Trends in episode based and person based rates of admission to hospital in the Oxford record linkage study area

Abstract

Most routine health service information systems are based on episodes rather than on people: data are collected and analysed as isolated events, such as individual hospital episodes, and events relating to the same individuals are not routinely brought together. Thus, for example, the Hospital Activity Analysis and the annual Hospital In-patient Enquiry can be used to count episodes of admission to hospital but not numbers of people admitted to hospital. While variation in rates over time or between different populations may reflect differences in the numbers of people treated or variation in multiple admissions per individual, or both, routine statistics cannot generally be used to make these distinctions. Rates of admission, measured as discharges from and deaths in hospital, have risen steadily in England for many years at an average increase of about 2% a year. There has been speculation about whether this rise has represented a true increase in the number of people treated by the National Health Service or whether it may simply reflect an increase in repeat admissions.12 In England the Oxford record linkage study may be used to link together successive episodes of admission to hospital for the same individuals and to distinguish between episodes of admission to hospital and numbers of individual people admitted to hospital in a large, defined population.3 In this paper we present some general findings comparing trends in episode based and person based rates of admission.