Hip fracture incidence and mortality in an English Region: a study using routine National Health Service data

Background We investigated the validity of routine hospital and mortality hip fracture data in one English Region and estimated trends in hip fracture between 1978-1981 and 1993-1995. Methods We identified from Hospital Episode System (HES) data for 1993-1995 all hip fracture episodes relating to individual patients aged over 65 years resident in Wessex. We determined from the discharge method code whether an individual had died during that admission. The number of individual regional and district admissions and deaths were compared with those presented in the Public Health Common Data Set. We compared regional admission rates with data for 1978-1981 from a previous study. Results National comparative indicators for hip fracture overestimated individual admissions in Wessex by 17 per cent (in health authorities by 1-56 per cent). National comparative indicators for hip fracture mortality underestimated individual deaths in Wessex by 48 per cent. Between 1978-1981 and 1993-1995 the age-sex-standardized hip fracture rates rose from 1.90 to 2.63 per 1000 per year for men and from 5.70 to 7.70 per 1000 per year for women. Rates increased in all age groups except those aged 65-69 years. There was also a small fall in absolute mean annual numbers in this age group. The rates also fell in females aged 70-74. Conclusions It is possible to adjust routine national HES data to take account of multiple episodes within a single admission. These methods should be applied to national comparative indicators for hip fracture admission and deaths. Hip fracture rates continue to rise in those aged over 70 years. There may be a cohort effect with those born after 1925 showing stable rates which needs further investigation.