Understanding and improving experiences of care in hospital for people living with dementia, their carers and staff: three systematic reviews

Reading and re-reading

Beginning at the full-text screening stage and continuing into the final stages of synthesis, two reviewers (RGJ and RA) read and re-read included papers during processes of familiarisation, coding, summarising and checking.

Identifying relationships between studies

During data extraction and the creation of tables summarising study characteristics, the same information about each study was documented in the same way, supporting the systematic identification of similarities and differences in study aims, location, design, interventions and findings. The initial process of coding also contributed to establishing relationships between studies.

Creating translations and identifying refutations

Translation and refutation of study themes within each review occurred throughout the review process. Relationships between review 1 study themes, and how these linked to the interventions reported in reviews 2 and 3, were discussed regularly between core reviewers (RGJ, RA, IL, JTC, MR). Study concepts were discussed more broadly with the PAG at the four whole-team meetings (February 2018, September 2018, May 2019 and July 2019) and at consultation meetings (see Stakeholder involvement in reviews 1 and 2).

Coding and concept development

During data extraction, one reviewer (BA) extracted themes for each study using Microsoft Word, and one reviewer (RGJ) uploaded a PDF file of each study into NVivo software. Where searchable PDF files were not available,42–48 the findings and discussion were typed into a Microsoft Word document and these were uploaded into NVivo. Findings and discussion sections were coded deductively according to a preliminary conceptual framework developed before the start of the review to support our application for funding (Figure 2). The model was put together from a scoping of the literature and drew on previous published models and frameworks of dementia care. 49–52 Coding was undertaken inductively where findings were not represented by this initial conceptual framework. Each code was organised using ‘participant/researcher subcodes’ that identified whose perspective the coded text represented (e.g. staff, people living with dementia, carer, researchers). Memos were kept about references made in papers to existing theory, interpretations of how papers linked with or refuted each other, and emerging categories/themes.

FIGURE 2.

Preliminary conceptual framework describing the nature and complexity of possible factors that may influence the experience of hospital care for someone with dementia.

Following coding, it was decided that the findings differed in significant ways by participant type and they were, therefore, synthesised separately. Noblit and Hare41 suggest using a pre-existing framework for conceptual development, for example by adopting the thematic structure from a key paper to guide the synthesis. However, we adopted the approach proposed by Spicer,53 who posits the development of concepts through an inductive process of interpretation across studies. RGJ, in consultation with RA, conducted translation of studies by further regrouping and refining concepts from the coded text by perspective [people living with dementia: subreview A (see Subreview A: findings from reviews 1 and 2 about the experience of care for people living with dementia in hospital – feeling afraid and insecure), staff: subreview B (see Subreview B: findings from reviews 1 and 2 about the experience of care for people living with dementia in hospital – feeling prevented from being able to give ‘good’ care); carers: subreview C (see Subreview C: findings from reviews 1 and 2 about the experience of carers of people living with dementia in hospital – feeling stressed and desiring inclusion)] to create subreview conceptual maps. These conceptual frameworks were then used to organise how the findings were communicated for both reviews 1 and 2.

Characteristics of prioritised papers and their relation to the remaining included papers in review 1

In review 1, we conducted a prioritisation process to manage the large number of included papers [see Prioritisation of papers in Review 1 (experience of care)]. We prioritised 21 studies66–68,79–82,86,87,94,96,99,104,107,108,110,118,120,122,123,127 reported in 29 papers65–71,79–82,86,87,94,96,99,102,104,107,108,110,111,113,114,118,120,122,123,125,127 as most able to meet the questions of review 1. Thirty-six papers43,46,47,57,60,61,63,64,73–76,78,83,84,88–91,95,97,98,100,103,106,109,112,115–117,119,121,126,128,131,132 were identified as medium priority, and 17 papers42,44,45,56,58,59,62,72,77,85,92,93,101,105,124,129,130 were identified as least able to provide answers to the review 1 research questions.

Following synthesis of the three subreviews, the contribution of prioritised studies was tabulated against the findings from medium-priority studies in relation to each subreview main theme and subcategories (see Appendix 7, Tables 6–8). During this comparison, it was found that the majority of medium-priority studies supported the structure and content of the subreviews. One study132 interpreted responsive behaviour as resistance, rather than unmet need as we have done in this synthesis. However, we considered these to be compatible interpretations. One study106 structured findings around psychoanalysis and infant theory. Although the conclusions of this work were interesting, because the methods were poorly reported, and the findings differed from those of the remaining studies, we did not incorporate these findings into review 1. It was determined unnecessary to compare prioritised papers with the studies judged as least likely to contribute to the review, as it has been found that such studies tend not to have an impact on syntheses because of their sparse or descriptive findings. 38

Intervention characteristics

Intervention characteristics are summarised in Report Supplementary Material 3 using the TIDieR checklist. 145 Most of the interventions involved a combination of components including institutional level support, therapeutic support to carers, information/education for carers, inclusive approach to carers, activities for people living with dementia, training off wards, training on wards, existing specialist knowledge utilised, documentation to improve individualised care, new approach adopted to guide caring for people living with dementia, specialist capacity added, non-specialist capacity added, and structural changes to ward environments (see Report Supplementary Material 4). The study with the smallest number of components involved two components;137 the most common number was four,133,135,136,139,141,144 and the greatest number was seven. 23 Interventions were categorised (see Categorisation of interventions for review 2) according to intervention focus into six categories: improving staff information, knowledge and skills (n = 5),136,139–141,143 increasing ward capacity (n = 2),138,144 activity-based or tailored interventions (n = 2),116,134 changes to ward environment (n = 2),133,137 support for carers (n = 2),48,135 and special care units (n = 1). 23

Identity: review 1 (experience of care)

Identity was addressed in three ways within included studies: respect for personal preferences, space being made for personal items and respect for dignity and personhood. Simple respect for personal preferences, for example what someone liked to be called, what songs or hymns they liked, or what their usual routines were, could mean a lot. 65,68,70,99 Being able to have personal items such as family photographs by the bed not only helped people maintain their sense of identity, but talking about a past that they could remember was easier than expressing their identity in the disorientating present. 80 Both carers and staff noted that it was often small personal things that could make a person living with dementia ‘feel like a person’; it might be someone spending a few moments with them or engaging them in conversation as they walked past, or often something as simple as being acknowledged. 68,69,81

Contrasting these positive experiences, people living with dementia sometimes felt unimportant, a nuisance or ignored. 87,110,127 In addition, there were observations of personal preferences not being sought or being actively ignored:69,87,110,122

I kept saying to them that, although he’s down as Lewis Brown, he’s always from a little boy been called Roger, so I always said his name is Roger . . . they always spoke to him as Lewis and I said well you’ll not get a connection.

Carer (author edits)70

A lack of respect for the personal preferences of people living with dementia was also shown when staff woke individuals to serve them meals at times that fitted into ward routine rather than suiting the person living with dementia, did not defer to food choices, or did not offer drinks outside the ‘drinks round’. 87,122,127 Similarly, mealtimes, the giving of medications and personal hygiene were areas of care in which personal autonomy was often ignored. 94,110,122 For example, there were instances where the choice of people living with dementia to refuse medication was actively circumvented by concealing medications within food. 94

Identity could also be compromised by a lack of respect for dignity. Limited spaces between beds in bays often resulted in a lack of privacy. The personhood of people living with dementia was ignored when staff communicated briefly or not at all with people living with dementia about what was about to happen or happening, or when staff carried out routine care while talking to each other and ignoring the person for whom they were caring:81,87,107,127

They literally charge in. Very rarely announce themselves let alone explain what they are or who they are . . . I think that they don’t want to know you because you’re going to be a nuisance.

Person living with dementia127

Delays to individuals’ personal care needs were also not uncommon, and instances were observed of these being actively ignored. 71,87,107,127 Such experiences are consistent with Kitwood’s146 suggestion that people living with dementia are sometimes interacted with more as objects than as people, with consequent harm to their already vulnerable sense of who they are.

Identity: review 2 (experience of interventions)

Interventions in included studies supported the identity of people living with dementia in two ways following from Kitwood’s description: becoming familiar with their personal preferences and histories, and respecting their dignity.

A number of interventions involved the use of a tool completed by carers or assessment by a specialist to gather personal information about people living with dementia that could then be used by staff. 87,133,143,144 Participants across studies talked about the benefits of knowing personal information about people living with dementia. For example, in the participatory music intervention, music that was meaningful and therefore most therapeutic to people living with dementia often linked to their past experiences. 134 Volunteers found that knowing personal information about a person living with dementia not only supported the development of rapport, but also could be used to soothe distress:

So we kind of redirected the conversation towards, ‘Well, I’m writing some reports. I know you used to write lots of reports [patient was former teacher],’ and then it led on to a discussion about her work and my work so we kind of distracted her away from her worries.

Volunteer, p90 (author edits)139

Staff found that by understanding the likes and dislikes of people living with dementia they were often able to resolve disruption without medication or restraints, because they were better able to meet the needs of people living with dementia. 140 Staff on the dementia-friendly ward133 felt that the changes to the environment made it easier to provide individualised care and to support emotional needs. However, individual needs meant that some of the environmental changes could actually cause distress:

The environment changes impact on patients differently . . . we had a patient with dementia and an acute delirium, who was really scared of the picture opposite her . . . she kept asking ‘who is standing there, is that my dog?’ She must have a pet at home.

Health-care assistant (author edits)133

Carers attending a carer support group135 perceived that the psychoeducation they received sensitised them to the wishes of the person living with dementia, and the authors of the study concluded that the group helped both carers and staff to reflect on the identity and personhood of people living with dementia. 135 Two of the studies explicitly discussed how interventions supported the dignity of people living with dementia. One study raised awareness in staff of the need to consider dignity through staff training,23 and found that carers perceived fewer staff failed to consider dignity on this special care unit. This suggests that training may have supported the dignity of people living with dementia. On the ward where access technology was installed,137 staff perceived that the technology fostered dignity for people living with dementia through privacy, autonomy and capability:

This [the chip card] helps me find my room, I know I can try out all the doors, so I can forget my room and still find it; I don’t have to worry and I don’t have to interrupt and ask the people in white to show me my room.

Person living with dementia (author edits)137

The privacy of their room also created a place where people living with dementia were free to keep personal possessions which were often strong prompts for past identity, for example by sharing photographs with staff, without the fear of another patient taking them.

Physical comfort: review 1 (experience of care)

A number of studies referred to physical discomfort from unmet needs due to pain, hunger, thirst or constipation/incontinence. 70,87,96,107,114 Discomfort was often linked to difficulties people living with dementia had with memory and communication. For example, authors of studies which focused on managing pain for people living with dementia in hospital observed that because of issues around memory, when asked questions, a person living with dementia was likely to only be able to respond according to his or her present state. 79,102 This meant that pain management could easily be inadequate because staff commonly relied on self-reports of pain. 79 Staff and carers said that it was possible to interpret pain levels of people living with dementia according to their gestures, posture, body movements, behaviour and/or metaphor, and that these could augment straightforward reports of pain. 79,123

Physical discomfort could also result when people living with dementia were left on their own or were not in their own bed, for example being left on a trolley for a long period of time despite acknowledging this was not good care. 70 Finally, discomfort occurred when staff either delayed care or overtly ignored the problem:

My mum’s lips kept sticking together because she wasn’t drinking and we had to constantly say ‘Can somebody please swab them and clean them?’ . . . and I feel, may be, if someone could have just been a bit more on the ball there.

Son70

Psychological comfort: review 1 (experience of care)

Despite observations from studies that suggested physical care was prioritised over psychological care of people living with dementia,69,87 some studies documented positive experiences of care for emotional comfort of people living with dementia. Staff were observed to respond sensitively when people living with dementia appeared in distress and to be sympathetic to those in pain or who were frail:66,70

Phyllis continued to cry . . . the housekeeper went over to Phyllis ‘Phyllis, now don’t cry. It does you no good love’ . . . The housekeeper wrapped both arms around Phyllis and rocked with her like a child, gradually slowing until the sobbing ceased.

Researcher observation70

A caring approach through touch was observed to calm or reassure people living with dementia, although comfort from touch was dependent on the relationship between the person living with dementia and staff,122 and in fact touch was not always comforting. 111 Some studies found that the general ambiance of a ward could improve emotional comfort by instilling a sense of security for people living with dementia, or add to a sense of insecurity. 68,81,123,127 Edvardsson et al. 81 attributed the behaviour of people living with dementia to the presence and ‘ways of being’ of staff members on the ward; when staff were present and engaged with people, and even when staff were present but not engaged, this supported a sense of security in people living with dementia. When staff were absent, people living with dementia became more anxious and this could easily be communicated to others in a sort of ‘collective escalation’. 81

Other aspects of ward environments were found to distress people living with dementia. Moving wards, or being moved for tests, and often being moved at very short notice or with little explanation, was seen to be challenging for people living with dementia. 107,118 Carers and staff across several studies reflected on the fact that the noise and general busyness from telephones, alarms, buzzers, surveillance equipment, staff talking (around people living with dementia and not to them), and patients calling out could be quite distressing to people living with dementia:87,107,118

Our environment isn’t good . . . in a big ward room there is a lot of movement and talk between the patients and noise from the TV, and there are many other unfamiliar sounds from buzzing signals etc., that can create anxiety, especially at bedtimes.

