Article Text

Download PDFPDF

The importance of identifying preferred place of death
  1. Maimoona Ali1,
  2. Margred Capel2,
  3. Gareth Jones2 and
  4. Terri Gazi2
  1. 1 Prince Charles Hospital, Merthyr Tydfil, UK
  2. 2 George Thomas Hospice Care, Cardiff, UK
  1. Correspondence to Dr Margred Capel, George Thomas Hospice Care, Ty George Thomas, Whitchurch hospital Grounds, Park Road, Whitchurch, Cardiff CF14 7BQ, UK; Margred.capel{at}


Objectives The majority of people would prefer to die at home and the stated intentions of both statutory and voluntary healthcare providers aim to support this. This service evaluation compared the preferred and actual place of death of patients known to a specialist community palliative care service.

Design All deaths of patients (n=2176) known to the specialist palliative care service over a 5-year period were examined through service evaluation to compare the actual place of death with the preferred place of death previously identified by the patient. Triggers for admission were established when the patients did not achieve this preference.

Results Between 2009 and 2013, 73% of patients who expressed a choice about their preferred place of death and 69.3% who wanted to die at home were able to achieve their preferences. During the course of their illness, 9.5% of patients changed their preference for place of death. 30% of patients either refused to discuss or no preference was elicited for place of death.

Conclusions Direct enquiry and identification of preferences for end-of-life care is associated with patients achieving their preference for place of death. Patients whose preferred place of death was unknown were more likely to be admitted to hospital for end-of-life care.

  • palliative
  • end-of-life care
  • preferred place of care
  • Service evaluation
  • advance care planning

Statistics from

Request Permissions

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.


Place of death is considered to be an important aspect of end-of-life care.1–4 Lay people, healthcare policymakers, commissioners and providers attach different significance to the place of an individual's death. An individual achieving their identified preferred place of death is significant to them and their immediate family or carers, and it is considered to be a quality marker of good end-of-life care. For service providers, it is often reflected in service-level agreements as a marker of performance. Commissioners are interested when considering distribution of resources and a method of making healthcare efficient while meeting the needs of service users. Policymakers refer to place of death as it reflects the value society places on the importance of achieving a ‘good death’.1 ,2 ,5 ,6

A service evaluation was undertaken to examine whether patients achieved their preference for place of death. The governance around the data use was agreed through the hospice governance processes. Ethical approval was not required. Studies suggest preferences for place of death may differ from place of care.7 Patients are known to change their preference for place of death during their illness trajectory.8 ,9 Opportunities to discuss these preferences were already part of communication with the patient. The dynamic care plan enabled the changes to be recorded and the comparison drawn against the last recorded preference for a place of death. A proforma was designed to capture the information related to the preference for place of death; all team members received training prior to its introduction.

George Thomas Hospice Care is a community-based hospice specialist multidisciplinary team without inpatient beds. It is located in the Welsh capital city of Cardiff, serving a population of over 300 000. It is within the geographical area served by a regional oncology centre and commissioned to accept all appropriate referrals from a discrete geographical area. Referrals were accepted from healthcare professionals or patients with the permission of the patient's general practitioner (GP) for patients with identified palliative care needs. Out of hours, the local district nursing service and GPs provided generalist support for these patients. District nurses were able to arrange non-qualified nurse ‘sitters’ to support patients and families at home for a maximum of two nights each week. Patients, families and healthcare professionals had access to specialist palliative medicine telephone advice at night. During 9:00–17:00 Saturday and Sunday, there was access to advice and face-to-face assessments from a clinical nurse specialist, and if required a palliative medicine specialist. Inpatient hospice admission (provided by a different charity) was also available locally.

The deaths of all patients known to the service in the 5 years from 1 January 2009 to 31 December 2013 were examined to compare the actual place of death with the preferred place of death previously stated by the patient. Preferred place of death was identified through discussion with the patient and amended as necessary during the illness journey. Place of residence was defined as their current place of living at the time of assessment, which included the patient's or their immediate carer's home, nursing home, residential home or prison. If patients were unwilling or unable to explore the subject of end-of-life care, the preferred place of death was recorded as an unwillingness to express a preference.

Additional information including the diagnosis, events which precipitated admission and the use of the anticipatory prescribing of medication were considered in an attempt to identify any trends in home, hospice and hospital deaths in this patient group to steer future service activity and education.


The key worker for each patient who died completed a ‘proforma’:

Information collected prospectively included:

  • Demographic details;

  • The diagnosis;

  • The preferred place of death (considered to be the last recorded preference in the case notes) and highlighted on the proforma if different from the current place of residence;

  • The actual place of death;

  • The GP practice;

  • Whether the death was anticipated;

  • Aspects surrounding end-of-life care, including the organisation of anticipatory medication, use of the Welsh Integrated Care Priorities (ICP), delays in starting syringe drivers and the provision of ‘sitters’ to support care.

