Oral bisphosphonate prescription and non-adherence at 12 months in patients with hip fractures treated in an acute geriatric unit.

A poorer functional status at the time of fracture is a predictor of non-adherence to oral bisphosphonates initiated after a hip fracture, and suggests further opportunities for optimization of secondary fracture prevention in this high-risk population.
Introduction: Low adherence to treatment is a problem in post-fracture secondary prevention. We aimed to analyze the prognostic factors (related and predictive) associated with non-adherence to oral bisphosphonate prescription for hip fracture due to bone fragility (HFBF) 12 months after discharge from an acute geriatric unit.
Methods: Prospective study of bivariate data analyzing related and multivariate factors predicting non-adherence of oral bisphosphonates at 12 months after treatment for HFBF. The statistical study was performed with SPSS 19.0.0.
Results: We attended 368 patients with HFBF. At discharge, oral bisphosphonates were prescribed to 226 (61.42%) patients. At 12 months, we followed up 160 (70.7%) patients, 104 (65%) of whom had non-adherence to oral bisphosphonates. Bivariate analysis (adherent vs. non-adherent): age (83.79 ± 5.82 vs. 85.78 ± 5.80, p = .029); Lawton and Brody Index (4.29 ± 3.40 vs. 2.67 ± 3.31, p = .004); baseline Barthel Index (BI) (85.89 ± 21.99 vs. 74.18 ± 26.70) (p = .004); BI at admission (18.84 ± 10.00 vs. 14.47 ± 11.71, p = .004); BI at discharge (34.20 ± 15.40 vs. 27.45 ± 16.71, p = .011); baseline Functional Ambulation Classification (5.66 ± 0.98 vs. 5.43 ± 0.99, p = .025). Multivariate analysis: BI 0.980 (0.965-0.995, p = .007). Discriminatory capacity of the AUC model (± 95% CI): 0.634 (0.545-0.722).
Conclusions: At 12 months, there was low adherence to treatment with oral bisphosphonates in our model. A lower BI prior to treatment is a predictive factor for non-adherence treatment with oral bisphosphonate.