Making the most of routine data in palliative care research--a case study analysis of linked hospital and mortality data on cancer and heart failure patients in Scotland and Oxford.

The research base of palliative care is growing rapidly, but despite methodological advances, some of the practical challenges of working with people at the end of life will persist. This means that analysis of routine data is arguably more important in studying palliative care than it is in other aspects of health services research. End-of-life researchers have been using the high-quality linked data from cancer registries for many years. This paper explores the value of a less well-known resource for palliative care research: linked mortality and hospital activity data. Two case studies are presented using information from Scotland (population 5.1 million) and the former Oxford region of England (population 2.5 million). The advantages and limitations of linked hospital and mortality data for research and service planning in palliative care are drawn out through analyses investigating hospital bed utilisation by people with cancer and heart failure and the influence of social deprivation on the use of hospital services in the last year of life. The use of such data deserves a higher profile.