Article Text
PDF
Abstract
Objectives People with dementia and their carers may experience poor end of life care, often having to navigate complex health and social care systems. The objectives of this study were to identify: (1) barriers for people with dementia and their carers in accessing good quality end of life care and (2) cost-effective ways of enabling improvements.
Methods Rapid participatory appraisal gains information from different sources across a single locality. Participants were carers of people who had died of dementia, health and social care staff. Care pathways were modelled. Barriers identified from qualitative work were triangulated with detailed case audit and economic analysis. Results were presented to the stakeholder group who generated feasible solutions.
Results Nine cases were audited (predominantly white British male subjects, 64–84 years with advanced dementia). Six resided at home, one in a care home and two in continuing care wards. Five died in their place of residence and four in the acute hospital. Care costs over the 6 months before death were higher in care homes or continuing care (£37 029) than for those living at home (£19 854). Synthesis of qualitative and quantitative data generated areas for improvement: patient pathway, impact of acute hospitalisation, economic costs, advance care planning, impact on carers, skills and training of health and social care staff.
Conclusions This method provided evidence for stimulating change in the health and social care system. Following the patient journey and identifying barriers to care allowed local context to be considered. Service providers generated solutions enabling ‘ownership’ of changes to service delivery.
Statistics from Altmetric.com
Introduction
Approximately 600 000 people in the UK have dementia (10% of those over the age of 65 years) but it is estimated that by 2026 this number will approach 840 000 rising to 1.2 million by 2050.1 Worldwide, prevalence figures are also set to rise steeply; currently an estimated 24.3 million people have dementia, and this will double every 20 years to 81.1 million in 2040.2 In the UK, one in three people over the age of 65 will die while suffering from dementia.3 Despite this high mortality, clinicians often fail to adopt a palliative or supportive approach,4 and many people with dementia have poor access to good quality end of life care;5,–,7 services have struggled to meet their needs and provide adequate support for professional and family carers.8
Improving end of life care for people with dementia requires a multidisciplinary and holistic approach. Interventions to improve care are necessarily complex, involving several interacting components. Attention must be paid to the community, and the health, social, political and economic contexts in which such interventions may take place.9
Health and social care organisations are complex adaptive systems. They are unique and interventions may not easily be moved from one system to the next.10 Rapid participatory appraisal11 is a technique where information is gained from different sources in a study community and the results combined to describe the present delivery of, and possible improvements to, service provision;12 researchers and participants work collaboratively to address issues in specific systems.10 The research may, in itself, facilitate change,11 and is thus a form of ‘participatory action research’. This method has previously been used to assess and improve palliative and primary care services for older people.11 13 Marie Curie Cancer Care launched its ‘Delivering Choice Programme’ in 2004 after identifying how poor end of life care was often related to inadequate service design, poor coordination of care, lack of communication between professionals and inadequate support for carers. It is a form of rapid participatory appraisal which works in partnership with the health service, voluntary sector, social care and other providers to develop patient-focused models serving local needs.14 As yet, the delivering choice programme has not focused on people with dementia.
In this study, our aims were to identify barriers for people with dementia and their carers to access good quality end of life care and to identify cost-effective ways of enabling service improvements based on these findings. A secondary aim was to assess the effectiveness of rapid participatory appraisal as a methodology for improving end of life care in this patient group.
Detailed findings from the qualitative study are presented in the accompanying paper. Here, we present the results of the case note audit, patient pathways, economic analysis and the overall outcomes of the project.
Methods
Study setting
The London Borough of Haringey has a total population of 225 000. People aged 65 years or over comprise 11.1% of residents and an estimated 1579 people have dementia.15 Haringey is a deprived and culturally diverse area. It is served by a number of acute hospitals, a mental health trust and care homes. Social services are provided by the local council.
Project management and oversight
The project was overseen by a steering group consisting of a range of local professionals and stakeholders including a palliative care consultant, district nurse, general practitioner (GP), adult social services and primary care trust commissioners, a representative from Alzheimer's Society, an expert by experience, community matron and a consultant from Mental Health Services for Older People. The group developed the project protocol and met four times during the course of the study.
Ethical issues
Ethical approval was sought from Barnet Enfield and Haringey Mental Health trust ethics committee. They designated this project as ‘service evaluation and assessment’. Therefore, ethical approval was not required under UK research governance arrangements. The study was endorsed by the social services of the local Council.
Sources of data
We conducted qualitative interviews with carers of people with dementia, focus groups with health and social care professionals (50 participants), a structured case note audit, mapping of patient pathways and economic data analysis. Methods for the qualitative study are described in the accompanying paper.
