Background: Crohn's disease and ulcerative colitis are chronic relapsing conditions. National hospital statistics in England only provide episode-based admission rates and cannot be used to identify multiple admissions per person. We used record linkage to provide a person-based profile of admission rates.
Methods: Age-standardized admission rates were calculated from the computer tapes of the Hospital In-patient Enquiry (HIPE) for 1970-1985 and the Oxford Record Linkage Study (ORLS) for the period 1970-1986. In addition, annual hospitalized prevalence, first hospital admission rates (as a proxy for incidence) and readmission rates were calculated for the Oxford population. Predictors of relapses were assessed by logistic regression models.
Results: Episode-based admission rates for Crohn's disease increased over time, both nationally and in Oxford. In Oxford, hospitalized incidence rates showed no increase for Crohn's disease. Neither episode-based nor person-based rates increased for ulcerative colitis. Relapses resulting in hospital admission were more common for Crohn's disease than for ulcerative colitis. They increased over time. We showed the influence of some sociodemographic characteristics (age, social class, district of residence, operations) on relapses.
Conclusions: Data on episodes of hospital care for Crohn's disease showed an increase over time while data related to individuals in receipt of care showed no increase over time. Unlinked episodes of hospital care are of limited value when studying temporal trends for chronic conditions with multiple admissions.