Objective The primary objective of the study was to determine the relative importance of place of death to patients with advanced cancer being treated at a cancer centre in England.
Methods Demographic data were collected from the electronic patient record system, with additional information obtained from the patients themselves. Patients were questioned about a variety of end of life issues, including preferred place of death (PPD), and ‘acceptable’ places of death. They were also asked to rank the importance of factors previously linked to a ‘good death’.
Results 120 patients participated in the study. 51 (42.5%) patients stated that ‘home’ was their PPD, while 80 (67%) patients stated that home was an acceptable place of death. Patients from areas with worse deprivation scores were less likely to want to die at home than patients from areas with better deprivation scores (p=0.03). The most important factors associated with a good death were ‘to have my pain/symptoms well controlled’, ‘to not be a burden to my family’ and ‘to have sorted out my personal affairs’ respectively. Place of death was ranked as the seventh most important factor.
Discussion Place of death is undoubtedly an important factor in achieving a good death for some patients and carers. However, for others a home death is either unimportant or to be avoided. The results of this study, and the results of similar studies, suggest that place of death may not be a good marker of the quality of end of life care.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
The End of Life Care Strategy was developed with the aim of ‘promoting high-quality care for all adults at the end of life’ in England.1 It identifies four domains associated with a ‘good death’—that is, ‘being treated as an individual, with dignity and respect’, ‘being without pain and other symptoms’, ‘being in familiar surroundings’ and ‘being in the company of close family and/or friends’. The publication of the strategy was followed by the publication of series of quality markers and measures for end of life care.2 However, the ‘vital sign’ or key measure of success of the strategy remains the proportion of deaths that occur at home, as opposed to in hospital.3
Studies undertaken in the UK and other developed countries suggest that the majority of people would prefer to die at home, although the majority of people actually die in hospital.4 5 For example, only 22.56% cancer-related deaths occurred at home in England and Wales in 2005.6 It should be noted that this statistic does not include people who died in places such as care homes and nursing homes, which are home for many older people. Nevertheless, studies undertaken in the USA suggest that place of death is relatively unimportant to patients with advanced disease in achieving a good death.7 8
The primary objective of this study was to determine the relative importance of place of death to patients with advanced cancer being treated at a cancer centre in England. In addition, the study attempted to prospectively identify factors associated with a preference to die at home or elsewhere.
The study was conducted at the Royal Marsden Hospital (RMH), which is a major cancer centre in England. The study was approved by the RMH Committee for Clinical Research, and a local research ethics committee. Patients were given an information sheet about the study, time to consider the study, the opportunity to discuss the study and asked to provide written consent before entering the study.
Subjects were recruited from both inpatient wards and outpatient clinics at the RMH. The inclusion criteria for the study were (1) diagnosis of locally advanced/metastatic cancer; and (2) age >18 years. The exclusion criteria for the study were (1) inability to give informed consent; (2) inability to complete questionnaire and (3) estimated prognosis <1 week.
Demographic data were collected from the electronic patient record system, with additional information obtained from the patients on matters such as accommodation and co-habitants. Post codes were used to calculate electoral ward-level deprivation scores (Carstairs scores).9 Carstairs scores are derived from four of the variables included in the 2001 census: (1) unemployment; (2) overcrowding; (3) car ownership; (4) low social class (social class IV and V). Carstairs quintiles are obtained by ranking the Carstairs scores; each quintile contains 20% population, with quintile 1 containing the least deprived areas and quintile 5 containing the most deprived areas.
Patients were asked to complete the Memorial Symptom Assessment Scale–Short Form (MSAS-SF).10 The MSAS-SF is a validated tool for the assessment of symptoms in patients with cancer; it can be used to derive a physical (PHYS) subscale score, a psychological (PSYCH) subscale score and a Global Distress Index (GDI) score. The PHYS subscale score is the average of the distress ratings of 12 defined PHYS symptoms; the distress ratings are: 0, symptom not present; 0.8, ‘not at all’; 1.6, ‘a little bit’; 2.4, ‘somewhat’; 3.2, ‘quite a bit’; 4.0, ‘very much’. The PSYCH subscale score is the average of the frequencies of six defined PSYCH symptoms; the frequencies are: 0, ‘symptom not present’; 1, ‘rarely’; 2, ‘occasionally’; 3, ‘frequently’; 4, ‘almost constantly’. The GDI is the average of the distress rating of six defined PHYS symptoms and the frequencies of four defined PSYCH symptoms.
Patients were also asked to complete a questionnaire about a variety of end of life care concerns, including preferred place of death (PPD), ‘acceptable’ places of death, possession of a last will and testament, possession of an advance statement, any discussions with family or the general practitioner or oncology team concerning end of life care choices, and any involvement with community specialist palliative care services (online appendix 1). They were also asked to rank the importance of factors previously linked to a ‘good death’7; these factors were ‘to feel my life is complete’, ‘to be at peace with my God’, ‘to be involved in decisions about my care’, ‘to be able to die at home’, ‘to have sorted out my personal affairs’, ‘to have sorted out my funeral arrangements’, ‘to not be a burden to my family’, ‘to not be a burden to society’, ‘to be mentally alert (until death)’ and ‘to have my pain/symptoms well controlled’ (Figure 1).
The study was conducted in a quiet section of either the inpatient ward or the outpatient clinic. The MSAS-SF and the questionnaire could be completed either by the patient or by the researcher, if preferred or felt to be more appropriate.
The data were analysed using SPSS (version 19) software. Univariate binary logistic regression was used to determine the relationship between PPD and other recorded variables (age group, gender, cancer diagnosis, marital status, co-habitant status, deprivation scores, involvement with community specialist palliative care services). A Mann–Whitney test was used to compare the non-normal continuous variables of the subscale scores: MSAS-SF PHYS, MSAS-SF PSYCH and MSAS-SF GDI. A p value of ≤0.05 was deemed to be statistically significant.