Nurse (reviewer edits)107

Included papers commonly described experiences of psychological discomfort. 69,79,81,87,96,107,113,118,122,127 Experiences of disorientation, discussed above, could create high levels of psychological discomfort for people living with dementia, and a lack of familiarity or meaningful connection was noted to add to fear, worry or anxiety. 66,99 Unmet needs relating to the other aspects of personhood, including attachment, inclusion, identity and occupation, could all create emotional distress.

People living with dementia appeared to express discomfort in a number of ways. Some examples included simply telling people they were unhappy – ‘I hate it [in here]’127 – but many authors linked a range of responsive behaviour to psychological and/or physical discomfort. Responsive behaviours included refusing food,66,108,122 refusing medication,122 refusing care (e.g. washing or toileting),96,108,110 removing/disconnecting medical equipment,105,107 aggression,81,87,96,104,108,120,122,123,127 agitation,66,81,96,107,118,120,127 vocalisation/verbal aggression/crying,68,80,81,87,96,108,122,123,127 undressing,96,104 walking/wandering,81,96,104,107,108,110,123,127 and/or rummaging/invading other patients’ space/moving furniture. 80,81,96,123 Finally, people living with dementia were perceived to express discomfort through withdrawal: by closing their eyes when staff approached or by closing their mouth and looking the other way when they did not want to eat or take medicines. 94,99,110

Comfort: review 2 (experience of interventions)

Findings from review 2 intervention studies linked to comfort in relation to companionship/relationship; meeting unmet needs; the senses; the structural environment, privacy and security. 116,133,138–140,144

Three studies found that providing companionship could support comfort for people living with dementia. 138,139,144 The two studies that focused on improving ward capacity by bringing in volunteers138,144 both found that the companionship provided by volunteers, whose primary role was to talk to people living with dementia, had a settling effect. Volunteers were also able to provide comfort during intensely vulnerable times when others were not able to:

She sat with the patient, was holding their hands and there were no relatives around, like they were abroad and they just couldn’t be there on time, and she was there holding hands while the patient was dying and with tears in her eyes.

Medical consultant138

Validation, Emotion, Reassurance, Activity (VERA) training for nursing students139 took an approach that drew on theories of PCC in that it emphasised the personhood of the person living with dementia and the importance of relationships between those providing care and people living with dementia. Students prioritised getting to know people living with dementia in order to promote comfort, and reported that getting to know patients by finding out about their personal histories, likes and dislikes helped them provide comfort, for example by knowing how to distract the person living with dementia from worries.

On a ward adapted to create a dementia-friendly environment,133 one of the adaptations was to remove the central nursing station; instead nurses completed paperwork on the wards in close proximity to the patients. Staff did not comment on how this change impacted their relationship with people living with dementia. The authors comment that the staff were not encouraged to change how they interacted with patients and carers, so while the opportunity for connections with people living with dementia was not optimised, it is possible that the constant companionship of having a nurse on the ward could have offered additional comfort to people living with dementia as was suggested in review 1.

Participants involved in three interventions referred to providing comfort through meeting unmet needs. 116,134,140 In the intervention using gentle persuasive approaches,140 staff were taught to interpret responsive behaviour as unmet need, leading to changes in practice [see Experiences of disorientation and responsive behaviour: review 2 (experience of interventions)]. Similarly, in the intervention Namaste Care,116 where people living with dementia were provided soothing sensory stimulation at the end of life, staff said that when the activities did not calm agitation, this led them to seek other causes for the behaviour. These staff viewed agitation as information that could help them meet the needs of the patient. Staff found that a Namaste Care approach supported comfort through touch:

There’s something about touching the skin. You are connecting with that person. I find it very soothing especially if you have a patient that’s very agitated.

Health-care support worker116

Stimulation of other senses could also be comforting. A health-care support worker trained in Namaste Care116 told about a patient who loved church hymns, and when he played them on his phone, ‘it calmed him completely. It took him back to a place where he was happier, content’ (p359).

Attachment: review 1 (experience of care)

Only a few studies referred indirectly to attachment and mainly through the sense of maintaining or fostering close relationships. The importance to the person living with dementia of maintaining family connections was observed across a couple of studies:96,99 either through the opportunities afforded or encouraged at visiting time or through families maintaining the familiarity of care practices:

I used to put me mum her nighty on [in hospital] and see to her and do her teeth and tuck her in . . . I think she felt better me doing that . . . It was more like being at home, when she stays with me.

Daughter or son (author edits)99

This extract suggests that familiar people in hospital are able to support a person living with dementia’s sense of security because they help maintain connections to prior routines, abilities and caregiving relationships. In one study, a person living with dementia was observed comforting a family member, giving the person living with dementia the opportunity to maintain their caregiving role to a loved one. In this extract Alma waited for a long period of time with her mother Patricia, a person living with dementia, during admission:

Even when I was standing next to her she’d say, ‘I bet your legs are really hurting you, because I couldn’t stand all that time’. And then she’d say to me, ‘Would you like to go and have a drink?’

Daughter114

Familiarity with staff could also develop into a sense of attachment. Consistency of staff was observed to foster more positive relationships with people living with dementia:

I think it may have been because . . . they did longer shifts and . . . my mother was under their wing so they developed a relationship to her which, to her, is very important. Whereas the other staff that I saw . . . they hadn’t got such a close relationship with her.

Daughter (author edits)70

A lack of opportunity for family members to maintain relationships with or be with the person living with dementia, either through limited space on the ward107 or through limited visiting hours, was noted in several studies. 96,99,118

Attachment: review 2 (experience of interventions)

None of the included papers in review 2 directly discussed supporting the needs of a person living with dementia to have contact with those with whom they had an attachment. A few studies, a special care unit23 that adopted a proactive and inclusive approach to carers and a ward fitted with access technology to which carers were given key cards,137 could potentially support attachment by allowing access to a person living with dementia for carers, however perceptions about this were not reported.

Inclusion: review 1 (experience of care)

Reference to inclusion in review 1 prioritised studies involved support for social interaction, or issues around respecting the rights of a person living with dementia to be involved in decisions about their own care. Direct inclusion by involving people living with dementia in decisions about their care was not commonly reported, but was referred to in one study; a clinician specifically asked that a person living with dementia be in the room while their care options were discussed with their family. 70 Inclusion was, however, experienced indirectly through interactions that helped create a sense of companionship and reassurance, for example when staff showed understanding of someone’s cultural beliefs. 68,69,87,99,110,123,127 Inclusion was also apparent through the togetherness found between people living with dementia. People living with dementia in shared living spaces were able to give support to and befriend each other:80,114

Mike walks back onto the bay. Alan calls to him ‘Wanna look at this Mirror [newspaper]?’ ‘Ay’ replies Mike ‘I’ll look at the local scandal. This football?’ Alan tells him ‘No, it’s local history’. ‘I’ll be in it then if it’s local history’ quips Mike.

Researcher observation, standard care ward87

They also provided comfort and gave physical assistance to each other, and called for nurses on someone else’s behalf as and when needed. 87

By contrast, the ward environment could also make people living with dementia feel excluded as a result of the physical location of their bed if they were isolated in a side room. 69,87,99 People living with dementia described not being asked to join in social activities that occurred on the ward,69 and talked about how the meals offered were not familiar to them or something they would usually eat. 127 Feelings of exclusion were also experienced by people living with dementia who perceived that they were not receiving the care that others were given or were not being asked their thoughts about the care options available to them:

. . . that physio he goes from room to room asking whether you want to do physio. Nobody asked me . . . Everybody do this or do that and I’m not asked [Larry looked very downcast, shoulders slumped, looking down].

Researcher observation of a person living with dementia (author edits)127

Decisions around discharge and assessment of capacity in particular seemed to be an area from which people living with dementia often felt excluded82,127 or, as in the following case, were likely to be excluded from if they did not agree with the team’s assessment of their capacity:

At follow-up, he expressed unhappiness because he felt ‘tricked’ by the social worker and doctors into accepting a trial discharge; but there had been no review or sign of any attempts to get him home.

Author82

Inclusion: review 2 (experience of interventions)

A number of interventions included in review 2 involved components that might potentially increase inclusion or a sense of belonging, for example those that encouraged interactions between staff or volunteers and people living with dementia,23,138–140,144 or between groups of people living with dementia and staff, for example through activities,23,116,134 social dining spaces133 and day rooms. 23 Participants from intervention studies rarely talked about aspects of inclusion explicitly. However, staff did notice how the mood and behaviour of people living with dementia improved because of interaction with volunteers:138,144

A young volunteer spent time with a patient with advanced dementia. The patient was seen to be responding in an animated way and the patient’s relatives commented later what good spirits the patient was in.

Researcher observation138

Carers noted that, as a result of a carer support group,135 people living with dementia were more engaged in social interaction because the carers got to know each other and as a result, got to know other people living with dementia. On the dementia-friendly ward,133 staff said that the artwork on the corridor walls – historical photographs of the area – created opportunities for staff, patients and carers to interact.

Occupation: review 1 (experience of care)

Boredom and a lack of meaningful activities for people living with dementia were noted across several studies. 87,96,99,127 People living with dementia spoke about the frustration of having nothing interesting to do and how this could result in them feeling as though they were simply wasting time:80,127

What are you doing during the day?

Well, there are some examinations. That’s it, you know . . . there is nothing here to do, nothing at all. I can’t come up with anything whatsoever.

Person living with dementia80

Some studies suggested that people living with dementia who had nothing to do were more prone to worry about their situation and that this could result in wandering or restlessness. 69 By contrast, people living with dementia responded positively when they were given something meaningful to look at or do, either involving others in the ward or one on one with a staff member:69,81

There were five patients sitting in the day room and one staff member was dressing the hair of one woman . . . The staff member involved all of the five patients . . . All of the patients present in the room expressed appreciation, interest and joy.

Researcher observation (reviewer edits)81

Occupation: review 2 (experience of interventions)

A number of interventions in included studies provided occupation for people living with dementia. As noted above in relation to inclusion, activities often created situations for social interaction and were linked to improved mood in people living with dementia:23,116,134,138 ‘Oh I enjoy it, [participatory music sessions are] like going to the pub’ (person living with dementia, reviewer edit). 134

Some staff involved with participatory music linked improved mood directly to enjoyment of the activity itself:

[Person living with dementia] requested her favourite piece of music and when it was played, her whole body reacted as she leant back, stretched her neck and closes her eyes. She was completely absorbed in the music and there were tears from her eyes.

Researcher observation (reviewer edit)134

Other staff attributed the responsive behaviour of people living with dementia to boredom and frustration at not being able to do anything. Similarly, some carers of people living with dementia who were not offered activities expressed dissatisfaction because they thought that this might have prevented responsive behaviour. 23 As well as providing enjoyment, participants understood activities as relieving boredom and related frustrations:

She was a completely different person when she’s done that music and that’s because a lot of her frustration comes from being bored and not being able to do anything because of her broken hip. That’s where her anger comes from.

Staff (author edit)134

Staff working on the ward that had access technology perceived that this technology supported people living with dementia in moving about the ward safely and self-regulating their need to rest. 137 The difficulties created by people living with dementia wandering on wards without access technology were noted in another study, where staff either returned patients to bed repeatedly or were obliged to walk with them. 23

Activities seemed to have the potential to powerfully and positively have an impact on people living with dementia, but the activities needed to be meaningful for that potential to be realised; for example, some people living with dementia found music difficult to enjoy134 and volunteers noted that individualising activities was important:

. . . every day you come on it’s different. You never get two days the same . . . Some days you might get four people up and have a game of bingo, and sometimes you might just wander round chatting . . . It all depends on the patients.

Volunteer (reviewer edit)138

Summary: the experience of care in hospital for people living with dementia

Overall, the experience in hospital for people living with dementia can be characterised by Feeling afraid and insecure, and review 1 (experience of care) studies linked optimal experiences of care for people living with dementia to PCC. Disorientation, fear and insecurity led to responsive behaviour that was an attempt to communicate these experiences of distress. Care that was focused on tasks and routines, and did not acknowledge personhood, tended to increase the fear and insecurity of people living with dementia. PCC met the needs of people living with dementia by providing for their psychological as well as physical needs. Although acknowledging the personhood of all individuals in hospital is important, PCC is particularly salient for people living with dementia because of the often intense experiences of disorientation that can be triggered by the confusing environment of the hospital on top of already-present dementia-related threats to their personhood.

In line with the findings of review 1, the factors that were found to improve the experience of care for people living with dementia in review 2 (experiences of interventions) were centred on relieving the intense fear and insecurity that can result from experiences of disorientation in hospital, and interventions in review 2 demonstrated a number of approaches that did this. These approaches included training staff, volunteers and/or carers to identify unmet need in response to the emotional distress of people living with dementia, and that encouraged them to interact with people living with dementia, particularly around learning personal likes and dislikes and respecting dignity and personhood. They involved adapting environments to relieve disorientation and to make it easier to establish familiarity, and encouraging interactions between people living with dementia and others. They involved organising activities that were meaningful to people living with dementia: not only were the activities a source of pleasure, but they created situations in which people living with dementia interacted with others. Participants on the whole perceived that these approaches supported security and relieved distress.

