Association of bone mineral-reducing medications with two-year complications following open reduction and internal fixation of ankle fractures: a propensity score-matched analysis

Bone mineral-reducing medications (BMRMs) are commonly prescribed, but their association with outcomes after ankle open reduction and internal fixation (ORIF) is unclear. We evaluated preoperative exposure to ten BMRM classes and two-year complications after ankle ORIF. Adults undergoing ankle ORIF at least two years before data extraction were identified in TriNetX. Exposure required at least one medication record 1–180 days before surgery for proton pump inhibitors (PPIs), glucocorticoids, loop diuretics, thiazolidinediones, calcineurin inhibitors, aromatase inhibitors, selective serotonin reuptake inhibitors (SSRIs), first- or second-generation antipsychotics, or antiepileptics. Exposed patients were 1:1 propensity score matched to patients unexposed to all classes. Risk ratios (RRs) for five two-year outcomes were adjusted for false discovery rate (FDR). Matched cohorts included 227–9,858 patients per group. Nonunion was more frequent with glucocorticoids (2.36% vs. 1.72%; RR 1.37), PPIs (2.68% vs. 1.79%; RR 1.50), loop diuretics (3.98% vs. 2.20%; RR 1.81), SSRIs (2.49% vs. 1.72%; RR 1.45) and antiepileptics (2.70% vs. 1.73%; RR 1.56); all 95% confidence intervals excluded 1 and all FDR q ≤ 0.023. Antiepileptics were also associated with revision ORIF, the primary outcome (3.55% vs. 2.65%; RR 1.34, 95% CI 1.13–1.59; q = 0.009). No class was associated with device infection or hardware failure. Outcomes for thiazolidinediones, calcineurin inhibitors and aromatase inhibitors were not estimable (too few events). Preoperative glucocorticoid, PPI, loop diuretic, SSRI or antiepileptic exposure was associated with higher two-year nonunion risk after ankle ORIF, and antiepileptic exposure with revision. These hypothesis-generating associations may reflect confounding by indication.