Case note audit
The medical records of nine people with dementia who had died in the preceding 9 months were reviewed; these were purposively sampled by gender and ethnic background and all had accessed Mental Health Services for Older People. Data were collected using a structured audit tool and included patient demographics, severity of dementia, other comorbid conditions and carer demographics. Information on all health and social care services received 6 months before death was collated from a range of sources: acute hospital notes (each hospital admission and outpatient contact), primary care trust (for community based services, that is, occupational and physical therapy and mental health trust clinical records, electronic and paper notes), social services (including private care agencies), nursing and care home records and one general practice. Where possible, we documented medications prescribed and home equipment provided. Despite receiving the necessary permissions some general practices would not grant access to primary care records. We also examined care plans created in the 6 months before death and the place of death.
Economic analysis
This information was gathered from the case note audit.
Unit costs
Inpatient hospitalisations were costed using Health Related Group (HRG) tariff payments which reflect the payments made by primary care trusts to hospitals during the 2008/2009 reference period.16 We have assumed that the tariff applied covers all costs during the hospital admission, including contacts with consultants and allied health professionals. While HRG tariffs do not reflect the actual cost of care, they are relevant for this analysis which seeks to understand costs from a commissioning perspective. Thus, they do not reflect the actual costs of care provided by the National Health Service (NHS) hospital but the cost to commissioning bodies and the UK Department of Health. Costs for community health and social services were estimated at 2007/8 values, and inflated to 2009 values using the UK Office for National Statistics health-specific inflation index. Medications were estimated at 2009 UK prices. Costs of equipment were obtained from the council (purchase and installation costs only).
Calculating cost of care
The financial cost of care was estimated for each patient over the 6 months prior to death, and was broken down into:
Health costs: hospital inpatient admissions, outpatient services, prescribed medications and community health services
Social services costs: accommodation (eg, residential home or continuing care ward), formal care (paid carers, sitters and day care attendance), equipment and community social services (social workers, occupational therapists).
Informal costs of care associated with patients living at home were included separately to estimate costs from a societal perspective. Costs of care are presented as weighted averages across all nine patients.
Synthesis of findings
Pathway of care
Pathways of care were modelled by combining findings from the case note audit (individual patient pathways) and the qualitative work. We mapped the health and social care professionals involved at each stage, common points of transition and the typical trajectory taken by patients over the last 6 months of life (figure 1).
End of life pathways for people with dementia. GP, general practitioner; PwD, person with dementia.
Identification of key themes and development of recommendations
Barriers identified from the qualitative study were triangulated with the findings of the detailed case note audit and economic analysis, ensuring that each finding was validated by data from two other types of data source. Face validity was ensured by presenting the themes at a meeting of the project steering group who then generated recommendations to improve care. Subsequently, local service providers attended a public meeting at which solutions for each recommendation were generated, thus ensuring that these reflected local needs were feasible and deliverable within available resources.
Results
Case note audit
Demographics and clinical characteristics
The majority of patients were white British males aged from 64 to 84 years. Most were in the advanced stages of Alzheimer's disease with multiple comorbid conditions, particularly osteoarthritis and hypertension; two also had cancer (table 1). Six people were cared for in their own home, one in a care home and two on a continuing care wards. Of these, five died in their usual place of residence. Most carers were spouses, sons or daughters. Carers were generally resident with the person if they were the spouse. For those residing in the same home, carers' input was substantial including the provision of physical care, emotional support, shopping, cleaning and control of finances.
Demographic and clinical characteristics
Services used
Two-thirds of the patients had a care plan documented in the 6 months prior to death; however, we found only one case where advance care planning discussions had occurred and none had completed any formal advance care planning documents. Two-thirds of participants were cared for in their own home. A wide range of health and social care services were accessed during the last 6 months of life (table 1). Many required specialist equipment, mostly supplied by social services (two required hoists, one a stair lift, two hospital beds and two families needed to fit downstairs showers).
Economic costs
The average cost of care over the 6 months prior to death was estimated at nearly £25 000 per patient (table 2). This represents the cost to the Department of Health and to Social Services for purchasing goods and care for these patients. Costs were higher for patients in care homes or NHS continuing care (£37 029) than for those living at home (£19 854). Average costs associated with informal (family) care provided to patients living at home are estimated at an additional £36 844 over the last 6 months of life. For patients residing in care homes or hospital wards, the bulk of the cost was related to their accommodation; whereas for patients residing at home, the majority of costs comprised formal paid care, hospitalisations and housing modifications. Hospitalisations represented almost 20% of the total average costs of care and occurred more frequently among patients cared for at home.