An area worthy of further attention to increase feelings of security is the benefit afforded to people living with dementia by having access to others with whom they feel an attachment, as this was mentioned in studies but not investigated in any depth.

Knowledge and experience of dementia: review 1 (experience of care)

Ten of the prioritised studies referred to the impact of staff knowledge and experience which ranged from a high level of expertise in caring for people living with dementia, to some staff who had very little or no previous professional or personal experience. Those who had past experience of caring for people living with dementia were more able to draw on their skills to interpret non-verbal cues102,123 and recognised the importance of ‘building a picture’ of the person living with dementia to inform their understanding of how to best care for that person. 66,68,70,110,123 This included recognising that responsive behaviours often signified an unmet need. 66,68,123

By contrast, many staff lacked experience or knowledge of dementia, and this could prevent good care. For example, hearing impairment was mistaken for difficulties with cognition by those staff with less experience. 107 A lack of knowledge could mean the use of inappropriate assessment tools79,107 or the use of force to complete routine tasks,108 and left staff feeling unsure about how to respond to individual behaviours of people living with dementia. 120 Not having specific knowledge of older people and their needs was described as ‘contributing to widening the gap between real and ideal care, making nurses “act without a map and compass” ’. 108 A couple of studies reported that staff and students did not feel equipped by their training to respond to responsive behaviours expressed by people living with dementia:107,108,123

You simply feel inadequate, I cannot, and I don’t have the knowledge. You try everything and anything and nothing seems to work. It’s like you improvise, make random long-shots, trying one thing after the other.

Nurse107

Teodorzcuk et al. 118 found that knowledge and skills gaps underpinned poor practice, and this was compounded further if colleagues modelled suboptimal practice.

Constructions about dementia: review 1 (experience of care)

Staff were observed to hold different ideas about dementia that could have an impact on how they provided care. Some staff understood that people living with dementia might have unmet needs that they expressed with responsive behaviours,66,68,123 while other staff thought that such behaviours were the result of neurological impairment,104 or interpreted these as people living with dementia being awkward or disruptive. 81,114 Linked to this, there were also different assumptions made about the ability to engage with people living with dementia. Staff in some studies spoke about the fact that people living with dementia could not communicate or ‘give back’, leading to dissatisfaction in care provision,67,68,71 and people living with dementia could end up being regarded as low priority compared with others on the ward and not offered the same level or aspects of care. 68,71,104,127 By contrast, some staff acknowledged that people living with dementia could share in acts of affection and humour,81,122 and could reciprocate. 123

Values/personal philosophies about care: review 1 (experience of care)

Different views and philosophies of care could have an impact on the care that staff provided to people living with dementia. The values of individual staff were closely connected to job satisfaction: when they were able to meet their self-expectations for providing care, their job satisfaction was high. 108 Although values are an individual staff characteristic, because of the close link to job satisfaction we discuss values further in Staff emotional well-being.

Characteristics of individual staff: review 2 (experience of interventions)

Intervention studies in review 2 made little reference to individual staff characteristics; however, two studies noted the impact when staff had personal experience of people living with dementia,139,143 and one study139 discussed how a lack of personal experience had an impact on student nurses. An intervention that supported junior doctors to become dementia champions143 noted that experience of dementia was important in determining whether or not a junior doctor volunteered to be a dementia champion. The link to the personal meant that the aims of the intervention were perceived to be particularly valuable, and acted to fuel the enthusiasm needed to volunteer for and to take initiative in the project:

. . . my granny had it . . . I remember when she was in hospital . . . Mum panicking . . . ‘Will anyone help her eating?’ . . . I could just imagine . . . how much nicer it would be for our family at home, to know that there was . . . such a scheme in the hospital.

Junior doctor (author and reviewer edits)143

Naughton et al. 139 found that student nurses who had known a person living with dementia previously were more resilient in the face of responsive behaviour.

Workplace structures and routines: review 1 (experience of care)

Workplaces were often structured for the efficient completion of physical caregiving tasks. Although an understandable approach, this could prevent the kind of interactions between people living with dementia and staff required for PCC because insufficient staffing numbers67,87,102 and high rotation of staff led to brief encounters between people living with dementia and individual staff, particularly senior staff. 107 Not only did this prevent staff from getting to know people living with dementia because of a lack of time, but, as a result, staff could feel reluctant to engage with people living with dementia:

We rarely have the same patients for very long, instead we are moved between different units . . . sometimes you tend to think that I’m only to have this patient for one day, and then you don’t get so involved.

Nurse107

Staff perceived that caring for people living with dementia required more time and that lack of time was a key reason why good care did not always happen. 67,87,102,104,107,108,120

Workplace structures and routines: review 2 (experience of interventions)

Despite interventions to improve care, hospital staff, volunteers and student nurses still perceived that there was a lack of time and resources to provide PCC to people living with dementia. 23,116,133,136,138,139 Even in the intervention involving a special care unit23 where staffing levels had been increased, because all of the patients had cognitive impairment staff perceived that caseloads were still too high.

Two intervention studies136,141 involving staff training found that low staffing levels and lack of time prevented implementation of the interventions. The intervention reported by Smyth et al. 141 was designed to be delivered on-ward so that it would be more relevant; however, it proved impossible to free up small groups of staff to be taught together on the wards. The time needed for one-on-one teaching was considerably more than had been anticipated and rendered the intervention unfeasible. In Horner et al. ,136 staff engagement was not high, despite staff being given the choice to complete training in hard copy or online. Only 6 out of 26 staff members completed the training; the authors suggested that this was partly because internet access was not made available to junior staff. Although an education officer was made available on the ward to answer questions, staff rarely approached the officer. According to ward leaders, lack of time, limited resources and difficulty retaining experienced staff were the greatest challenges.

In another intervention study,133 staff perceived that new techniques, resources and structural changes to the ward could not benefit people living with dementia when there was no time to put them to use:

You just need to give time to them (patients with dementia) and unless we have enough staff . . . let’s be practical how can we? Sometimes you just end up frustrated.

Health-care assistant (author edits)133

By contrast, in five studies, staff talked about how interventions provided capacity that met the needs of people living with dementia and/or carers in ways that they otherwise felt unable to do owing to limited time. In the carer support group, staff felt more confident about carers’ understanding and skills, leaving them more open to carer involvement: ‘A lot of [carers] are more understanding after going [to the program] . . . so we can incorporate them into the care more’ (nurse, author edits). 135 In an intervention48 in which Alzheimer’s Society volunteers offered information and advice to carers, staff expressed appreciation for this service because it worked to meet carer needs. Staff recognised that carers needed support, but felt required to prioritise patients. Staff on a ward that made the structural change of adding access technology137 perceived that it provided people living with dementia with a protected and private room and the use of points of reference, and that it supported ‘learning trails’ because people living with dementia were unable to access any room but their own. This technology meant that people living with dementia were able to become familiar with the ward environment more quickly than before, and were more calm and so needed less support. It also supported safer wandering so that staff did not need to walk with wandering people living with dementia, thus freeing up their time. In both interventions138,144 that increased capacity on the ward by allocating volunteers to interact with people living with dementia, staff commented on the fact that volunteers freed up nurses’ time. In one of these studies,144 staff also perceived that volunteers made their jobs easier because care from volunteers resulted in reduced responsive behaviours.

Ward priorities: review 1 (experience of care)

A number of studies concluded that PCC would remain challenging as long as institutions organised themselves around the routine delivery of medical tasks. 67,94,96,107,118,122 Studies described staff who wanted to provide PCC but felt unable to do so because of ward cultures that required them to prioritise routine care. 87,94,110,122,123 Some studies found that, because of the ward culture, staff perceived that they should prioritise the needs of patients other than those living with dementia. 71,104,107 For example, from observations on a cardiology ward, Nilsson et al. 107 concluded that disease was the organising care principle, which meant that people living with dementia did not fit into the system of care in the unit:

I don’t think that older people with cognitive impairments fit in here with us . . . it’s difficult to combine cognitive impairments with acute care. And we should ask ourselves to what extent we should treat people with dementia.

Nurse (reviewer edits)107

One study104 found that concern over safety was prioritised over psychological well-being and dignity for people living with dementia. Staff in this study, establishing whether perceived confusion was due to an acute delirium or dementia, responded to acute delirium by working to reduce confusion and improve well-being, but responded to dementia by shifting the focus to simply minimising harm to the patient, co-patients and staff:

Doctors have the skills but they are too busy and these patients are at the bottom of the list. There are so many patients with competing priorities – patients with dementia are just labelled and deemed not able to be helped . . . they want to treat the acute problems.

Clinical nurse consultant (reviewer edits)104

It was also suggested that ward policy had a key role to play in fostering the provision of PCC. To give good care, staff emphasised the need to understand the preferred routines of and personal information about people living with dementia. 66,79,108,123 Some staff talked about the importance of the role of carers in this, either by providing information or tips that helped staff understand certain behaviours99,102 or by their presence alongside the person living with dementia and their ability to physically help when staff time was limited. 99,114 However, although it was often recognised that carer involvement could inform good care, it was rare for there to be a clear strategy or ward policy for involving them,104 and staff could differ in their approach to carer involvement within wards as well as across wards. 99 In a study about the use of deception with people living with dementia to manage their emotions, Turner et al. 120 noted a lack of policy to guide staff on this difficult issue.

Ward priorities: review 2 (experience of interventions)

In addition to lack of time, ward cultures were explicitly discussed as a barrier to PCC due to prioritisation of acute health needs over psychological needs in three studies. 23,134,140

Observations on a special care unit, where multiple changes had been made to increase the attention paid to psychological needs (see Report Supplementary Material 4), found that care most commonly given to people living with dementia remained focused on tasks and routine despite improvements in a number of areas. 23 It was observed in this study that allied health professionals, mental health nurses and activities co-ordinators attended to psychological needs in the activities room, but that, on the ward, although nurses sometimes interacted with people living with dementia while providing physical care, PCC remained less common than task-focused care.

Staff trained to perceive responsive behaviour by people living with dementia as unmet need140 said that they were less likely after their training to restrain people living with dementia physically or pharmacologically as a means of addressing behaviour, and were more likely to attempt to redress the reason for distress. However, participants spoke of other staff who had not received training tending to prioritise acute issues, and how this acted as a barrier to their providing more PCC.

In an intervention that provided a weekly musical activity session,134 the sessions were interrupted a number of times so that physical care could be provided to people living with dementia. Staff eventually adapted their approach by placing a ‘do not disturb’ sign on the door to prevent such interruptions.

Roles and hierarchies: review 1 (experience of care)

Another area of institutional culture explored in included studies was the impact that roles and hierarchies could have on care, which was highlighted in five studies. 66,104,118,122,123 Role in this sense referred to both the type of professional field (domestic, health-care assistant, nurse, physician, allied health professional) and the perceived hierarchies within and across these roles. Owing to the importance of the psychological well-being of people living with dementia in hospital, those who knew personal information about patients were particularly helpful in guiding decisions about their care. These could often be the people perceived as lower on the ward hierarchy – cleaners, porters, health-care assistants – but were also allied health workers (occupational therapists and physiotherapists, social workers), nurses and volunteers. A number of studies suggested that the valuable information these people had could be ignored, or they did not think that it was appropriate or they did not feel empowered to speak up, despite their expert knowledge of the person living with dementia. 66,118,120,122,123 Teodorczuk et al. 118 suggested that this could result in a feeling of powerlessness, which could then stifle practice:

I’m walking past and somebody’s sitting there and their bowl of soup is there getting more and more cold and I think you know I’m a 55 year old woman I’m capable of feeding someone a bowl of soup, I’m not allowed to do it.

Domestic118

Some roles with the capacity to meet the psychological needs of people living with dementia did not facilitate PCC. Nurses who were ‘specials’, whose role it was to keep people living with dementia safe, were not expected to interact with people living with dementia, despite spending hours next to them. 104

Roles and hierarchies (review 2: experience of interventions)

Three intervention studies138,143,144 demonstrated the importance the consideration of role could have to the success of an intervention. The issue of role was particularly important to the two studies138,144 that introduced volunteers onto wards. Authors of Wong Shee et al. 144 surmised that staff perceptions about the potential for the volunteer role to overlap their own created a reluctance for some staff to engage with volunteers. Nurses asked that the volunteers not wear the same colour shirts in order to differentiate between them, and expressed reluctance to share information about patients. Volunteers in the study felt they needed to know more about the functional abilities of people living with dementia to be able to interact better with them and to choose activities. They experienced reluctance by staff to share information as a barrier to fulfilling their roles, and thought that by involving staff in the earlier stages of the intervention training, some of the problems could have been prevented. By contrast, staff in the study by McDonnell et al. 138 did not perceive a threat to their role. Over time, as the volunteer role was clarified, staff differentiated the role of volunteers as focused on ‘chat’ and activities with people living with dementia. Common to both studies was a process where volunteers gained confidence as they became familiar with the ward routines and their role within them.

