Table 2 Details of Included Studies

Tay et al., (2020) [50]

Prospective cohort

Singapore

The objective of this study was to identify modifiable factors associated with the comfort of dementia patients dying at home and families’ satisfaction with care.

202 deceased people with dementia, with a median age of 88.

A palliative home care programme consists of a skilled MDT, e.g. trained ACP and in the use of dementia-specific tools to assess patient’s symptoms and QoL.

Relevant outcomes included the Comfort Assessment in Dying with Dementia scale which assessed dying patients’ comfort. The Satisfaction with Care at the End-of-Life in Dementia scale evaluated family caregivers’ satisfaction 2 months after bereavement.

Independent factors associated with families’ satisfaction with care were comfort [β (95% CI) =0.149 (0.012-0.286), P =.033] and honouring of medical intervention preferences (96.0%) [β (95% CI)=3.969 (1.485-6.453), P =.002].

Tilburgs et al., (2020) [24]

RCT- single blinded cluster

Netherlands

To explore the effects of an educational intervention for GPs.

38 Dutch GPs across different practices. 140 PwD across the selected practices.

Intervention group GPs were trained in ACP, including shared decision-making and role-playing

exercises. Control group GPs provided usual care. at 6 month follow-up participants medical records were analysed using random effect logistic and linear regression.

Primary outcome was ACP initiation. Secondary outcomes included the number of medical and non-medical preferences discussed with the GP.

Compared to the ‘care as usual’ GPs, the GPs in the intervention group discussed a mean additional 0.8 medical preferences (95% CI: 0.3, 1.3; p =.003), and 1.5 nonmedical preferences (95% CI: 0.8, 2.3; p <.001) per PwD.

Bryant et al., (2019) [22]

SR (relevant included study = Pre/Post test)

Mixed, (1 relevant = US)

To examine the effectiveness of interventions in increasing

ACP for PwD.

4 studies, 1 relevant

Advanced Care Treatment-Plan intervention

with caregivers of PwD involving weekly educational sessions across 4 weeks delivered by advanced

practice nurse facilitators. Intervention took place in adult day care centres.

Knowledge about dementia and EoL treatments,

Self-efficacy regarding ACP and decisions to develop an ACP for PwD.

Findings strongly

suggest that the ACT-Plan intervention is feasible and appropriate for

caregivers. The intervention group showed an increase in outcomes under study compared to the control group. Those in the intervention group also developed more ACP.

Hum et al., (2019) [49]

Prospective cohort study

Singapore

To explore patients’ symptoms and quality-of-life, and their association with eternal feeding. To evaluate the impact of the programme on these parameters and examine family caregiver burden.

458 patients were enrolled in the homecare service, of which 254 were included in the analysis of this study

Homecare service with an MDT of doctors, nurses, and medical social workers who make regular home visits and phone calls during office hours. During visits the homecare team assessed patients’ needs to manage pain and behavioural symptom. Caregivers were also given education on a range of topics.

Pain (Pain Assessment in Advanced Dementia).

Patient symptoms (Neuropsychiatric Inventory Questionnaire).

Quality of life (Quality of Life in Late-Stage Dementia).

Family caregiver burden (ZBI).

This integrated multidisciplinary palliative homecare team which is trained and accessible at all hours addressed the needs of people with advanced dementia living in the home, improved the QoL, and supported families to care for their loved on at home.

Jennings et al., (2019) [53]

Retrospective cohort

US

To evaluate the effect of a dementia care co-management model on subsequent end-of-life care

PwD treated by the service who later died (N-322). Medical records of last 6 months looked at retrospectively.

.

The model involved dementia service nurse practitioners partnering with primary care providers and community organisations to provide comprehensive dementia care to PwD (any stage or severity).

Place of death.

Discussions of future care.

The majority had a SPC discussion or consultation in their last 6 months of life, 69% had SPC input at the time of death, and 66% died at home.

PwD with an advance healthcare directive within the service were more likely to have had a discussion about SPC (78% vs. 64%; P = .01), die with SPC input (74% vs. 62%; P = .03), and die at home (70% vs. 59%; P = .04).

Miranda et al., (2019) [15]

SR (8 relevant: RCT=5; case-control=2; cross-sectional=1)

Mixed (8 relevant: US=4, Italy=2, UK=1, Japan=1)

To examine evidence on home palliative care interventions in dementia in terms of effectiveness on EoL care outcomes.

Eight studies were included in the review, all home based.

A range of home based palliative care interventions involving specialist and non-specialist palliative care interventions.

A range of outcomes including patient death at home, institutionalisation, functional status, behavioural symptoms at EoL, pain, satisfaction and resource use.

The interventions which commonly focused on optimal symptom management, continuity of care and psychosocial support, showed the potential benefits of the interventions in improving EoL care outcomes.