Additional retrospective information was collected and obtained if necessary from healthcare professionals in the hospice, secondary care and family members to establish the precipitant for admission when patients had been admitted for end-of-life care to the hospice or hospital if this was not their preferred place of death. Categories for admission were grouped into:

  • A new condition (including seizure, haemorrhage, obstruction, urinary retention, acute dyspnoea, falls or other sudden events);

  • An exacerbation of an existing symptom or uncontrolled symptoms;

  • Patient choice;

  • Carer choice;

  • Patient or family perception of a lack of support at home;

  • Professional choice (GP, district nurse or specialist).


In total, 2176 patients known to the service between 1 January 2009 and 31 December 2013 died. The mean and median number of days patients were known to the service was 162 and 84, respectively. The majority of deaths, 1921 deaths (88%), were patients with a malignant diagnosis, which reflects the bulk of the referrals to the service. The most common cancer diagnoses were lung, colon, breast and pancreatic cancer which are reported in table 1.

Table 1

Cancer diagnoses of patients included in the service evaluation 2009–2013

In total, 255 deaths were those of patients with non-malignant disease with the most common non-malignant diagnoses reported in table 2.

Table 2

Non-cancer diagnoses of patients included in the service evaluation 2009–2013

Where patients died

Home was the preference expressed for place of death for the majority of patients known to the service. More patients known to the service died at home than other locations (table 3).

Table 3

Patient's preferred place of death compared with their actual place of death 2009–2013

Figure 1 demonstrates the actual place of death of patients known to the service who expressed a preference.

Figure 1

Actual place of death of patients who expressed a preference.

Table 3 compares the preferred place of death with the actual place of death during the service evaluation expressed as an option conversion rate. Seventy-three per cent of patients who expressed a preference about their preferred place of death were able to achieve this preference; 69.3% of patients who wanted to die at home were able to achieve this preference; and 79.1% of patients who wanted to die in the hospice were able to achieve this preference.

In total, 255 patients with non-malignant disease known to the palliative care service died between 2009 and 2013, of the 190 who expressed a preference, 132 (69%) achieved their preferred place of death.

During the course of their illness journey, 9.5% of patients changed their preference for place of death.

Patients whose preferred place of death was unknown

Thirty per cent of patients either refused to discuss or no preference was elicited for place of death (table 3). Patients who refused to discuss preferred place of death or whose preferences were unknown were more likely to be admitted to hospital for end-of-life care (figure 2).

Figure 2

Actual place of death of patients where preference was unknown to service.

Management of death in place of residence 2009–2013

For the majority of patients (80%) who died at home, their end-of-life care was supported with anticipatory medication prescribed by their GP.

The use of the ICP to support end-of-life care was recorded in 472 (50%) individuals.

People who did not achieve their preferences for end-of-life care

The precipitants for admission for those patients who expressed a preference to die at home, but were subsequently admitted to the hospice could be grouped into symptom control or family/carer struggling to cope with the patient at home. Not all reasons for admission were identifiable.

A number of patients were admitted to secondary care and were then transferred to the hospice for further care where they subsequently died. It is unknown whether this reflects a change in the individuals preferred place of death.

The main precipitants for admission for those patients who were admitted to the hospital for end-of-life care who had previously expressed the preference to die at home were newly identified problems or an exacerbation of an existing problem. Falls are a significant trigger for admission to hospital. Forty-seven falls were recorded (7.5% of hospital admissions).

A perception of lack of support or social care was a significant trigger for admission to both hospice and hospital (figures 3 and 4).

Figure 3

Precipitants for admission to hospice in people who had previously expressed a wish to die at home 2009–2013.

Figure 4

Precipitants for admission to hospital in people who had previously expressed a wish to die at home 2009–2013. EOL, end-of-life; POC, package of care.

During the course of the service evaluation, there were 13 incidences of a patient being admitted into secondary care due to a lack of an available hospice bed. However, this mainly related to the year 2013 when the inpatient hospice beds were reduced in number due to refurbishment.


A proforma completed on patients’ death enables data to be gathered to support the clinical activity of the palliative care team and target areas for future activity and education.3 Positive experience of this has encouraged joint working with the other local charities and information technology support from Public Health Wales to develop the proforma and data analysis. Uniformity of proforma users and data entry is essential to enable this information to remain uncompromised and accurately influence health service provision. The outcomes from ongoing service evaluation, for example, looking at the events, which trigger admission can enhance our understanding of our local service user's experiences and how services could be shaped to support them.8 ,10

The majority of diagnoses included in this service evaluation are malignant diseases, though the number of deaths from individuals with non-malignant disease has increased since this service evaluation was last conducted reflecting the increasing number of referrals to the service and need in this community. Despite the variable illness trajectory of non-malignant disease, the majority of these individuals known to the service were able to achieve their preferred place of death.