Junior doctors who became dementia champions perceived a change in their role as they learned more about the needs of people living with dementia, from a more administrative role focused on establishing physical history to a more holistic role concerned with the lives of people living with dementia and their carers. The authors identified the adoption of the role of dementia champion, combined with permission from ward leaders to take this role, as key to the doctors’ ability to make such changes to practice. The doctors talked of being role models to other doctors and nurses, as they were asked for advice about caring for people living with dementia:

. . . other people ask me about it . . . at the moment it’s only me in orthopaedics out of us four [junior doctors], and the others, they do all their own stu? themselves but if they don’t know something they’ll say ‘‘you’re the dementia champion, should I be doing this?’’

Junior doctor (reviewer edits)143

The doctors also valued the role because it created an opportunity for them to develop leadership skills and to report on the project at conferences, offering them valuable experience and a sense of achievement.

Leadership at the ward and institutional levels was shown to be important to the success of interventions. McDonnell et al. 138 had support at strategic and ward levels, which may explain how the initial conflict around roles between nursing staff and volunteers was resolved during this intervention. In the special care unit that reallocated specialist mental health nurses to the unit,23 staff felt motivated and encouraged by the supportive and approachable leadership on the ward. However, these staff also perceived that senior management’s emphasis on measurable aspects of care demonstrated a lack of understanding of the nature of, and support for, PCC, and these perceptions undermined changes to practice.

Staff perceived that training made available across team members was an expression of concern for staff well-being and safety by administrators, despite acknowledging there were additional reasons for training:

I was really pleased that this has become more of an interest to the corporation—that they want to . . . reduce the number of restraints used in a more proactive way as opposed to just telling us that we shouldn’t be using [restraints].

Hospital staff (reviewer edits)140

These findings suggest the perception by staff that institutional- and ward-level leadership supported their professional development and changes to practice were a key factor in intervention success.

Documentation and communication: review 1 (experience of care)

A number of studies found that systems for sharing information that fostered PCC, such as personal preferences and backgrounds of people living with dementia, and individual approaches to managing responsive behaviour were non-existent or were not consistently maintained. 79,94,107,108,120 In such a case, each member of staff was required to re-establish the same information ‘from scratch at every shift’. 94

Documentation and communication: review 2 (experience of interventions)

As was described in subreview A [see Identity: Review 2 (experience of interventions)], a number of interventions involved the instigation of an information-gathering tool to record personal information about people living with dementia. However, in addition to documentation, intervention studies found that the extent to which participants engaged with one another and with others outside the intervention was important to how well the intervention was perceived and implemented. Systems of documentation that made information accessible supported communication between staff members and between staff/students/volunteers, whereas a lack of communication created misunderstanding and prevented interventions from working as intended.

In an intervention increasing ward capacity by allocating volunteers to provide company to, and support activities with, people living with dementia,144 lack of trust between hospital nursing staff and volunteers led to reluctance by staff to share patient records which were necessary to guide volunteers in providing person-centred activities. Training for volunteers included confidentiality and privacy training, and volunteers had signed confidentiality agreements, so these experiences might have been prevented had staff been adequately informed about the scheme before the intervention started. In another intervention allocating volunteers to an acute ward,138 the authors found that initially staff did not always welcome volunteers. However, staff perceptions and their communication with volunteers improved over time as staff came to understand the nature of the volunteer role. Staff who worked on the wards linked to the carer support group135 perceived that their lack of knowledge and involvement was a barrier to a programme they valued and they would have liked to know more about it. 135 These studies demonstrate the importance of familiarising existing staff with and including them in new interventions any incoming staff.

A number of studies found that the impact of staff interactions, whether positive or negative, were important to the way interventions worked. Staff expressed appreciation for staff with specialist knowledge of dementia brought in as part of interventions because of the help they were able to provide. 23,143 In an intervention providing PCC training to nursing students before placement,139 positive response from staff members was found to be important. When staff modelled the ways of communicating with people living with dementia that students had been taught, this approach was legitimised:

What I’ve realised . . . is that a lot of the nurses do this framework without realising it . . . I was watching my mentor and she just knew exactly the right thing to say . . . She could calm them down instantly but she was using things that we’ve been taught.

Student nurse (reviewer edits)139

However, when other hospital staff did not provide PCC, this acted as a barrier to those wanting to provide it. 139

Finally, five studies23,138,140,143,144 that involved staff training and/or peer learning particularly noted that a sense of team developed as a result of the intervention. In one study143 the authors found that the peer-led nature of the scheme was an important driver for junior doctors’ motivation, and that peers were easy to ask and receive support from in the early stages of the project when there was a lot to learn. These authors conclude that space for reflection, providing opportunities for development and being part of a team meant that the dementia-carer scheme at that was markedly effective. In two studies that increased ward capacity by allocating volunteers,138,144 volunteers also spoke about the importance of peer support and their ability to learn from each other, as well as being appreciative of the support offered by practitioners and/or researchers who were part of the intervention.

The ward environment (review 1: experience of care)

Staff participants from a number of studies perceived that the atmosphere in an acute care environment was not suitable for people living with dementia. The busy environment, the noise, the rapid pace, lack of staff knowledge and continuity were perceived by many to be less than ideal in relation to what these patients needed. 108,114,118,123 Staff participants also described the unsuitability of wards for fostering good care for people living with dementia because of design that prevented interaction with others, for example because people living with dementia were alone in a room, sitting alone or kept in bed. 96,99,107 The physical structure of wards communicated their purpose and focus: resources (e.g. provision of social space), equipment and furnishings were often there to promote physical care, with little provision for systems to support the sharing of knowledge about personal information of people living with dementia, or that facilitated communication and interactions between staff, carers and people living with dementia. 96,99,104,118 By contrast, one study found that a ‘home-like’ environment alone was inadequate to create the experience of being at home, but rather such an experience required personal interaction that created feelings of safety, connection and welcome. 81

The ward environment: review 2 (experience of interventions)

Two studies133,141 mentioned the way the physical layout of wards impacted an intervention. In a staff training intervention,141 lack of communal spaces made it difficult for anything but one-to-one training. The authors concluded that the inability to work in groups prevented the benefits of role modelling and group teaching. The ward layout also prevented patients from interacting and engaging in activities.

In a ward redesigned to be dementia-friendly,133 staff perceived that the non-clinical environment of the ward meant they provided better care to people living with dementia and carers, because it created a more relaxed space that supported collaboration. On this ward, one of the adaptations was to remove the central nursing station; instead nurses completed paperwork on the wards close to the patients. Most staff commented on the fact that working alongside rather than separately to patients enabled them to keep a better eye on them, making it easier to prevent falls and injuries, although one staff member perceived that distraction from people living with dementia prevented them from completing documentation as well as they had before.

Relationships with carers: review 2 (experience of interventions)

In review 1, hospital staff referred to the value of carers as a resource for understanding people living with dementia [see Ward priorities: review 1 (experience of care)], and acknowledged the need carers had for support [see Good care versus the carer: review 1 (experience of care)]. In review 2, a few interventions supported relationships between carers and staff. The special care unit included the introduction of a policy supporting carer involvement, however some staff perceived carers as demanding, and some felt that lack of time prevented them from interacting with carers. 142 Durepos et al. 135 attributed the wider success of their in-hospital carer support group to increased relationship-centred interactions from carers. Carers described feelings of significance, purpose and achievement, and staff perceived them to be more engaged in caring on the wards:

To me [the carer support group] changed the culture on the unit, the way the families are more involved . . . all the team members are always asking me about it, before that was just social work.

Support group leader135

The authors suggested that staff involvement from the beginning of the intervention might have supported more relationship-centred actions among carers, staff and people living with dementia.

Staff emotional well-being: review 1 (experience of care)

Many studies highlighted the varied and changing emotions they experienced as a result of caring for people living with dementia in the acute setting. Caring could change from moment to moment, and staff experienced happiness, laughter, compassion, tenderness, fulfilment, frustration, stress, anger, exhaustion, powerlessness, inadequacy, offence and fear. 66,67,107,108,122,123 Berg et al. 123 suggested that caring involved ‘sharing everyday life as people’ in both its positive and its negative dimensions. This meant that, in positive moments, nurses could feel physical and personal closeness, warmth, tenderness, ‘give and take’, laughter and humour that, with continuity, could engender deep relationships. At other times, it could be much more negative. This could be due to the lack of feedback from people living with dementia, the difficulties in handling the emotions of people living with dementia, communication difficulties and the challenges of coping with responsive behaviours such as physical and verbal aggression, and the ‘impossibility of caring for a number of people living with dementia at the same time’. 123

There were many descriptions among the studies from staff with different roles talking about the feeling of not having done a ‘good job’ in their care for people living with dementia, as highlighted here by one nurse:

. . . you don’t get to give as good of care because they don’t understand what you are doing and why you are doing it . . . It upsets me. It makes me feel very, I don’t know, like I’m not doing a good job. It’s stressful.

Nurse (reviewer edits)67

This extract is an example of a conflict in care, whereby the nurse perceived that communication difficulties between her and a person living with dementia prevented her from giving good care. Included studies refer to a number of types of conflicts in care, described below, and these created situations in which staff felt it impossible to provide good care, which led to moral distress and affected their emotional well-being. Although staff acknowledged that ‘good care’ for both people living with dementia and carers involved a cost to them in terms of emotional burden, nonetheless they perceived that, when they were capable of providing it, this kind of care was better for their own personal well-being. It fulfilled their values around good care, which usually involved not only physical caring but also psychological caring. Therefore, caring for people living with dementia could be fulfilling or could lead to moral distress and eventually cynicism and burnout, depending on whether or not staff were able to provide what they understood to be good care.

Staff emotional well-being: review 2 (experience of interventions)

Review 2 intervention studies addressed staff emotions, but primarily in relation to feelings of fear and/or inadequacy due to a lack of knowledge and skills in how to care for people living with dementia, which is discussed below as a conflict in care between good care and the person living with dementia. The exception to this is an interesting finding from the intervention study involving access technology,137 where the authors found that access technology provided a level of physical safety because people living with dementia and others were unable to enter unauthorised spaces. Staff perceived that people living with dementia felt safer, and the authors concluded that staff themselves felt more secure because nurses did not have to worry about patients wandering off of the ward or into co-patients’ spaces. Staff also perceived that people living with dementia picked up on their mood in a way that emphasised the value of staff not feeling stressed:

. . . sometimes, when we are under-staffed, the patients all go wild on us . . . I guess they just pick up on our moods . . . So I think we just try, when we know we’re going to have a busy day, to relax and project a feeling of security.

Hospital staff (reviewer edits)137

This suggests the possibility for positive and negative spirals of interaction between staff and people living with dementia, whereby responsive behaviour by people living with dementia creates feelings of stress in staff, who then de-prioritise PCC, which increases responsive behaviour. The quotation above suggests the opposite possibility also holds, that a sense of calm and security in staff can communicate itself to people living with dementia, increasing their feelings of security and reducing responsive behaviours, thus supporting feelings of calm in staff.

Good care versus the institution/role: review 1 (experience of care)

As highlighted earlier (see Institutional and ward cultures and environments), ratios and ward expectations of staff often underestimated the complex care that people living with dementia needed, and meant that staff sometimes experienced conflict in the care they were able to provide due to conflicting priorities66,94,122 and/or not having enough time:67,71,108

I’m starting to take care of my patients the way the hospital is dictating to me to take care of them because that is the way it is. Inside that doesn’t feel good, it angers me and I can’t change it.

Nurse67

Nurses and health-care assistants in particular were seen to experience the conflict between providing ‘physical care’ (cure) and ‘emotional care’ (care). 114,122 One nurse who had previously worked in a nursing home surmised:

Juggling responsibilities is a challenge – hospitals are about cure rather than care. Here we cure, in the nursing home we cared.

Nurse114

Bailey et al. 122 characterised this experience as an implicit and often unacknowledged conflict between opposing discourses: one around the nature of medical care and one around the nature of PCC. Another study found that the personal and professional integrity of nurses was often compromised in caring for people living with dementia as, despite having a greater need for time and attention, staff faced being unable to meet the needs of people living with dementia because they had very limited possibilities to do so:108

It is eating me away not getting the time and peace to be present; you know you are not doing a good job.

Nurse108

Good care versus the institution/role: review 2 (experience of interventions)

As was true for review 1, staff participants in a number of intervention studies experienced conflict between providing physical care and PCC, where it did not seem possible to provide both, despite interventions to improve the experience of care for people living with dementia. None of the studies framed these experiences in terms of staff emotional or moral distress as was done in review 1 studies; however, the description of situations is similar, and therefore fits this conceptual framework.

Participants perceived that understaffing and lack of time prevented staff and/or students from being able to connect with patients, leaving study participants feeling helpless and disillusioned in the face of systemic barriers. 23,116,139 Student nurses trained specifically to communicate with people living with dementia139 described a range of care provided by hospital staff to people living with dementia on the wards, from excellent practice to staff who were less engaged:

People who do that [ignore people living with dementia], I ask myself have they always done that or have they just been around these people so long that they’ve just gotten used to it and it’s just easy for them now to do it that way and just ignore them.

Student nurse139

Student nurses said that such experiences tainted their views about working with people living with dementia in future, despite largely enjoying their placement experiences.

