OBJECTIVES: To establish and validate a method of linking data from the Minimum Data Set (MDS) and Medicare hospital claims, to estimate hip fracture incidence rates for Medicare beneficiaries aged 65 and older in Washington State, and to compare the incidence rates of hip fractures in nursing home and non-nursing home residents.
DESIGN: Retrospective analysis of Medicare population-based enrollment, hospital claims, and nursing home administrative data sets.
SETTING: Nursing home and non-nursing home setting.
PARTICIPANTS: Medicare beneficiaries in Washington State residing in the community or in skilled nursing facilities.
MEASUREMENTS: Crude age- and sex-specific and standardized age- and sex-adjusted hip fracture incidence for persons residing and not residing in nursing homes.
RESULTS: From October 1, 1993, through September 30, 1995, 7,812 Medicare beneficiaries aged 65 or older were hospitalized for hip fractures (6,566 fractures for 1,155,234 person-years of exposure in non-nursing home residents and 1,246 fractures for 42,986 person-years of exposure in nursing home residents). The standardized age- and sex-adjusted hip fracture rate of nursing home residents (23.0 per 1,000 person-years) substantially exceeded that of non-nursing home residents (5.7 per 1,000 person-years) (incidence rate ratio = 4.0, 95% confidence interval = 3.7–4.5).
CONCLUSION: The incidence of hip fracture in nursing home residents far exceeds that in noninstitutionalized older people. Linkage of MDS and Medicare hospital claims data is a useful tool for epidemiological surveillance regarding events in nursing homes that are likely to result in hospitalization.