Moore et al., (2019) [27]

SR (5 relevant: RCT=5)

Mixed (5 relevant = US, Hong Kong, France, Russia, Peru)

To explore whether interventions incorporating education regarding the progressive nature of dementia increased carers’ understanding of dementia and improved mental health burden.

11 studies were included, 5 were relevant

All were multicomponent interventions which included education/information on progressions of dementia as a core component.

Burden (Zarit Burden Interview/ Family Caregiving Burden Inventory).

Knowledge of dementia (self reported on a visual analogue scale).

Depression (Beck Depression Inventory).

There was no significant evidence found to support or disprove the effectiveness of education on progression of dementia on carers knowledge and mental health.

Sternberg et al., (2019) [51]

Pre/post quality improvement study

Israel

To examine clinical and health services outcomes of a quality improvement pilot project to provide

home hospice care for older people with advanced dementia.

People with advanced dementia (N =20) being treated in the Maccabi Healthcare Services

homecare program,

Study participants received home hospice care as an extension of their usual care for 6–7 months (or until they

died) by an MDT who were available 24/7. The MDT included nurses, occupational therapists, social workers, spiritual care. MDT members took part in workshops to improve communication between teams and enhance dementia knowledge.

Family members were interviewed using validated questionnaires, i.e. Satisfaction with care (Volicer validated tools).Caregiver burden (Zarit Burden Interview) scale.

Advanced dementia prognostic tool (ADEPT).

Functional and cognitive status (FAST tool).

Hospitalizations prevented and medications discontinued, were determined by medical record review and team consensus.

Results suggest that home hospice care for people with advanced

dementia can improve symptom management and caregiver satisfaction, while decreasing caregiver burden,

preventing hospitalisations and discontinuing unnecessary medications.

Dixon et at,. (2018) [19]

SR (2 relevant: retrospective cohort = 1; cross-sectional=1)

Mixed, (2 relevant = US)

To review existing evidence concerning the effectiveness of ACP in improving EoL outcomes for PwD.

18 studies total, 3 relevant

ACP interventions, preserving identity, planning for ACP, treatment limiting advance directive & written advance directive

Depression (CSDD), QoL (BASQID), Coping (IMMEL), Self reported anxiety (CSDD), Medicare spending, hospital death, ICU use, place of death, comfort in dying (CAD-EOLD)

Having ACP and advanced directives in place can lead to a number of improvements for PwD and carer, however, there is a need for more randomised designs.

Harrison et al., (2018) [55]

Prospective cohort study (pilot)

UK

This paper describes and discusses an innovative partnership between Dementia UK and a UK hospice, and outcomes from year one evaluation and sets out future plans.

50 referrals were received between December 2016 and September 2017 (N =39 PwD, N =11 carer).

Pilot study introducing the first EoL care Admiral Nurse to a Specialist Community Palliative Care Team to identify and support PwD who required palliative and EoL care.

Level of functionality (Barthel Index).

Satisfaction with end-of-life care (SCW-EOLD).

Preferred place of death.

Inappropriate hospital admissions.

Of the 12 deaths within the Admiral Nurse caseload during year 1, most PwD (n =10) died in their usual place of residence i.e. community, and preferred place of death.

Harrop et al., (2018) [55]

Cross sectional

UK

A pilot study to introduce the first EoL care Admiral Nurse to specialist community palliative care team to identify and support PwD who require palliative and EoL care.

Total of 35 surveys were completed (20 HSCP, 9 current carers, 6 bereaved carers).

An innovative service for palliative care of EoL care for PwD introduced into a UK hospice which assessed the addition of an Admiral Nurse to the team.

Number of referrals to hospice palliative care team.

Number of referrals of PwD being care for in their home.

Family satisfaction

75% increase in referrals of PwD to the hospice palliative care team.

287.5% increase in referrals of PwD being care for in home since the pre-project year (n =8) and first year of the project (n =31).

98% of carers rated the service extremely helpful (n =10) or quite helpful (n =4).

69% of carers indicated an improvement in their knowledge, confidence and practical skills (n =11).

Spilsbury et al., (2017) [48]

Retrospective cohort study

Australia

To determine if community-based palliative care was associated with reduced hospital costs in the last year of life, and to compare people dying from caner and those from other conditions.

Cohort of all who died between Jan 2009 and Dec 2010 in Western Australia from life limiting condition including dementia.

A palliative nurse consultancy service was available to residential care facilities. The service included advise, assessment, staff education and telephone follow up. Client dates of enrolment and disenrollment were used to define periods of time receiving community-based palliative care.

Day-specific hospital costs: mean day-specific hospital cost as being a cohort average hospital cost, mean day specific hospital cost averaged over only decedents who were in hospital on a particular day.

