In a population survey, 2652 respondents aged 15+ years reported their preferred place of death, if dying of 'a terminal illness such as cancer or emphysema', to be home (70%), a hospital (19%), hospice (10%), or nursing home (<1%). The majority of respondents in all socio-demographic categories reported a preference for dying at home, with the greatest majorities occurring in younger age groups. After weighting to the age-sex distribution of all South Australian cancer deaths, 58% in our survey declared a preference to die at home, which is much higher than the 14% of cancer deaths that actually occurred at home in South Australia in 2000-2002. Multivariable analyses indicate that predictors of preferred home death include younger age, male, born in the UK/Ireland or Italy/Greece, better physical health, poorer mental health, and fewer concerns about dying at home. Predictors of preference for death in a hospice rather than hospital include older age, female, single, metropolitan residence, having higher educational and income levels, paid employment, awareness of advanced directives, and interpreting 'dying with dignity' as death without pain or suffering. Investigating the differences between preferred and actual places of death may assist service providers to meet end-of-life wishes.

Quality of care from the patient's perspective is an increasingly important outcome measure for cancer services. Patients' and carers' perceptions of cancer care were assessed through structured telephone interviews, 4-10 months post-discharge, which focused on experiences during the most recent hospital admission. A total of 481 patients with a primary diagnosis of cancer (ICD-10 C codes) were recruited, along with 345 carers nominated by the patients. Perceptions of clinical care were generally positive. Less positive aspects of care included not being asked how they were coping, not being offered counselling, and not receiving written information about procedures. Results also highlighted inadequate discharge processes. Carers were more likely than patients to report negative experiences. Percep- CANCER AFFECTS ONE IN EVERY THREEAustralians by the age of 75 years, 1 and cancer care constitutes 5% of health care expenditure in this country. 2 Survival outcomes in Australia are among the best in the world, 3 and continue to improve, however quality of life and quality of care are increasingly being viewed as equally important outcome measures for cancer control.Recently, calls have been made to reform cancer services in Australia. Optimising cancer care, a consultative report prepared by the peak cancer organisations in 2003, recommended adopting an integrated, multidisciplinary approach, with more emphasis on psychosocial care, to replace the traditional model of a single specialist offering care. 4 These recommendations have been echoed in the Commonwealth Government' s National service improvement framework for cancer (2006) 5 and several state cancer plans. [6][7][8] Cancer Australia is currently implementing the national strategies for "Strengthening Cancer Care", including the establishment of cancer services networks (CanNET). 9 The goal of these What is known about the topic? There is increasing interest in measuring patients' perceptions of care to assess quality of cancer services. What does this paper add?This study measured patients' and carers' perceptions of cancer care in the tertiary setting, with the aim of identifying aspects of care that require advocacy or action. What are the implications for practitioners? While patients and carers perceived clinical care to be of high standard, cancer care could be further enhanced with more attention to information provision, psychosocial support and better discharge procedures. Giving greater consideration to meeting carers' needs is also required.