Bisphosphonate (BP) users have decreased mortality, but this could be due to channeling bias. National healthcare data on hip fracture showed lower mortality in patients who were treated prior to fracture or began treatment after fracture. Reduced mortality after only one prescription filled points to the importance of patient factors. INTRODUCTION: Use of bisphosphonates has been found to be associated with decreased mortality even when adjusted for sex, frailty, bone mineral density and comorbidity, but BP may chiefly be initiated in patients with osteoporosis whose life expectancy is judged to be good. Our aim was to investigate the association between BP initiated before or after a hip fracture with mortality, and any modifying effects of comorbid conditions and recurrent fracture. METHODS: This register-based cohort study used prescription and mortality information for Danish patients born ≤1945 experiencing a hip fracture between 1/Jan/1999 and 31/Dec/ 2002 (N = 42,076). Patients who began BP after hip fracture were compared with hip fracture patients who remained alive at the time when their matched index case began treatment. RESULTS: Patients who used BP prior to their hip fracture (4.6 %) had significantly lower 3-month mortality (adjusted odds ratio, OR, 0.68; 0.59-0.77). Patients who began BP after the fracture (2.6 %) had significantly decreased mortality, both for patients who filled only one prescription (adjusted hazard ratio, HR 0.84; 0.73-0.95) and for patients who filled multiple prescriptions HR 0.73 (0.61-0.88). There was a significant interaction by gender with no significant risk reduction in men. CONCLUSION: This national dataset shows significantly and substantially improved survival in women who receive BP before or after their hip fracture. However, the observation of a reduction in mortality in patients who filled only one prescription for a BP suggests that patient factors may account for a considerable part of the survival advantage observed with BPs.
Osteoporosis International, 2012, Vol 24, Issue 1, p. 245-52