Community based SPC was associated with hospital cost reductions across multiple life-limiting conditions including dementia.

Rosenwax et al., (2015) [46]

Retrospective cohort

Australia

To describe patterns in the use of hospital emergency departments in the last year of life by people who died with dementia and whether this was modified by use of community-based palliative care.

All people who died with dementia in Western Australia over 2 years (N =5261).

A comparative cohort of people who died from other conditions amenable to palliative care (N =2685).

Looking retrospectively at the effect of PwD who received community based palliative care and the number of emergency department admissions in last year of life, versus those who did not receive it.

Hospital admissions

Only 6% of the dementia cohort used community-based palliative care compared to 26% of the other cohort.

For PwD, the provision of community-based palliative care was associated with a significantly reduced daily rate of ED visits over the last year of life, with a six-fold difference at the EoL.

Toye et al., (2015) [26]

Pre/Post study

Australia

Action research study conducted to trial a strategy intended to support a consistent palliative approach for PwD nearing death.

130 eligible staff took part.

30 families took part.

Note:

Numbers differed for different parts of intervention

(educational sessions, educational booklet etc.).

Two plans were focused on in this paper which were providing information and education to staff and supporting families. Educational sessions and information booklets were mainly used, with the intervention being informed by numerous experts.

Knowledge (Dementia Knowledge Assessment Tool, Palliative Approach Questionnaire).

Wilcoxon Signed Rank Test comparing pre and post scores from staff educational session.

View of practice change collected by ‘end of study’ interviews with family carers.

Staff support and information sessions and resources were rated positively by staff but the actual impact in on knowledge, views, and confidence were small. Family feedback for the sessions and resources ‘supporting families’ was primarily positive.

Chang et al., (2010) [25]

Cross sectional

Australia

To investigate the application

of a new dementia information booklet for family caregivers in both the community and residential care.

672 information booklets were distributed to family carers (n =129 across the dementia services). 233 carers completed the evaluation questionnaire (not specified how many from community only).

The booklet, Information for Families and Friends of People

with Severe and End Stage Dementia (Palliative Care Dementia

Interface: Enhancing Community Capacity Project,

2006), formed the fundamental material of this project

Carer satisfaction with the booklet, Knowledge about dementia. Outcomes were measured by questionnaires and qualitative data gather from interviews.

72% indicated that their preference was to receive information either at the onset of dementia or at the time of the diagnosis of dementia, or soon after diagnosis. 97% of the carers found the booklet to be helpful and that the booklet

contained sections that were useful (96%).

86% of carers felt it should be provided freely.

Treloar et al., (2009) [41]

Exploratory retrospective

Cohort

UK

Detailed interviews of key carers who had supported PwD

at home were carried out, with the aim of identifying major factors which make such

care feasible.

Key carers interviewed more than three months after the death of PwD (N =50)

A novel service supported persons with advanced dementia to live and die at home. The service is run by a psychiatrist of old age and includes an advanced nurse practitioner acting as a key worker to coordinate care for the PwD, get necessary equipment, support the carer. Staff are available by phone out of hours.

Interviews were completed with carers, to discus satisfaction with care etc. Authors didn’t specify specific outcomes.

This study demonstrates that good, home based palliative care of dementia can be

achieved with very positive outcomes. Bereavement may be helped by the process of caring at home till death.

Key factors for success include the right equipment, expertise around relevant medication, food, and

social care needs, as well as understanding and support for funding care commissioning and informal care.

Haley et al., (2008) [37]

RCT

US

To examine the joint effects of bereavement and caregiver intervention on caregiver depressive symptoms.

Primary carers (N =254) who cared for their spouse at home

Enhanced bereavement support which comprised of three components: two individual and four family counselling sessions, weekly support session for caregivers, and ‘ad hoc’ counselling available for duration of study.

Depressive symptoms and carer resilience (Geriatric Depression Scale)

Lower depressive symptoms were found in the intervention group compares to the control group both before and after bereavement.

Shega et al., (2008) [39]

Retrospective cohort

US

To evaluate the impact of hospice enrolment i.e. SPC, on the care of PwD and describe the symptom burden PwD experience.

Family members of persons who have died with dementia (N =135).

PwD enrolled to a hospice programme compared to non enrolees. The researchers were blinded to the patients hospice enrolment status.

Location of choice for PwD.

Hospital deaths for PwD.

Pain and burdensome treatments for PwD (Verbal Descriptor Scale).

Caregiver satisfaction with care (interviews, pats adapted from the Toolkit of Instruments to Measure EoL Care).

PwD enrolled in hospice were significantly more likely to die in their location of choice, and less likely to die in hospital, compared to those who weren’t enrolled.

Caregivers of enrolees were more satisfied with the care their loved one received than non-enrolees.