Unit of Health Care Epidemiology
England’s integrated National Health Service makes it ideally suited for the study of routinely collected health data for public health benefit. The University of Oxford has a long history of collecting and linking such data for epidemiological and health services research. With over 50 years’ worth of linked hospitalisation records and mortality statistics, the UHCE database covers a uniquely long period in the history of population health under the NHS.
The anonymised linkage of individuals’ successive records helps to trace the health of population subgroups and follow de-identified individuals over time. This bestows research capabilities in both population-based epidemiology and health services research.
The unit has two continually growing datasets of hospitalisation records and mortality statistics.
The first is a national dataset covering the entire population of England, comprises nearly half a billion hospitalisation records (Hospital Episodes Statistics and its national predecessors) going back to 1968, multiple-cause-coded mortality records from 1995, and is fully linked (hospitalisations to mortality) from 1999 to the present.
The second is the Oxford Record Linkage Study, which commenced in 1963 and covers a regional population of approximately 4 million people. This dataset comprises approximately 20 million hospitalisation records and a million deaths, is linked throughout its history, and includes the longest run in England of multiple-cause-coded mortality data, starting in 1979.
Through extensive collaborations with clinical experts and public health scientists as well as with academics and students, a wide range of epidemiological and health services research is conducted into areas such as
- disease-specific hospitalisation trends, distinguishing numbers of admissions (a measure of the healthcare burden of disease) from numbers of people (a measure of the epidemiological burden of disease);
- mortality rates and trends in multiple cause coding on mortality records;
- geographical variation of diseases or operations;
- uptake of operations in individuals with specified disease history;
- outcomes after admission for a particular disease or operation;
- and studies of mother-infant pairs including associations between perinatal factors and subsequent disease in childhood.
Research into these areas, which might otherwise take many years to complete, be excessively costly or resource-intensive can be conducted extremely efficiently.
The Unit of Healthcare Epidemiology (UHCE) is part of the Nuffield Department of Population Health and is supported by Oxford NIHR Biomedical Research Centre, Public Health England, and Health Data Research UK.