The impact of palliative care on cancer deaths in Hong Kong: a retrospective study of 494 cancer deaths.

OBJECTIVES: To study the utilization of public health care by advanced cancer patients in their last 6 months of life and their end-of-life process within the last 2 weeks of life. METHODS: This was a retrospective study on 494 cancer deaths from four public hospitals in 2005. This sample was selected from all in-patient cancer deaths by the ratio of one in four. Data were collected by review of charts and an electronic data base. RESULTS: A total of 494 cancer deaths were analysed. The mean age of all cancer patients (n = 494) was 72.6 years. Two-thirds of cancer patients received palliative care and half died in palliative care setting. Patients were categorized into three groups according to palliative care coverage and the place of death. The first group comprised of patients who received palliative care service and died in palliative care units (PCS-PCD group, n = 247); the second group of patients who received palliative care service within the last 6 months of life but died in non-palliative care wards (n = 86); and the third group of patients who never received palliative care and who died in non-palliative care wards (NPCS-NPCD group, n = 161). Differences among groups were tested by one way ANOVA. During the last 6 months of life, patients in the PCS-PCD group had less admission to acute care wards (P = 0.012), shorter duration of stay in acute care wards (P = 0.003), and less admission to an intensive care unit setting (P < 0.001). Within the last 2 weeks of life, the PCS-PCD group had fewer interventions initiated (P < 0.001); had higher number of symptoms documented in patient's record (P < 0.001); and were more likely to receive analgesics (P < 0.001), adjuvant analgesics (P < 0.001) and sedatives (P < 0.001). Patients in PCS-PCD group were more physically dependent in the last 2 weeks of life (P < 0.001), but mentally more alert at 72 hours before death (P < 0.001). Patients in the NPCS-NPCD group had fewer patients with a do not resuscitate order present (P < 0.001), and more patients with cardiopulmonary resuscitation performed (P < 0.001). CONCLUSION: Our results suggest that palliative care service has played a role in improving end-of-life cancer care in Hong Kong.