During the service evaluation period, 73% of patients who expressed a wish about their preferred place of care were able to achieve their preference. In total, 69.3% who wanted to die at home were able to achieve this preference which compares favourably with available literature. Recent studies5 ,9 ,11 ,12 suggest that involvement of community specialist palliative care teams increases the possibility that patients may achieve their preference for end-of-life care at home.

Advance care planning (ACP) including exploring end-of-life care and place of death with patients increases the probability an individual will achieve this preference.9 ,10 Other factors which influence preferred place of death include age, gender, living with relatives, low functional status and cancer diagnosis.9 The support through involvement and awareness of preferences by the GP also increases the possibility of the patient achieving their preference.13 ACP is recognised to impact on the quality of end-of-life care and is associated with decreased hospital admissions.14 The oral discussions are not always formally recorded in writing. Uniform use of the same approach including how and where this information is held and facilitating access to other health professionals suggests this could lead to a reduction in avoidable hospital admissions and greater satisfaction.14 ,15

The Welsh ICP was not used to support the deaths of the majority of patients in this service evaluation partly reflecting local barriers to its use in primary care in addition to the negative press associated with the Liverpool Care Pathway. Although these two pathways are distinctly different, this is not always recognised reflecting the importance of ongoing education provision by specialist services. Pathways acting as frameworks of good practice are useful tools to guide end-of-life care for generalist professionals providing care in any environment.16 End-of-life care is theoretically a less frequent experience for the generalist in comparison to the specialist working in this area, and it is not possible to draw any conclusions from the lack of its use in this service evaluation. Careful consideration is required to determine how to support and aid communication between the plethoras of health professionals who can be involved in an individual's end-of-life care in the community. Use of multiprofessional home-based documentation would enable the various health professionals who visit patients’ homes to communicate their assessments and thoughts in a timely manner which may enhance patient care; future changes to the Welsh ICP should reflect this.

For those patients known to specialist teams whose health is clearly deteriorating, anticipatory medication is requested from their GPs. The medication requested enables immediate management of the principal symptoms experienced in end-of-life care. Inadequate symptom control was a significant precipitant for admission to both hospital and hospice. The absence of medication was not identified as a precipitant during this 5-year period, indeed the majority of home deaths did have anticipatory medication in place. Many different systems exist across the UK,1 ,9 ,16 using a framework to support best practice appears effective in achieving quality and location of death, for example, in care homes.17 The University Health Board supports designated pharmacists in each locality to stock such anticipatory medication to avoid delays in access.

During the course of their illness, 9.5% of patients changed their preference for place of death, which is comparable with that reported in the literature.6 ,8 ,9 The reasons for these changes were not recorded in this proforma and may be explored in greater depth in qualitative studies. In the community, health professionals from primary care, the voluntary sector, the ambulance service and hospital teams may be involved in the care of an individual, and access to dynamic care plans which enable changes in preferences to be recorded need to be used uniformly. Success of such measures has been reported in other parts of the UK.15 Currently preferences expressed are recorded in the community hospice notes, recorded on the shared palliative/oncology systems database, conveyed verbally, discussed in primary care meetings (where these occur) and are passed onto the GP out of hours service.

It is possible that individuals who were admitted to either hospice or hospital and subsequently died there had changed their preference for place of death to that location. The current dynamic care plan and proforma does not enable this to be recorded, as family/carers or other health professionals involved are not asked if the patient changed their preference. Rapid discharge for end-of-life care at home is an option for patients in this geographical area. The close proximity of a new clinical event leading to admission and the patient's subsequent death suggests some of these events may be part of the dying process and may represent a missed opportunity to diagnose dying in patients in the community. Potential future working with the hospital teams on this aspect remains an option. Feedback to admitting professionals from the outcome of these admissions at local education events may have a role. Support through education such as the locally provided palliative introductory module for GPs may influence this.

A significant number of triggers for admission to both hospice and hospital relate to family/carer choice and may reflect the perceived or actual psychological burden caring for someone at home causes.18 This aspect requires significant examination of the home care services in Cardiff and comparison with successful other services in the UK with the potential option of remodelling to plug any deficiencies identified.

A significant number of triggers for admission relate to symptom control—either in relation to an exacerbation of existing symptoms or new symptoms. In these situations, either the primary care services available felt it appropriate to admit the patient or the patient felt the situation was sufficiently acute to over-ride their previously identified preferences and justify admission.18 This later possibility was not captured on this proforma. If these admissions are to be avoided, it is necessary for the primary care services to consider the event may be the individuals final end-of-life event in which case this may need to be communicated sensitively with patient/family/carers and appropriate support enlisted at home to avoid admission. Identifying what access from specialist support will enable these acute scenarios to be managed in their place of residence will enable patients to achieve their preferences in a symptom controlled and supported manner for family/carers.

