Warfarin therapy and risk of hip fracture among elderly patients.

STUDY OBJECTIVE To ascertain the relationship between warfarin therapy and subsequent hip fracture in a large elderly population. DESIGN Retrospective, population-based cohort study. SETTING Population-based health care administrative databases for Ontario, Canada. PATIENTS Elderly patients receiving warfarin (52,701 patients), thyroid replacement therapy (40,555), an oral corticosteroid (43,915), or a proton pump inhibitor (60,383). The proton pump inhibitor group served as controls. MEASUREMENTS AND MAIN RESULTS The association between warfarin therapy and subsequent hospitalization for hip fracture in elderly patients was examined by researching administrative data from January 1, 1994-March 31, 1999, for the elderly population of Ontario. Relative to patients receiving proton pump inhibitors, patients receiving warfarin (adjusted risk ratio [aRR] 0.94, 95% confidence interval [CI] 0.81-1.09) or thyroid replacement therapy (aRR 1.02, 95% Cl 0.89-1.18) incurred similar risks of hip fracture. As expected, patients receiving oral corticosteroids incurred an increased risk (aRR 1.44, 95% CI 1.21-1.70) relative to patients receiving proton pump inhibitors. CONCLUSION Warfarin was not associated with increased risk of hip fracture.