SM support for people with diabetes and cognitive impairment are under way but these focus on intellectual disability150 and learning disability151 rather than dementia. Although the results are not available, the protocols do provide information about components of the intervention. For example, the intervention being evaluated by Walwyn et al. 151 involves establishing participants’ daily routines and lifestyles, identifying all supporters and helpers and their roles, setting realistic goals for change and monitoring progress against agreed-on goals, while the Taggart et al. 150 study involves evaluating an adapted version of DESMOND (The Diabetes and Self-Management for Ongoing and Newly Diagnosed patients with T2D programme). 152 The adapted version includes longer sessions, increased use of pictures in the educational information, greater involvement of carers, more repetition and interactive sessions, and a strong focus on celebration and fun,153 components that would appear appropriate for people living with dementia.

Context–mechanism–outcome 1 summary

Interventions for people with dementia and diabetes need to address the stigma that this group faces. There is evidence that people living with dementia have the potential to be involved in SM, although most research currently focuses on SM skills in people with mild or early-stage dementia, and there is little evidence relating to SM in people with dementia and a comorbid condition such as diabetes. However, the literature identified important components of interventions that are likely to be transferable to people with dementia and diabetes; for example, focusing on strengths and abilities, being emotion focused rather than problem focused, respecting autonomy and working to build confidence and empowerment. The involvement of family carers in programmes is key, but it is important to balance the needs of the people living with dementia with those of the carer to ensure that people living with dementia are not disempowered.

Programme theory

If delivery systems promote a person-centred and partnership approach to care, allowing HCPs to understand the individual needs and abilities of people living with dementia and diabetes and their families (C), then (1) HCPs feel confident that they are acting in the best interests of people living with dementia and diabetes and their families (M), and (2) this generates trust between HCPs and people living with dementia and diabetes and their families (M), leading to a better fit between care planning and patient and carer needs, and (potentially) a lessening of the burden of medicalisation experienced by people living with dementia and diabetes and their families (O).

Identifying patient and carer priorities

People living with dementia and diabetes have two chronic conditions with different trajectories. Dementia will generally ‘have a progressive or stepwise pattern of progression in illness severity, symptoms and disability over time, which will require continual adaptation to new issues or limitations’,154 whereas diabetes may have a more constant course with longer periods in which to adapt, but the trajectory of each is likely to have an impact on the other. Supporting people living with dementia and diabetes, and their families, with these ambiguous trajectories along the route from early-stage dementia to end-of-life care is a difficult clinical enterprise. Delivering appropriate and sustainable care for people living with dementia and diabetes requires a change from a curative, biomedical strategy to a more person-centred approach whereby patient priorities are at the forefront. 155 The Institute of Medicine defines patient-centred care as being ‘respectful of and responsive to individual patient preferences, needs, and values, and ensuring that patient values guide all clinical decisions’, and is a hallmark of high-quality care. 156

Maintaining independence and engagement with day-to-day activities was a clear priority for participants in all groups (e.g. people with dementia and diabetes, older people with diabetes, people with dementia). 59,70,100,113,114,132 For example, a qualitative study100 of older patients with type 2 diabetes found that their primary goal in diabetes SM was engagement with their day-to-day activities, rather than maintaining biomedical parameters. The authors concluded that HCPs have to prioritise each patient’s life experience when providing advice about diabetes management. This was supported by our stakeholder interviews, as seen in the following quotation from a general practitioner (GP):

But actually at this stage [referring to when people have complex health needs] people are interested in autonomy, mobility you know, retaining as much function and independence as they can, being a burden on their families you know, so all the normal things and they’re often much, much more important than a lot of the medical stuff.

