Family perceptions of care at the end of life in UK nursing care homes

Abstract

Background

Over a fifth of the population of developed countries die in care homes. While studies are emerging on the outcomes of care in the last few weeks of life, few report on the experience as perceived by the family members.

Methods

As part of a wider study to improve the delivery of end-of-life care, bereaved relatives of residents who had died in a care home/hospital were sent the Family Perception of Care Scale questionnaire to evaluate their experience of care provision for their relative in the last month of life. The Family Perception of Care Scale questionnaire was posted to bereaved relatives, from 37 nursing care homes in south-east England, 3–6 months following the resident’s death. The questionnaires were posted over a 14-month period from 1 October 2009 to 31 November 2010.

Results

A total of 869 questionnaires were posted, with a 42% response rate. A global question within the Family Perception of Care Scale looking at the overall satisfaction with the quality of end-of-life care (Q24) indicated that bereaved relatives were satisfied with the care provided. Qualitative responses from family members highlighted some excellent care, although issues in relation to medical input, professional teamwork, last days of life and spiritual care remain problematic. Results provide an important insight into care provision at the end of life within these care homes.

Conclusion

While some issues can be addressed through education, relationships and value-based issues are likely to be more difficult to address in light of increasing pressure of healthcare support for UK care homes.

Introduction

Quality care at the end of life in care homes remains a critical health policy issue internationally (Costantini and Lunder, 2012; Department of Health, 2013). Despite increased use of hospice services and developments in end-of-life care in nursing care homes (NHs) in the USA, Canada and UK, the quality of care given to residents in the last month of life is not widely researched (Hockley et al., 2010; Li et al., 2013). The media coverage of UK care homes is rarely positive, with considerable pressure in relation to the high turnover of care home staff and managers (DEMOS, 2014; Keogh, 2014). With 56% of residents dying within a year of admission in UK NHs, care provision at the end of life is an important consideration (Kinley et al., 2013). In this paper, end-of-life care refers to the last weeks of life.

Older people in care homes comprise a diverse population with a burden of chronic and terminal conditions, including high rates of dementia (Alzheimer’s Society, 2013). Consequently, there are a number of challenges to the provision of good end-of-life care. These include identifying prognosis (Brown et al., 2012), pain and symptom management (Husebo et al., 2011; Kaasalainen et al., 2012), advance care planning (Froggatt et al., 2009; Stone et al., 2013) and appropriate medication at the end of life (Kinley and Hockley, 2010). As a result of such factors, unnecessary hospitalisation is common (Harrison et al., 2016). Educational gaps in the training of care home staff coupled with communication problems between healthcare providers, family members and residents represent major challenges to providing quality end-of-life care (Brazil et al., 2006; Oliver et al., 2004; Zheng and Temkin-Greener, 2010).

While there is increasing research in relation to end-of-life care and care homes, there is little reported research evaluating end-of-life care from the perspective of family members. The importance of capturing bereaved people’s views on the quality of care provided in the last months of life is highlighted by a number of national surveys that have taken place in England using the Views of Informal Carers – Evaluation of Services (VOICES) questionnaire (Burt et al., 2010). While VOICES was developed to evaluate quality of end-of-life care in a number of different settings, it has not been validated in care homes for the UK. A systematic review looking at validated tools to assess satisfaction with end-of-life care in care homes recommended the Family Perception of Care Scale (FPCS) (Parker and Hodgkinson, 2010). The FPCS, developed in Canada, contains 25 Likert-type questions relating to four domains: resident care, family support, communication and rooming (physical environment) (Vohra et al., 2006). It is considerably shorter than the revised (short-form) version of VOICES – VOICES-SF (Hunt et al., 2014) – and validated for use in care homes, albeit in Canada.

User satisfaction has become a key attribute to measure the quality of care provided (Healthwatch, 2017; Jeon et al., 2012). For researchers, policy makers, regulators and commissioners, assessing ‘satisfaction with care’ may be used as an evaluative outcome. For care home directors, regular measurement of user satisfaction can assist organisations in meeting regulatory requirements. User satisfaction information is becoming the basis for comparing service delivery and improvement (Brazil, 2009). In care homes, user satisfaction with care is rarely undertaken to measure quality of end-of-life care.

