Longitudinal change in hip fracture incidence after starting risedronate or raloxifene: an observational study.

This study examined patients' risk profiles and adherence to treatment in relation to the effect of risedronate and raloxifene on hip fracture incidence. Administrative billing data were used to follow two cohorts of women aged 65 and older after starting therapy with either risedronate (n = 86,735) or raloxifene (n = 37,726). The fracture risk profile was described using a 6-month history period before starting therapy. Effectiveness of each therapy was evaluated by comparing the incidence of hip fractures during the first 3 months with the subsequent 12 months among women adherent (medication possession ratio >80%) compared with those non-adherent to treatment. At the start of therapy, the raloxifene cohort was younger than the risedronate cohort (median age 73 vs. 76 years) and had fewer prior fractures (p < 0.01 for both). In the first 3 months of therapy, hip fracture incidence was lower in the raloxifene group (0.51 per 100 person-years) compared with the risedronate group (0.94 per 100 person-years). In the subsequent 12 months, the incidence of hip fractures decreased among patients adherent to the risedronate regimen [relative risk (RR) 0.70, 95% CI 0.59-0.84, p < 0.01] and did not change significantly among patients adherent to the raloxifene regimen (RR 1.02, 95% CI 0.73-1.44). In poorly adherent patients, neither drug decreased hip fracture risk. Risedronate treatment in adherent patients rapidly decreased the risk of hip fractures, whereas raloxifene treatment did not.