Good care versus the person living with dementia: review 1 (experience of care)

For some, conflict arose from their interactions and expectations about people living with dementia. Perceptions around communication issues could result in staff resorting to physical force to complete required tasks, or to walk away from the person living with dementia as soon as the ‘task’ was complete. 70,108 Even with understanding that responsive behaviour likely represented an unmet need, such behaviours could be quite challenging for staff and impact on their ability to deliver the care they wanted to give:108,114

You become so frustrated that you have to leave the room, it feels like you cannot cope with this, it is too difficult when you are pinched, hit or have your hair pulled. I have certainly walked out of showers and felt ‘no way, someone else needs to take over.’

Nurse108

Some new nursing students described feeling intense distress in the face of responsive behaviours of people living with dementia:

This sounds awful but I used to be terrified of patients who would scream because I’ve never experienced that. I’d be so scared to go in the room and try and calm them down because I’d think they’d start getting aggressive.

Student nurse139

Good care versus the person living with dementia: review 2 (experience of interventions)

Intervention studies described changes in approach to providing care to people living with dementia by staff, students and volunteers following training and new experiences of providing care on wards.

Before receiving training about PCC, staff reported feeling uncertain about how to support people living with dementia clinically; for example, they prioritised staff and patient safety over PCC and used directive language with people living with dementia. 140 This was echoed in other studies; for example in a standard care unit compared with a special care unit,23 carers felt that the discomfort of people living with dementia was exacerbated by staff who did not understand how to care for them:

[Health-care assistant] kept shouting at him, turn over, turn over I can’t get to you. So eventually I opened the curtains and said that man’s confused he can’t understand you. [The health-care assistant] knew I was sitting outside the curtain and it didn’t deter her, she was really shouting.

Wife of person living with dementia, author and reviewer edits23

Such experiences served to highlight the changes to practice that staff described following training. Following their learning experiences, many staff, students and volunteers expressed greater confidence in their ability to understand the needs of people living with dementia and to provide care for them even in the face of responsive behaviour. 23,139,141,143,144 However, not all experiences of training were straightforwardly helpful, depending on the experience of the care providers. One study described a process of emotional distress developing into personal resilience as staff learned to care for people living with dementia. When nursing students felt that they were able to make a connection with people living with dementia, this increased their feelings of self-efficacy; when student nurses responded to distress but were unable to make a connection, some were left feeling uncertain and inadequate. However, student nurses who had experience of interacting with people living with dementia were often more accepting of experiences of disconnection, and were not left feeling inadequate:

You just let them, because you can’t change it anyhow. You try everything. They’re so in it and you just think, ‘What can you do?’

Student nurse139

The authors of this study suggested that nurses need to develop personal resilience in the face of responsive behaviour, as was already the case for the students who had experience of interacting with people living with dementia. In this same study, students caring for people living with dementia who were also aphasic, deaf or blind or spoke English as a second language felt underprepared for such complex situations, highlighting a limitation of the VERA training,139 which is centred around verbal interaction. At the same time, the students acknowledged that these experiences challenged them and that sometimes they were able to adapt their approach in positive ways. 139

Good care versus the carer: review 1 (experience of care)

Conflict for staff relating to the carer arose in three main ways. First, staff could feel disturbed because they were not able to care for the carers of people living with dementia when they wanted to. 65–67 Nurses recognised that many carers needed support, but felt that they had no time to help:

We are shortchanging the patients and we are shortchanging the patient’s family. That just doesn’t sit well with me . . . When you have got six, seven patients we don’t have the time to do what we need to do and its frustrating.

Nurse (reviewer edits)67

Second, there were tensions about whether or not to involve carers in care for people living with dementia. While some studies described staff who welcomed carer involvement and proactively sought to establish relationships with carers,66,71,122 others highlighted instances where staff could be less enthusiastic,65,122,123 and this could vary between staff members on the same ward talking about the involvement of the same carer. 99 In the study by Bailey et al. ,122 there was an ambivalence towards carers that was shared by many staff. The authors concluded that carers experienced distress because of the strange environment of the hospital and because they needed to hand over care of people living with dementia to staff. This in turn generated additional tension for staff to deal with. 122 When it came to choices about adopting artificial nutrition and hydration care at the end of life, the relationship between staff and family carers ranged from close and supportive to distant and hierarchical. The type of relationship appeared to follow from the extent that carers’ opinions about treatment differed from those of staff. 65

Finally, some nurses spoke of conflict between standing up for people living with dementia and taking the concerns of carers into account. In the next quotation, a nurse describes how she felt obliged to advocate to stop artificial nutrition and hydration because she believed that this best met the desires of people living with dementia, while at the same time she understood the carers’ desire to help:

. . . my main duty is to stand up for the patient . . . But on the other hand – and this makes it so dual – you have a supporting role towards the family . . . you experience internal conflict . . . you have to fulfil two different roles that sometimes really complicates the situation.

Nurse (author and reviewer edits)66

Although these experiences related specifically to decisions at the end of life, the potential for conflict between supporting the needs of people living with dementia and supporting the needs of family members more generally was repeated across the six studies cited in this section.

Intervention studies in review 2 involved components aiming to improve experience of care related to carers; however, they did not report any staff experiences of conflict in relation to the carer.

To lie or not to lie? Review 1 (experience of care)

Two studies highlighted the conflict staff experienced in deciding whether or not to tell the truth, and the impact that this could have on them and on people living with dementia. 94,120 In Turner et al. ’s120 study, in response to difficult questions, to manage behaviour, to provide personal care or to share medical information, staff told the truth, passed the buck, or distracted or lied to people living with dementia. Although most staff said that telling the truth was their preferred option, they also said they thought that telling the truth was inappropriate because it undermined their relationship with the person living with dementia, because of their responsibilities on the ward or because of their ethical beliefs. 120 Distracting was considered to be the best option across participants, as it also allowed staff to avoid giving upsetting information and to avoid lying. When this did not work, staff considered lying. When relatives were present, staff were more likely to tell the truth; when patients with dementia were significantly agitated, staff were more likely to lie as a result of their previous experiences of patients who had become physically aggressive.

In another study observing the delivery of oral medicines to people living with dementia on an orthopaedic ward, concealing medicine by giving it to the patient while assisting them with eating was observed to be prevalent. 94 This often followed unsuccessful attempts to give medication and was compounded by a lack of information shared between staff and between shifts. It was observed to be a contentious issue for staff and one in which the autonomy of the person living with dementia could be ignored because staff focused on the necessity of the task at hand. 94

No studies in review 2 addressed the use of deception.

Review 1 (experience of care)

Four studies67,71,108,122 highlighted the impact of providing of good care and experiencing job satisfaction. Along the continuum of care, it seemed that the more staff were able to deliver good care (i.e. PCC), the better their well-being and the better they felt about both their personal and their professional integrity:

You can go home and think I’ve done a good shift, I’ve done a good job, but you don’t actually get any satisfaction, do you know what I mean? All you can do is as I’ve just said, you’ve done a good job, you’ve done your job right, but I just love it 100 per cent.

Nurse122

This quotation demonstrates the complexity of caring for people living with dementia; this nurse referred to the fact that people living with dementia do not get better as ‘but you don’t actually get any satisfaction’. Nonetheless, when she felt that she had done a good job, she loved work ‘100%’. PCC supported job satisfaction because staff perceived that the care they were giving was of high quality. 71 The more staff were made to focus on routine or task-orientated care at the expense of not supporting the psychological well-being and autonomy of the person living with dementia, the more stressful and less satisfying their work became. Nilsson et al. 108 concluded from their observations that the further away actual care was from ideal care, the more staff felt as though they were not doing a good job. A nurse in the study by Byers et al. 67 summarised similar feelings:

You finally get off from work and you don’t really feel like you have accomplished anything. When you care for these patients, when you have run all day and you didn’t get done what you think you needed to get done to care for your patients . . . There is not any accomplishment. You can’t say, I really helped this person today.

Nurse67

Job satisfaction: review 2 (experience of interventions)

A number of intervention studies noted that PCC increased job satisfaction. 23,48,135,143,144 However, people had different experiences. Wong Shee et al. 144 interviewed two volunteers who said that they had left the project in the study because they had felt a lack of purpose when interacting with people living with dementia. The authors suggested that managing volunteers’ expectations of interacting with people living with dementia was an important part of training, for example by suggesting that non-verbal cues signal a beneficial exchange. Volunteers said that the role had provided them personal satisfaction because they felt that they were helping others, and that their self-confidence had grown as a result. Perhaps more surprisingly, hospital staff said that the work of the volunteers had increased satisfaction in their job because they were reassured that people living with dementia were engaged in the kind of social interaction and companionship that staff thought they needed. 144 The junior doctors who became dementia champions also commented on how this role gave them a sense of achievement. 143

Stressors to carers: review 1 (experience of care)

Caring for people living with dementia at home was demanding, but it was particularly stressful in the time leading up to their hospital admission because of the illness or injury, and carers were physically and emotionally exhausted by the time admission took place. 96 During hospitalisation, rather than having a break from care, carers found the disruption to roles and routines and time needed to travel back and forth to hospital stressful. 96 Some carers described feeling welcome on the ward;86,96 however, others felt excluded as a result of not being welcomed or involved in care,99,107 not being given information about the health of the person living with dementia unless they took the initiative of asking,96,107 being spoken to curtly or impatiently71 and having restrictive visiting hours. 96,99,118 By contrast, when staff welcomed carers and kept them informed, carers felt reassured and involved:86,96

They are community types not a type apart. When you walk in there you don’t kind of creep in and look all round the place. You walk straight in and ‘Hi how are you doing’ and they treat me the same as her.

Husband of person living with dementia86

Kelley et al. 99 found that lack of communication with carers by staff, restricted visiting hours and meal policies, and the way that people living with dementia were often kept in bed or sitting down so their mobility was hidden meant that carers lost touch with the physical and emotional health of people living with dementia while they were in hospital. This made it harder for them to contribute to decisions about, and to cope after, discharge.

In a study exploring respite care in hospital, Gilmour et al. 86 observed that some nurses were willing to attempt to maintain home caring routines, and that the carer was positioned by these nurses as an expert about the personal information needed to care well for the person living with dementia. The presence or absence of such behaviour by staff was a crucial factor in whether or not carers were able to relinquish their caring responsibilities to staff and receive respite. When nurses did not seek to understand how care was given at home, carers felt the need to visit regularly to check on the care being given, rather than receive respite. Arrangements for discharge was an issue staff more commonly asked carers questions about,96,102 and some carers reported carefully planned discharges,96 whereas others described a lack of consultation and poor planning.

All three studies that focused on carer experiences interviewed carers after people living with dementia had been discharged home. 86,96,99 Carers reported increased confusion,99 reduced food/drink intake and/or weight loss,86,99 decreased mobility86,96,99 and constipation,86 which they attributed to the hospital admission:

. . . walking, if you don’t use it you lose it . . . and then I go into hospital and she would be sitting in bed and I would think ‘oh all that hard work and two days in bed for her and that’s the length of the hall gone’.

Wife of person living with dementia (reviewer edits)86

Such deterioration left carers wary about future admissions to hospital. 86,96 Carers also reported that these changes added to the levels of care that people living with dementia needed on arrival home,86,99 and in one study a person living with dementia experienced termination of social services while they were in hospital for 2 weeks, which left the carer to make new arrangements and required the person living with dementia to become accustomed to new community caregivers. 96

Stressors to carers: review 2 (experience of interventions)

As in review 1, review 2 intervention studies found that caring at home was stressful, and that hospitals could introduce additional stressors for carers. 23 As was also found in review 1, carers in review 2 were surprised that they were not asked for personal information about people living with dementia, which left them feeling neglected and ignored, and added to existing anxieties:

I did have to ask to find out what was going on, and I know the ward was busy and you don’t want to interfere with people when they’re working sort of, but . . . when you’re feeling that anxious, you just want that little bit more reassurance.

Carer23

One intervention in particular,135 which sought to improve the experience of care for carers through an in-hospital support group, was found to reduce stress by fostering an increased feeling of inclusion, improved carer well-being, acceptance of and increased knowledge about dementia, and increased capacity to cope with caring but also with bereavement:

[People outside the] don’t want to know that I’m not fine . . . that drives them away . . . [In the] they are living it and they want to hear.

Carer (author edits)135

Stressors for those in the support group were also reduced because the members helped each other:

. . . it’s like a pipe-line . . . [group members] will tell me how my wife is doing when I’m not there . . . [which] takes a lot of pressure off me.

Carer (author edits)135

The peer support initiated during hospital-based carer support group meetings was found to transfer to the wards as carers got to know each other, and carers reported that these peer relationships were then found to support better relationships between carers and staff:

[I felt] . . . closeness . . . [and] rapport with the nursing staff and with the social workers and the coordinators on the floors and the doctors.

Carer (author and reviewer edits)135

Carers also reported improved relationships between themselves and other people living with dementia on the ward because of new relationships with each other.

Despite such benefits, carers also spoke of difficulties associated with the support group, including dwelling on negative aspects of their situation and feeling frustrated with group dynamics and conflicting views. In addition, attendance at the support group was not possible for everyone. The authors interviewed two carers who chose not to become involved because they had other dependants at home, either children or a parent, and they both worked. The authors concluded that holding the group during daytime hours could be a barrier to involvement, particularly of younger carers.

