A randomised controlled trial to evaluate the impact of a human rights based approach to dementia care in inpatient ward and care home settings.

Dementia in society

There are currently > 850,000 people living with a diagnosis of dementia in the UK. 1 As well as the biological changes associated with a dementia process, which can lead to a range of cognitive difficulties, dementia is associated with numerous psychological and social consequences. Sixty-one per cent of people living with dementia have reported feeling anxious or depressed, 40% have reported feeling lonely and just over one-third do not feel part of their community. 1 This poses a major threat to the quality of life of a large number of people in our society and is in direct conflict with the National Dementia Strategy, the aim of which is to help people with dementia to ‘live well with dementia’. 2 The cost of dementia to the UK each year is estimated to be £26B. 3 We live in an ageing population and the issues associated with dementia will continue to increase. The status quo is unsustainable. Providing good-quality care to people with dementia will continue to be a concern over the coming years.

In 2007, the Alzheimer’s Society published Dementia UK. 4 This report stated that:

Dementia must be made a publicly stated national health and social care priority. This must be reflected in plans for service development and public spending.

Dementia UK. 4 © Alzheimer’s Society 2007

In 2009, Living Well With Dementia: A National Dementia Strategy2 was published, outlining the government’s plan for providing quality services in dementia care. Dementia has been highlighted as a government priority. In 2015, the prime minister launched a programme of work5 that aims to deliver major improvements in dementia care and research by 2020. The focus of this is improving the services provided for people with dementia, with the view that England could become the best country in the world for dementia care and support, for people with dementia, their carers and families to live, and for undertaking research into dementia.

Dementia is widely feared in society6 and, traditionally, people with dementia have been among the most devalued, experiencing the double stigma of old age and cognitive impairment. Kitwood7 suggested that personhood (i.e. the state of being a person) is bestowed on us by the treatment of others. The stigma and misperceptions surrounding dementia and the resulting reactions of people towards those living with dementia have led to care practices that can undermine the humanity and personhood of an individual with dementia. 7 The literature highlights issues such as removing all choice and personal autonomy from people with dementia,8 restraint, and restrictions on ‘wandering’. It is clear that there are occasions when human rights for people with dementia are unnecessarily limited and their application is not routinely considered in clinical decision-making. 9 It is essential that approaches are adopted that maintain the humanity of an individual and challenge the stigma associated with dementia that people often report feeling.

Human rights

Human rights are brought into UK law through the Human Rights Act. 10 They represent the fundamental ways in which a person can expect to be treated simply by virtue of being human. Although human rights are based on values held for centuries, they became formalised following the atrocities of the second world war, in particular the Holocaust. It was acknowledged that human beings can inflict dreadful suffering on each other and that explicit statements on our rights as human citizens were required. The articles of the Human Rights Act are broad-ranging, covering physical, psychological and social issues, and they represent the minimum standard of treatment that we should expect. Table 1 outlines the articles of the Human Rights Act.

The United Nations adopted the Universal Declaration of Human Rights in 1948. 11 The European Convention on Human Rights,12 created in 1950 by the Council of Europe, was the first post-war attempt to unify Europe and institutionalise the shared values of democracy, human rights and the rule of law. The UK was among the first states to ratify the Convention and British jurists were highly influential in its design. The Human Rights Act10 incorporates most of the Convention rights into UK law. It came into force across the UK in October 2000.

Human rights law, including the rights composing the Human Rights Act, can be understood through the FREDA (Fairness, Respect, Equality, Dignity and Autonomy) principles. 13 The FREDA principles are not law in and of themselves. They are the values that run through the rights protected by the Human Rights Act and are at the heart of high-quality health and social care.

Human rights in health care

Where, after all, do human rights begin? In small places, close to home – so close and so small that they cannot be seen on any map of the world. Yet they are the world of the individual person . . .

Eleanor Roosevelt, 195814

Human rights are diagnosis neutral and compel us to treat everyone as human beings regardless of the difficulties they may be experiencing. They also recognise, however, that, in certain complex cases, a balance may need to be struck in order to meet competing rights of different individuals or to protect an individual from unwarranted risk, and rights may need to be limited. A human rights based approach describes the process of using the articles of the Human Rights Act in a very practical way to influence daily life. 15 A human rights based approach to care both allows for that balance to be considered and provides a lens through which such difficult decisions can be made. Failure to take the human rights of the service user into account can also lead to legal suits that impose an additional financial burden and undermine public confidence in services. 16,17 The NHS Constitution for England states that the NHS:

. . . has a duty to each and every individual that it serves and must respect their human rights.

Department of Health. 18 © Crown copyright. Contains public sector information licensed under the Open Government Licence v3.0

Not only is it unlawful for NHS organisations to work in a way that is incompatible with human rights, but the application of a human rights based approach establishes minimum standards of care that help to safeguard individuals, particularly those who are vulnerable. They also remind us that individuals require a great deal more than safeguarding in order to maintain their self-respect and sense of dignity. The culture of organisations has led, on occasion, to staff delivering task-orientated, risk-averse care that fails to consider the human rights of an individual. 16 Human rights, in this context, can therefore be viewed as codifications of how relationships can be understood and the social obligations we hold as human beings. 19

The Human Rights Act10 is law; however, in health-care settings, it needs to be translated into a clear set of principles that guide everyday practice, bridging the gap between the legal system and good-quality health care. 20 Human Rights in Healthcare21 achieves this translation by outlining the key ingredients of a human rights based approach. An alternative but similar construction is found within the PANEL principles. 8 The PANEL principles are participation, accountability, non-discriminatory, empowerment and legality, and these are defined more fully in Table 2. They represent the guiding principles for organisations to follow to maximise the chances of the services they deliver aligning with a human rights based approach.

Making the link between law and ethical practice is not the only step required. There is also a need to translate the concepts of a human rights based approach into practical strategies that can facilitate the everyday decision-making of staff; in other words, there is a need to make ‘choices guided by values’22 and by the more practical elements of a human rights based approach, such as proportionality (i.e. responding to situations in a way that is appropriate in magnitude and degree), the fit with other legal frameworks such as the Mental Health Act23 and Mental Capacity Act,24 proactive strategies (predicting responses to events through knowledge of the person and responding before a negative event occurs) and balancing rights and risks to make sensible decisions.

The disability model of dementia

Discussions around dementia and the difficulties it causes to individuals have historically been dominated by a medicalised notion of dementia, in which there is no cure and nothing can be done other than watch the person decline. 25 More recent social movements to recognise dementia as a disability26 have opened up opportunities to frame dementia within a rights based approach. The United Nations’ Convention on the Rights of Persons with Disabilities27 aims to ensure that, ultimately, people living with a disability do not experience discrimination and that their rights are maintained and promoted. The UK has ratified the Convention, which means that all UK laws and policies should be compliant with it. 28 As a result, people living with dementia should be able to utilise the Convention to protect and promote their rights.

Human rights and dementia

Although there is still limited empirical work being carried out specifically in the area of dementia and human rights, the last few years have seen an expansion of this topic as an area of focus. Several charters of human rights have been produced29,30 that aim to influence policy related to dementia. Literature also exists considering some of the major issues that may threaten an individual’s human rights. Laird31 has provided examples of how fundamental human rights can be violated in health-care settings:

[S]ituations cited by British Institute of Human Rights include failure to change soiled bed sheets, neglect leading to pressure ulcer development, not helping people to eat when they are too frail to eat themselves, excessive force used to restrain people and washing or dressing people without regard to dignity.

Laird,31 p. 6

It is notable, however, that the majority of publications are focused at a policy level32 or are discussion papers reviewing a concept33 as opposed to attempts to apply a human rights based approach in practice and evaluate its effectiveness. In 2016, Dementia Alliance International launched The Human Rights of People Living with Dementia: From Rhetoric to Reality. 34 Although this was a move to ensure that people living with dementia are aware of their rights, it stopped short of outlining the specific applications of a human rights based approach. In addition, the Dementia Engagement and Empowerment Project (DEEP) has worked alongside people with dementia to produce Our Dementia, Our Rights. 35 Although this was a real attempt to raise the issue of rights in the collective minds of society, and to produce a document in an accessible format, it did not evaluate the impact that the practical application of human rights law could have on the day-to-day lives of people living with dementia.

Since the work of Tom Kitwood,7 it has been widely accepted that person-centred principles are important in the provision of high-quality dementia care. These principles have, however, been criticised for being vague and difficult to research and enforce. 36 There are high levels of congruence between the fundamental principles of person-centred care and a human rights based approach, such as empowerment and inclusion. 37 A human rights based approach gives backbone and a legal framework to person-centred principles,37 potentially making them clearer to operationalise and more accessible to rigorous research.

Human rights training

Although there are various models of training to promote human rights awareness,38 there is limited evidence of their efficacy in terms of behavioural change. 39 These models include:

• values and awareness model – this focuses on transmitting basic knowledge of human rights

• accountability model – this assumes that participants will already be involved in the protection of individual and group rights and focuses on professional responsibility in relation to this

• transformational model – this is geared towards empowering individuals who have previously experienced human rights abuses to both recognise human rights abuses and commit to their prevention.

The suggested common themes in these models are fostering and enhancing leadership, coalition and alliance development and personal empowerment. 38

Attitudinal change for staff through human rights awareness training may be more effective when there is emphasis on staff’s emotional responses and defences and the impact of organisational culture. 39 Reflections on rights awareness training in both dementia and intellectual disabilities services suggests that change might be achieved through placing ethical decision-making centrally. This has been termed ‘dilemma-based learning’. 37

Human rights evaluation

The need to evaluate human rights initiatives is often overlooked and there is no real consensus about how to evaluate them. 40 It has been argued that the evaluation of human rights based approaches is problematic for a number of reasons, including a belief that legal concepts should be monitored rather than evaluated, a fear that evaluation will lead to legal ramifications and a distaste for quantifying the extent of human misery and abuse when rights are not being upheld. 37

Donald40 provides a clear framework for evaluating human rights based approaches in health-care services. This framework encourages the exploration of human rights knowledge and understanding; skills in applying human rights based approaches; attitudes, perspectives and values; and, ultimately, the outcome and impact of applying the approach for the realisation of human rights. There is an argument that this is more palatable, as it allows researchers to directly assess the process and impact of the human rights based approaches rather than attempting to quantify abuses.