Diab 12

Person-centred approach to self-management in people living with dementia and diabetes

A recognition of patient motivators and goals, and negotiation of a mutually agreed management plan, could improve adherence to SM regimens. 98,102,118,125,157 For example, in a qualitative study of older adults with diabetes (aged ≥ 65 years), participants reported a variety of motivators for maintaining SM once a routine had been established. HCPs needed to understand what these motivators were and work with patients towards their SM goals. Feedback from participants included the observation that HCPs did not listen to, or were dismissive of, their concerns and, as a result, this reduced the trust that they had in their HCP. 118

A Cochrane review137 looked at SM interventions for people with diabetes and severe mental illness. Searches, which were conducted in 2016, identified only one RCT that met the inclusion criteria. This was a small study (n = 64) involving adults with type 2 diabetes and schizophrenia or schizoaffective disorder (average age 54 years). The intervention, which involved a 24-week education programme, had no significant impact on glycaemic control, although there were small improvements in diabetes knowledge and self-efficacy. The authors suggested that the lack of impact could, in part, be explained by the lack of a person-centred approach within the programme.

We found only limited evidence that evaluated the impact of our identified context on glycaemic control or adverse events. Of the six included sources that focused on people with dementia and diabetes, only one59 evaluated an intervention. This was a controlled study evaluating the provision of personalised education sessions for diabetes and cognitive impairment. The study showed an increase in self-efficacy and a short-term effect on glycaemic control (HbA_1c levels initially declined but had returned to baseline after 6 months). The authors suggested that this supported a need for long-term connection and maintenance programmes for this group.

Trusting relationships

A key element of achieving a person-centred or partnership approach to care is building a long-term trusting relationship with the people living with dementia and diabetes and their family69,70,72,73,79,102,125,142 that involves listening and valuing the patient’s subjective health experiences:

It’s allowing a two-way exchange of information, isn’t it, about how different conditions might affect things.

Res 1

The impact that a trusting relationship between HCPs and people with dementia can have on patient care is illustrated in a qualitative study87 exploring GPs’ perceptions of what is meant by a ‘timely’ diagnosis of dementia and how this differs from an early diagnosis. The timely diagnosis was seen as requiring a more patient-centred approach that recognises the issues that families go through in reconciling the stigma of diagnosis and consequences for the future. The authors argue that a ‘diagnosis’ of dementia was not a discrete act ‘. . . but a collective, cumulative, contingent process’. 87 The GPs in this study saw the diagnosis as a journey for both them and the patient and their family in reaching a nuanced and contingent resolution, leading to a co-constructed diagnosis. This study, like others,39 highlights how vital relationship continuity can be in creating trust and helping to secure appropriate and timely care for people living with dementia.

A feasibility study122 explored the use of patient-centred consultations for improving medication adherence and management among older people with diabetes. The intervention included the use of dialogue tools whereby the patient was asked to describe their day living with diabetes, to ‘. . . encourage participants to reflect, engage in dialogue, and verbalize their experiences’. 122 The findings showed that people living with type 2 diabetes felt listened to, which generated a trusting relationship with their HCP. Participants also reported feeling more capable of managing their diabetes, although, as this was a feasibility study, the authors did not measure the impact of the intervention on glycaemic control. HCPs reported that the process supported patient-centred partnership but the authors cautioned that the dialogue tools are, by themselves, not enough to ensure patient-centeredness; adequate communication skills are also required to prevent the process becoming a tick-box exercise.

We know that for people who have previously self-managed their diabetes (often for many years), the transition to needing SM support can be difficult and distressing. 39,142 Although we found some literature on enablement for people living with dementia (see Context–mechanism–outcome 1: embedding positive attitudes towards people living with dementia), we found no evidence about how (or whether or not) a patient’s previous diabetes knowledge or SM strategies are acknowledged or used by HCPs.

Shared decision-making

The management of people living with dementia and diabetes represents a complex process with difficult trade-offs to be made, so methods of involving patients and families in shared decision-making are important aspects of achieving person-centred care. In a multicentre cluster RCT of older people with type 2 diabetes (not dementia), Branda et al. 71 introduced a decision aid that encouraged shared decision-making between participants with diabetes and HCPs. There was no difference in glycaemic control between the decision-aid intervention group and the control group. However, the decision aids did facilitate patient-centred practice and co-construction of decision-making related to diabetes management. The participants also felt more engaged in their diabetes management despite the lack of clinical outcomes. Similarly, Mathers et al. ,111 in a RCT of a decision aid for patients with type 2 diabetes (adults aged between 42 and 87 years), found no statistical effect on glycaemic control (HbA_1c decreased in both intervention and control groups), but decision-making by patients was more autonomous and promoted more realistic expectations and greater knowledge. The authors suggested that this can lead to better long-term SM practices.