One hundred and twenty patients participated in the study. The median age was 61 years (range 32–85 years), and there were 66 men and 54 women. Table 1 contains further demographic information about the study population.
The patients recorded a median of 12 symptoms (range 1–29 symptoms) each on the MSAS-SF. The most common symptoms were lack of energy (90%), pain (77.5%), feeling drowsy (70.8%), lack of appetite (61.7%) and dry mouth (60%). The median score on the PHYS subscale was 1.07 (range 0–3.27), the median score on the PSYCH subscale was 0.77 (range 0–3.47), while the median score for the GDI was 1.19 (range 0–3.24).
Eighty (67%) patients reported regular contact with community-based specialist palliative care services. Ninety (75%) patients had completed a last will and testament, but only 12 patients had an advanced statement or similar document. Forty-nine patients had discussed their wishes about place of death with their family, although few patients had had this discussion with their general practitioner (n=9) or their oncology team (n=7). Forty-four patients had discussed their wishes about medical treatment in the event of losing capacity with their family.
Table 2 contains information about PPD, and also ‘acceptable’ places of death.
At the time of writing, 4 months after the end of the study, 71 (59%) of the study population had died, with 34 dying in a hospice, 20 dying at home and 17 dying in a hospital. Data on current status/place of death were derived from the electronic patient record system and telephone correspondence with the patients' general practitioners and community palliative care teams. Overall, 32 (45%) patients had died in their PPD, with a further 21 (30%) patients dying in an ‘acceptable’ place of death (table 3). Thus, 16 (22%) patients had died in an unacceptable place of death. There was no statistical difference between the number of patients achieving their PPD of home and the number of patients achieving their PPD of a hospice (p=0.162). Table 4 shows the patients' rankings of the importance of factors linked to a ‘good death’.
There was a significant association between the Carstairs quintiles and the PPD; patients from areas with worse deprivation scores were less likely to want to die at home than patients from areas with better deprivation scores (p=0.031). However, there was no association between the PPD and age, gender, cancer diagnosis, marital status, co-habitant status, involvement with community specialist palliative care services, MSAS-SF PHYS subscale score, MSAS-SF PSYCH subscale score, or MSAS-SF GDI subscale score (table 5).
The concept of a ‘good death’ varies across cultures,11 12 and also within the same culture.7 13 Thus, studies have reported important differences between the views of patients, carers and healthcare professionals.7 8 14 15 These differences relate both to factors linked with a good death, and also to the relative importance of these factors.
In this study the most important factors associated with a good death were ‘to have my pain/symptoms well controlled’, ‘to not be a burden to my family’, ‘to have sorted out my personal affairs’, ‘to be involved in decisions about my care’ and ‘to be mentally alert until death’, respectively. Nine patients ranked ‘to be able to die at home’ as their most important factor, and overall patients ranked this domain as the seventh most important factor. The data from this study on the relative importance of dying at home are consistent with the data from analogous studies undertaken in patients with advanced disease from North America.7 8 16
It is of note that, Payne et al14 reported that none of their English patients with advanced cancer mentioned place of death when asked for their thoughts about a good death. In contrast, place of death was an important factor for the majority of healthcare professionals interviewed in the same study. Similarly, Vig and Pearlman17 reported that none of their American patients with advanced cancer or heart disease mentioned place of death when asked for their thoughts about a good death. However, when subsequently questioned about dying at home, 23% patients rated this as ‘very’ important, and 19% patients rated this as ‘quite’ important.
The End of Life Care Strategy endorses the recording of a patient's preferred place of care and their PPD. It should be emphasised that these are very different parameters, and that both of these parameters can alter over time.18 Indeed, several studies suggest that the number of patients with cancer wanting to die at home decreases as the disease progresses.18,–,20 One of the interesting things to come out from this study is that while patients usually have a PPD, they also often have alternative ‘acceptable’ places of death when questioned. Of note, only 80 (67%) patients stated that home was an acceptable place of death, although 97 (81%) patients stated that a hospice was an acceptable place of death.
The study has a number of limitations. In particular, it was conducted at a single institution, which may have influenced the responses about PPD. Thus, 10% of patients stated that the oncology centre was their PPD and 64% of patients stated that the oncology centre was an acceptable place of death. Similarly, the study was conducted in a mainly white population that generally lived in privileged, as opposed to deprived areas. Nevertheless, as discussed above, the data from this study are in keeping with the data from analogous studies undertaken in patients with advanced disease from North America.
Place of death is undoubtedly an important factor in achieving a ‘good death’ for some patients and carers. However, for some a home death is either unimportant, or is something to be avoided. Indeed, Parkes has commented that ‘home can be the best place or the worst place to die’.21 Furthermore, Gomes et al22 have pointed out that there is little evidence to suggest that those who die at home experience better care than those who die in hospital. Hence, we would suggest that the key measure of success of the End of Life Care Strategy should be the proportion of people who achieve a ‘good death’, as defined by the relevant individual, rather than the proportion of people who die at home or in hospital. In addition, we would recommend that patients are asked about acceptable and unacceptable places of care and death, and not simply about preferred place of care and death.
The authors would like to acknowledge Dr Yolanda Barbachano and Kabir Mohammed for providing statistical support for the study. They would also like to thank their patients for taking part in the study.
Funding The study was funded by the Palliative Care Research Fund 345 at the Royal Marsden Hospital. The authors would like to acknowledge the support of their charitable donors, particularly Joe Goodall and the other family and friends of Debora Matthews.
Competing interests None.
Ethics approval The study was approved by St Mary's Hospital REC.
Provenance and peer review Not commissioned; externally peer reviewed.