Cultures of care

It may be comforting to assume that the human rights of the most vulnerable people in our society are routinely upheld and promoted by those tasked with caring for them. Unfortunately, the sad truth is that this is not always the case. The Francis report,41 arising from the lack of care provided at Mid-Staffordshire NHS Foundation Trust, highlighted the importance of creating the:

. . . right culture of care . . .

Francis. 41 © Crown copyright. Contains public sector information licensed under the Open Government Licence v3.0

to ensure that people are treated in ways that promote dignity and respect.

The Care Quality Commission (CQC) routinely uncovers practices that threaten the human rights of people living with dementia. For example, in one care home, inspectors noted that many residents stayed in bed all day for no apparent reason. When the inspectors questioned staff about this practice, they were told, ‘One side [of the house] we get up Monday, Wednesday and Friday. The other side we get up Tuesday, Thursday, Saturday’. 42 This is obviously unacceptable and in direct conflict with the principles of the Human Rights Act. 10

When considering the moral imperative we all hold to protect the vulnerable, it has been highlighted that ‘compassion is the basis for all morality’. 43 If we want to develop cultures of care in which person-centred care is a reality, then it has been suggested that ‘the NHS must be a fertile soil for meaningful caring relationships’. 44 The work of Buber and Smith45 encouraged viewing relationships as ‘I-Thou’, thereby engaging on a human-to-human level with the people we provide care and support for, as opposed to ‘I-It’, which adopts a detached task-orientated approach whereby people are viewed as jobs to be done and tasks to be completed. It has been suggested that, in many care settings, ‘the gap between the rhetoric and the reality remains uncomfortably wide’46 when we are considering models of person-centred care. There is no obligation to carry out person-centred care other than knowing that it is the right thing to do. With their statutory weight, human rights approaches can strengthen person-centred approaches47 and maximise the chances that they will be adopted.

Training and care

The training currently given to care providers does not automatically feel congruent with the aim of producing compassionate, person-centred cultures of care. It is recognised that there are major failings in the training of staff who provide care, particularly those who work in the care home sector. The CQC found that, of those care homes told to improve after a visit, 71% had significant training gaps, with dementia care, safeguarding and the Mental Capacity Act faring worst. 48 This is particularly worrying given that > 70% of people who are residents of care homes are living with dementia,1 and that the very fact that a person resides in a care home increases the chances that there will be queries about their capacity.

In providing training we are assuming that we are equipping people to make complex clinical decisions on a day-to-day basis. In reality we are often training people to become task orientated and driven. Models of training that include real-life situations tend to produce better outcomes with more emotional attachment to them. 37

Care planning

The availability of a good-quality person-centred care plan does not automatically ensure that good-quality person-centred care is provided, but it does provide a template by which the standard of care can be judged. In 2017, NHS England stated that:

. . . care planning is a crucial element in delivering improved care for people living with dementia.

Reproduced from NHS England. 49 Contains public sector information licensed under the Open Government Licence v3.0

The National Institute for Health and Care Excellence (NICE)’s50 quality standard statement 4 for dementia requires that each person has a personalised care plan. There are models of good-quality care planning, such as enhanced care planning, but these are often not adopted. Traditional care planning approaches adopted in NHS services, such as the care programme approach (CPA), do not always lend themselves to the full involvement of people living with dementia because of the somewhat restrictive nature of their content and a focus on risk assessment. 51 It has been suggested that the CPA maintains:

. . . a system which too often defines people by their diagnosis and medication . . .

© The National Survivor User Network52

and

. . . finds it difficult to recognise the whole person and the unique individual . . .

© The National Survivor User Network52

Any model of quality dementia care recognises the centrality and importance of an in-depth knowledge of the person, their wishes and their preferences in providing support that is person centred and, therefore, is more likely to uphold their human rights. 53 It follows, therefore, that a good-quality care plan should be a vehicle for collating this detailed knowledge about a person and their care.

Decision-making in care settings

It is recognised that ‘making decisions that concern people’s health and quality of life creates complex ethical dilemmas, and one has to choose among alternatives’. 54 This can lead to decisions that have an impact on an individual’s human rights. For example, Robinson et al. 9 explored the area of balancing risks and rights in relation to wandering. They highlighted that staff often act in particular ways, such as having a locked-door policy, because they are worried about being viewed as negligent. The implementation of a human rights based approach may provide staff with a more comprehensive and robust framework in which decisions can be made, drawing on human rights principles, particularly proportionality, least restrictive practice and proactive strategies, rather than relying on the most risk-adverse approach. 37

Rationale

We are existing in systems where the care provided to some of the most vulnerable people in society is failing to meet their complex needs. Additionally, we are not equipping our workforces to meet these needs because of woeful lack of investment in their development.

This project will build on the existing literature exploring how the human rights of people living with dementia can be undermined and unnecessarily restricted within traditional models of care but expand the focus to look at an operationalised model of providing care that embeds a human rights based approach. The proposed intervention aimed to put human rights at the heart of care planning and service delivery. A human rights based approach was chosen as the appropriate focus for this project because not only does the NHS have a legal requirement to uphold the human rights of service users, but it is recognised that quality care is both person centred and respectful of an individual’s human rights. 55

Embedding a human rights based approach through the application of the ‘Getting It Right’ assessment tool56 and training package aimed to maximise quality of life and well-being for people with dementia and provide a framework for staff to make decisions about care within a human rights based approach, using the principles of proportionality, proactive strategies, positive risk-taking and use of least restrictive practices.

Aim

To establish whether or not the application of a human rights based approach to health care leads to significant improvements in the care and well-being of people with dementia in hospital inpatient and care home settings.

Specific objectives

1. To investigate whether or not the application of a human rights based approach to health care, as opposed to treatment as usual, leads to significant improvements in the quality of life of people with dementia in hospital inpatient and care home settings, as measured by scores on the QOL-AD scale. 57

2. To explore whether or not training on the application of a human rights based approach to health care leads to identifiable improvements in the quality of staff decision-making, as measured by vignette-based interviews with staff.

3. To explore whether training in the application of a human rights based approach to health care, and the use of the ‘Getting It Right’ assessment tool,56 as opposed to the standard care planning procedure, leads to identifiable improvements in the person-centred quality of service users’ care plans, as measured by care plan audits.

4. To explore whether the application of a human rights based approach to health care leads to changes in the well-being of family carers of people with dementia who are in hospital inpatient and care home settings, as measured by the Warwick–Edinburgh Mental Well-Being Scale (WEMWBS)58 and the Zarit Burden Interview (ZBI). 59

5. To validate a novel human rights and well-being questionnaire for dementia inpatient care.

6. To explore the costs and consequences of human rights training for staff looking after people with dementia in hospital and care home settings in terms of patient-reported well-being, care plan development, staff stress, family member well-being and overall quality of care, compared with usual patient management.

Intervention package

The intervention package being applied was a novel human rights based intervention package that had previously been piloted within the host trust (Mersey Care NHS Foundation Trust). It consisted of three linked elements.

1. A one-day training package delivered to staff from the intervention unit at a time and place that was convenient to the site. The training was delivered by co-applicant Sarah Butchard, who jointly developed the intervention package and is an experienced clinical psychologist and senior clinical teacher. It was based on dilemma-based learning, utilising clinical scenarios that commonly occur in dementia services. It incorporated both direct learning about a human rights based approach and its utility in dementia and the practical application of the human rights based assessment tool (‘Getting It Right’56).

2. The completion of ‘Getting It Right’56 (see Appendix 1), which was based on person-centred principles and on the learning from enhanced care planning. 53 The aim of using the tool was to build up a person-centred care plan that was explicitly linked to the FREDA principles. Each unit was given multiple copies of the assessment tool following the training and requested to complete the assessment with both new and existing residents on the unit. There was no stipulation made as to how many assessments needed to be competed at each unit. It was emphasised that any member of staff, not just those who were qualified, could complete the assessment with residents, and that it was more important that it was completed by someone who had a good relationship with that resident.

3. Monthly booster sessions were delivered by Sarah Butchard to address issues arising from the application of the assessment tool. Three booster sessions were offered, one per month, following the training. These adopted a consultation model and allowed staff to reflect on any difficulties they had in applying the assessment tool.

Inclusion/exclusion criteria

The inclusion criteria were broad and are outlined below in relation to both sites (clusters) and individual participants at these sites.

Setting

The research was conducted in dementia inpatient wards within NHS trusts and in care homes. Table 3 shows the sites that participated in the study and their basic characteristics.

Although the initial aim was to recruit 10 NHS wards and 10 care homes, practicalities resulted in eight NHS wards and 12 care homes being recruited. In reality, however, far more people living with dementia are care home residents than are admitted to specialist dementia wards. It is estimated that one-third of people with dementia live in care homes. 60 It is harder to obtain specific figures on those accessing specialist dementia wards, but figures for the local regions where the study was carried out suggest that only 1.5% of people living with dementia will need support on a specialist dementia inpatient ward. 61 It is, therefore, reasonable that more care homes than wards were included if the figures are to represent the population of people living with dementia.

Sample size

The sample size was based on the primary outcome measure, the QOL-AD scale,57 and on conservative figures on several parameters.

Recruitment procedure

Initial expressions of interest to be involved in the study were invited from local NHS trusts and care homes via existing networks and contacts. A decision was made to recruit initially within the north-west of England owing to logistical and financial constraints.