Both of these studies involved the use of a decision aid in a single consultation between a clinician and a person with diabetes. In the Mathers et al. 111 study, the mean duration of the consultation was 15.31 minutes. People living with dementia and diabetes are likely to need interventions that involve more frequent contact and include repetition and reinforcement. Decision-making involving people living with dementia is likely to be complicated by issues of consent, concordance and the appropriateness of treatment for people with dementia.

Context–mechanism–outcome 2 summary

Self-management of diabetes in people with dementia is likely to be contingent on the development of a trusting relationship between the HCP and the person living with dementia and diabetes and their family, involving understanding and incorporating patient priorities and how these may change over time. This in turn facilitates a person-centred approach to care planning and diabetes management. There is currently little research that looks at person-centred approaches to diabetes management in people living with dementia, and further research is needed to develop interventions that support partnership working and incorporate the consideration of the risk-benefit balance for different treatment options.

Programme theory

If HCPs are expected to develop skills that enhance the delivery of individualised and tailored care to people with dementia and diabetes (e.g. enablement rather than management, listening/communication/negotiation, shared decision-making), then this legitimises the work and creates the expectation in patients and in HCPs that diabetic care for people living with dementia is important (M), leading to the provision of more tailored diabetes care (O) and better engagement with SM by people living with dementia and diabetes and their family carers (O).

Tailored and flexible

The importance of providing tailored care for older people with diabetes was emphasised in the literature. In a meta-analysis of SM programmes in older people living with diabetes, Sherifali et al. 126 concluded that the strategies most effective at reducing HbA_1c levels were those that involved tailored interventions or psychological support. Tailored interventions consisted of patients receiving advice and support that met their particular needs and goals for diabetes SM and the psychological support emphasised coping strategies for depression and distress. The authors did not, however, unpick what it was about tailoring or psychological support that might have triggered these different responses. A RCT to assess barriers to SM experienced by older people with type 2 diabetes demonstrated the importance of assessing the specific self-care barriers faced by this cohort and then individualising diabetes management. 116 Individualisation was aided by a diabetes educator making telephone contact between clinic visits.

Current clinical guidance also highlights the need for individualised and flexible care for older people with dementia and diabetes. Clinical guidelines on diabetes recommend that target HbA_1c levels be relaxed for older people who are frail or have comorbidities and/or dementia. 26,28,29,158 A summary of recommendations for glycaemic control in older people (including those with dementia) can be seen in Table 5.

For example, a best clinical practice statement28 by a multidisciplinary national expert working group recommends that, for people with dementia, clinicians should aim to achieve fasting blood glucose of 6–9 mmol/mol (range HbA_1c 53–64 mmol/mol; 7–8%). Although guidance for general diabetic populations recommends levels of 6.5–7 mmol/mol, it also recommends that these targets be relaxed for those with complex health needs or those who are frail. 158 Despite this, evidence from observational studies suggests that many older people with diabetes are potentially overtreated. 25 This could be partly because these targets do not ‘fit’ with current performance measures159 (such as QOF targets in UK primary care), a factor suggested by a number of our stakeholders and illustrated by this quotation from a diabetes specialist:

. . . we encourage people to set agreed targets with the patient . . . that may well be . . . higher than the general population target which is a key message we get across to the GPs because they’re so driven by QOF.

Diab 9

Flexible diabetes care is not just about adjusting target HbA_1c levels. A case study117 involving the use of continuous glucose monitoring to measure hypoglycaemia in older adults with diabetes found that simply relaxing HbA_1c goals may not be adequate to protect frail older adults against hypoglycaemia. They found ‘an unexpectedly high frequency of hypoglycaemic episodes in older adults with poor glycaemic control’;117 this was true of those with type 2 diabetes as well as of those with type 1 diabetes. The authors concluded that there is a need for treatment regimens that better match patients’ self-care abilities, although they did not explore further what such regimens might look like. 117

Developing appropriate skills

Research suggests, however, that HCPs do not always have the skills needed to provide flexible individualised care for people with diabetes and/or dementia. 39 This was echoed by our stakeholders, as the following quotation illustrates:

I’ve seen very very few examples where it’s done well, any of this, any of this sort of self-management, shared decision-making, anything. It works well I think in a few types of situations where people feel they have the autonomy to act in the interests of the patient and are able to make decisions that might not necessarily follow the rules but they can see the patient as an individual and understands their circumstances, otherwise I think health professionals are possibly becoming themselves much more risk-averse and not wanting to suggest things that aren’t perceived as being healthy or might not be the right answer.

