Who receives specialist palliative care in Western Australia and who misses out

Executive summary
Our research describes the provision of palliative care services at a time of transition. Typically, palliative care has offered a holistic, non-curative focus with an emphasis on improving the quality of life of people with life-limiting conditions. Traditionally it has aimed to improve the conditions of people who were dying of cancer. Palliative care now seeks to extend a holistic, team-based and family-centred approach to people with other life-limiting conditions. Until now the degree to which this aim was reflected in the actual provision of health services has been unclear. There has been a paucity of population-based data on which to base equitable healthcare decision-making regarding the extension of palliative care to people suffering from a range of life-limiting, complex and painful conditions. Our study provides such data for Western Australia and provides an insight into who is missing out on palliative care services in the last year of life. T

he study was a longitudinal, population-based, retrospective cohort study of 26,882 people who died in Western Australia between 1 July 2000 and 31 December 2002. A secondary analysis was conducted of data from the Silver Chain Nursing Association and two administrative health databases of the Western Australian Data Linkage System: the Mortality Register and the Hospital Morbidity Data System.

Three possible palliative care constituencies (that is, groups of people who would benefit from palliative care) were proposed in order to model three scenarios of service provision:

(a) a minimal constituency. That is, a disease-based potential palliative care constituency composed of people who died of any of the following 10 conditions: cancer, heart failure, chronic obstructive pulmonary disease, renal failure, Alzheimer's disease, liver failure, Parkinson’s disease, motor neurone disease, HIV/AIDS, and Huntington’s disease;

(b) a mid-range constituency. That is, a condition-based potential palliative care constituency composed of people who were hospitalised in the last 12 months of their lives for a condition listed as an underlying cause of death on their death certificate; and

(c) a maximal constituency. That is, a condition-based potential palliative care constituency composed of people who died from all causes except those that produced sudden deaths, defined as poisoning, injury, or originating during pregnancy or the puerperium.

Key Findings1
The minimal constituency represented 50% of all deaths during the 2.5-year study period, 54% of which were attributed on the death certificate to cancer. Four out of five of people in the minimal constituency were aged over 65 years; one in 50 was Aboriginal; half were married and a third were widowed; seven out of 10 lived in a major city; and a third were socioeconomically disadvantaged.

About half of the people in the minimal constituency died in hospital, about a third in their usual place of residence, the remainder in a residential aged care facility or elsewhere. Use of specialist palliative care services reduced the likelihood of dying in hospital or in a residential aged care facility.

The rate of use of specialist palliative care services by people dying of malignancies (68%) was much higher than the rate for people dying of non-malignant conditions (8%). In addition, people were less likely to receive specialist palliative care services if they were aged 84 years or over; female; Aboriginal; living in a remote area; or socioeconomically disadvantaged.

In the last 12 months of life, one person in four in the minimal constituency had a hospital admission for specialist palliative care services, whereas nine people out of 10 had a hospital admission not involving specialist palliative care services. Our findings suggest that the use of specialist palliative care services potentially reduces the demand on other hospital beds.

Slightly less than one person in three received home-based specialist palliative care services in the last 12 months of life. People who received home-based specialist palliative care services were more likely to receive home-based non-palliative services. People who received homebased non-palliative care services, and did not receive home-based specialist palliative care services, had more visits and more home hours of home-based non-palliative care services than people who received home-based specialist palliative care services and home-based nonpalliative care services.

People who received specialist palliative care services at home and/or in hospital spent more time in hospital than people who did not receive any specialist palliative care services. However, people who received only home-based specialist palliative care services spent less time in hospital than people who did not receive any specialist palliative care services. Referral to home-based palliative care therefore potentially ensures the best possible, and most appropriate use, of the various kinds of home-based services.

The annual cost of providing hospital-based specialist palliative care services in Western Australia for the study period was calculated at $9.2M. If one half of all people in the minimal constituency had received hospital-based specialist palliative care services, this cost would have been doubled to $18M. Providing home-based specialist palliative care services to one half of this constituency would cost $9.1M, based on current home-based costs.

Conclusions
Our study identified those people who are the most disadvantaged in respect to receiving specialist palliative care services. A significant proportion of people within one of our potential palliative care constituencies, and who died in the period July 2000 to December 2002, did not receive specialist palliative care. These people may have received other forms of health care. They may also have benefited from the holistic, family-oriented services provided by generalist palliative care services. Not surprisingly, and given the traditional palliative care focus on cancer, those people dying of cancer were well serviced in comparison to people dying from other complex and painful conditions. Nevertheless one third of the people who died of cancer did not receive specialist palliative care.

Other measures of disadvantage have been noted that are reflective of broader disadvantage within Australian health care. To be Aboriginal, elderly, living in a rural or remote area and to be socially disadvantaged lessens one's chance of receiving appropriate, and apparently cost effective, services in the last weeks and months of life. By pointing to the areas of most critical need, this study assists policy and planning in the provision of health services.

Our study focuses on people who are not receiving specialist palliative care services, while also providing a unique window into Western Australian palliative care services. Once clients have been referred to palliative care, they are provided with a cohesive service by the network of local providers. Those who receive specialist palliative care services in hospital are more likely to receive home-based palliative care and the reverse is also true. This group of people are also more likely to die in their own homes, a common request of dying people that has been reported in the literature and anecdotally by health care providers.

By providing a comprehensive and empirically grounded picture of services used in the last year of life, this study has sought to develop and test a universally accepted normative benchmark for defining palliative care needs at population level. It makes a strong case for improving the range and level of palliative care provided to terminally ill people. Extending palliative care to accommodate even a minimal constituency is a costly exercise. Findings from our study allow informed advocacy to occur on behalf of vulnerable populations in a country that prides itself on universal access to high quality health services.

1 Only results from the minimal constituency are reported in the Executive Summary.