The research team also worked closely with the National Institute for Health Research (NIHR) ENRICH (Enabling Research in Care Homes) programme to identify care homes that identified themselves as willing to participate in research and to support care homes in being involved in the study. Initially all care homes in the north-west area that had been identified as research ready were approached and invited to take part in the study.

Characteristics of sites

The sites recruited varied in terms of their size, current levels of occupation and proportion of residents living with dementia. Table 4 outlines these characteristics at both baseline and follow-up. It is evident from these figures that even if a care home was not branded as exclusively for people living with dementia, a high proportion of residents were living with this condition.

Informed consent

Obtaining informed consent is always an ethical dilemma when working in dementia care, and particularly when working with people in the later stages of dementia. The very fact that people are living in care homes or on a dementia ward means that they are likely to be in the later stages. The team acknowledged that people in the later stages, and particularly those without carers, are vulnerable to potential abuses of rights, and, therefore, it was important that they were included in the study. Every attempt was made to obtain informed consent from every potential participant, in line with the Mental Capacity Act. 24 Experienced clinical staff assessed the capacity of each potential participant, in line with best practice in research governance and the recommendations of the Mental Capacity Act,24 and individuals gave (or withheld) consent if they were able to do this themselves. If a person was not able to give informed consent, they were not asked to complete the self-report measures. Although the QOL-AD scale57 was chosen specifically because it is claimed to be suitable for people in the later stages of dementia, it was felt reasonable to assume that if a person was unable to give informed consent, then completion of the measure would be too cognitively complex for them. There was no reason for people without a family caregiver to be excluded from the study if they were able to give informed consent to participate.

When possible, when a person was unable to give consent, and therefore unable to complete the self-report measures, someone (either a family carer or a staff member) was invited to complete a proxy QOL-AD scale.

If a person was unable to give informed consent, and so they were not included in the self-report element of the study, they could still be included in DCM in cases when a nominated consultee could be identified. When it was not possible to identify a nominated consultee, or if the consultee advised against including the individual in the study, the person was not included in any aspect of the research. Informed consent was sought at both baseline and follow-up. Figure 2 presents a flow chart outlining these issues.

FIGURE 2.

Flow chart to illustrate the process of consent and participation.

Allocation concealment

A web-based system was used to cluster randomise each recruited site. A complete-list randomisation was used, meaning that an even number of sites were entered into the system and allocated, at random, half to the usual-care group and half to the training group. The result was not seen until the allocation process was complete, and it was seen only by those who had access to the system. It was also possible to provide a blinded allocation report for the people who needed to be blinded to group allocations; groups were named group 1 and group 2 rather than control and intervention.

Implementation

Following recruitment, the web-based system generated the random allocation sequence. Sites were enrolled by the trial team, specifically the trial manager, Sarah Butchard, after which the web-based system completed the assignment of sites to the control or intervention group. If a service user was willing and able to consent to participate in the study, then they were included. Consent was obtained directly from the service user if possible; if this was not possible, then a proxy, usually a member of staff, was asked to provide consent on the service user’s behalf.

Blinding

Service users, research assistants who were collecting the data and the trial statistician were blinded. Service users received daily care and did not know whether or not staff had received the training. Staff members were obviously unblinded at follow-up as they knew whether or not the unit had received the training. Research assistants attended the sites to complete assessments and did not know to which group a site had been assigned. The trial statistician was able to see the data labelled as group 1 and group 2. The unblinding of the trial results occurred at a results meeting attended by members of the independent monitoring committees.

Primary outcome measure

Secondary outcome measures

Development of outcome measures specific to the study

Issues of specificity

It is important that the outcome measures utilised allowed exploration of the specificity of the intervention in improving care and well-being over and above the application of general training. This has been addressed in a number of ways.

• The care plan audit measured the documented standard of care that a person should be receiving but also tapped into increases in human rights based language and concepts that would suggest that the human rights based nature of the intervention had an effect over and above simply providing generic training.

• The completion of human rights knowledge and attitude questionnaires measured changes in these areas pre and post training but did not look at the impact that this had on staff in their everyday working lives and how it affected service user well-being.

• Staff interviews were conducted to explore whether or not the introduction of a human rights based approach leads to differences in their decision-making processes when considering care issues. Again, this would be evaluated through the identification of key phrases and concepts in the transcripts that would specifically indicate that a human rights based approach had a direct influence on daily decision-making.

• The FREDA-based questionnaire enabled the team to explore whether or not service users felt that their human rights were respected and upheld more after the intervention.

Taken together, these elements allowed an evaluation of the proposal that the human rights based approach outlined had benefits that would not be seen by generic training.

Development of the FREDA assessment tool

Although there is recognition that violations of human rights can occur in health-care settings, little has been done to attempt to quantify the extent to which this occurs. To this end, work was undertaken to develop and begin validating a questionnaire measure based on the FREDA principles in order to assess how well individuals subjectively experience their human rights as being upheld.

The FREDA principles have been used elsewhere in health care to aid individuals’ understanding of their human rights. 70 However, the validity of these constructs has not been empirically tested. Therefore, the initial stage of this tool development was to consult with service users and their carers.

Items for the FREDA questionnaire were first generated from focus groups with people living with dementia and their carers. Participants came to one of two focus groups to discuss the care that they had received in relation to their human rights. The main aim of the focus groups was to investigate whether or not the FREDA principles adequately covered areas relevant to dementia care, along with eliciting examples of when such principles were valued or disregarded. All participants consented to the data generated by the focus groups being used in relation to the development of the human rights agenda.

People in the later stages of dementia are often excluded from consultation because of the increased communication and comprehension difficulties that can arise as the condition progresses. Given that this measure would be exploring the potential violations of an individual’s rights, it felt important that this group of people who may be vulnerable to having their human rights undermined were included in the consultation. A method developed by Kate Allen71 was utilised, which involves showing the person living with dementia a picture of an unknown person and asking them to reflect on how that person would feel in a particular situation and what advice they would give them. It is suggested that this elicits more information than asking direct questions about the treatment that they have received. This method was used on a dementia inpatient ward within Mersey Care NHS Foundation Trust and the information elicited was incorporated into the data collected from the focus groups.

Following the focus groups and ward interviews, the information was themed and statements were developed that reflected these themes. Grouping these statements together revealed four overarching themes: identity, dignity, empowerment and autonomy. The developing questionnaire was, therefore, named the IDEA questionnaire.

This resulting questionnaire was piloted with a group of people living with dementia in the community. As a result of this piloting phase, some changes were made to the structure and phrasing of items on the measure (e.g. removing any double negatives from the questions).

Missing data

There are two types of missing data possible in this data set: missing items within questionnaires at a time point and missing measures at a time point.

In the case of items missing within a questionnaire, the following approach was taken. If a missing value rule existed for a questionnaire, then this was utilised. Over and above this, if ≤ 25% of the items in a questionnaire were missing, then these were replaced with a pro-rated individual item score.

It was expected that there would be participants missing at follow-up who were present at baseline and vice versa and so the analysis model was influenced by this. The data were assessed for differences between those present at both baseline and follow-up and those present only at baseline for possible predictors to be included in the sensitivity models.

Baseline characteristics

Participant demographics, including age and gender, were reported, and split by allocated group, for baseline and follow-up. The type of dementia patients were living with was also included, when appropriate. There was no statistical comparison of the data for the two groups.

Interim analyses

No interim analyses were planned or scheduled to be completed. During the trial, no additional analyses were identified or requested by the DMEC.

Primary effectiveness analyses

The original model of analysis was planned to be a multilevel analysis of covariance model. Owing to the very nature of the wards and care homes, it was understandable that, for a number of reasons, participants present at baseline might not be present at follow-up. Therefore, a linear mixed model was used to assess the effect of time (baseline or follow-up) and group (control or intervention) and interaction of time and allocated group. The model also included site as a random effect. The main effect of interest was the group effect.

As it became evident that the ability to collect self-report data on the primary outcome measure was limited, proxy data were collected in the absence of self-report data. An additional term (self-report vs. proxy) was added to the model to assess the importance of this difference. If it was found that this term was significant, then separate analyses of self-report and proxy data were completed. This understandably affects the number of data available for the analysis and would have implications for the power of the study. The alternative was to include a self-report versus proxy and condition (group) interaction. This assesses whether or not there is a consistent difference between self-report and proxy data in both groups. The former model of analysis was chosen to allow simpler, more intuitive understanding of the data. Either way, the power of the models that could be applied would have been affected by the implications of using a mixture of proxy and self-report data.

Secondary effectiveness analyses

A linear mixed model was applied for all secondary outcome measures when appropriate. For the knowledge and attitudes questionnaire, data were collected pre and post training for the intervention group and only at baseline for the control group. This precluded the use of the linear mixed model to establish a group effect, and therefore a paired sample t-test was used to establish whether or not there was a difference in score before and after training.

Additional analyses

As indicated the significance of covariates, age, gender, DCM score, dementia type and whether or not the person had a carer were investigated by adding these to the linear mixed model.

It was also noted that one site had a different follow-up time from the other sites, with only 11 weeks in follow-up rather than the established 16 weeks. This nuance was investigated by allowing the time variable to vary for this site.

Economic analyses

Based on the Medical Research Council’s guidelines for the evaluation of complex interventions,72 our standard operating procedure for economic evaluation alongside pragmatic randomised controlled trials (RCTs) and experience in the conduct of economic evaluation alongside trials of psychosocial interventions in dementia care,63,73 we, from a public sector, multiagency perspective:74–79

• fully costed the human rights staff training programme, distinguishing between set-up/training costs and running costs, with the former amortised over 3 years

• collected service use data using an adapted Client Service Receipt Inventory, which recorded participants’ frequency of contacts with hospital services and selected community-based services at baseline and follow-up; participants’ medication usage was also recorded and service use was costed using national unit costs for the price year 2014–1580,81

• conducted a cost–consequences analysis in which the consequences included patient-reported well-being (QOL-AD57), family member well-being (WEMWBS58 and ZBI59) and overall quality of care (ASCOT).