Res 1

One consequence of this is that care is not tailored appropriately. In a RCT116 involving older people with diabetes, the authors wrote that ‘the study team felt that many patients with multiple conditions were on complex regimens that were clearly beyond their coping abilities’. In a qualitative study with older adults with diabetes and other comorbidities (not dementia), participants reported that they felt overwhelmed when dealing with their diabetes and other conditions and that they felt that their HCPs did not understand the difficulties they experienced with balancing their comorbidities so that they could achieve a reasonable quality of life. 70 The following quotation from a GP highlights how ideas about assessing a patient’s abilities to cope with treatment plans are not yet part of routine care. The GP is referring to the ‘choosing wisely’ initiative from the American Board of Internal Medicine Foundation,160 which aims to promote conversations between clinicians and patients to prevent the use of unnecessary tests and procedures:

. . . choosing wisely American stuff you know, I think we’re all warming up to this agenda but I don’t think anyone’s quite cracked you know, it’s not mainstream yet.

Diab 12

In addition, Kennedy et al. 103 conducted a RCT of SM support for people with long-term conditions in primary care in the UK. The results of the RCT103 and accompanying process evaluation104 and qualitative study60 suggest that SM support did not fit with a biomedically focused ethos and, as a result, practices did not give it the priority needed to embed it in the day-to-day work of primary care. A qualitative study conducted in the USA also found that an overemphasis on biometrics and medicalisation by HCPs was a barrier to SM,100 and a study looking at SM pathways in primary care in the UK concluded that primary care is currently failing to support SM. 119 This is illustrated by the following stakeholder quotation:

. . . for the general population, self-management . . . is not working particularly effectively . . . translate that to a much more delicate and fragile group . . . who have other comorbidities and have dementia . . . then those types of responses are likely to be even less effective [of SM support strategies].

Diab 2

Several studies70,98,134 suggested that there is ‘goal divergence’ between patients, carers and HCPs, particularly when there is medical uncertainty, as is generally the case with the dynamic nature of living with multiple conditions. In a realist synthesis, Yardley et al. 134 suggested that, to reconcile discrepancies between the goals of patients, carers and HCPs, there is a need for less emphasis on a ‘diagnostic-cure model’. Patient-centred communication and collaboration between HCPs and patients/families appear to be key to achieving individualised care70,98,134 (see also Context–mechanism–outcome 2: person-centred approaches to care planning).

Context–mechanism–outcome 3 summary

We found evidence that, to be able to provide flexible and individualised care for people living with dementia and diabetes, HCPs need to prioritise communication, negotiation and partnership working. They need to be provided with appropriate training and support so that they have the confidence to focus more on quality of life and patient abilities and less on biometrics and clinical targets. However, currently the evidence to link our proposed context with glycaemic control or a reduction in adverse diabetes-related events is limited.

Programme theory

If HCPs maintain regular contact over time (e.g. face to face, by telephone or by e-mail) with the person living with dementia and diabetes and their family carer, monitoring and anticipating needs throughout the dementia trajectory (C), then HCPs feel more equipped to meet patients’ needs (M), and people living with dementia and diabetes and their families believe themselves to be supported through the transition from functional independence to functional dependence (M), leading to improved diabetes management (O).