This paper reports data from bereaved relatives’ satisfaction with end-of-life care through their completion of the FPCS. Data are taken from a wider study evaluating different ways of facilitating a UK practice development programme in end-of-life care in NHs (Hockley et al., 2013; Kinley et al., 2013, 2014). In the UK, NHs have on-site nurses working alongside care workers but they rely on external medical support.

Methods

A cross-sectional postal questionnaire survey was undertaken (October 2009–November 2010) to evaluate bereaved families’ perception of the provision of end-of-life care across 37 NHs in south-east England who were in the first year of the 2-year Gold Standards Framework programme (www.goldstandardsframework.org.uk). The next of kin was sent the FPCS questionnaire by the NH manager 3–6 months after the death, as part of a wider study (Kinley et al., 2014). Individuals were excluded if the relative had physical/mental health problems preventing completion of the questionnaire, the next of kin had already had a bereavement of another relative within the NH during the study, or the next of kin would be unable to complete the questionnaire in English.

The questionnaire and supporting documentation (letter signed by NH manager, information sheet, consent form, ‘decision not to participate’ form, and details of a social worker external to the research team who could give support if necessary) were sent with a stamped addressed envelope to the next of kin. Instructions were given to return either the questionnaire or the ‘decision not to participate’ form to the research team; when either were not returned within 4 weeks, one reminder letter was sent from the NH manager. A ‘thank you’ card was sent to those who returned completed questionnaires.

Ethics approval was granted by the joint UCL/UCLH committees on the ethics of human research in 2009, REC reference number 09/h0715/74. The larger study was registered on the ISRCTN registry (trial number: ISRCTN76029577).

An adapted version of the FPCS was used in this study. The original questionnaire consists of 25 items in which bereaved relatives indicate their agreement as rated on a seven-point Likert scale in which a score of 1 is least effective and a score of 7 is most effective (Vohra et al., 2006). The FPCS also asks respondents to identify three items as priorities for providing quality end-of-life care. An opportunity is given to include written comments. For the purposes of this study, some changes to the original FPCS were made to fit the UK context, as follows:

• Minor changes were made to the phrasing of seven questions.

• Q19, ‘My family member was placed on an appropriate floor/unit’, was removed, with permission from those who validated the original form, as it was not applicable to this study – all NH residents remained in their own rooms.

• The FPCS questionnaire was posted to all bereaved relatives, irrespective of the resident’s place of death.

• A ‘not applicable’ option was also provided.

• The total possible score ranged from 24 (negative assessment) to 168 (highly positive assessment).

• Bereaved relatives were also asked to identify the place of death, their relationship to the deceased resident and their age.

Data analyses of the quantitative data were undertaken using IBM SPSS version 21. Data are presented as frequencies and mean (standard deviation (SD)). Overall respondent satisfaction was calculated by adding the respondent responses to all FPCS questions. Thematic analysis was used to analyse respondents’ hand-written comments. All hand-written comments were typed up and stored in a database. The comments were read, and words, phrases and concepts were highlighted to develop codes. An initial list of codes inductively derived was developed. When new concepts emerged they were added to a revised coding framework. Similar and related codes were organised into higher-level categories, which were restructured into themes to form a coherent whole. JK and JH engaged in a reiterant process of discussing areas of agreement and disagreement to enhance analytical rigour and achieve consensus.

Results

Quantitative

The FPCS questionnaire was sent to 869 bereaved relatives from 37 NHs, generating a response rate of 42% (366/869). A total of 77% of residents (277/366) died in the NH. Most respondents were sons or daughters of the deceased resident (60.0%) (see Table 1).

Table 1.

Characteristics of those returning the Family Perception of Care Scale (FPCS).

	n (%)
Bereaved relativesa (n = 366)
Gender (% female)	242 (66.1)
Age (years)
30–49	28 (7.9)
50–59	117 (33.1)
60–69	122 (34.6)
70–79	49 (13.9)
80+	36 (10.2)
Relationship to resident
Son/daughter	218 (60.0)
Spouse/partner	67 (18.5)
Sibling	14 (3.8)
Friend	10 (2.8)
Son-in-law/daughter-in-law	8 (2.2)
Parent	1 (0.3)
Other	45 (12.4)

^aDenominators vary among variables due to missing responses.