The EQ-5D-3L was included for participants with mild to moderate dementia to allow comparison with other published studies, and with previous trials, but a cost–consequences approach rather than cost–utility analysis was undertaken because of the range of relevant outcomes spanning the person with dementia, their family members, hospital and care home staff and objective measures of care quality.

Dementia care mapping

Dementia care mapping is an observational tool. A trained observer (mapper) records the behaviours of several participants for a specified amount of time (in this case 6 hours) to gain an insight into participants’ day-to-day experience. Owing to ethical reasons, observations can take place only in communal areas. After a 5-minute period, the mapper records a Behaviour Category Code that indicates what the individual was doing. Alongside this, a mood and engagement (ME) value is recorded, indicating how engaged the individual was and if their mood was positive or negative. Table 6 summarises the definitions of each score for ME.

Dementia care mapping is an established approach that looks at person-centred care in practice. This measure was completed for all sites at both baseline and follow-up. The study focused on one aspect of DCM that records the ME levels of up to eight participants living with dementia within a 6-hour time frame. The ME score for each unit at both baseline and follow-up were recorded and compared to look for changes in the quality of care provided.

Staff interviews

Data from the staff decision-making interviews were initially analysed as one data set using thematic analysis, as outlined by Braun and Clarke. 83 An inductive, or ‘bottom-up’, approach85 to data analysis was taken. An inductive approach assumes that the themes are derived directly from the data86 as opposed to imposing the data onto a pre-existing model.

From this analysis, themes were identified related to staff decision-making strategies. Themes were not combined as fully as they would usually be in thematic analysis, as it was felt important to identify specific, rather than general, decision-making strategies in this context. The interviews were then reanalysed to identify the frequency with which these strategies were discussed in each group (i.e. intervention baseline, intervention follow-up, control baseline and control follow-up).

Sites

In total, 50 sites were contacted and offered the opportunity to be involved in the study (care homes, n = 34; NHS wards, n = 16). Twenty-eight sites declined to participate, and so 22 were randomised to either the intervention or the control arm. Between randomisation and data collection, two sites opted out of the study as they no longer wanted to be involved. These sites had previously been randomised and they withdrew because of a change of management. Of these sites, one had been randomised to the control arm and one had been randomised to the intervention arm. During data collection, two of the intervention sites did not complete training; therefore, they were classed as intent to treat (their data were still included in the intervention group’s analysis). The reason given for not completing the training at both sites was that there was not enough time or resources to release staff to attend the training. One of these sites accepted a ‘booster’ session and wanted to use the ‘Getting It Right’ assessment tool. 56 The recruitment of sites to the study is outlined in Figure 4.

FIGURE 4.

Number of sites approached to take part in and recruited to the study.

Overall recruitment

In total, the study recruited 439 people living with dementia, with 213 in the intervention arm and 226 in the control arm. These people may have completed self-report measures, had proxy measures completed, been observed in DCM or been involved in the care plan audit. They may have completed one, or more than one, element of the study. Table 8 summarises the number of aspects of the study that participants were involved in.

New participants were able to be recruited at follow-up. Sites rather than participants were randomised to receive control or intervention, and participants did not receive control or intervention. This was an expected feature of recruitment, as it is was possible that participants would not be at the same site from baseline to follow-up. Recruitment from each site, for baseline and follow-up, was based on the individuals present at the time of the visits. In addition, 245 staff members were recruited to the study.

Staff completing the training

The aim was to recruit as many staff as possible from each intervention site to the training day. In practice, this was extremely difficult, and site managers reported this to be the case even when multiple training days were offered. Table 9 summarises the number of staff trained at each intervention site and the number of booster sessions that each site accepted during the intervention (i.e. the 3 months following the training).

If the two sites that did not engage with the training are excluded, then an average of 8.88 staff per unit were trained, which equates to 28.7% of staff being trained. There was a large range in the proportion of staff trained at a particular site (11.6–52.4%).

It was initially envisaged that the booster sessions would take place every month and last 1–2 hours, and that staff who had either attended the training or who had not been able to attend the training but were completing the assessment tool could share their experiences, and that any difficulties or concerns could be addressed. This was to be similar to a consultation-style session and to be supportive in nature. In reality, only two sites accepted all three booster sessions offered. Booster sessions were most often rejected because of lack of time. They were also often not utilised in the way that was initially expected. In the majority of the sessions, the team member met with the manager of the unit alone, and the manager reflected on the assessment tool. This did not meet the aims of allowing the staff to explore difficulties that were arising and to cement the learning from the training session.

Service users completing the primary outcome measure

One hundred and forty-nine service users were able to self-complete the primary outcome (QOL-AD57) and a further 256 proxies were recruited to complete it. Figure 5 outlines the flow of participants completing the QOL-AD at each time point.

FIGURE 5.

Time points and number of service users/proxies completing the QOL-AD. a, Some participants completed only the baseline QOL-AD, some participants completed only the follow-up QOL-AD and some completed both the baseline and follow-up QOL-AD.

The most common reasons for not completing the follow-up visit after completing the baseline visit were death and being discharged from the care home or hospital ward. Differences in demographics and outcome results were tested for between the groups who died and those who did not die, and between the groups who were discharged and those who were not discharged.

The mean age of the participants who died (87.0 years, standard deviation (SD) 7.4 years] was significantly higher than the age of participants who did not die [81.4 years, SD 7.6 years; t(320) = –2.867; p = 0.004]. The QOL-AD score of participants who died (28.93, SD 5.35) was significantly lower than the QOL-AD score of the participants who did not die [32.85, SD 7.21; t(242) = 2.002; p = 0.046].

The mean age of the participants who were discharged (78.4 years, SD 7.7 years) was significantly lower than the age of those who were not discharged [82.5 years, SD 7.4 years; t(320) = 3.840; p < 0.001]. There was a significant difference in the proportion of men and women in the not-discharged and discharged groups [χ(1) = 9.219; p = 0.002]. The reason for this is that the group who were not discharged had a higher proportion of women (63.0%) than men (37.0%), whereas the discharged group had a higher proportion of men (58.1%) than women (41.9%).

Staff interviews

A total of 245 staff members were interviewed during the study: 101 members of staff completed an interview at baseline only, 88 members of staff completed an interview at follow-up only and 56 members of staff completed interviews at both time points. This yielded a total of 301 interviews. Table 10 summarises the number of interviews completed by staff at both baseline and follow-up in the control and intervention groups.

Care plan audits

At each site, 50% of eligible service users were randomly selected to participate in the care plan audit aspect of the study. Figure 6 outlines the number of service users involved in the care plan audits at each time point, while Table 11 summarises the reasons that care plans were not selected for audit.

FIGURE 6.

Time points and number of service users involved in care plan audits.

Dementia care mapping

At each site, eight service users were randomly selected to participate in DCM. Figure 7 shows the number of potential participants and those involved at each time point.

FIGURE 7.

Time points and number of service users completing DCM.

Maintenance of ‘blind’ assessments

Although the intention was that the attending researchers would not be aware of a site’s allocation while they were completing the assessments, in some cases this was unavoidable. Unblinding occurred three times at baseline and four times at follow-up, so in total seven sites were not blinded to the researchers by completion of the assessments. Table 12 summarises the reasons why researchers became unblinded to group allocation and when this unblinding process took place.

Secondary outcome measures

There was no evidence of a significant difference for the allocated group for either of these outcomes. The IDEA questionnaire scores demonstrated trends towards non-normality, and a square root transformation was applied to rectify this. There was no indication that time and allocated group were significant; this is displayed in Table 24.

For the care plan audit there was a significant effect of time [F(1,220.19) = 22.093; p < 0.001].

Owing to the limited number of family carers present providing data for the IDEA proxy questionnaire, the WEMWBS58 and the ZBI,59 it was not possible to analyse these with any stability using a linear mixed model. Descriptive statistics have been included in Appendix 2.

There was a statistically significant difference in both the knowledge [t(30) = –7.02; p < 0.001] and the attitudes questionnaires [t(55) = –53.87; p < 0.001] demonstrating an increase in both immediately post training. These figures are summarised in Table 25. The difference between the intervention and control group at baseline was also assessed and no statistically significant difference was seen [knowledge, t(66) = –0.914, p = 0.364; attitude, t(97) = 1.532; p = 0.129]. Given the short time span and no control data over the same period, these results must be interpreted carefully, but there is some evidence of an improvement in these domains post training.

Additional covariates

The additional covariates included in the linear mixed models were age, gender, DCM, dementia type and whether or not a person had a carer (Carer _ YN).

For the QOL-AD measure,57 none of these covariates was statistically significant. For the IDEA questionnaire, there was an indication that gender was statistically significant [F(1,70.43) = 5.96; p = 0.02]. The results of all these models are given in Appendix 2.

For the CPA, the additional covariate added into the model was the completion of the QOL-AD57 (self-reported, not completed and proxy completed). This was statistically significant, indicating that there were differences in the CPA scores.

Variation of follow up

One of the sites had only an 11-week follow-up period rather than the 16-week period stipulated. Allowing the time difference to feature in the linear mixed model by adjusting the time variables to have three levels (baseline, 11-week follow-up and 16-week follow-up) did not change any of the statistical interpretation of the models. Therefore, it was deemed that there was no impact of this difference in follow-up time on the outcomes observed.

Intraclass correlation coefficient

The ICC was calculated for the QOL-AD. The difference between pre and post results was calculated first. Then the variance component was calculated with site as a random effect. Using this method, the ICC was found to be 0.09.

Dementia care mapping

At each site, the ME scores for participants were grouped together to give an overall score at baseline and follow-up. The scores were then separated into intervention and control, and an overall average of the ME scores was calculated. The results of the mapping exercise are displayed in Table 26.