Anticipating needs

Regular contact between HCPs and the people living with dementia–carer dyad appears to be an important contextual factor for HCPs to anticipate transitions and help people living with dementia and diabetes and their family carers to manage changes in function and SM capabilities. 142 This is particularly important for people living with dementia for whom the dementia may progress in an uneven pattern of decline,92 and for whom the transition from autonomy to delegation or caregiver-led management may be particularly difficult. 39,79,118 Regular contact may have particular advantages for people living with dementia, as illustrated by this quotation from a stakeholder interview:

. . . if it’s set up on a regular basis, so the person knew, you know, like Tuesday afternoons when I speak to my diabetic nurse, that can be put in their diary.

Dem 4

Penn et al. 119 (looking at diabetes but not dementia) suggested that if SM support is offered at regular points throughout a person’s trajectory, rather than just focusing on the point of diagnosis, then HCPs are more likely to pick up problems with SM and be able to offer appropriate support. In a RCT116 focused on improving diabetes management in older people, the authors found that older adults were reluctant to make changes to their medication between clinic visits. However, regular telephone contact from a diabetes educator encouraged people to adjust their insulin dosage, leading to better glycaemic control. Regular contact is likely to be even more important for people with dementia, and guidance and commentary on the management of diabetes in people living with dementia recommends regular reassessment to identify additional care needs. 28,83 Older people with diabetes (including those with dementia) are particularly likely to be in need of support during vulnerable periods, such as after a period of hospitalisation. 116,142

Current care pathways, however, lack the capacity to consistently assess SM capabilities or provide SM support. 119 This is illustrated by the following quotation:

. . . since I’ve been in the care of the diabetic clinic everything else has gone out the window. When I was in the care of the specialist nurse at the GP’s, I would have a regular sort of every 6-month check on my feet . . . and the amount of protein in my urine, all those tests have now ceased, I’m now only looked at from a point of view of sugar levels.

Person with type 2 diabetes

What sort of contact and with whom?

The quotation above also illustrates another important point. It is not just about having regular contact but also about who that contact is with. The quotation suggests that care at the GP surgery, where the patient was known, was preferable to the less personalised (and less comprehensive) care received at the diabetic clinic. Clearly this is only one person’s experience; however, many of the studies we included39,59,62,64,66,116,130 support the link between regular contact with the same HCP, the development of a good relationship and improved SM practices. For example, a controlled study59 of diabetes SM education for people with cognitive impairment found that regular contact with a single diabetes educator led to the development of a rapport, which was important in helping participants to develop strategies to improve SM (e.g. remembering when to take their medication).

Of the studies we included, 38 described the development or evaluation of an intervention; two included people living with dementia and diabetes, 16 included people living with dementia (not diabetes), 14 included people with diabetes (not dementia) and six included other groups such as older people or those living with long-term conditions. This includes a diverse range of interventions, although many were aimed at improving SM. Interventions were delivered by a range of HCPs, most commonly GPs, nurses, OTs, psychologists and, in US studies, certified diabetes educators. Regardless of who was delivering the intervention, studies consistently highlighted the importance of the continuity and quality of the relationship. A number of studies looked at the development and delivery of interventions delivered remotely using technology (e.g. text messages). This is explored further in Context–mechanism–outcome 5: family engagement.

Although the quality of the relationship is important, people with dementia and diabetes need to be managed by practitioners with appropriate expertise. Stakeholders highlighted the particular challenges HCPs face when caring for people with dementia and diabetes, and the different skill sets that different specialists are likely to have. Dementia as a comorbidity may challenge a diabetes specialist, and a dementia specialist may lack appropriate diabetes knowledge. From our stakeholders we found examples of roles that involved someone with expertise in both mental and physical health, although this was the exception rather than the rule:

. . . I have a very good colleague . . . who is a specialist physical health-care nurse and a mental health nurse, and that is a, I’ve often thought of this person as a really interesting model for the future . . .

Dem 7

One paper described a programme that included embedding a psychiatrist within an integrated care pilot for people with diabetes in inner London. 123 The involvement of the psychiatrist raised awareness of patient mental health problems among other team members and meant that the team was able to identify if poor SM might be a result of mental health issues. The author said that clinicians became more engaged when they were able to ‘really think about the person in their entirety’. However, the paper provided no data on the impact of the intervention on patient outcomes.

