1 School of Social and Cultural Studies, The University of Western Australia
2 School of Occupational Therapy, Curtin University of Technology
The Western Australian Data Linkage System was used to study patterns of specialist palliative care delivery during the last 12 months of life, comparing use between different socio-demographic groups and causes of death. In the two and half year study period, 8007 people had either cancer alone, or cancer and another condition considered amenable to palliative care recorded in part one of the death certificate. Of this group, 56.3% received some form of specialist palliative care. Particular groups of people were disadvantaged in relation to access to specialist palliative care, with those single or widowed, aged over 85 years, or those living in a region other than a major city significantly less likely to receive specialist palliative care services. The vast majority of people in the last year of life had at least one hospital admission. Those receiving specialist palliative care community-based services were more likely to die in their own homes. The population-based data in this study provides a unique picture of palliative care provision in WA, suggests areas of need and helps to lay the groundwork for future research.
Population-based research in Australian palliative care is still in its infancy, though the national body, Palliative Care Australia, has proposed using this approach to guide future service provision.1 Much can be learned from tracing the provision of palliative care services and documenting exactly who does and does not receive these important end-of-life supportive services. However, this kind of epidemiological research is complex and reliant on methods which can be time-consuming and frustrating, as access to reliable records is not always feasible. Our Western Australian study was able to deal with these methodological problems by using the unique Data Linkage System available in WA, which spans 20 years and contains core datasets including cancer registrations, inpatient hospital morbidity data and death records. The linking of a large number of administrative health records enables patterns of healthcare service use to be investigated in relation to a variety of socio-demographic variables. For the purposes of our study we also linked residential and community specialist palliative care services (SPCS) to the other relevant core datasets, enabling us to describe all SPCS use across the whole WA population.2,3
The population was identified from the Australian Bureau of Statistics Mortality Register as all people who died in the period 1 July 2000 to 31 December 2002 in WA and whose underlying cause of death, as recorded on part one of the death certificate, was from either cancer or selected non-cancer conditions considered amenable to palliative care (heart failure, renal failure, chronic obstructive pulmonary disease, Alzheimer’s disease, liver failure, Parkinson’s disease, Motor Neurone Disease, HIV/AIDS and Huntington’s disease).4 The outcome variables in the study were cause of death, place of death, use of SPCS in the last 12 months of life and demographic information (age, gender, marital status, Aboriginality, geographical remoteness and socio-economic disadvantage). The following discussion focuses upon those who died of cancer (ICD 10 codes C00 – D 48), as recorded on part one of the death certificate, as this group constitutes the vast majority of people who have symptoms and needs warranting SPCS.
Specialist palliative care services in WA, like other states and territories, provide a coordinated, integrated and patient-focused model of care, including respite and other supportive services to carers.5 In WA these services are provided by tertiary and non-tertiary public hospitals, hospices and community-based agencies. Silver Chain provides the vast majority of specialist palliative care community service in WA.
In the study period, 8007 people had cancer as one of the conditions listed on part one of the death certificate (29.8% of all deaths). Of this group 7399 people had cancer only and 608 had cancer as well as at least one other condition considered amenable to palliative care. Table 1 shows the kinds of service use for these two groups. Although palliative care is now a well recognised and important subspecialty, many of those who died did not access SPCS. This was particularly so for those who had cancer and one or more co-morbidities considered amenable to palliative care. The study also found that once a patient was referred to specialist palliative care they were more likely to receive other SPCS. In other words once they were in the system they were well serviced, indicating that referral is crucial to specialised service provision. Of course not all people dying of cancer require or request specialist palliative care. Nevertheless service use is one way of assessing potential unmet need.
Our study used logistic regression to estimate the likelihood of receiving specialist palliative care in the last 12 months of life. Table 2 indicates those who may be disadvantaged by missing out on SPCS and not surprisingly tends to mirror the broader disadvantages in Australian healthcare. Those who died of cancer were significantly less likely to receive SPCS if they were single or widowed, aged over 85 years, or lived in a region other than a major city.