In accordance with one of the study objectives, it was hypothesised that an increase in ME scores would be seen at sites where training had been delivered. However, Table 26 highlights that, regardless of whether sites were intervention or control, there was no improvement in ME scores. In addition, for the majority of sites there was a decrease from baseline to follow-up.

Outcome measures

Descriptive statistics for the outcome measures are shown in Table 27 for baseline and Table 28 for follow-up. As can be seen, completion rates of the carer measures (ZBI59 and WEMWBS58) were low, precluding any further analysis.

The mean EQ-5D-3L self-report score at baseline was 0.74 (SD 0.267), which is similar to the mean UK population score of 0.78 (SD 0.26) for the age group of 65–74 years. We note that proxy scoring of the EQ-5D-3L was much lower (mean 0.35, SD 0.337), which is a similar finding to other studies involving proxy reporting for people with dementia. 63 The lower proxy scores of the EQ-5D-3L were driven by lower ratings in the self-care item.

Service use

Most health and social care contacts at both baseline and follow-up were home visits. The mean number of contacts was highest for general practitioners (GPs) and community nurses. A more detailed breakdown of the community- and hospital-based services used during a 4-month period is available in Appendix 2.

Medication

At baseline, medication records were collected for 254 participants. Medications were categorised by British National Formulary therapeutic class, obtained from Prescription Cost Analysis88 data, and are shown in Table 29. The most frequently prescribed category of medication was for the central nervous system, which includes dementia medication, analgesics, antidepressants and hypnotics and anxiolytics. A total of 101 out of the 254 participants for whom information was available were prescribed a dementia drug. Eight of these people were prescribed more than one dementia drug.

A more detailed breakdown of the medications prescribed is available in Appendix 2.

Cost of the intervention

There were no intervention design costs as the intervention was developed before this study took place. Nine training sessions were delivered on site to staff. Staff who received training at hospital sites ranged from NHS band 3 support workers to NHS band 7 ward managers. Staff who received training at care home sites ranged from care home managers to care home assistants. Each training session lasted 5 hours and was delivered by a NHS band 8a trainer. Sites were given two copies of the training manual and 12 copies of response booklets. In total, 71 members of staff were trained at a cost of £7157, or £101 per staff member. The intervention costs are summarised in Table 30.

Staff decision-making interviews

The thematic analysis identified 21 themes from the combined data set (301 interviews). The interviews were then reanalysed to identify the frequency of each decision-making strategy within the data and for each group (i.e. control baseline, control follow-up, intervention baseline and intervention follow-up). This information is shown in Table 31.

There were similar total numbers of decision-making strategies in each group. Table 31 shows that the most common theme in all groups, with the exception of the control group at follow-up, in relation to decision-making strategies was working as a team or consulting with other staff. The most common strategy described in the control group was hierarchical decision-making, whereby responsibility for making a decision was passed to another member of staff who was considered to be in a higher position and therefore more able to decide. It can be noted, however, that there was a drop in the reliance on hierarchical decision-making in the intervention group (although it remained the second most popular strategy in this group).

Although there are some variations in decision-making strategies across the groups, and it is acknowledged that any conclusions drawn would need to be cautious, it does appear that there are some differences in the intervention group at follow-up. Staff in this group were less likely to cite common sense as a strategy that they would employ to help them make decisions, whereas this strategy actually increased in the control group. The intervention group was also less likely to cite fear, the need to follow rules and ‘it’s not my job’ as methods by which they made a decision. The intervention group was also more likely to explicitly mention human rights (such as dignity, respect and positive risk-taking) and person-centred care. It is perhaps not surprising that the intervention group was more likely to mention specific human rights based approaches, given the training that they had undertaken and their knowledge by this time that the site was involved in a human rights study. Given that it has been suggested that person-centred principles are closely linked to the key elements of a human rights based approach,37 it is interesting to note that there was also more discussion of person-centred principles in the intervention group at follow-up, despite the principles not having been explicitly taught in the training.

Follow-up interviews

Follow-up interviews were offered to all intervention sites that completed the training. Some sites were unable to accommodate the interviews because of changes in management and lack of staffing. In total, six of the eight sites that completed training completed interviews designed to look at acceptability and effectiveness of the intervention. Interviews were conducted on an opportunistic basis and included staff who had attended the training, staff who had not attended the training and, when possible, the unit manager. The interviews were completed by a research assistant (not the same research assistant who had visited the site to collect data) and a member of the PPI reference group, who was either a person living with dementia or a carer.

On average, four staff per site were interviewed. Of the eight sites that had completed the intervention, four had a new manager by the time of these interviews. Table 32 shows the staff who were interviewed at each site and the percentage of staff trained who remained at the site at this point (approximately 6–10 months after the intervention). Sites here are listed as 1, 2, 3, etc., so that they are not identifiable. Given that the information gathered at individual sites is later linked to management styles, it was felt important to preserve the sites’ anonymity.

Three themes were derived from the data on the use and acceptability of the intervention. It is fully acknowledged that the questions posed to staff guided them to discuss the acceptability and effectiveness of the intervention and that the themes generated are therefore not surprising. The overarching themes and subthemes related to acceptability and effectiveness are summarised in Table 33 and described in more detail below.

Suggestions for improvements

During the interviews, staff were also asked to suggest improvements to the approach. The following suggestions were made:

1. training more staff from each unit so that knowledge could be shared and the whole team could appreciate the approach

2. providing more booster sessions to cement learning

3. digitalising the ‘Getting It Right’ assessment tool56 to make it easier to use.

Management style

As management support was such a key feature in many of the interviews, the data gathered were subsequently analysed using a deductive approach and mapped on to the characteristics from Bass and Avolio’s87 management style. It is acknowledged that the information gathered in this way was not specifically about management style and any conclusions made are tentative.

Bass and Avolio87,89 describe transactional leadership as a relationship in which both the leader and follower are working to meet their own self-interests. In this type of leadership, the leader gives clear direction as to what the follower must do in order to be rewarded. Whereas transformational leadership is described as moving beyond pure self-interest. In this style of leadership, the leader works to enhance motivation towards a shared vision. The model breaks down each leadership style into different characteristics. Transformational leadership comprises intellectual stimulation, individualised consideration, charisma and inspiration. By comparison, transactional leadership comprises leadership styles such as passive management by exception (i.e. waiting for problems to arise before taking corrective action), active management by exception (i.e. monitoring performance and taking corrective action if the follower fails to meet standards), laissez-faire and contingent reward. Table 34 outlines the characteristics associated with both transformational and transactional leadership styles.

It has been suggested that a transformational management style can enhance commitment, involvement, loyalty and performance and help individuals deal with stress. 90 In contrast, a transactional management style is likely to induce stress in staff.

When the data from the interviews were deductively analysed in line with the management styles outlined above, certain patterns were noted. At sites 2 and 5, more codes were identified that related to a transformational management style, and to two characteristics in particular: intellectual stimulation and individual consideration.

Intellectual stimulation refers to a managerial style whereby the manager encourages their staff to behave in creative ways, generating novel solutions for old problems. It is suggested that this leads to a working environment that values people questioning their own assumptions and those of the organisation and tolerates differences in views and opinions. 89 Four codes appearing to relate to intellectual stimulation were identified in the interviews.

1. Managers actively seeking ways in which the team can improve:

   We heard about the study from another care home that is actually a dementia care home and we sort of asked could we join. We were desperate to be involved.

2. Managers being open to criticism:

   If I have got problems in my home can you tell me. If you can do something better that is great. It affects everybody.

3. Managers showing evidence of problem-solving and adapting the human rights based approach and the ‘Getting It Right’ toolkits:

   They still use the books, but the pages have been printed and put on the walls in the bedroom. This makes them much easier to use for staff and patients.

4. Managers actively disseminating information about a human rights based approach:

   Well basically we just got an e-mail saying listen, we are going to start trying this, see if it works, see if it is helpful and they brought in the paperwork and went through it with us and we just started implementing it on the ward.

Individual consideration is characterised as helping others to develop, letting others know how they are doing and giving personal attention to those who seem rejected. Two codes were identified that related to this aspect of transformational style.

1. Managers being active in motivating staff to attend training and to utilise the assessment toolkits:

   But I am the one who gave the staff the support and said ‘how are you getting on? Have you been through it?’ We supported each other. Do you understand it? I think at first they said we just have to concentrate on the red things and that was it, you can leave the others. And I said it is not that you can leave them it is just that initially when they come into the building, they are the things you need to give them, safe care but everything is important. And they were like OK we will carry on doing it. But we worked together with that one. I was integral. I am really lucky with my staff.

2. ‘Understanding staff’ and their motivations:

   For them to have volunteered for it, nobody ever volunteers for training. They would be reluctant at first. If I hadn’t driven it, would they have done it, probably not? I do that with any of the training because if you didn’t they would say if you can’t be bothered they can’t be bothered. Put it that way. I do think someone has to give them that drive. If you explain it to them they are like OK.

At sites 1 and 3, there were more codes relating to a transactional management style. These mainly fit into the laissez-fare style of management, although one code fits into the passive management by exception category.

Laissez-faire describes a manager who is content to continue functioning in the way the organisation has worked: someone who does not encourage staff to develop, embraces the status quo and asks no more of staff than what is essential. Four codes relating to a laissez-faire style of management were identified in the interviews.

1. Lack of management support and encouragement:

   I think it covered every aspect but I don’t think we have enough support and staffing levels. A lot of staff on the floors.

2. Failure in communication between past and present management. The following quotation was from a new manager who had not been in position when the study took place at their unit:

   I only heard about the study from this evaluation.

3. Failure to communicate to staff:

   No they just got put in the office and we never got round to it. I don’t know where they are because our manager has left.

4. Using a lack of resources as a reason not to make changes:

   Always resources stop me from doing these things. You always want more than what you have got. I don’t feel as though profit and care are good bedfellows really. When certain things are counted and continent aids are counted that does not lead to respect or rights. That is a major issue.

