Incidence and Predictors of Multiple Fractures Despite High Adherence to Oral Bisphosphonates: A Binational...

Oral bisphosphonates (BPs) are highly effective in preventing fractures and are recommended first-line therapies for patients with osteoporosis. We identified the incidence and predictors of oral BP treatment failure, defined as the incidence of two or more fractures while on treatment (≥2 FWOT) among users with high adherence. Fractures were considered from 6 months after treatment initiation and up to 6 months after discontinuation. Data from computerized records and pharmacy invoices were obtained from Sistema d'Informació per al Desenvolupament de l'Investigació en Atenció Primària (SIDIAP; Catalonia, Spain) and Danish Health Registries (Denmark) for all incident users of oral BPs in 2006-2007 and 2000-2001, respectively. Fine and Gray survival models using backward-stepwise selection (p-entry 0.049; p- exit 0.10) and accounting for the competing risk of therapy cessation were used to identify predictors of ≥2 FWOT among patients having persisted with treatment ≥6 months with overall medication possession ratio (MPR) ≥80%. Incidence of ≥2 FWOT was 2.4 (95% confidence interval [CI], 1.8 to 3.2) and 1.7 (95% CI, 1.2 to 2.2) per 1000 patient-years (PYs) within Catalonia and Denmark, respectively. Older age was predictive of ≥2 FWOT in both Catalonian and Danish cohorts: subhazard ratio (SHR) = 2.28 (95% CI, 1.11 to 4.68) and SHR = 2.61 (95% CI, 0.98 to 6.95), respectively, for 65 to <80 years; and SHR = 3.19 (95% CI, 1.33 to 7.69) and SHR = 4.88 (95% CI, 1.74 to 13.7), respectively, for ≥80 years. Further significant predictors of ≥2 FWOT identified within only one cohort were dementia, SHR = 4.46 (95% CI, 1.02 to 19.4) (SIDIAP); and history of recent or older fracture, SHR = 3.40 (95% CI, 1.50 to 7.68) and SHR = 2.08 (95% CI: 1.04-4.15), respectively (Denmark). Even among highly adherent users of oral BP therapy, a minority sustain multiple fractures while on treatment. Older age was predictive of increased risk within both study populations, as was history of recent/old fracture and dementia within one but not both populations. Additional and/or alternative strategies should be investigated for these patients.