Being married increased the likelihood of accessing SPCS, both in hospital and in the community. Married people with cancer used more community-based services, indicating that the ability to be cared for in their usual place of residence may ease the cost of hospital services.
Those aged over 85 years dying of cancer were less likely to receive SPCS. Compared with those aged 75-84 years, this group was proportionally less likely to have accessed SPCS in both hospital and community settings. It could be argued that this group was more likely to reside in aged care facilities and that a generalist level of palliative care was provided for in this setting. However, this assumption should not mask the broader question the findings suggest: are older people with cancer adequately serviced by palliative care providers?
It is well known that Australians residing in rural and remote areas are less well serviced in relation to healthcare. Our study provides evidence to show that this is particularly so for SPCS provision.
Although numbers were small, when compared with non-Indigenous people, Indigenous Australians were more likely to have used SPC, despite 80% of this group living in rural and remote areas. However, these figures need to be understood in relation to place of death, with only 26 of the 98 dying in their usual place of residence. Of the 72 who died in a place other than their usual residence, 63 died in hospital. As it is culturally very important for Indigenous people to die in their ‘own country’, the place of SPC delivery is of prime importance.6
In the last 12 months of life 96.5% of people who died of cancer, or cancer and another condition considered amenable to palliative care, had at least one hospital admission. Of the 3.5% of people dying of cancer who were not admitted to hospital in the last 12 months of life, the vast majority died in their usual place of residence, be it a residential aged care facility (35.0%) or some other location (58.8%). Table 3 summarises hospital admissions in the last year of life for those people who were: (1) hospital admission only; (2) specialised palliative care admission only; and (3) both specialised palliative care and hospital admission. Those admitted for both SPC and hospital admissions had a lower mean number of hospital admissions (6.3) than those who had hospital admissions only (6.9). More importantly those admitted for both had a lower mean length of hospital stay (29.0 days versus 33.5 days). In Table 3, the large standard deviations in comparison to the means are a reflection of the skewed distribution of the data.
Very little is known about where people die despite growing evidence to indicate that many people prefer to die in their usual place of residence.7 As the goal of palliative care is to allow patients to die comfortably and, if possible, in a place of their choosing, data about actual place of death is of interest to palliative care practitioners. Our study shows that of those who died of cancer either with or without another condition considered amenable to palliative care: 29.6% died in their usual place of residence; 17.1% died in a hospice; 48.6% died in a hospital; 1.7% died in a residential aged care facility; and 3.0% died in some other place. Multiple logistic regression was used to examine the effects of receiving specialist palliative care on place of death. When considering the effect of receiving SPCS, after adjusting for other factors, there is a seven times higher chance (95% CI 6.1-7.9) of dying in the usual place of residence if the patient received community-based specialist palliative care only, when compared to no specialist palliative care from any source. Receiving SPCS in the home, in particular, potentially allows people to die in a place of their own choosing.
Our study has provided unique information on the healthcare use of WA people in the last year of their lives. While reliant on administrative databases, the data provides a good baseline to determine obvious inequities and areas of need. It is heartening to find that palliative care services are now acknowledged and their influence noticeable, but it is also concerning to find that over 40% of people who died of either cancer or cancer and another condition considered amenable to palliative care did not receive specialist palliative care services. Our study has not investigated non-specialist palliative care service and many people may be receiving a variety of supportive services from general practitioners, other specialists and community organisations. Nevertheless, the holistic, multidisciplinary and family focused care provided by SPCS is a service that should be offered to families as they negotiate the difficult last months of life.
The study was funded through a National Health and Medical Research Council Palliative Care Program grant. Thanks are extended to Janine Calver, D’Arcy Holman, Ellen Nightingale, Gill Lewin, Dimitris Matsakidis and the palliative care community of Western Australia.
5. Health Department of Western Australia. Palliative care in Western Australia: The present and the future. Response of Palliative Care Advisory Group to Clinical Services Framework. Perth (Western Australia): Department of Health; 2006.