Passive management by exception is described as dealing with an issue only if it arises. It describes a manager who is happy if staff are meeting agreed standards as long as things are working and as long as individuals know the standards. One code was identified that related to this style of management.

1. Passive support:

   We didn’t use it but I can’t see what would prevent us. I was wanting to have a word with the manager about it rather than just go ahead off my own back. I would rather ask, check with the manager to check that it is OK.

At sites where there were more descriptions of managers being active or transformational in adopting a human rights based approach, staff described utilising the ‘Getting It Right’ toolkits and they felt generally more supported and understood and experienced the worth of prioritising them to provide person-centred care. Staff provided examples of service users’ improved well-being, and of engagement and increased trust with services users and their families.

At sites that had more themes related to transactional management, there was a smaller degree of uptake of the approach. Although staff reported that they were excited about the approach (‘I just remember it all being really exciting. I was buzzing’), a lack of management support meant that the approach was not applied: ‘No they just got put in the office and we never got round to it’.

Well-being measures

As in previous studies,94 QOL-AD scores were lower when the measure was completed by a proxy than when it was completed by the person living with dementia. Previous studies have shown that agreement between proxy and self-report on the QOL-AD is higher when examining observable functions such as physical health and disability and relatively poorer for more subjective measures. It has also been shown that disagreement between proxy report and self-report increases as the severity of dementia increases. 95 This is interesting in the context of this study, as many of the people involved will have been living with severe dementia. Similarly, previous work has found that whereas care home residents’-reported QOL-AD scores correlate highly with levels of depression, carer-completed reports on the same measure correlate highly with levels of dependence and behavioural issues. 96 Therefore, as dementia progresses and people living with it become more dependent, proxies would see those living with dementia as having a lower quality of life, whereas there is evidence that levels of depression decrease as severe dementia progresses97 and so these same people living with dementia may rate their subjective quality of life higher.

In this study, self-reported quality of life, as measured by the EQ-5D-3L, was just as high as the reported norms of people of a similar age without dementia. This again challenges the assumption that living with dementia has a negative impact on self-reported quality of life. It is often assumed that the reason for a difference between self-report and proxy report measures is that people with dementia lack the required insight into the impact of their condition to be able to accurately report on their quality of life, but it is of course possible that proxies are unable to imagine what it is like to live with dementia, particularly in the later stages, and as such may assume that the quality of life will be poor. This may not be the reality for the person living with dementia. As such, it is possible that the self-report and proxy report versions of the same measure are actually measuring different concepts. Pickard and Knight98 identify two distinct proxy perspectives. The ‘proxy-patient’ perspective requires proxies to project themselves into the patient’s internal state, whereas the ‘proxy-proxy’ perspective is based on the proxy’s judgement. The authors found that there tends to be less discrepancy between self-report and proxy report when a proxy is explicitly asked to adopt a proxy-patient perspective. There is also evidence that a proxy’s own quality of life has an impact on their judgement. 99 It could, therefore, be questioned whether or not it is appropriate to use proxy measures to capture the subjective well-being of people living with severe dementia.

Although it has been suggested that the QOL-AD57 can be usefully completed with some people with a MMSE score of as low as 3100 (although it was originally suggested to be valid for use with people with MMSE scores of > 1057), it quickly became obvious that the majority of people living with dementia in the care homes and wards visited were unable to complete the measure, even with assistance from skilled clinicians. Of course, this does not mean that information about quality of life cannot be elicited from people with more severe dementia101 or that the team did not think that this was important. Indeed, processes undertaken to elicit information from people in the later stages of dementia were described earlier in this report (see Development of the FREDA assessment tool). 71 Rather, the use of the QOL-AD and the heavy reliance on proxy measures would raise the question of whether or not it was an appropriate measure to use in this context. It may have been more useful to interview residents about their experiences of quality of life to elicit their views in a more naturalistic way.

Figure 8 summarises the domains of well-being covered by the QOL-AD. 57 Given the items included in the measure, it is worth considering why an intervention would be expected to bring about change in these domains. The current intervention under investigation, even if successful, is not likely to change a person’s financial issues, their ability to do things around the house (although a person-centred approach may encourage people to do more) or their marriage.

FIGURE 8.

The constructs included in the QOL-AD.

It may be more useful to look at other concepts that have been shown to have links to subjective well-being in dementia. Studies have indicated that care home residents with greater perceived control (e.g. regarding decisions about food, social activities, daily routines, privacy and sense of self) experienced significantly greater psychological well-being. 102 Similarly, a study of care home residents in Singapore identified five predictors of quality of life in this population:103 comfort, dignity, food enjoyment, autonomy and security. Taken together, this evidence would suggest that a more innovative measure, designed to tap into concepts that actually matter to people living with dementia, is needed if we are to accurately capture the subjective quality of life of people living with dementia, particularly those with more severe dementia.

Dementia care mapping

Dementia care mapping was utilised to observe the care provided at each unit. In particular, the ME scores from each unit were compared at baseline and at follow-up to uncover any differences in care following the intervention. However, it is important to be cautious about how much can be extrapolated from the overall ME scores. Although DCM was originally created as a clinical tool to help improve the quality of care in clinical settings, it has been used effectively in a number of research studies to gain a more objective measure of service user well-being. 104 There are potentially a number of factors, over and above the hypothesis of no change to care, that could have accounted for no change or a decrease in ME scores.

It was apparent that at baseline the majority of staff were not aware that the research was being carried out; therefore, at this time several staff initially believed that they were being observed and were uncertain about the motives of the research assistants. As a result, levels of engagement with service users may have been influenced and potentially increased, particularly given that there was a clear finding within the study that care plans improved when people felt that they were being monitored. Generally, at sites at follow-up, staff appeared to be more relaxed and it was felt that the behaviours observed were potentially a more accurate representation of the day-to-day care being provided. Given this finding, it could have been useful to complete a pre-baseline set of measures to allow the staff to become accustomed to being observed. This would, of course, have had financial and time implications for the completion of the research.

Between baseline and follow-up, there were a large number of service users who had been discharged or had died, which, at times, resulted in a different service user being observed at follow-up. According to the Bradford Dementia Group,105 it is difficult to see change at separate time points when observing different people, and therefore any changes that were hypothesised to occur at intervention sites may not be reflected in the overall ME scores.

Throughout the study, it was noted that there was a large turnover of staff at sites from baseline to follow-up. For a number of sites, there was a larger proportion of agency staff being employed at follow-up than at baseline. Although the use of agency staff is beneficial for meeting the desired staffing levels, it does not necessarily result in continuity of care for service users. A lack of consistent staff could potentially be unsettling for some service users,106 and this could be a reason why ME scores decreased at particular sites at follow-up. In addition, at some sites agency staff were not being used and hence staffing levels were extremely low. This is highly likely to have had an impact on ME scores, as staff members were able to meet only the basic care needs of service users.

In some sites in particular, the environment could have had an adverse impact on ME scores. More specifically, the layout of the communal areas where mapping took place was not ideal. At times, individuals could not be observed, which meant that they were given a behaviour category code of Q (quit: individual was not in mapping area) for a number of time frames. At these particular sites, there may have been several observations that could not be reported, and, owing to this, the overall ME scores for these sites may not be a true representation as behaviours and levels of engagement could have been missed. This was, of course, true at both time points.

A general observation made during DCM was that, at follow-up, some service users’ cognitive abilities had declined, and in some instances their physical health had deteriorated. This could result in individuals not being able to participate or have difficulty participating in behaviours that generate higher ME scores. This may account for no change or a decrease in the overall ME scores obtained. However, given the client group involved this would not be unexpected after a 4-month follow-up.

During DCM, the team observed a number of representatives from external organisations facilitating various activities with service users. At one site in particular it was noted that, at baseline, activities took place throughout the day, but at follow-up this did not occur. As activities can potentially have a positive influence on an individual’s ME score, this may account for the decrease in ME scores at this site.

For these reasons, it is important to think about the scores obtained during DCM in context rather than in isolation.

Recruitment

The study recruited more participants than had been originally planned. Although initially it was feared that the recruitment rate of people living with dementia might be low because many might not be able to complete the self-report measures, the introduction of proxy report measures ensured that a high proportion of people at each site could be included in the study. Research assistants strove to recruit as many people as possible at each site and the only reasons that they did not recruit everyone at each site were lack of consent, the person not being available during the data collection period, the person not having a dementia diagnosis and time pressures on staff to complete proxy measures (see Figure 5). It was considered important that as many people as possible at each site were given the opportunity to be involved in the study. Similarly, as it could not be known until analysis whether or not there were significant differences between proxy and self-report scores on the QOL-AD;57 therefore, it was important to obtain as many of each category of questionnaire as possible. It was felt that the development of good working relationships at each site between the research assistants and the staff at the unit contributed considerably to the high recruitment figures. Reflection by the TSC and in research team meetings focused on how important it was that the research assistants had previous experience working in dementia care settings, as this allowed the development of a trusting working relationship with staff at the care homes that was built on mutual respect and understanding.

Study design

The study involved a rigorous investigation of the intervention that allowed clear claims about its effectiveness and utility to be made. Without a rigorous framework for evaluation and a clear conceptual framework (outlined in Figure 1), claims about effectiveness could have been made without giving consideration to the impact on well-being of the people living with dementia. It would have been possible to say that training improved knowledge and attitudes, changed decision-making strategies and, without considering a control group, improved care plans. Without looking at the elements of care and well-being, it could have been concluded that the intervention was a success. The translation to improvements in care and well-being is not always considered in such approaches. 107

Patient and public involvement

There were high levels of engagement with people living with dementia and carers in all aspects of the study through the PPI work, and their contributions both guided the study and influenced its direction. In total, 79 people living with dementia or carers gave input into some element of the study design, delivery or the subsequent public engagement work. The involvement of people living with dementia and those who support them was felt to be paramount in this study because of its congruence with the underlying principles of a human rights based approach to care. PPI is concerned with a shift to carrying out research and development with or by members of the public and service users rather than on or for them simply as participants. 108 This helps to uphold and promote the rights of each person to have ‘a right to voice’, with suppression or denial of that voice being a potential violation of rights. 109 Similarly, the nature of this study and the fact that it was funded by NIHR makes this even more important,110,111 as it could be considered a core democratic principle that:

. . . people who are affected by research have a right to have a say in what and how publicly funded research is undertaken.

