Intended for healthcare professionals


Evidence based palliative care

BMJ 1999; 319 doi: (Published 21 August 1999) Cite this as: BMJ 1999;319:462

There is some evidence—and there needs to be more

  1. Irene J Higginson, professor of palliative care and policy
  1. New Medical School, King's College, and St Christopher's Hospice, London SE5 9PJ

    In the first decades of the 21st century much healthcare spending will be concentrated on the end of life. Predictions for the year 2025 show an ageing population, with more people worldwide dying from chronic or progressive illnesses rather than acute conditions.1 Indeed, this revolution is already upon us. In the United States end of life expenditure through Medicare consumes 10-12% of the total health budget and 27% of the Medicare budget.2 Among older people healthcare expenditure for those in the last year of life was 276% higher than for people of similar age.3 Cartwright and Seale estimated that in England about 22% of hospital bed days were taken up by people in the last year of life.4 Although discussion of death is still taboo in our society, all health and social professionals must now be assessing the best way of caring for a person with a progressive illness and their family. This underlines the importance of palliative care.

    Palliative care is a person centred approach concerned with physical, psychosocial, and spiritual care in progressive disease. It focuses on both the quality of life remaining to patients and supporting their families and those close to them. Throughout the world specialist palliative care services have grown, though their distribution is uneven. In 1999 there were over 6560 hospice or palliative care services in 84 countries, with 933 services in the United Kingdom, almost 1200 in the other 36 countries in Europe, 3600 in North America, and 350 in Australia and New Zealand.5 And within countries the provision of palliative care varies geographically and between patient groups. In some parts of the United Kingdom 70% of patients with cancer are cared for by a palliative care team, and over a third die in a hospice, while in other areas only a few receive specialist care.6

    Moreover, some sections of the community lose out on specialist care. They include people with progressive diseases other than cancer: in 1995 in the United Kingdom only 3.3% of new referrals to inpatient and 3.7% of those to community palliative care services had a diagnosis other than cancer.7 Some data suggest that patients in socially deprived areas, those from minority groups,8 and older patients from all sections of the community9 have limited access to palliative care. Palliative care services operate in different ways, varying in funding (voluntary or NHS), team composition, staff to patient ratio, out of hours care, and treatment regimens used. Yet in order to promote greater use of clinical protocols, best practice guidelines within services, and guidance on the most effective models of care we need more research evidence. There are, however, difficulties in applying an evidence based approach to palliative care.

    Firstly, an absence of evidence does not mean that a service or treatment is not effective, just that we do not know. Outcomes such as the quality of care, quality of life measures including quality of death, and the best resolution of bereavement are hard to measure, especially when patients are frail and ill. Thus, many studies exclude quality of life as an outcome variable, or include only patients who can complete questionnaires. The challenge is to ensure that those aspects of care that are hard to measure do not become a lower priority than aspects—such as survival or function—that are easy to measure.

    Secondly, palliative care presents particular problems for the researcher While the randomised controlled trial remains the gold standard to determine the efficacy of treatments, some palliative care cannot be investigated by a traditional trial. There have now been three systematic reviews of the evaluation of palliative care services Rinck et al examined 11 randomised-controlled trials: all had methodological problems.10 In two studies the problems were so severe that no results were reported. Problems were associated with recruitment of a study population in 10 studies, homogeneity in six, patient attrition in four, defining and maintaining contrast between interventions in six, and selecting outcome variables in four.

    Another analysis considered 18 prospective comparative studies, retrospective and observational studies as well as randomised trials.11 When specialist multidisciplinary care was compared with conventional care, four of the five randomised controlled trials and most of the comparative studies indicated that the specialist, coordinated approach resulted in similar or improved outcomes in terms of patient satisfaction; patients being cared for in the place of their choice; family satisfaction; and control of family anxiety, patient pain, and symptoms. Those studies that examined costs showed a reduction in hospital inpatient days, more time spent at home, and equal or lower costs. While two studies, one in the UK12 and one in the US,13 suggested that pain was less well controlled at home than in a hospice, a recent study of home care palliative services in England and Ireland has reported better pain control at home.14

    Particular problems exist in measuring the quality of life of a patient with a progressive illness. In reviewing different models of palliative care Salisbury et al concluded that inpatient hospice care resulted in better pain control than conventional care, but there were few other differences.15 However, they found that quality of life was inadequately measured and the evidence for hospital and home palliative care teams very limited.

    Reviews of evidence also need to cover the changing problems for different groups of patients in the last year of life and the factors that appear to precipitate admission from home. This approach has been successful in evidence compiled for the National Council for Hospice and Specialist Palliative Services in their guidelines on care in the last days of life.16 In this review most of the studies were observational, quasiexperimental, or retrospective. The difficulty here is to account for the effects of biases in non-randomised comparison groups. Nevertheless, care shortly before death is often found lacking in the health Ombudsman's reports, and guidance on the best management of pain, symptoms, and emotional, social, and spiritual problems is needed for day to day practice.

    Clearly difficulties remain in the rigorous gathering of evidence about many aspects of palliative care. But these difficulties should not be allowed to stand in the way of applying palliative care where its efficacy, patient and family satisfaction, and cost-effectiveness have been shown. This is particularly the case for multiprofessional home care teams and inpatient hospices. Up to now the voluntary sector has led the way in expanding and developing palliative care, particularly through the hospice movement, and there now needs to be more public investment. For their part, hospices must pay greater attention to the need for standards in palliative care and form partnerships with the NHS to evaluate care and identify best practice. The priorities are to develop standards that will apply to all services and to evaluate those aspects of palliative care which remain unevaluated, such as hospital support teams, day care, carer support, complementary therapies, and care in the last days of life.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.
    16. 16.