Patterns and costs of hospital-based cancer care in the last year of life: A national data linkage study in England.

Abstract

e18769 Background: Delivery of high quality cancer care is associated with rising costs, both in earlier stages of the illness trajectory and at the end of life. A significant portion of the costs and health care utilisation occurs in the last year of life. Most publications to date have focused on costs in hospital. Little is known about the costs of care for cancer patients across the entire health service. The aim was to examine primary, secondary and acute health care utilisation and cost in the last 12 months of life and how these differ by cancer diagnosis and other patient characteristics among decedent patients aged 60 and over. Methods: We conducted a retrospective cohort study of people aged 60 years and over (N=26,077) who died in England between 2010 and 2017. We used routinely collected and linked data from primary care (Clinical Practice Research Datalink (CPRD) secondary and acute care, (Hospital Episode Statistics (HES), and death data (Office for National Statistics (ONS)). This provided a nationally representative sample of the English population. We analysed of healthcare utilisation and resource use amongst decedents by gender, primary cause of death, age, geography, socio-economic status and comorbidities. Results: Overall, 90.2 % of the cancer decedents were admitted in the hospital at least once in the last 12 months of life. About 50% of patients we admitted to hospital in the last month of life with 37.6 being admitted to the hospital more than once in the last month of life. The health care utilisation and costs increased sharply in the last month of life. life. The mean number of hospital admissions in the last year of life was 3.7 (SD, 5.8). Those dying of haematological cancers (N=2093) had highest number of hospital admissions (mean:7.2, SD:10.8) and longer average hospital stay (mean:36.7, SD:33.0) (mean:12.0, SD:14.4). Use of outpatient services was highest in the group dying of haematological cancers (mean:12.0, SD:14.4) whereas those dying of prostate cancer (N= 2197) had higher number of emergency (mean:2.0, SD: 1.9) and GP visits (mean:30.8, SD: 20.7). Healthcare costs were highest among haematological cancers and lowest among those dying of breast cancer. Proximity to death and comorbidities were the main contributors of end-of-life care health care utilisation and costs. Conclusions: This study uses large linked datasets (linked to the whole spectrum of hospital episode statistics) providing a comprehensive picture of healthcare services accessed by cancer patients at end of life in England. There is significant variation in use and cost of care for cancer patients in the last year and month of life. Further analysis of variation according to hospice, palliative, and social care service provision may identify strategies to address this variation.