One other intervention,48 offering information and advice from Alzheimer’s Society volunteers on the hospital site, mentioned reduction of stress. Carers said that they appreciated receiving information about dementia, and also that it was valuable for them to make contact with others to whom they could turn in times of stress. One carer said that talking to the Alzheimer’s Society volunteers had ‘taken a load off’. 48

Unique potential of carers to facilitate person-centred care in hospital: review 1 (experience of care)

Prioritised studies from review 1 that focused on carers found that carers were in a unique position to provide emotional support and the kind of personal information staff needed to provide better care in relation to the special vulnerabilities of people living with dementia. This ability at times led carers to either provide helpful information to staff about care or provide care on the ward themselves. However, the extent to which staff embraced this potential to improve care varied within and across wards, and many carers described experiences of having their expert knowledge ignored79,86,96,99 and/or their willingness to provide help refused. 87,96,99

Carers needed to be interacted with as individuals. Kelley et al. 99 found that carers varied in their knowledge of caring for people living with dementia and in their willingness to include people living with dementia in decisions and, therefore, in their ability to interact with people living with dementia and staff in constructive ways:

If they have reached crisis point . . . you can see their irritation levels with that person are obviously very high . . . it’s not beneficial for anybody when they are irate with each other in the day room.

Hospital staff (author edits)99

Most carer participants in included studies were devoted, conscientious and knowledgeable about the personal lives and care routines of people living with dementia who had been admitted to hospital, but this may have been partly due to self-selection bias, and it is important to acknowledge, as the quotation above suggests, that not all carers had these characteristics. An internal audit in a prioritised study exploring processes of decision-making around discharge found that carers were available only half the time when staff sought to discuss discharge with them, that some were older and had dementia themselves, and that family conflicts, domestic violence and economic hardship created complications for some carers. 102 These findings suggest that it is important to acknowledge that some people living with dementia did not have carers to offer the kind of support described below, or that their carers were not able to provide such support. However, as the authors of one study exploring carer experiences emphasised,99 carers who had less capacity to provide PCC were not a reason for less carer involvement, but, instead, involving them was an opportunity for support and upskilling. A few studies concluded that carers would benefit from hospital staff supporting and informing them according to their needs. 69,99

Close relationships: review 1 (experience of care)

The advantage carers had because of their intimate relationships with people living with dementia aligns with Kitwood’s concept of ‘attachment’, which is addressed less often in studies focused on people living with dementia than other aspects of PCC (see Subreview A: findings from Reviews 1 and 2 about the experience of care for people living with dementia in hospital – feeling afraid and insecure). The studies focused on carers suggest that this is an area worth further exploration because of the benefits to experiences of care that close relationships were found to potentially offer carers, people living with dementia and staff. Carers were often able to fulfil a particularly valuable role by supporting the fundamental need of people living with dementia to establish a feeling of security in hospital:

The trolleys really are side by side so you really haven’t got much room at all . . . I stroked her hair and made sure that she was alright.

Daughter of person living with dementia (author edits)69

People living with dementia could be more amenable to receiving care from their carers than from staff, and some studies described carers taking on some of the work of health-care practitioners:87,96,99

They used to ring me up . . . ‘She won’t take it’ . . . So I used to go down and I used to give her the medication . . . When they wanted to wash and change her . . . again they used to have to ask me.

Carer (author and reviewer edits)99

Some carers expressed support for staff96 and tried to help them by spending time with people living with dementia in order to reduce demands and offering support to co-patients:

. . . the problem with him . . . was that he wouldn’t sit still . . . he was up and down the ward walking . . . I think [the nursing staff] found this quite troubling. So if I could sit with him and try and get him to stay put that was something for them.

Wife of person living with dementia114

However, different individuals felt differently about becoming involved in care; for example, some carers said that they needed a break from providing care while people living with dementia were hospitalised for illness/injury, but they helped on the ward because they thought it was expected of them. 96 This corresponds to the need Kelley et al. 99 identified for staff to interact with carers as individuals.

Close relationships: review 2 (experience of interventions)

Three interventions facilitated the carers’ ability to maintain close relationships with people living with dementia. 23,135,137 The in-hospital support group135 provided therapeutic support to carers for their relationships with people living with dementia, but it was also found to support the development of relationships with staff and other people living with dementia on the ward. Carers spoke of feeling empowered by these group meetings, and this led them to be more active in navigating the hospital system:

These gentlemen [physicians] are professionals, uh and, in my day and age you always took a step back . . . but . . . I’ve learned . . . it’s ok to talk with staff . . . the group says go ahead and uh that’s how sometimes problems are solved.

Carer (author edits)135

Although the aims of this support group (i.e. to increase carers’ well-being and their knowledge about and skills for coping with dementia) seemed to be met, the impact was greater than on the individual carer, and, through the relationships that were formed, staff and people living with dementia across the ward benefited.

The special care unit23 had a policy of involving carers on the ward, and carers expressed appreciation when they were included in the care of people living with dementia. Finally, the ward that was structurally adapted to have access technology meant that carers could be given access cards that allowed them to visit people living with dementia 24 hours per day. 137

Expert knowledge: review 1 (experience of care)

In subreviews A and B it was identified that information about the personal preferences, background and mannerisms of people living with dementia supported the provision of PCC in hospital. Included studies exploring the experiences of carers 86,96,99,102 found that, because of their regular provision of care to people living with dementia, carers were often experts in such knowledge, but this knowledge was not always utilised.

Carers expressed surprise that staff did not routinely ask for information about people living with dementia considering the difficulties that people living with dementia have with memory, making sense of the present, and communicating:86,96

I would have thought with dad having dementia that the first time somebody went onto that ward . . . there was a system in place where they came and asked you things, because they must know as well as we know that dementia patients doesn’t remember things.

Daughter of person living with dementia (reviewer edits)96

Personal information was important for staff to deliver care:

I had explained to them about her meals . . . Just give her bread, no butter, and jam . . . a cup of tea . . . But nobody would listen . . . and then they are getting upset because she’s not eating.

Carer99

Some carers described staff who sought personal information about people living with dementia from carers, and valued it:

. . . they have asked the right questions before going, so have had a little history there, um, and something to work on while he has been with them.

Son of person living with dementia (author edits)86

However, staff often were not aware of personal histories and likes and dislikes,114 and, even when carers made them aware of these things, it was not uncommon for carers to perceive that information about care routines was not used by hospital staff. 86,99 Personalised information from carers supported staff to make connections with patients and to provide meaningful activities, 99 as well as helping them to interpret what the statements or actions by people living with dementia meant. 99,102

Carers sometimes worked quite hard to continue existing routines, for example by providing daily newspapers, favourite clothing or personal items that were important to people living with dementia,99 and these supported the development of familiarity and security in the hospital environment, as well as creating prompts for conversation for those who knew the person living with dementia less well, such as staff members. Carers also brought items that related to interests or hobbies meaningful to the person living with dementia,99 supporting greater occupation while in hospital.

The three studies that focused on carers all found that acknowledging the value of personal knowledge held by carers about how to care for people living with dementia was key to positive carer experiences, as well as to improved care for people living with dementia.

Expert knowledge: review 2 (experience of interventions)

In review 1, it was found that carers sometimes needed support with and upskilling in dementia, as well as having expert knowledge about the preferences and personal histories of people living with dementia. Two intervention studies referred to these issues: one involved upskilling carers135 and one instigated a system for recording the preferences and personal histories of people living with dementia. 23

The in-hospital support group135 developed carers’ understanding about dementia and helped their ability to accept it:

I think the presentations gave us some insight in what to expect . . . like when I got a phone call that my wife was choking . . . it was another stage but I was prepared for it.

Carer (author edits)135

The special care unit23 adopted the use of a personal information tool on which carers could record information about people living with dementia that might be useful on the ward. In discussions with carers about whether staff had engaged with them about patients’ backgrounds and interests, positive and negative comments were noted from the respondents. Half of carers (n = 10) on the special care unit commented that they had been approached by staff to complete the personal information tool:

I filled one form in I answered, you know, her interests, what she enjoyed doing, I do think it’s a good idea. The girl [nurse] that gave me the form said it was, to help them understand the person, to get to know the lady in the bed.

Carer23

Carers’ perceptions about the quality of care: review 1 (experience of care)

Participant carers were very concerned about the quality of the care that people living with dementia received in hospital. 86,96,99 It could be difficult for carers to ascertain care quality because people living with dementia were not always able to reliably describe care. 96 Carers responded by visiting at different times of the day, asking questions of staff, advocating for people living with dementia and working to fill in the gaps in care that they perceived. 86,96,99

Carers judged quality of care in different ways according to individual expectations, with these following from their knowledge of the person living with dementia and their previous experiences of care. 86,96 In general, they expected individualised care that met both the physical and the psychological needs of the person living with dementia, for example to ensure that the person was well fed, hydrated and clean and that staff took care over safety, interacted personably and kindly, dealt competently with disorientated and/or responsive behaviour, and provided appropriate medical care. 96 Carers expressed appreciation for the care that staff gave people living with dementia,86,96 but dissatisfaction with care was common. For example, in one study,96 every participant talked about at least one issue with which they were dissatisfied. Carers perceived a lack of attention to personal interaction, failure to maintain personal cleanliness, failure to include people living with dementia socially and an inability to manage disorientation and distressed behaviour as a sign of low levels of staff competence. 96

When noticing shortcomings, some carers rationalised them, for example by attributing poor communication to stress or inexperience of staff. 96,99 Some carers sympathised with the situation staff faced, and worked to support them, whereas other carers simply judged that care was poor. 69 When a judgement of poor care was made, carers felt angry and frustrated, and when carers experienced end-of-life care that they perceived to be lacking, their feelings of anger and injustice were particularly intense. 96 Jurgens et al. 96 conceptualised this as a process during which carers’ views developed over repeated incidents, and carers who began by rationalising staff behaviour eventually blamed staff for poor outcomes when these outcomes continued. 96

Perceptions about the quality of care: review 2 (experience of interventions)

Studies of interventions in review 2, as was the case with review 1, found that carers were highly concerned about the kind of care that people living with dementia received. In a study evaluating a special care unit,23 the authors also interviewed carers of people living with dementia on standard acute wards to provide a comparison. Some carers perceived care to be good and some perceived care to be lacking on both wards, concerning knowledge of staff about dementia; occupation for and activities of people living with dementia on the wards; maintenance of personal cleanliness; the ward environment; and communication between staff and carers. The carers of people living with dementia on the special care unit responded generally with more satisfaction about all of these issues, except communication between staff and carers, which most carers from both groups desired to be better. When asked how their experiences might have been improved, similar numbers of carers (special care unit, n = 9; standard care unit, n = 10) suggested improvements to communication. For example:

I would like it if [staff] came and introduced themselves. So if they haven’t seen you before, then you’re sat by your mother’s bed, they should come over and say, well, I’m the ward sister, or I’m the daily nurse who’s looking after her.

Son of person living with dementia on special care unit (author and reviewer edits)23

Staff on the special care unit but not on the standard care units were provided with training about caring for people living with dementia, and specialist capacity was added by bringing mental health nurses and activity co-ordinators to the special care unit, suggesting that the training that staff received and access they had to role models may have had a positive impact on experiences of care for people living with dementia and their carers.

Two carers interviewed in a study that provided training to staff136 said that they evaluated care according to the quality of physical care and by whether staff treated people living with dementia as people or as medical problems, and that the way staff interacted with people living with dementia, including providing reassurance and re-orientation, was important in their judgements about whether or not the care was of good quality: ‘There is no substitute for a listening ear and a comforting and accepting presence for the patient. It shows an important level of humanity’ (carer of person living with dementia). 136 With parallels with findings from the special care unit, the two carers interviewed in the study by Horner et al. 136 had different experiences of care, one positive and the other negative, and the negative experience included inappropriate communication.

Carer perceptions were not always wholly negative or positive; sometimes the same carer had positive and negative experiences of communicating with different staff members:

We saw [the consultant], who was excellent, he was informative, he was helpful, he was sympathetic, but there was one nurse that came across as abrasive and therefore you’re a bit wary about asking too many questions, but the auxiliary nurse was lovely.

Son of person living with dementia23

Carers of people living with dementia on the special care unit23 noted a lack of communication from staff about discharge. Around half of the carer participants described discharge as delayed, hurried and undignified. Negative evaluations about the quality of care provided to people living with dementia was an additional stressor for carers. 116

The experience of training for hospital staff, students and volunteers

Staff who had knowledge about caring for people living with dementia built a picture of how best to care for individual needs,66,68,96,110,123 including interpreting non-verbal cues,102,123 and perceived that responsive behaviour from people living with dementia represented unmet needs. 66,68,123 By contrast, staff who were not knowledgeable about caring for people living with dementia were much less able to interpret and respond constructively to behaviour. 79,107,108,118,120,123 In such situations, restraint through physical or pharmacological means was more common. 104 The evidence suggests that training for all staff – including health-care assistants, cleaners and porters [see Roles and hierarchies: review 1 (experience of care)] – is needed to enable the delivery of PCC.

The experience of ward cultures

Elements of ward culture identified in review 1 as problematic for staff who wanted to provide PCC to people living with dementia included the prioritisation of task-focused/routine, physical and/or acute care, a lack of continuity of care, and perceptions that care for people living with dementia was less important than care for other patients.