Copyright INVOLVE February 2012. 110 This publication is the copyright of INVOLVE. We are happy for our materials to be downloaded, printed out, copied, used and distributed free of charge for non-commercial purposes. Where our materials are copied, used or distributed, the source of the material must be identied and the copyright status acknowledged. No alteration or adaptation of any of the materials may be made without obtaining the permission of INVOLVE. Reproduction or distribution of materials either in whole or in part for a commercial purpose is prohibited without the prior written permission of INVOLVE

Alzheimer Europe recently issued a position paper112 regarding involving people with dementia in research through PPI. In this paper the authors recommend that people living with dementia be involved in all aspects of research, including the initial generation of research ideas. An important output from this study is the development of a PPI reference group comprising people living with dementia and carers who have developed an interest in promoting the area of human rights in dementia. Although the group’s initial role was to support the current study, the members felt that it was of utmost importance that the output would be accessible information that increased public awareness about the human rights of people living with dementia and how easily these can be undermined in society. This PPI reference group will continue to exist after this study ends and will form the basis for the development of further research in the area.

Innovative methods

Innovative methods were used throughout the study to try to ensure that information was elicited in ways that would be as ecologically valid as possible. It has previously been discussed in this report (see Development of the FREDA assessment tool) that every effort was made to include people living with later stages of dementia in the development of the assessment tool. 71

A key element of the study was the exploration of staff decision-making strategies. To assess this, staff were provided with clinical vignettes that were designed specifically for this study and that combined potential human rights violations with person-centred principles. Vignettes have long been used to investigate a number of phenomena in the social, behavioural and health sciences. 113 In this study, staff were read a number of clinical vignettes and asked what they would do in each particular situation. On each occasion they were asked supplementary questions of ‘how would you come to that decision?’ and ‘what would help you in coming to that decision?’ It was felt that asking staff to reflect directly on their own experiences and how they make decisions may feel threatening to many. Similarly, it is often unethical or impossible to assess clinical decision-making experimentally with real clinicians and service users in health-care settings. Therefore, although vignette designs may be an alternative method for investigating how health clinicians make decisions that affect service user care, it has been questioned whether or not a written stimulus, and a participant’s responses to it, can accurately represent certain aspects of what happens in the real world. It has been suggested that three conceptually distinct, but functionally inter-related, factors contribute to validity. 114

1. Vignettes must simulate aspects of real-world scenarios and bear some resemblance to situations encountered by the participant (construct validity).

2. Vignettes, and the differences between vignettes, elicit some kind of effect that is hypothesised to exist independently in the real world (internal validity: the degree to which changes in the dependent variable can be accurately attributed to changes in the independent variable).

3. Vignette studies should produce results that generalise to real-world situations encountered by the participants and others like them (external validity).

It is important to note that vignettes are not intended to recreate real-world situations. 113,115,116 Rather, they are designed to ‘approximate, isolate, manipulate, and measure key aspects of the decision-making processes that individuals use in real world situations’. 116 Similarly, participants’ behaviour in a vignette study is not intended to be interpreted as representative of their behaviour in the real world, but rather viewed as strong predictors or proxies for such behaviour. 115,116 It is, therefore, more useful to think about whether or not the mental and behavioural processes used in responding to the vignettes are activated in a way similar to the manner in which these processes are used in real life.

In considering these factors, the vignettes used in this study were developed collaboratively with staff working in dementia care settings so that they accurately reflected clinical situations that may be encountered on a regular basis. The vignettes were refined during the pilot phase of the study and they elicited a wide range of distinct responses that reflected a variety of decision-making strategies. It cannot be claimed, however, that the responses reflected proxies to behaviour, as little, or no, behavioural change was noted as a result of the intervention despite changes in decision-making strategies.

National dementia initiatives

The approach was also congruent with a number of national initiatives related to dementia. There is a clear commitment to reducing the prescription of antipsychotic medication to people with dementia. 92 The principles of a human rights based approach, particularly proportionality, least restrictive practice and proactive strategies together with a detailed understanding of the person with dementia, through a person-centred care plan, are key in finding alternatives to antipsychotic prescribing in challenging behaviour. A key to managing challenging behaviour is the understanding of the need being expressed through this behaviour. 117 An in-depth knowledge of the person is obviously essential in helping formulate this. It was proposed that the application of this approach would lead to more detailed person-centred care plans that allow a more detailed knowledge of the person.

The National Institute for Health and Care Excellence (NICE),50 in its quality standard for dementia, states that:

People with dementia have an assessment and an ongoing personalised care plan, agreed across health and social care that identifies a named care coordinator and addresses their individual needs.

NICE. 50 © NICE 2010 Dementia Support in Health and Social Care. Available from www.nice.org.uk/guidance/QS1. All rights reserved. Subject to Notice of rights. NICE guidance is prepared for the National Health Service in England. All NICE guidance is subject to regular review and may be updated or withdrawn. NICE accepts no responsibility for the use of its content in this product/publication

The proposed intervention, if applied, would clearly address this issue.

Many of the concepts covered in the ‘Getting It Right’ assessment tool56 map directly onto the standards of the Quality Network for Older Adults Mental Health Services118 and allow wards to meet the required standards more readily. It was hoped, therefore, that the approach may be acceptable to sites as it would help them to meet other requirements.

Effectiveness and acceptability of the intervention

Despite staff and managers stating that they found the approach useful and simple to use – ‘It was explained simply in layman’s terms’ – there was limited uptake of the approach in practice. It was hard to get staff to attend the booster sessions and releasing staff for training seemed extremely difficult. Although NHS wards were harder to recruit, once recruited they had clear mechanisms in place to enable staff to attend the training day. In contrast, care homes were relatively easier to recruit but were often unable to protect the time required for staff to attend the training. There were multiple occasions when the trainer would visit a care home to find that no staff had been given dedicated time to attend the training and so staff were taken off active duty to attend, resulting in additional pressure on the other staff who had to cover their work. The fact that 2 out of 10 intervention sites were unable to undertake the training in the required time frame, despite having prior knowledge of the requirements, demonstrates the difficulties or level of willingness sites had in releasing staff for training and development. Similarly, several staff from care homes reported attending the training on their days off as they were interested in the topic but had not been given the time to attend during working hours. Although this shows a positive commitment from these staff, it calls into question the commitment of those care homes to the approach.

Although it was initially envisaged that as many staff as possible would be trained at each site, in reality only 8.88 members of staff (28.7%) at each site were recruited to the training. The proportion of staff trained at each site ranged widely, between 11.6% and 52.4%. Similarly, the protocol suggested that three ‘booster sessions’, which were designed to support the implementation of a human rights based approach, should be delivered to each site. In reality, this was not possible. There is obviously a major difference between training the majority of staff at each site and facilitating booster sessions to embed the approach (the expectation), and training an average of 28.7% of the staff with minimal follow-up (the reality). However, staff who had attended training were positive about it, and felt that it was tailored to their needs and had an impact on their ability to complete their job successfully: ‘The tutor lady was fab. Simple’. Given the reluctance of sites to engage with the training and booster sessions, it is interesting that a theme from interviews with staff in the intervention sites was that they would have liked more training, training for more staff and more booster sessions.

There was also evidence that the ‘Getting It Right’ assessment tool56 was not routinely completed at participating sites. When it was used, however, it was considered to be helpful (‘the training and the handouts and the paperwork that they gave us has definitely helped’) and it had a positive impact on the work people were carrying out. Unfortunately, there was no mechanism within the study to specify the number of assessments that needed to be completed at each unit, and in hindsight this may have improved the completion rate.

These somewhat mixed findings about the acceptability and effectiveness of the intervention seem to reflect the finding that uptake varied across sites and that a number of factors influenced uptake. However, it was clear that the majority of staff enjoyed the training and felt that it was targeted at a level appropriate to their skills and knowledge. They also valued the ‘Getting It Right’ assessment tool,56 but a variety of factors meant that they were not always able to use it in their setting.

Ultimately, despite positive reports from staff about the approach, there was also evidence of difficulty engaging teams with the training, problems getting staff to attend follow-up sessions, limited use of the assessment tool and the approach having no impact on the care and well-being of people living with dementia. Therefore, the intervention could not be concluded to be effective. This will be considered when offering training to the control sites (in line with the original protocol) and different models will be considered.

A further limitation of the study related to the intervention was the lack of any fidelity assessment. No attempts were made to assess whether or not residents found the application of the ‘Getting It Right’ assessment tool56 useful. The study may have been improved by adding this to the evaluation framework.

On reflection, the evaluation mechanism used to assess the effectiveness of the intervention was completed with a population once removed from the direct intervention. Although staff received the human rights training, completed the assessment tool and attended the booster sessions, the primary outcome measure (QOL-AD57) was completed with the residents at each unit. The team still maintain that it was of utmost importance that the well-being of people living with dementia, who were of course the recipients of the care from staff who had received the intervention, was assessed. Without improvements to the care and well-being of people living with dementia, there is little use in applying the intervention. However, as discussed in Appropriateness of outcome measures, Well-being measures, it may have been more useful to consider other ways of measuring the impact of the intervention on people living with dementia.

Theory of planned behaviour

A major finding of the study is that although the intervention has changed the knowledge staff have about human rights, the attitudes that they hold and their reasoning about decision-making in relation to clinical care, it has not altered their behaviour when delivering care. The theory of planned behaviour126 (Figure 9) may provide some insights into this finding. This theory proposes that the chances of an individual behaving in a particular way is influenced by their attitude towards the topic, their perceptions of other people’s (their peers’) attitudes to the issue and their perceived level of control over the situation.