All NHs were from within the independent sector – 70% of NHs were run ‘for profit’. In total, 54% of NHs were part of large cooperatives (see Table 2). The average number of beds was 59 (range 22–160). The number of GPs supporting the NH varied. While 43% of NHs were supported by one GP practice, eight NHs (22%) were supported by four or more GP practices.

Table 2.

Nursing care home characteristics.

NH	Mean FPCS score for Q24: ‘overall satisfaction with the end of life care’ – where 7 is the highest	Number of beds	Type of home
			Small single private provider	Part of cooperative
1*	6	50	✓
2	5	24	✓
3	4	60		✓
4	4	88	✓
5	4	56	✓
6	7	46	✓
7	5	92
8*	6	44	✓
9	6	22	✓
10	6	55		✓
11	5	39	✓
12	5	110		✓
13	5	62		✓
14	6	85		✓
15*	7	48	✓
16	7	45	✓
17	7	59		✓
18	6	100		✓
19	7	30		✓
20	6	28	✓
21*	6	78		✓
22	7	30	✓
23*	5	67	✓
24	7	34		✓
25	6	74		✓
26	6	51		✓
27	6	56		✓
28	4	22	✓
29*	7	75		✓
30*	7	28		✓
31	6	160	✓
32	7	110	✓
33*	7	60		✓
34	7	75		✓
35*	5	38		✓
36*	6	47		✓
37	6	50	✓
	* = not-for-profit NHs

NH: nursing care home; FPCS: family perception of care scale.

All analyses on the FPCS were collapsed across NHs. Overall, the quantitative scores reported that bereaved relatives were satisfied with the ‘quality of end-of-life care provided’ (Q24), revealing a mean score on the adapted FPCS of 135 (SD 30.17) from a possible total score of 168 (see Table 3). The majority of the not-for-profit care homes (n = 10/11) had a mean FPCS score of 6 or 7, nine of which had 60 beds or fewer; 75% (n = 9/12) of NHs with a bed number of 60 or fewer had a mean score of 7.

Table 3.

Family Perception of Care Scale (adapted).

Question	Item	n	Mean (SD)
1	The staff were friendly to me	361	6.5 (1.0)
2	The staff treated my family member with dignity	360	6.2 (1.2)
3	The staff spent enough time with my family member	352	5.7 (1.5)
4	The staff provided comfort to my family member	359	5.9 (1.4)
5	The staff were sensitive to the needs of my family member	358	5.9 (1.4)
6	The staff kept me informed about my family member’s health	359	5.9 (1.5)
7	The staff kept me updated based on what I wanted to know	359	5.9 (1.5)
8	The staff spoke to me in a way that was easy to grasp	355	6.2 (1.1)
9	The staff described what to expect as my family member came closer to death	297	5.3 (1.9)
10	The staff informed me about care options during my family member’s last days	279	5.2 (1.9)
11	The staff involved me in the planning of care	319	5.4 (1.7)
12	There was a plan of care tailored specifically to the needs of my family member	326	5.6 (1.6)
13	The staff took note of an action I thought necessary and we worked things out appropriately	295	5.6 (1.6)
14	My family member’s pain was eased to the greatest extent possible	321	5.9 (1.4)
15	Other symptoms were eased to the greatest extent possible	317	5.8 (1.5)
16	The staff informed me when they thought that my family member was dying	297	5.6 (1.9)
17	The staff welcomed me to stay with my family member	308	6.0 (1.5)
18	The staff helped me to be involved in the care of my family member	318	5.7 (1.6)
19	My family member’s room offered privacy	349	6.4 (1.0)
20	Spiritual support was at hand for my family member	272	5.2 (1.7)
21	The staff asked about and met important rites and rituals of my family member/us	248	5.0 (1.7)
22	There was someone there for my family member to talk to	316	5.7 (1.6)
23	There was enough staff to deal with my concerns	338	5.6 (1.7)
24	Overall, I am satisfied with the end-of-life care that was given to my family member	340	5.9 (1.6)
Please list, in order, the three questions in this survey that you think are the most important for excellent end of life care
1. Question # ______
2. Question # ______
3. Question # ______
	Minor change to phrasing of question

While all questions were positively rated by bereaved relatives (mildly to strongly agree), the lowest ranked questions (mildly agree) clustered around communication and family support (Q9, Q10, Q11). The other area of care in which bereaved relatives were only mildly satisfied was with the availability of spiritual support (Q20, Q21) (see Table 3).