Many studies exploring the experience of people living with dementia described their distress because of emotional and/or physical discomfort (see Comfort). Hospital staff often felt prevented from providing psychological support by institutional targets and ward cultures that prioritised physical care, routines and risk management. 67,87,99,102,104,107,108,118,120 Even when staff understood responsive behaviour as unmet needs, many perceived that they did not have time to respond to it with PCC because task-focused/routine care took all the time they had. Hospital routines tended to be prioritised over the personal preferences of people living with dementia, for example when staff woke individuals to serve them, did not defer to food choices, continued to call a person the name on their patient notes regardless of the name that the person used, the giving of medications, personal hygiene and/or refusing to provide drinks outside the ‘drinks round’. 87,94,96,110,122,127

Hospital staff spoke of their emotional distress from witnessing the realities of dementia, including the responsive behaviour of people living with dementia,66,67,107,108,122,123 and experienced moral distress when they were unable to provide good care. 67,94,96,108,122,125,150 Our conclusion from review 1 was that to complete the physical care that is crucial to a caring role, a person-centred approach is also crucial. It alleviates the psychological distress of people living with dementia;68,80,81,87,94,96,102,107,108,110,118,122,127 prevents ‘difficult’ behaviour as much as it is possible to do so,81,87,94,96,102,104,107,108,110,118,122,123,127 freeing staff to do their work; and represents closely the values nurses describe that relate to good care, thereby reducing moral distress and fostering job satisfaction. 67,96,108,122 Although conceptually the argument was made that PCC takes less time because it prevents the escalation of responsive behaviours, staff perceived that PCC took longer. Further studies are needed to explore the extent to which psychological care and usual task-/routine-focused care can be provided together; and whether or not PCC does, in fact, take longer.

Because the evidence suggests that the core of the problem for people living with dementia in hospital is the intense fear and insecurity they experience as a result of disorientation, continuity wherever possible is an aspect of ward culture that would be important to foster. One approach suggested in review 1 studies is to support familiarity through relationships [see Experiences of disorientation and responsive behaviour: review 1 (experience of care)].

Hospital staff in a number of prioritised studies from review 1 perceived that care for people living with dementia was of lower priority than care for other patients on the ward, and that people living with dementia were not offered the same level of care as other patients. 68,71,104,127 Some staff who lacked training perceived that people living with dementia could not be helped, or, for some of those who did have the skills to help them, people living with dementia were ‘at the bottom of the list’ of those who were helped. 104 Such perceived low status may underlie some of the perceptions staff had that there was no time to meet the needs of people living with dementia; it may have been perceived that there were more important things to be done. 104 When adequate resources were unavailable to staff to enable them to meet the needs of people living with dementia, as was often the case in the included studies, staff implicitly received from hospital management a message that the priority of the needs of people living with dementia was indeed lower.

We conclude that increased recognition of care for dementia through investment in training and workforce capacities would be needed to transform ward cultures that currently do not prioritise care for dementia. Such changes are likely to improve job satisfaction for staff, who want to provide good care to people living with dementia but feel helpless to do so in the face of acute ward cultures that prioritise task-/routine-focused care.

People living with dementia outcomes

Two studies in the tailored management strategies174 and following care protocol171 subcategories assessed the comfort and symptom control of people living with dementia (see Appendix 15, Table 19). The study of an intervention introducing an individualised care plan to control discomfort174 in patients with dementia/delirium found a small, yet, according to the study authors, not clinically or statistically significant, reduction in discomfort levels from admission (time 1) to 24 hours before discharge (time 2) (p = 0.58). Although patients receiving the care plan had significantly less discomfort at time 2 than at time 1 (p = 0.041), the difference did not persist when time 1 discomfort scores were controlled for (p = 0.075). The cluster RCT171 assessed the effectiveness of a care guide programme at improving the comfort of and quality of end-of-life care for older adults with dementia and other conditions, with three outcomes. The condition of the care recipient during the dying process was assessed separately by nurses and carers. There was evidence that the implementation of a care guide had a beneficial effect on patient comfort while dying compared with standard care when assessed by nurses (mean difference in comfort, baseline-adjusted: 4.30, 95% CI 2.07 to 6.53; p < 0.001) but not when assessed by carers (p = 0.82). Findings for the measurement of symptom management assessed by both groups did not show a significant difference in symptom control between the groups. However, nurse assessments using another measure showed that the intervention group had significantly better scores on the Palliative Care Outcome Scale than the control group (mean difference in Palliative Care Outcome Scale, baseline-adjusted: –2.62, 95% CI –4.96 to –0.71; p = 0.009), indicating fewer symptoms and care needs in the last 3 days of life. 171

Levels of agitation were reported in one study,173 the trial involving training staff in individualised non-pharmacological strategies to manage challenging behaviour. There was no evidence to support the effectiveness of the intervention on agitation, and both the intervention and the control groups improved over time (Cohen’s d = –0.28, 95% CI –0.75 to 0.19; p = 0.24).

Given the small samples and the potentially different dementia stages targeted in the studies described above, there is currently limited evidence of the effectiveness of staff interventions in improving comfort and symptom control or behaviour of people living with dementia in hospital.

Carer outcomes

Three studies assessed the impact of interventions under the following care protocol171 and the tailored management strategies172,173 subcategories on carers (see Appendix 16, Table 20). Two randomised studies171,173 assessed satisfaction with care provided to patients. In the intervention implementing a care guide programme for end-of-life care,171 there was evidence for significantly less satisfaction with care in the intervention group than in the control group post intervention (reported mean difference in satisfaction scores, baseline-adjusted –4.00, 95% CI –7.87 to –0.12, p = 0.04; Cohen’s d = –0.74), although wide CIs indicate some uncertainty about this effect. In the study by Mador et al. ,173 staff training in individualised non-pharmacological strategies to manage challenging behaviour did not translate into improved carer satisfaction with the nursing and overall care their relative received. A closely linked outcome was assessed by Luxford et al. 172 around carers’ satisfaction with clinicians’ communication. Although there were insufficient data to calculate effect sizes, authors reported that when the TOP 5 tool was implemented carers showed higher satisfaction ratings than when carer evaluations of hospital admissions were undertaken without the TOP 5 tool in place (p < 0.05). Quality of communication between staff, carers and patients was also an outcome of the care guide programme for end-of-life care intervention. 171 This study concluded that there was a lack of evidence that the care guide programme was effective in improving communication between staff and patients or communication between staff and carers.

Staff outcomes

Eleven studies in this category described interventions to improve staff information, knowledge and skills, and assessed staff-related outcomes (see Appendix 17, Table 21). Confidence in providing care was the most commonly assessed outcome, reported in nine studies (three controlled139–141 and six uncontrolled before-and-after studies165–169,172). Differences in study design and insufficient data for calculating effect sizes meant that studies in the ‘confidence’ outcome category could not be meta-analysed (i.e. uncontrolled before-and-after studies did not always provide the paired pre–post comparisons needed for meta-analysis) and these are therefore synthesised narratively below. The controlled before-and-after studies provided mixed evidence: one study140 reported a large effect size on staff confidence in delivering PCC after an educational programme following the gentle persuasive approaches principle (Cohen’s d = 1.30, 95% CI 1.14 to 1.46; p < 0.0001). However, in the other two controlled studies there was a lack of evidence for a beneficial effect of training programmes for staff141 and nursing students139 on their confidence in providing care. Additionally, training based on the VERA framework to develop nursing students’ dementia communication skills139 had a positive effect on their ability to identify person-centred responses (Cohen’s d = 1.21, 95% CI 0.26 to 2.15; p = 0.01), although wide CIs indicate some uncertainty of this effect. In the same study, training did not increase confidence in dementia communication,139 as assessed with a bespoke questionnaire (Cohen’s d = –0.10, 95% CI –0.99 to 0.80; p = 0.83).

Of the six uncontrolled studies, three167,168,172 provided evidence that staff had higher levels of confidence in providing care after attending educational programmes to better care for people living with dementia, with small to medium effect sizes (see Appendix 17, Table 21). A fourth study169 assessing the impact of two levels of training in PCC (foundation level, half-day; intermediate level, 3 days’ training) on a small number of staff found that caring efficacy significantly changed after staff completed the intermediate-level training compared with baseline (Cohen’s d = 0.92, 95% CI 0.25 to 1.58; p = 0.01), but not after the foundation-level training (Cohen’s d = 0.27, 95% CI –0.25 to 0.79; p = 0.30). The remaining two studies assessing confidence did not provide sufficient data for effect sizes to be calculated. In the study by Asomaning et al. 165 the authors reported a non-significant increase in self-efficacy scores at 6 months after an educational programme to enhance staff members’ ability to care for dementia patients with challenging behaviours. During pilot testing of the ‘dementia-friendly hospitals’ programme,166 staff’s confidence in their ability to care for patients with dementia was evaluated in four community hospitals. The post-programme overall confidence level was significantly higher than the level at baseline (p < 0.001), and at 4 months the authors reported that confidence scores remained stable in three of the hospitals but had significantly dropped compared with the immediate post-programme scores in one of the hospitals (p = 0.02). It should be noted that the study had very low response rates (14%),166 which limit the generalisability of the findings. Overall, mixed evidence from nine non-randomised studies suggests that dementia training/educational programmes designed to promote to better care for people living with dementia may be linked to a moderate short-term increase in staff confidence in providing care.

Two studies evaluated the impact of two-level PCC training169 and a brief psychosocial141 training intervention on staff attitudes towards people living with dementia. Although staff attitudes improved over time, the significant change was already evident after the half-day foundation training in the two-level training study169 (baseline to foundation level Cohen’s d = 0.79, 95% CI 0.24 to 1.31, p = 0.002; foundation to intermediate level Cohen’s d = 0.56, 95% CI –0.16 to 1.28, p = 0.12). Inadequate reporting meant that we could not calculate the effect sizes of the findings of the brief psychosocial training intervention, but the authors reported positive trends yet inconclusive results for attitudes, which can probably be attributed to the relatively small sample size. 141 Mean measures across groups for burnout, assessed using the Maslach Burnout Inventory, in the same study141 indicated low scores but inadequate reporting, and an unsuccessful attempt to obtain additional information from the corresponding author rendered the results inconclusive.

Satisfaction in caring was an additional outcome of the two-level training for hospital staff. 169 Our pairwise comparisons indicated that basic training did not lead to significant positive changes (mean difference baseline-foundation level 0.2, 95% CI –0.05 to 0.45; p = 0.11), but the intermediate-level training over 3 days was found to have a beneficial effect on staff satisfaction in caring (mean difference baseline-intermediate level 0.34, 95% CI 0.03 to 0.65, p = 0.03; Cohen’s d = 0.82, 95% CI 0.07 to 1.57). However, the study authors169 reported a significant main effect of time (p < 0.001) and estimates suggesting significant improvements in satisfaction at both training levels. A second study173 measured satisfaction in caring, testing the effectiveness of staff training on individualised non-pharmacological strategies to manage challenging behaviour. Although not clearly reported, the findings indicated no evidence of an effect of the intervention on nursing staff satisfaction measured at discharge (p = 0.50). There was also a lack of evidence of an intervention effect on appropriateness of psychotropic medication prescribing (intervention 91% appropriate vs. control 100%; p = 0.06) and total daily doses of antipsychotics (p = 0.82) or benzodiazepines (p = 0.73) administered in the same study. 173 In a second study172 assessing medication prescribing, the introduction of the TOP 5 communication tool was followed by a significant reduction in use of antipsychotics in a major metropolitan hospital, and a decrease in the use of Risperidal Quicklets (Janssen-Cilag Ltd, High Wycombe, UK) in a principal referral hospital (p < 0.10).

Communication among clinical staff was evaluated in the cluster RCT assessing the effectiveness of a care guide programme at the end of life. 171 Based on nurses’ assessments, doctors of patients in the intervention group were more likely to be informed about the impending death of the patient than those in the control group (odds ratio 2.51, 95% CI 1.06 to 5.95; p = 0.04), although the wide CIs add some uncertainty to that estimate. There was a lack of evidence of an effect on other communication items, such as whether the doctor or other care staff were contacted after a patient’s death (see Appendix 17, Table 21).

Overall, studies indicate a moderate increase in staff confidence in providing care, at least in the short term, following dementia training/educational programmes. Evidence for other staff-related outcomes including attitudes towards people living with dementia, satisfaction in caring, well-being, staff communication and medication prescribing is poor and examined in a small number of studies.

Effectiveness of activity-based interventions for people living with dementia

Six studies134,175,177,179,180 evaluated the effectiveness of activity-based interventions in patients with a dementia diagnosis or mild to moderate cognitive impairment. 176 Behavioural disturbances were clearly reported as the reason for admission in three of the studies. 176,177,179 Despite including different activities or programme structure, all six studies were essentially driven by the idea that behavioural problems of people living with dementia represent unmet emotional or social needs, and cultural or social activities have the potential to address these needs by increasing engagement or re-engaging people in their environment and meaningful activities, promoting communication with and connection to others, and improving well-being. Interventions included 30-minute music therapy134,175 and art viewing,180 or multiactivity programmes such as a number of individualised social activities,176 a tailored activity programme177 starting from goal-oriented to repetitive and more passive, sensory-based activities, and participation in a psychodynamic therapeutic community for 6 hours per day. 179 The comparator in the only included RCT176 was usual care, which shared components with the intervention group except the individualisation of social activities. The remaining studies used baseline pre-intervention time points as the comparator.