Some patients (30%) either refused to discuss or no preference was elicited for their preference of place of death. Patients who refused to discuss or whose preference was unknown were more likely to be admitted to hospital for end-of-life care (figure 2). This suggests that hospital admission is the default option for care in any circumstance and reflects the perception that as long as life-sustaining interventions exist some individuals will choose them in preference to accepting the reality of their situation.19–21 Evidence also suggests patients who receive chemotherapy in the last months of life are more likely to die in hospital.20 Future changes to the proforma could readily collect this information and provide information to assist the palliative care team in understanding this group of individuals. At present, it is not possible to draw conclusions whether these two groups overlap and if they do how communication can be adapted to facilitate decision-making.

Most interventions target healthcare, but to truly have an impact on this patient group interventions targeted at society are also important. Initiatives such as ‘Dying well matters’ or the current vogue for ‘Big Conversations’ in these areas may subliminally open the door, so that healthcare professionals can then discuss these important topics with patients and their families/carers.20 ,22 ,23

Society has changed in the UK and people in the city do not necessarily live in the extended family network with potential informal carers in close proximity. Society in general is also less familiar with home deaths.19 ,21 If patients live alone, they are more likely to be admitted for end-of-life care.9 ,24 If home is not an option, it is disingenuous to offer them a choice of location of preference without being able to enhance the community support available for these individuals when required. Female family members have in the past automatically assumed the care provider role, it is unsurprising therefore that if they become ill they are more likely to be admitted and not have a home death.19 ,25 Female family members are more likely to be in paid employment than a century ago which can provide challenges for families and carers to balance this caring role with work and other commitments, particularly in the face of an uncertain prognosis.16 It is a challenge for employers in enabling their employees to support family members/carers during terminal illnesses, a challenge which has not yet been addressed by government policy.

People with cancer in their last year of life experience at least two admissions, which are more likely to be as an emergency rather than an elective event.26 This is logical; in general, patients are less likely to require elective intervention and admission if health services are responding appropriately to their clinical situation and illness trajectory. If the individuals overall health is deteriorating, services should be reviewed with the potential to consider how future admissions could be avoided. ‘Avoidance of admission’ is a more accurate description rather than considering these admissions as ‘inappropriate’.19 If the patient recovered from the acute event, they may well perceive the admission worth the effort.26 Distinguishing these events from the end-of-life care is the challenge, which often presents itself to the generalist. A varying proportion of individuals with terminal illness will present as an emergency, and potentially this will be their final admission.27 This is not explored in the raw statistics presented in the Bevan report,26 and these individuals are unlikely to be known to community palliative care teams. Some unexpected clinical events will trigger admission irrespective of the services in place to support the individual, for example, unexpected haemorrhage.

The Bevan report identified our University Health Board as one of the lower rates of emergency admissions within Wales, but once patients in our catchment area were admitted, they experienced significant length of admissions. These crude statistics require further dissection, but this provides opportunity for community and hospital services to consider how services can be shaped to facilitate discharge and then maintain the individuals at home. Conversely how to identify and support such individuals at home to prevent avoidable admissions requires further work and analysis.17 ,19 The information collected in our proforma suggests that some patients were admitted to hospital and recovered from the acute precipitant for admission, but subsequently died while waiting for discharge. These individuals appear to have surprised their family members/carers and have fallen outside the scope for the existing rapid discharge for home, which the health board does offer.

In the process of shaping these services thought to the provision of both in and out of hours rapid support at the interface between health and social care is essential to relieve both the practical and psychosocial burden. These individuals can be identified as they are becoming increasingly physically dependent on others for care and their symptom burden may be simultaneously increasing.18 ,19 It is noted that when inpatient hospice admission beds are unavailable, admissions to hospitals increase. Patients are only able to choose from the options available to them and if commissioners wish to encourage alternatives for hospital admissions then comprehensive alternatives need to exist.28 In reality, primary care working with specialist palliative care are unable to prevent all avoidable admissions.18


The authors would like to thank Mary Coakley, Allen Bryant, Kath O'Connell, Karen Smith, Lynnette Williams, and Cara Hinchliffe.



  • Contributors MA entered the data, wrote and submitted the paper. MC and TG planned the service evaluation. MC undertook the data analysis and conclusion and re-drafted the paper. GJ prepared the figures and tables and contributed to the data analysis. TG redrafted the paper.

  • Funding The hospice received funding from commissioners which include Cardiff and Vale University Health Board and the Welsh Government.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.