Bereaved relatives were asked to identify three items in the FPCS they felt were most important for excellent end-of-life care (Table 4). ‘Staff treated my family member with dignity’ was the highest ranked by bereaved relatives, followed by ‘effective pain control’ and being ‘sensitive to the needs of my family member’ including ‘comfort’ care and ‘effective symptom management’.

Table 4.

Items from the Family Perception of Care Scale that bereaved relatives identified as priorities for excellent care.

Item, n (%)
The staff treated my family member with dignity, 163 (50.8)
My family member’s pain was eased to the greatest extent possible, 119 (37.1)
The staff were sensitive to the needs of my family member, 94 (29.3)
The staff provided comfort to my family member, 53 (16.5)
Other symptoms were eased to the greatest extent possible, 52 (16.2)

Qualitative

Out of the 366 returned questionnaires, 30% (110/366) next of kin had written comments on the FPCS. Considering the poor light in which care homes are generally seen, there are some very positive comments written by bereaved relatives alongside the negative (see Table 5). While the majority of not-for-profit NHs and homes with smaller numbers of beds had high quantitative scores, the majority had received some negative qualitative responses (not-for-profit NHs (n = 6/11); homes with 60 beds or fewer (n = 7/23). A number of respondents specifically stated they found filling in the FPCS helpful.

Table 5.

Qualitative comments at the end of each Family Perception of Care Scale questionnaire.

NH	Number with qualitative response	Positive comments	Negative comments	Comment was a statement unrelated to experience in NH	Mixed response
1	7/9 (78%)	3	3	1
2	0/1 (0%)
3	3/7 (43%)		3
4	10/14 (71%)	4	4	1	1
5	4/6 (67%)	2		1	1
6	1/2 (50%)	1
7	5/17 (29%)	1	2	1	1
8	0/1 (0%)
9	4/10 (40%)	2	1	1
10	1/1 (100%)	1
11	6/9 (67%)	3	1	1	1
12	12/16 (75%)	2	2	3	5
13	7/11 (64%)	1	3	2	1
14	6/11 (56%)	1	1	2	2
15	7/11 (64%)	3		1	3
16	2/4 (50%)	2
17	4/7 (57%)	2		1	1
18	12/18 (76%)	3	3	3	3
19	0/3 (0%)
20	2/8 (25%)	1	1
21	10/20 (50%)	3	4	3
22	2/8 (25%)	1			1
23	6/9 (67%)	1	2	1	2
24	5/10 (50%)	2		2	1
25	7/12 (58%)		5	1	1
26	5/8 (63%)	2		2	1
27	19/33 (58%)	6	3	7	3
28	1/2 (50%)				1
29	6/17 (35%)	3		3
30	5/10 (50%)	1	2	1	1
31	3/7 (43%)		2	1
32	3/9 (33%)	2		1
33	6/11 (56%)	2		3	1
34	6/20 (30%)		2	3	1
35	7/10 (70%)	1	3	3
36	7/12 (64%)	2	1	4
37	5/8 (63%)	3	2
Total		61	50	53	32

Four themes (both positive and negative) were identified from the free-style comments written by the bereaved relatives, as follows.

Communication and relationship-based care

Good communication with both the relative and their family member was seen as paramount to good end-of-life care. Families wanted a friendly human face – people who were ‘warm, caring and genuine’.

It is very important that the relatives and home have a good relationship. I had absolute confidence in the staff at [NH]. [R. 18]

Given the challenges of providing high quality end-of-life care in care homes, as identified earlier, this was provided when care home staff had acknowledged its importance alongside the need to form good relationships with the families of residents (despite some being many miles away):

End-of-life care for my mother was outstanding. What stood out were: kind, sensitive staff; full 24-hour access and privacy; easy communication at all stages; totally professional at all times; outstanding care. [R. 21]

… when the hospital said Mum had to go into care I was worried. I live in Yorkshire she in [SE England] and didn’t know what to expect. [The NH] was wonderful. Mum had the best care possible and they were so friendly and helped me no end … didn’t mind if I stayed all night near the end and even though Mum was not there long before she died, they felt like extended family. [R. 31]