Sample sizes across studies were small, and participant numbers varied depending on the outcome or subscale measured, ranging from 4 to 76. Five out of the six studies had ‘weak’ global quality ratings, primarily based on poor data collection methods, low response rates post intervention and the absence of assessor and/or participant blinding. Six studies measured a number of outcomes for people living with dementia that can be collectively described as ‘aspects of well-being’, including quality of life,176,180 well-being,180 patient engagement175,177,179,180 and emotional state/mood. 134,175,177 Two of the studies assessed the impact of such interventions on staff-related outcomes. 134,177 Differences in study design or insufficient data to calculate effect sizes meant that studies under the same outcome category could not be meta-analysed. However, effect sizes were calculated to aid the interpretation of findings when sufficient data were available.

People living with dementia outcomes

Appendix 19, Table 23, shows the means and SDs for control/pre-intervention and intervention/post-intervention groups for each aspect of well-being outcome (where these were reported or could be estimated). One study180 assessed the impact of a visual arts programme on eight domains of well-being measured by the research team using an observation tool and compared with an unstructured social activity not involving art. The study was conducted across three settings, one of which was NHS hospital wards providing care to 23 participants with dementia. Although some of the assessed domain scores improved at the 2-week and 3-month time points compared with the baseline activity scores (e.g. attention, sadness, disengagement, negative affect), overall there was a lack of evidence to support a beneficial effect of the programme on well-being. Quality of life was also assessed in the same programme180 along with a second study trialling the effectiveness of individualised social activities176 in older adults with cognitive impairment. Neither of the interventions was found to be effective at improving proxy-180 or self-reported176,180 quality of life.

Emotional states of people living with dementia were assessed in three studies. The authors of the music therapy study175 reported significantly higher frequency of positive mood ratings (general alertness and pleasure; p = 0.01) and lower observances of negative mood states (anxiety, anger, sadness; p = 0.045) during music sessions than during sessions without music. However, only a small number of participants were observed (n = 25), and the study did not provide additional details about the reasons for admission or other conditions that the patients may have been exposed to while on the unit. Gitlin et al. 177 assessed the same emotional states in 15 people living with dementia, comparing observations during a baseline standardised activity with the TAP-H intervention sessions. There were insufficient data to calculate effect sizes as the authors compared only the average percentage of time that participants engaged in behaviours: patients showed increased pleasure, and decreased alertness and negative mood states in intervention sessions compared with baseline. In the third study,134 observational data indicated that participants’ happiness scores increased by the end of each participatory music-making session, and the impact on engagement, distraction and relaxation was also consistently positive (statistical comparisons not reported).

Four studies175,177,179,180 provided evidence regarding the effect of activity-based interventions on patient engagement. In the TAP-H intervention,177 patients showed increased positive gestures but decreased positive statements compared with baseline behaviours. A decrease in negative statements and non-verbal behaviours was also observed (e.g. repetitive statements, verbal aggression, motoric or facial disturbances). Increased constructive and passive engagement (e.g. motor or verbal behaviours in response to the activity) and decreased self- or non-engagement (e.g. purposeless behaviour involving engagement with self, staring into space) was also observed during music sessions compared with sessions without music in the study evaluating a creative music therapy intervention. 175 A third study assessing the impact of a psychotherapeutic day hospital programme179 using a time series design found that the intervention was associated with better clinical progress in group therapy across the different time points (β = 2.01; p = 0.044). There was a lack of evidence of a positive impact of the visual arts programme180 on communication, measured with a scale covering a range of behaviours related to engagement such as conversation, awareness, pleasure, humour and responsiveness. Patients’ communication actually deteriorated between baseline and the 3- and 6-month time points, with study authors reporting significantly more difficulties in communication. 180 Calculated effect sizes also indicate a detrimental effect on communication (3 months Cohen’s d = 0.64, 95% CI –0.05 to 1.34, p = 0.07; 6 months Cohen’s d = 0.76, 95% CI –0.06 to 1.58, p = 0.06), but the wide CIs suggest imprecision of the effect estimate, possibly because of the small number of participants (see Appendix 19, Table 23).

Behavioural outcomes were assessed in three studies. The trial examining the effectiveness of individualised social activities176 showed lower scores on a scale measuring BPSD but there was no significant difference between groups post intervention (Cohen’s d = –0.45, 95% CI –0.99 to 0.11; p = 0.11). However, a psychotherapeutic day hospital programme179 including music, movement, psychodynamic and sociotherapy was associated with statistically significant reduction in neuropsychiatric symptoms across time points from admission to discharge (linear regression β = –4.21; p < 0.001), particularly anxiety and apathy. Observational data at the start and the end of the participatory music-making intervention134 indicated consistently positive effects, with reduced agitation among participants, although the authors did not provide additional comparative data.

Overall, there were mixed trends of low-quality evidence of the effectiveness of activity-based interventions to improve the experience of care for people living with dementia as reflected by aspects of well-being measures during their stay in hospital settings (see Appendix 19, Table 23).

Staff outcomes

In terms of staff-related outcomes assessed in two activity-based intervention studies, the TAP-H intervention177 reported an improvement in certified nursing assistants’ readiness to use tailored strategies during the course of the programme. In the participatory music-making intervention by Daykin et al. ,134 a reduction in staff absences was observed for the 2-month period with music sessions compared with the 2-month period without music sessions on the ward. However, the authors reported slightly more staff absences on the actual day of the music activity between the two periods. Nevertheless, the reported information is too limited to allow us to estimate what other factors could have contributed to these differences. Ward-level data from the same intervention showed an approximately 4% decrease in the number of patients prescribed antipsychotic drugs during the intervention time period (time B) compared with the usual care period (time A). A 28% decrease in the number of antipsychotic drugs was also observed on the day of the music activity (Tuesday, time B) compared with time A. However, the number of patients taking antipsychotics during their stay was greater at time B. 134 Insufficient reporting, lack of detail about measurement tools used and lack of comparable data in both studies has compromised the quality and applicability of these findings (see Appendix 20, Table 24).

People living with dementia outcomes

One RCT163 assessed the effectiveness of a specialist medical and mental health unit (MMHU) for people living with dementia/delirium compared with acute general or geriatric medical wards on a number of outcomes with mixed findings for the hospital experience. Based on direct observations, MMHU patients were more often in a positive mood or engaged (79% vs. 68%; p = 0.03) and less often in a negative mood or disengaged (11% vs. 20%; p = 0.05) than those in standard care. They also spent more time in an active state (82% vs. 74%, p = 0.10) and engaging in social interactions (47% vs. 39%; p = 0.06), but these between-group differences were not statistically significant. There was evidence that MMHU patients experienced more staff interactions that met their psychological and emotional needs (personal enhancers Cohen’s d = 0.75, 95% CI 0.31 to 1.17; p < 0.001), although the CIs for a reduction in personal detractors included negligible effects (Cohen’s d = –0.43, 95% CI –0.84 to –0.007; p = 0.08). There was a lack of evidence of an improvement in quality of life (self- or proxy-reported) measured with two different tools in the same trial. 163 However, the controlled study assessing the impact of the Care for Acute Mentally Infirm Elders (CAMIE) dementia unit150 reported beneficial effects for quality of life and well-being compared with a conventional geriatric ward. Quality of life (Cohen’s d = 0.66, 95% CI 0.36 to 0.96; p < 0.001) and well-being (Cohen’s d = 1.31, 95% CI 0.99 to 1.62; p < 0.001) increased during the hospital stay in the dementia unit patients compared with control patients, whereas ill-being scores decreased (Cohen’s d = –1.06, 95% CI –1.37 to –0.75; p < 0.001).

The quality and quantity of staff–resident interactions were measured in the study by Skea and Lindesay178 comparing two dementia care units (unit 1, a community hospital ward with enhanced version of traditional care; unit 2, a partnership scheme prioritising social care philosophy emphasising patient choice and independence) with a long-stay mental health hospital ward for people living with dementia. A series of observations was performed and interactions were coded in unit 1 at 12 and 24 months, in unit 2 at 6 and 12 months, and in the comparison mental health ward at baseline. There were insufficient data to calculate effect sizes; however, the authors reported a number of statistically significant differences between the units. There was a significantly larger number of ‘positive social’ interactions in unit 1 at 12 months than at the long-stay hospital ward (p < 0.001), and, after encouragement to improve further, the total number of interactions in unit 1 also increased significantly at 24 months (p < 0.001), mainly owing to increased ‘positive care’ interactions. 178 ‘Positive social’ interactions were also increased compared with baseline (p < 0.05). In unit 2, ‘positive care’, ‘positive social’ and total number of interactions were higher at 6 months than they were in the long-stay ward at baseline, and were further increased at 12 months (all p < 0.001). Between-unit comparisons showed a larger number of ‘positive social’, ‘positive care’, ‘neutral’ and total interactions in unit 2 at 12 months. Low rates of negative interactions were observed in both units. 178 Overall, quality of life was better in both units than in the long-stay hospital ward, and patients in the partnership scheme unit (unit 2) had a better quality of life than those in the community hospital (unit 1) (see Appendix 21, Table 25).

Patients with advanced Alzheimer’s disease cared for in a special unit using a palliative care approach were compared with those in a traditional long-term care unit in the study by Volicer et al. 182 The outcome of interest was discomfort assessed over 3 months. A large effect size was found in this study (Cohen’s d = –0.95, 95% CI –1.31 to –0.59; p < 0.001), indicating a positive impact of the dementia special care unit philosophy on patient comfort.

Three studies reported behavioural outcomes. The RCT assessing the effectiveness of the MMHU found no statistically significant difference in Neuropsychiatric Inventory total scores between MMHU patients and those in standard care at 90 days. 163 The study by Skea and Lindesay178 comparing two dementia care unit approaches measured ‘aggressivity’ as part of a behaviour rating scale. At baseline, ‘aggressivity’ was higher in patients in the unit that emphasised patient choice (unit 2) than in patients in the ‘enhanced’ traditional care hospital unit (unit 1), and slightly increased over the study period, although no significant differences were reported between units. By contrast, patients in the controlled study assessing the impact of the CAMIE dementia unit150 demonstrated significantly lower levels of agitation post intervention and compared with patients in the conventional geriatric ward (Cohen’s d = –1.26, 95% CI –1.57 to –0.94; p < 0.001) (see Appendix 21, Table 25).

Carer outcomes

One study163 in this intervention category evaluated the effectiveness of the MMHU on three outcomes for carers (see Appendix 22, Table 26). Carer well-being and strain were assessed 90 days after randomisation. There was no evidence of a beneficial effect on the carers of patients randomised to the MMHU compared with the carers of those in standard care (well-being Cohen’s d = 0.03; 95% CI –0.22 to 0.28, p = 0.81; strain Cohen’s d = –0.03; 95% CI –0.28 to 0.22, p = 0.48). Carers in the special care unit group were significantly more satisfied than those in standard care with overall care and specific care dimensions such as feeding and nutrition, the unit meeting confused patients’ needs, treating patients with dignity and respect, and discharge arrangements. Despite that, and as noted by the authors,163 both groups had a large number of unsatisfied carers. Dissatisfaction in the standard care group was twice as high for several care dimensions, including feeding and nutrition, treating patients with dignity and respect and meeting patients’ needs.

Staff outcomes

The study by Skea and Lindesay,178 comparing a special care unit emphasising resident choice and independence with a unit providing enhanced traditional care, reported two staff-related outcomes (see Appendix 23, Table 27). Job satisfaction in both units had improved at 12 months compared with baseline. Satisfaction scores in unit 2 (partnership scheme) were higher than those in unit 1 (community hospital ward), but the large effect size had wide CIs, including negligible effects (unit 2 vs. unit 1 at 12 months: Cohen’s d = 0.82, 95% CI –0.06 to 1.69; p = 0.06). Although there were not enough data to calculate effect sizes, the authors reported a non-significant decrease in staff well-being scores (measured using the General Health Questionnaire) in both units at 12 months compared with baseline. 178

People living with dementia outcomes

The comfort of people living with dementia was assessed by proxies using the Symptom Management at the End of Life in Dementia Scale; there was no evidence of a beneficial effect of the advanced dementia consult service on patient comfort (Cohen’s d = –0.01, 95% CI –0.97 to 0.95; p = 0.98) (see Appendix 24, Table 28). 181

Carer outcomes

Carer satisfaction with care was measured with the Satisfaction with Care at the End-of-Life in Dementia Scale, and communication with hospital providers was measured using the Quality of Communication Scale (see Appendix 24, Table 28). Both carer outcomes scores increased after the advanced dementia consult service was tested, but there was a lack of evidence to support a significant change (satisfaction Cohen’s d = 0.18; 95% CI –0.79 to 1.14, p = 0.71; communication Cohen’s d = 0.16; 95% CI –0.80 to 1.13, p = 0.74). These findings are not surprising as the sample size was very small and therefore the study did not have adequate power to detect differences in outcomes between groups. 181