Context–mechanism–outcome 4 summary

Continuity of care and regular contact are important for both people with dementia and those with diabetes, but it is likely to be even more critical for those with both dementia and diabetes. This is important so that professionals can recognise times of transition (e.g. worsening symptoms of dementia having an impact on diabetic control and increased risk of hypoglycaemia). At such times support may need to be ‘geared up’ but may be able to be ‘geared down’ once the transition has been successfully dealt with. Ensuring that all professionals have expertise in dementia and diabetes would be difficult, and collaborative practice is likely to be necessary for people with both conditions, particularly for more complex cases such as people who are insulin dependent or those with advanced dementia.

Programme theory

If family carers are routinely involved in care planning and information sharing, and are given the support they need to take on the tasks associated with managing diabetes in people living with dementia (e.g. medicine management, recognition of hypoglycaemia), (C) then family carers will feel supported and that their contribution is recognised and appreciated (M), leading to the development of effective SM strategies on the part of the family carer (O).

There is a great deal of evidence that family members often provide significant SM support for people with long-term conditions such as diabetes,45,69 particularly when dementia has an impact on a person’s ability to undertake self-care-related tasks. 142 A mixed-methods study exploring the impact of dementia on access to non-dementia services found that family members were often proactive in facilitating continuity of care and negotiating access to services for their relatives with dementia and diabetes,39,142 and a review of medicine use in people living with dementia found that administering medicines was a huge component of being a caregiver of a person living with dementia. 79 Managing the needs of a family member with dementia and diabetes raises particular anxieties for carers because of the risk of hypoglycaemia and other adverse advents associated with diabetes. People with dementia are likely to forget to eat (or to eat too much), and ensuring that they eat appropriately and that this is co-ordinated with their medication is a source of great concern for carers, particularly if they live at a distance. 39 This is complicated by a lack of flexibility in service provision, such as district nurses being able to go in only at certain times of the day, or social services carers being unable to oversee medication administration for people with diabetes. 39,142

Family carers may also play an important part in supporting or motivating their relatives in their SM activities. A quasi-experimental study conducted in the USA (adults with diabetes but not dementia) found that involving a care partner may make a person with diabetes more likely to participate in an automated telephone SM system. The authors suggest that the mediating mechanisms were emotional support leading to improved ability to regulate one’s own behaviour, direct assistance with diabetes problem-solving provided by the care partner, and reinforcement of adherence. 63

Despite this, family carers often feel undervalued or excluded from decision-making, and they may be ill-prepared to take on responsibility for SM. 39,42,64,79 The situation is often further complicated by the fact that they may take on SM-related tasks only once there is a crisis or a failure to adhere to medication. 42 The need for appropriate support for carers is highlighted by studies that found that problematic medication management practices may persist despite the involvement of a family carer. 79,107 For example, a systematic review looking at challenges to optimal medicine use in people living with dementia found that family carers may adopt strategies that are not always safe and effective, and that it was difficult for them to make decisions about when to withhold or to give medicines. 79 A tailored problem-solving intervention to maximise medication management practices among caregivers of people with memory loss found no significant difference between the intervention and control groups, although both groups showed a significant reduction in the number of medication management problems. 107 The authors suggested that the lack of difference between the groups is because the caregivers in the control group participated in a face-to-face baseline assessment by a study nurse or social worker, which included questions about their knowledge of medications and approaches to managing them, medication reconciliation and medication-related resources.

Interventions aimed at supporting people with dementia to manage their diabetes should take into account the education and support needs of family carers as well as those of the person living with dementia,39,64,79 support that needs to include the issues arising from both conditions and the impact that dementia is likely to have on diabetes management. 39 We found no studies evaluating structured interventions to provide education and support to family carers of people with dementia and diabetes. Although stakeholders did talk about involving family members in education, they also recognised that this was not done in a structured or comprehensive way:

. . . I think we could probably do a lot more . . . supporting families and carers and to give them the confidence, I think they’re so worried, it can be so . . . frightening . . . to have both conditions . . .

Diab 13

. . . patients are educated one to one or through diabetes-structured education, again I’ve never heard of a patient education for carers and those with dementia to support them . . .