Some relatives remarked, however, that there was too much rushing around at times for the resident to be respected and listened to; as one person put it, they wanted ‘more tender loving care’. In some NHs, it was relatives who shaped the care their family member received and it appeared to be a constant struggle for relatives to get the appropriate care they believed their family member required:

Each time it was up to the family to bring these factors to the forefront so as to be dealt with accordingly – but should have already been in place automatically. [R. 56]

Communication in relation to advance care planning was sometimes imposed by staff rather than discussed with both the resident and family being present. This resulted in one case where the true wishes of the resident had not been established:

I was very sad to hear that the nursing staff had discussed with my father that he should stay with them to die. He then asked me when he was going home and expressed a wish to die in his own home. I felt desperately sad and inadequate as I had not been consulted and it was then too late to implement his dying wish. This I will never forget. [R. 32]

Professional teamwork

Where staff worked well together with the inclusion of the GP, the care home was praised. Unfortunately, this was not always the case and examples were given of poor professionalism and teamwork:

I asked the nursing staff if a doctor would be attending and they confirmed an emergency doctor had been called. I later learnt that he did not attend but merely held a telephone conversation with someone on the nursing staff. [R. 26]

Again, while eight families specifically mentioned the support they valued receiving from specialist palliative care teams, there were situations when specialist palliative care advice was not carried through:

‘We had stand-up arguments with some of them as she writhed in pain in her bed Despite written advice from [the hospice] in front of them, I had to get them to call [the hospice] …’ [R. 32]

It was important for relatives to see good professional teamwork continue throughout the whole process – even after the death. If humane care didn’t continue after the resident had died, then good care given prior to the death was in danger of being discounted:

I was dismayed that within one and a half hours of his death, my father had become ‘the body’ to be removed immediately … This lack of sensitivity nearly obviated my appreciation of the compassion which had been exhibited to my father and to me during the previous few days. [R. 47]

Aspects of dying

There were comments from 16 relatives who specifically talked about the ‘good death’ that had occurred in the NH.

As it was she died in my arms at the NH. It was a ‘good death’… . In the last week I was there for hours at a time – and welcomed by the staff. [R. 43]

Many family members mentioned how they wanted to be there at the death – some were disappointed not to be called in time. However, it was clear from 11 relatives’ comments that staff in some of the care homes either didn’t really understand the process of dying or felt unable to speak about it to relatives. This was still the case even when the relative knew their family member was dying:

They seemed reluctant to accept that she was dying preferring rather to lecture me on the importance of getting her out of bed ‘to prevent contractures’. [R. 121]

Pain and breathing difficulties were the main symptoms that relatives commented on during the process of dying. Breathing difficulties were frightening and in many instances appeared to come on suddenly, often precipitating hospital admission. If anticipatory medication for symptoms during the last days of life were available then often residents remained in the NH and died peacefully. However, one relative spoke passionately that medication at the end of life had made their family member confused and distressed. Interestingly, of the 110 questionnaires that contained qualitative comments, 23 relatives (21%) commented that their family member died ‘suddenly’:

My mother-in-law was nearly 100 years old when she died. This was quick and without pain; she had her hair done in the morning and looked ‘lovely’, [R. 35]

Those residents whose disease trajectory deteriorated as a result of an ‘acute’ episode (falling and fracturing a bone; a stroke) were more likely to be admitted to hospital. In these instances a few respondents used the FPCS to comment negatively on the hospital care:

My mother suffered a massive stroke and although she was admitted into the hospital nearly two hours before I could get there, I had to watch her die on a gurney in a cramped dingy A&E cubicle. It was horrible. [R. 22]

A total of 21 relatives (19%) specifically mentioned that there were not enough staff, and sadly one incident was related to dying:

My wife passed away while I was with her but it took some time to find a member of staff to tell her she had gone. [R. 18]

These shortages were often reported in relation to care provision at nights or weekends – but one person talked about a lack of nurses generally in the NH.

Spiritual care

Spiritual care was very important for many relatives, a majority of whom wanted more input, especially when someone was dying. One person in particular mentioned how the priest had visited the family member in the NH for well over 5 years. Most relatives appeared to organise spiritual care themselves:

Our Methodist minister came to see my husband. There was no clergyman who visited [the home] regularly. [R. 28]

Many relatives spoke about the respectfulness with which their family member and indeed themselves were treated following the death.