FIGURE 9.

Theory of planned behaviour.

Although the current intervention may change an individual’s attitude to human rights (as demonstrated by significant changes in the attitudes questionnaire scores) and their intention to behave in a way more congruent with a human rights based approach (as evidenced by improvements in care plans and different decision-making strategies), the actual behaviour, as measured by DCM, remains unchanged. The theory of planned behaviour would suggest that this discrepancy between actual and intended behaviour may be accounted for by considering staff’s perceived level of control and perception of others’ views about human rights. Although the current study did not directly collect information on these areas, some tentative conclusions could be drawn from the available data.

Although the initial intention was to train as many staff as possible in a human rights based approach, practicalities meant that an average of < 30% of staff at each site were trained. Similarly, there was limited uptake of the booster sessions that would have provided staff, both trained and untrained in the approach, with the opportunity to discuss the application of a human rights based approach to care. As a result, there may have been fewer opportunities for staff members to assess their peers’ attitudes to and understanding of such an approach. This is further backed up by information from the follow-up interviews at intervention sites, in which staff suggested that having more opportunities to discuss the approach with other staff at the unit may have been helpful in applying it.

It has been hypothesised that perceived control relies on two elements: locus of control and self-efficacy. Whereas locus of control refers to a person’s beliefs around whether people in general can achieve positive outcomes and avoid bad outcomes through their own actions or whether external factors control these outcomes, self-efficacy refers to the perception that a person themselves has the skills and abilities to enact effective responses. A common theme across all groups, derived from the staff decision-making interviews, was the use of hierarchical decision-making. This implied that staff felt a lack of control in making autonomous clinical decisions and instead relied on seeking consent to act in a particular way from others, which may, in turn, have led to staff feeling that they lacked the self-efficacy to bring about change. Similarly, the follow-up interviews at intervention sites highlighted that staff applied the model more effectively at units where managers encouraged them to be independent thinkers and adopted a more transformational approach to leadership.

Given that a key finding of many care home studies is a lack of uptake of new interventions, future research could explicitly use the theory of planned behaviour to investigate human rights based approaches and their application.

Management style

The results of the follow-up interviews demonstrated the importance of management style in maximising the chances of this intervention being adopted at any particular site. Although there is insufficient evidence to make wide-ranging claims about this element of the study, initial findings showed that sites with a more transformational manager were more likely to adopt the approach and think of ways to adapt it, whereas sites with a more transactional manager were less likely to adopt the approach and more likely to dismiss its relevance. 90 It is perhaps not completely surprising that management style has such an impact on uptake given that hierarchical decision-making was one of the most popular decision-making strategies. If managers are not supportive and other staff look to them to make decisions, then the approach will not be adopted.

An issue highlighted in the study was that there was a high level of turnover of managers at the units. By the time of the follow-up interviews, half of the sites at which the intervention was delivered had new managers, and there was evidence of poor communication between past and current management: ‘I only heard about the study from this evaluation’. There is recognition that care homes that provide person-centred, high-quality care tend to have managers who focus on leadership rather than just management127 and that emotional intelligence is the key to good-quality care. 128 A Joseph Rowntree Foundation report129 exploring ways of improving the quality of care in care homes highlighted management as a key driver in developing and implementing improvements. The report proposed that a change culture must start with managers who are willing to embrace change (in this study the transformational managers) and that these managers must empower staff to participate in decision-making about both clinical care and the more organisational aspects of the unit. This fits well with the finding that staff felt reluctant to make decisions independently and instead sought advice from more senior colleagues. Linking to the theory of planned behaviour, it would appear that managers are key in enabling staff to feel that they have behavioural control over any initiatives and facilitating staff ability to implement these independently. This maximises the chances that initiatives will actually be implemented.

Given these findings, it would seem vital that managers with the correct skills and the ability to develop effective teams are recruited to these important positions. There has been a move within the NHS and other services to utilise values-based recruitment. 130 Values-based recruitment aims to select candidates for roles based on their individual values and behaviours and how closely these align with those of the organisation. It may be that selecting managers based on their values, particularly those related to an ethos of person-centred care, may improve the care provided at a unit.

Cultures of care

Culture has been defined as ‘the basic assumptions, values and norms shared by and influencing how members of an organisation behave and interact’. 131 Too often, substandard care is accepted and unquestioned. It is expected that quality of life will be low in care homes and that little else is aimed for. This can lead to care that is task orientated and driven by goals and objectives rather than by resident-orientated factors. 129 It is notoriously difficult to change cultures of care and, although many initiatives are introduced, a culture of care continues to exist that falls short of the optimal standards.

It is unacceptable that the culture of care in care homes remains substandard and unable to meet the basic needs of the people to whom it applies. A recent UNISON report132 found that < 20% of care home staff who responded to their questionnaire felt that they had enough time to provide care that upheld the dignity and individuality of residents. This led to residents not receiving treatment such as assistance to go to the toilet, efficient monitoring of their condition and meeting of personal care needs, and 88% of workers reported that they did not have time for a conversation with their residents.

A large-scale study looking at what constitutes an excellent culture in care homes (Care Home Organisations Implementing Cultures for Excellence – CHOICE)133 identified seven factors that contribute to a positive culture in care homes:

1. shared purpose in providing the best person-centred care

2. a sense of community between all involved in the care home

3. managers ensure that external pressures do not have a negative impact on service delivery

4. staff are empowered to take responsibility for residents’ well-being by active management processes

5. openness to change for the benefit of residents

6. using the care home environment to the benefit of residents

7. person-centred activity and engagement is integral to care work.

Again, these factors mirror those of this report in that they highlight the importance of cementing person-centred practice in care, not just talking about it; the centrality of management in ensuring that staff are empowered to take responsibility for their own practice and are protected from external pressures that may have an adverse impact on their ability to provide the best-quality care; and the necessity of cohesion among staff and managers in developing a positive culture at the unit.

This study aimed to utilise a human rights based approach to bring about cultural change. Previous work by the Equality and Human Rights Commission found that public authorities continued to lack a culture of respect for human rights, despite some examples of good practice. 134 They identified a number of barriers that they suggested could contribute to this finding. These included ‘initiative overload’, 135 with staff feeling that there is always something new to implement; a lack of autonomy among staff; the hierarchical and target-driven nature of health-care settings;136 and a low awareness of human rights among staff, service users and members of the public. It has been suggested that the use of human rights language in conversations with organisations, staff and members of the public is ineffective because of this lack of awareness. 137 In contrast to this belief, the current study has demonstrated that training could raise awareness of human rights, improve attitudes towards human rights and, in the context of decision-making, change staff’s use of language about human rights.

The CQC has integrated the FREDA principles into the inspection framework. When rights are fully integrated into policy and practice within services, the impact has been anecdotally described as:

. . . a magnet pulling services in the direction that best supports the dignity, respect, equality and autonomy of those that use them.

Equality and Human Rights Commission. 134 © Equality and Human Rights Commission 2009. The copyright in the document this publication has been adapted from, and all other intellectual property rights in that material are owned by, or licensed to, the Commission for Equality and Human Rights, known as the Equality and Human Rights Commission (‘the EHRC’)

This study does not endorse this finding. Instead, it suggests that although a human rights based intervention can change knowledge, attitudes and decision-making, it does not influence care delivery.

Training

The findings of this study call into question the efficacy of training as a medium for improving care and changing culture. Although the training appears to significantly increase knowledge about human rights and positive attitudes towards human rights immediately following the training, it does not translate into improvements to care practices and in the reported well-being of people living with dementia.

There is a current government mandate through Health Education England to provide effective, relevant dementia education and training for the entire workforce. 138 However, it is recognised that there is limited evidence about what effective dementia training and education for this diverse workforce looks like. 139

It seems important to consider what any training programme is aiming to achieve. Kirkpatrick’s model for the evaluation of learning140 identifies four levels for evaluating the efficacy of training.

• Level 1: reaction – learners’ reaction to and satisfaction with the programme.

• Level 2: learning – the extent to which learning has occurred included increasing knowledge, skills, confidence and attitude change.

• Level 3: behaviour – the extent to which staff behaviour or practice have changed and whether or not participants are applying their learning in practice.

• Level 4: results – examines what results have occurred because of the training, in this case the impact on people living with dementia.

The remit of this study was to train staff in a human rights based approach, with the aim of assisting staff to make decisions in complex clinical situations. It also trained staff in a new way of care planning and provided them with a tool to complete more person-centred assessments. What it did not do explicitly was teach people the skills to behave in a more person-centred way. Given that there are low levels of knowledge of more basic issues related to dementia observed, could we have expected people to make this transition independently? There was low uptake of the booster sessions, but, even if the uptake had been higher, the sessions would not have been detailed enough to fulfil this role. To start meeting these goals, we would be looking at a far more intensive model of coaching and consultation; there are examples of this such as the Newcastle model of challenging behaviour,141 but they are far more time and labour intensive and, therefore, are likely to also have financial consequences.

Preliminary results from What Works in Dementia Training and Education? A Critical Interpretive Synthesis of the Evidence139 suggest that there are a number of elements that contribute to training to bring about changes to people living with dementia. These are that the training:

• uses face-to-face delivery, discussion and activities that support the application of learning to practice

• is delivered by an experienced trainer

• is > 1 hour in duration

• is designed for a specific service setting

• provides training on a structured tool or delivery manual that assists the application of learning in practice.

This study utilised all aspects of this framework but failed to embed the approach into care. Given the results from this study it would be useful to add that management support is also required both to attend the training and, more importantly, to apply the approach to clinical situations.

Language

It was considered important that all language used in this report was in line with the underlying principles of the study, namely upholding the human rights of people living with dementia. To this end, every effort was made to adhere to the language guidance provided by DEEP, which was compiled by people living with dementia. 145