Diab 11

A number of studies linked support and education for carers to an improvement in SM. For example, qualitative studies looking at SM for people living with dementia115 and for people with diabetes (not dementia)69 argued that it is important to involve carers in the development of SM skills alongside the person they care for. This was also highlighted in the stakeholder interviews. For example:

We need to sort of normalise the situation where it is completely normal and expected that close family members will be involved in any decisions and there will be partnership.

Researcher: SM of long-term conditions

Context–mechanism–outcome 5 summary

Self-management for people with dementia and diabetes needs to be conceptualised as an activity that frequently involves not just the person with dementia but also their family members. Interventions need to take into account the needs and capabilities of family carers and the anxieties associated with managing medication and diet and preventing adverse events such as hypoglycaemic attacks. Including the family carers of people with dementia and diabetes should be the default option, and they should be included early, when the people living with dementia still have the capacity to make decisions and before SM breaks down.

Programme theory

As the dementia trajectory progresses, assistive technology (AT) needs to be tailored and adapted to the needs and requirements of people living with dementia and diabetes and their families (includes social, environmental and cultural needs) with the focus on maintaining autonomy for the people living with dementia and diabetes (C), leading to people living with dementia and diabetes and their families gaining an understanding (awareness) of the usefulness of AT in their management of dementia (M) and to a more effective and sustained use of AT to maintain autonomy and diabetes SM strategies (O).

The definition of AT used for this CMO is ‘any product or service designed to enable independence for disabled and older people’. 161 This broad definition has the merit of including both high- and low-technology devices, ranging from objects that may not be considered ‘technological’, such as notice boards and calendar clocks, to devices that are clearly technology based, such as GPS (Global Positioning System) locators, blood-glucose monitors, reminders for medical management, and the internet. 135 Gibson et al. 91 further subdivided AT into telecare and telemedicine. Telecare involves remotely monitoring people in their own homes and communicating with them at a distance via telephone and the internet, while telemedicine is technology-supported medical or nursing tasks that assess biometrics sent from the patient and instructions returned to the patient from their HCP. The use of AT in health and social care is seen as an important element in enabling older people to live independently, and is part of the Department of Health’s Prime Minister’s Challenge on Dementia. 162 Despite this, a review of the use of AT for people living with dementia suggested that the technology industry has limited awareness of the needs of people living with dementia and their carers, and needs to view the ‘dementia market’ as an attractive option. 105

We included 17 papers39,59,61,63,68,90,91,93,97,99,101,105,112,120,127,133,135 (five of which were reviews) in this section. Two39,59 relate to people with dementia and diabetes, seven61,90,91,93,101,105,127 relate to the use of AT by people with dementia, four97,120,133,135 concern the use of AT by older people and four63,68,99,112 involve people with diabetes (not dementia). There is clearly a large body of literature relating to the use of AT by people with diabetes (in particular telehealth and telecare), but most of this was beyond the remit of this review. We found only one study59 that looked at the use of AT to help manage diabetes in people with cognitive impairment, but there is a lot of evidence on the use of technology by people living with dementia. 91,97,120,135 More details of the included studies can be seen in Table 6.

Tailoring assistive technology to the needs of people living with dementia

In the UK, AT for people living with dementia is provided by a variety of sources, including health and social care services and the private sector and AT that has been adapted by users for their own particular needs. Qualitative research suggests that the bulk of provision involves the direct purchase of ‘off-the-shelf’ technology, which, if not specific to user needs, is adapted by family carers. 91,97 A review of AT for people living with dementia found that family carers were most likely to use technology that is not specific to caring or to dementia, such as tablets, baby monitors, smart phones and light sensors. 105 Health and social care services appear to provide inadequate support because they have limited provision of appropriate technology and/or they do not have the requisite knowledge base. 91,118 Moreover, provision may be variable:

In one local authority, we went to one recently, a dementia kind of carers group and sat with them and showed them a list from somewhere else and some of the devices on that weren’t on theirs, you know? It’s not equitable . . .