Cost of care over the last 6 months prior to death
▶ The steering group recommended that strategies should be generated to promote cost-effective interventions in the community (such as schemes to avoid acute hospital admission) rather than in the acute hospital.
Acute hospital admissions
This was common in the last 6 months of life with five of the patients requiring one or more unplanned emergency hospital stays (length of stay 2–40, median 12 days) (table 3).Pneumonia and dehydration were the commonest causes of admission.
Acute hospital admissions during the last 6 months of life and care received at death
Impact of acute hospitalisation
Over half of the patients underwent repeated admissions as death approached. Carers described how the person with dementia often experienced a marked deterioration during admission and how hospital staff had difficulty managing people with dementia.
▶ The steering group recommended more pro-active case management for people with dementia who are becoming increasingly frail to avoid unnecessary acute hospital admissions.
The dying phase
Four patients died in the acute hospital and the remainder in their usual place of residence (two on continuing care wards and three at home). No record was found of the palliative care team being involved in any case and only one person was on an end of life care pathway. None had any documentation of their spiritual needs or an advance care plan. Because of difficulties in accessing GP records, we were only able to obtain cause of death in four people (three died of pneumonia, one of chronic kidney disease).
Patient pathway
The care pathway experienced by people with dementia during the last 6 months of life was complex, involving a wide range of services, professionals and informal family carers (figure 1). This was complicated by medical and social crises which often precipitated an unplanned emergency admission to the acute hospital.
▶ The steering group recommended that a multi-agency dementia care pathway should be developed within the borough.
Advance care planning
In only one case did we find evidence of advance care planning discussions. The qualitative work highlighted how health and social care staff felt inexperienced in supporting this process. They lacked confidence in adhering to advance directives and resuscitation orders, which were mostly completed in the acute hospital but were not seen as ‘transferable’ to the community.
▶ The steering group recommended that staff in residential and nursing homes should receive additional training and support in the use of advance care plans.
Impact on carers
The case studies revealed how some patients in this study still resided at home and that their family care givers provided a range of social, physical and emotional support. The range of specialist equipment required highlighted the severity of physical disability. Our qualitative work illustrated how both family care givers and professionals reported difficulties and confusion in accessing services and how there was little support available to families and care home staff after bereavement.
▶ The steering group recommended that the borough's Admiral Nurse (carer support) Service should be more widely publicised and that bereavement support should be provided to nursing home staff.
Skills and training of health and social care staff
The case studies demonstrated the range of health and social care staff that came into contact with people with advanced dementia. The qualitative work illustrated how they realised that they needed to improve their skills. In particular, nursing staff in the acute setting had limited knowledge or resources to cope effectively with behaviours associated with dementia.
▶ The steering group recommended improved training for acute hospital staff in caring for people with dementia.
Areas of good practice
The qualitative work identified a number of strengths in services. The community matron for care homes had improved links between health and care home staff, resulting in better communication and easier access to medical advice. The Gold Standards Framework is being piloted in four care homes in the borough and this training was valued by care home staff. Carers spoke highly of services such as the Memory Clinic and Admiral Nursing.
Generating solutions
To complete the research cycle, a public meeting was held with key stakeholders including health and social care commissioners, GPs, acute hospital and mental health staff, past and present carers, the project steering committee and project team. A number of the recommendations made by the steering group were subsequently implemented. A working group was established to develop a dementia care pathway across the borough. Acute hospital geriatricians, community staff, including matrons, GPs and old age psychiatrists have set up a weekly ‘virtual ward round’: a conference call where frail older people whose health is deteriorating are discussed and action taken to avoid inappropriate hospital admission. This service is now being evaluated. Implementation of the Gold Standards Framework has been extended to include all nursing homes in the borough and training, commissioned by the primary care trust, is available to support this. Bereavement counselling has been made available to all staff working in local authority care homes and staff in the acute hospital are receiving training in dementia care.