Tea and biscuits were there if needed. She was respectfully treated when doctors/ funeral people came. [R. 36]

Staff attending the funeral was appreciated by relatives. This was most often seen when a rapport had been built up over time with the staff. However, when a staff member had said they would be attending the funeral but for some reason didn’t, there was disappointment. One respondent, who was a nurse specialist in palliative care, mentioned that she was very impressed by the care that other residents received following the death of her mother and the openness when a resident died.

Discussion

When evaluating service delivery and provision of care, few studies account for both the outcome and the experience of care delivery (quantitative and qualitative data) (Seers et al., 2012). In the present study, the quantitative analysis reports a largely favourable experience concerning the end-of-life care provision for participants' family members in NHs. However, the overall qualitative analysis, while supporting the positive quantitative data, does reveal nuances of underlying issues. Like all qualitative studies, the data highlight important detail; without such data it is difficult to know what to target in terms of education and improving care. Considering the poor light in which care homes are generally seen, the positive qualitative comments are encouraging, and there needs to be more overt praise for the work staff do in care homes.

A strength of the FPCS is that it is relatively short and specifically validated for use within a care home setting (Vohra et al., 2006). Within our study, there was a good response to the questionnaire (42% response rate) considering the relative had been bereaved within the previous 3–6 months. Compared to the VOICES-SF, which is twice as long with 58 questions (Hunt et al., 2014), the FPCS in this study received nearly a 10% higher response rate. This makes the FPCS potentially useful as an audit tool (Levy et al., 2016a, 2016b), especially if it can be validated for use in UK care homes. Facilitating audit in the care home sector, as a means of improving end-of-life care, may encourage financial support from commissioners (Hockley and Kinley, 2016).

The FPCS captures the experience of care when residents cannot provide a personal account of their experience. One of the main concerns of relatives highlighted in the qualitative comments on the FPCS was the lack of medical support. Medical input to UK NHs is intermittent and insufficient (Handley et al., 2014; Kinley et al., 2014). This has been the case since the closure of NHS long-stay geriatric wards in the 1990s, when funding was given to social services to provide care for frail older people in their own homes and care homes, with little consideration of medical input (British Geriatric Society, 2011). The majority of older people now admitted to NHs are frail with multiple morbidities and die within a year of admission (Kinley et al., 2013). This may suggest the need for a new model of medical provision. Alternative medical models for NHs do exist: retainer fees for GPs to attend the NH more regularly (Handley et al., 2014; Hockley et al., 2016); dedicated primary care team for NHs in the city of Glasgow (Hannah et al., 2005), although now disbanded because of equity issues; and even the idea of care home medicine (Baar, 2002; Gordon, 2015). A report from the British Geriatric Society (2011) recommends a combination of ‘enhanced primary medical and nursing care with a dedicated input from departments of old age medicine, mental health services, and other specialisms such as palliative care and rehabilitation medicine according to local needs’ (p. 4).

In the USA there has been a tradition of the development of teaching nursing homes being aligned to local universities to achieve high quality care (Hockley et al., 2016; Mezey et al., 2009). Nationally, further improvements in medical provision may occur through the recommendations emerging through the six care home vanguard sites (NHS England, 2016), the NHS 5-year forward plan (NHS England, 2017) and their associated sustainability and transformation plans (https://www.england.nhs.uk/stps/), and innovative suggestions such as the increased use of information technology. The use of technology may offer more than just greater access to medical support and staff training, providing a means of gaining user feedback about the experience of care.

Re-thinking different models of care provision is not just about medical care. Spiritual care was important for residents in their last month of life in the NH, as reported in the FPCS. Whereas 20 years ago chaplain support would have been part of the provision of end-of-life care in geriatric hospitals, this is not routinely provided in NHs. A new concept, parish nursing, currently emerging in the USA and the UK, might offer a solution (Wordsworth, 2014). Even in cancer care the emergence of validated tools to identify, measure and improve such care is just emerging (Vivat et al., 2017). Engaging care homes with those developing such tools would be one way to better ensure that the care home population is also included.

Relationships among the care staff, nurses, NH management, GPs, the resident and their family were extremely important to those responding to the FPCS. There is considerable discussion about person-centred care in NHs, which has tended to be interpreted as resident-focused care. Establishing instead a culture of relationship-based person-centred care, not only with residents and families but across the whole team, is emerging (Koloroutis, 2004; McCormack and McCance, 2017).