Res 2

A number of commentators argue that it is important to involve people living with dementia in the development of supportive technology. A systematic review127 examining the involvement of people living with dementia in developing information technology (IT) applications found that people living with dementia wanted to, and could, contribute to IT design. The authors argued that the involvement of people living with dementia and their family carers can help to ensure the development of appropriate AT that is ‘fit for purpose’ and that it can also enhance feelings of empowerment in people living with dementia. 127 However, even when included in the design of AT, people living with dementia are likely to need ongoing support. In an ethnographic study133 looking at methods to support the involvement of older people in the coproduction of AT, participants suggested that they would need ongoing support and assistance once the technology had been provided and that family members would need to be involved throughout the provision process. Coproduction involves the sharing of power between professionals and members of the public, working together in equal partnership. 163

Support in the use of assistive technology

One of the ways in which AT needs to be tailored to people living with dementia and diabetes is to ensure that interventions include appropriate support, for example from a family carer or a health or social care professional:

. . . technologies are great but you need to think is it the right thing for the patient . . . is there enough support around it to implement it and respond to it . . .

Res 2

The role of support in the use of AT, most often from family members, was a common theme. 68,90,91,112,120,133,142 For example, a mixed-methods study68 on older people with diabetes and a qualitative study91 on people living with dementia highlight the importance of family carer assistance in the use of AT. Family carers were found to invest a substantial amount of effort in embedding AT to help their relatives to maintain independence and reduce potential risks, which in turn minimises carer anxiety. 91 Family support was also thought to increase motivation to use, and participation in the use of, automated SM support. 63 However, a feasibility study of diabetes self-care support suggested that family ties are complex and family involvement in SM is not always helpful. 112

Technology in itself is unlikely to solve the problem of independent living for older people,97 particularly those living with dementia. 105 Knapp et al. 105 write that ‘successful interventions need to recognise the high value that many PLWD [people living with dementia] and their carers place on face-to-face service contacts . . . technologies that seek to reduce these contacts are unlikely to be acceptable or used (or, if they are, could exacerbate problems associated with social isolation and loneliness)’. AT, such as telehealth or telecare, appears to be most effective when technology augmented or involved face-to-face contact. 59,90 For example, a controlled study59 evaluated the impact of a tailored educational intervention, delivered via Skype on an iPad (Apple Inc., Cupertino, CA, USA), for older people with both type 2 diabetes and mild cognitive impairment. There was an initial decrease in HbA_1C levels post intervention but this was not sustained at 6 months. The authors account for this through the fact that contact with the HCPs ended, and that their input had been instrumental in sustaining participant adherence. Participants in this study had few problems using their iPads and wanted to continue using them after the trial to maintain social networking, retrieve information and communicate via e-mail. The authors suggest that using such AT with trained volunteers over the longer term may enable people living with dementia and diabetes to receive effective and efficient medication management. However, it should be noted that participants were paid to take part in the study and were given the iPads (neither of which is likely to be possible outside a research setting).

The importance of contact with a HCP was supported by a qualitative study61 exploring the use of a SM programme for carers of people living with dementia; participants preferred blended care (face-to-face and online modules) as they valued personal contact with a professional. As one participant said, ‘people experience emotions, whilst a computer is just an object’. 61 The assumption was that personal contact could increase motivation and adherence, although the mechanism for achieving this was not identified in the paper. This is supported by an evaluation of web-based appointments for people with diabetes in Newham, East London. Older people were less likely to have broadband at home and only 11% of those aged 70–79 years agreed to participate in Skype consultations, compared with 82% of those aged < 50 years. 164 In addition, a RCT of a cloud-based diabetes management programme aimed at improving SM (not dementia) suggested that connectivity with a coach helped people to feel less anxious and more motivated. 99

Context–mechanism–outcome 6 summary

Evidence suggests that, in order to make AT usable for people living with dementia and diabetes and their families, AT needs to be focused on the needs of the user (e.g. maintaining autonomy), involve people living with dementia and their carers in its development and include family carers in installation and training. Telehealth and telecare should be designed to involve (as the default option) care partners such as other members in the family network. Appropriate support in the form of face-to-face contact appears to be an important contextual factor that may lead to improved motivation and adherence. Service providers need to recognise the high value that people living with dementia and their family carers place on face-to-face contacts.