Discussion
Main findings
Although this study was conducted in a health and social care locality, many of the difficulties we found have been identified at a national level. For example, the lack of clear integrated dementia pathways has been highlighted in the English National dementia Strategy,6 the UK Alzheimer's Society have recently reported on the poor quality of care received by people with dementia in the acute hospital and advance care planning for people with dementia has not yet become part of routine practice in the UK.17 18
Our estimates of the economic costs of caring for someone with dementia over the last 6 months of life (£24 278 average cost, this being £19 854 for those still residing in their own homes compared with £37 029 for those in care homes or continuing care wards) are comparable with those reported by the Alzheimer's Society in 2007 which estimated the total annual cost per person with severe dementia in the community was £37 473 and £31 296 for people in care homes.1 Although those estimates were made over a year, we found that most costs are incurred in the last 6 months of life because care requirements and hospitalisations increase in the months prior to death. Informal care and accommodation made up the bulk of total costs and these costs are generally unavoidable; however, a number of costly hospitalisations (average £4356) were due to ambulatory care sensitive conditions such as pneumonia or dehydration which if detected promptly, could have been treated within the community. Because of the sample size and case selection these costs are indicative rather than definitive but suggest that caring for someone in their own home appeared to be less expensive than a care home or continuing care ward. However, this difference may be attenuated by the costs of acute hospital admission which were more frequent in those residing in their own homes.
Methodological considerations
The use of different research methodologies (individual qualitative interviews, focus groups, health economics and detailed case note review) enabled us to ‘triangulate’ our findings, ensuring they were robust and representative. The whole systems approach following the patient journey through a range of provider organisations identified barriers to providing good quality care. Involving local stakeholders, many of whom such as health and social care commissioners or ambulance staff may not traditionally be included in dementia or palliative care research, increased the sense of ownership of the findings and encouraged them to generate their own solutions. As with other researchers in this field, we found that the qualitative interviews provided the richest data.12 Using rapid participatory appraisal as a framework was key in trying to effect sustainable changes; this method allows local context and needs to be considered and is a more dynamic process than traditional health services research techniques.19
Limitations
Our study was limited by the numbers of carers from whom data were collected (seven in total). We reviewed only nine sets of case notes. This was due to difficulties retrieving case notes of deceased patients both in secondary and particularly in primary care. Thus, we mainly had access to the notes of people with dementia who had, at some stage, been cared for by the mental health trust and consequently our sample contained a higher proportion of patients who had been on continuing care wards (22%) than would be expected. This number of subjects is typical for similar projects,12 and representative of the older population of the borough. Our data for the qualitative arm were more robust with a total of 50 participants. We were careful to triangulate all of our findings with at least two other sources. We were unable to ascertain the views of people with advanced dementia, relying on the proxy views of their carers. We also had great difficulty in accessing general practice notes, despite having the required permissions, and therefore may have underestimated costs or missed important events.
Translating research into changes in service provision and practice
End of life care involves complex interactions among patients and carers, primary, secondary and tertiary health and social care providers, allied health professionals and ambulance staff, care home providers and health and social care commissioners.9 While setting up the project we ensured that we engaged local opinion leaders and providers, including healthcare commissioners; this significantly reduced the barriers to organisational change.20,–,23 The steering group made recommendations to improve care but it was local service providers who generated the solutions to these recommendations at a public meeting. Thus, there was ‘ownership’ of changes to service delivery.
We found different types of data were of use to different service providers; for example, the qualitative interviews that illustrated distress in nursing home staff when a resident died led to social services providing bereavement counselling while economic data were essential in informing discussions with health and social care commissioners who ‘purchase’ care.
As with many types of action research, we cannot be sure exactly which elements of our project may change outcomes. Rapid participatory appraisal is a dynamic process which is ongoing;19 the project steering group has developed into a working group which continues to review and improve the local dementia care pathway, with a plan to repeat the rapid participatory appraisal in 2012 for this patient group. It should be noted that at the time of this project there were wider changes in health policy and guidance, in particular the release of the English Government's National Dementia Strategy.6 The fact that many of our recommendations were concordant with this did, however, strengthen the changes that were made, ensuring that they were relevant at a local level. There are also ongoing changes within the national policy context which must be considered. In particular, the UK government's proposed changes to abolish primary care trusts and introduce general practice based commissioning highlight the need for ongoing review.
Conclusions
This work has illustrated that people dying with advanced dementia have complex medical and social needs. Using a rapid participatory appraisal framework, we involved family carers and a wide range of professionals from health and social care sectors to generate ideas regarding feasible and realistic service improvements, mostly within existing resources. The ongoing nature of this work allows constant local comparisons to be made enabling a continuous cycle of improvements to be considered.
Acknowledgments
The authors thank the staff and carers who participated in the project and the steering group for their support.
References
Footnotes
-
Competing interests None.
-
Ethics approval Approval provided by the Barnet Enfield and Haringey Mental Health Trust.
-
Provenance and peer review Not commissioned; externally peer reviewed.