Recognising dying is fundamental to managing the last day/s or week/s of life (Chapman and Ellershaw, 2015). In the past, diagnosing dying has been the role of the doctor. When NH staff were inadequately supported by GPs, confidence in acknowledging dying was lacking. This situation is compounded by the fact that the dying trajectory of very frail older people is often difficult to identify (Murray et al., 2005). Given such challenges, it could be argued that most of the NHs in this study were perceived as giving high quality end-of-life care by those who returned the FPCS questionnaires. Since this study, the ability to do this well has been further enabled by the publication in the UK of the nationally recommended five priorites of care for the dying person (Leadership Alliance for the Care of Dying People, 2014). An early audit reports further positive experience when GPs and NH staff work together to provide care to residents in their last days of life (Coleman et al., 2016).

Limitations

There are a number of limitations in this study. First, the FPCS is a validated tool in Canada but has not been validated in the UK context. One question (Q19) was removed because of its inappropriateness in the UK NHs taking part. Tools evaluating the families’ perception of care at the end of life of residents in care homes are limited, and the FPCS is one that is reported as the most appropriate for the care home setting. More international validation of the FPCS is important. Another limitation is that the study only involved NHs and not care homes without on-site nursing. The NHs who took part were also likely to have been enthusiastic about end-of-life care (and perhaps were already doing well in this area of practice) as the study was done within the first year of the 2-year Gold Standards Framework programme (www.goldstandardsframework.org.uk). Any results are therefore likely to be more positive than would be those from care homes reluctant to improve their end-of-life care practice. However, it is important to be able to report positive comments in a culture in which care homes often receive negative press. A final limitation is that the NHs were based in one regional area of south-east England and thus the generalisability of findings to other care settings across the UK is unknown.

Conclusion

This study reports the experience of end-of-life care as perceived by bereaved relatives completing the FPCS across 37 NHs in south-east England. NH organisations use satisfaction surveys for many aspects of care in order to evaluate the quality of care provided, but rarely specifically evaluate the last month of life. The FPCS was well received, with a response rate considerably higher than other tools. Quantitative scores were generally satisfactory. The qualitative comments highlighted some excellent care being delivered in many of the NHs, although attention to the medical input to care homes, to the last days of life and to spiritual care still needs addressing. Further validation of the FPCS for care homes across countries is important. Once validated, commissioners of care for frail older people might be interested in the tool for audit purposes.

Key points for policy, practice and/or research

• The quality of end-of-life care is rarely audited by nursing care homes.

• Bereaved families are a good proxy for end-of-life care in nursing care homes.

• Bereaved families are keen to be involved in evaluating end-of-life care in nursing care homes.

• Bereaved families would like more medical input to nursing care homes to support the care of family members.

• Recognising dying is fundamental to managing good care in the last weeks of life.

Biography

Julie Kinley is the Nurse Consultant for Care Homes working at St Christopher's Hospice in London. She was appointed into this post in 2013 to lead and manage the Care Home Project Team (CHPT). As well as having a research, audit and education role this team provides practice development to over 150 care homes. The research studies she has previously undertaken inform the CHPT and others model of practice.

Jo Hockley is a senior nurse researcher within the Primary Palliative Care Research Group at the University of Edinburgh, UK. Her work is entirely related with the important work undertaken by staff in care homes and quality of end-of-life care for frail older people.

Louisa Stone started her career within the oncology field. However, she quickly realised that palliative care was the area she wished to focus. Louisa initially worked as a deputy manager at St Christopher's Hospice but then moved roles to become the research assistant working on a study identifying models of facilitation to support implementation and sustainability of the Gold Standards Framework within care homes with nursing. She continues to work with care homes supporting them to implement and sustain end-of-life care programmes.

Kevin Brazil holds the appointment of professor of Palliative Care in the School of Nursing and Midwifery, Queen's University Belfast. He is also a Fellow in the Centre of Social Innovation and Evidence, Queen's University Belfast, where he is the Lead for the Chronic Illness and Palliative Care Research Programme. He holds an honorary appointment in the Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Ontario Canada.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Ethical approval

Approval was granted by the joint UCL/UCLH committees on the ethics of human research in 2009, REC reference number 09/h0715/74.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This research study was funded by The Wives Fellowship and St Christopher’s Hospice.