Pre-fracture medication use as a predictor of 30-day mortality in hip fracture patients: an analysis of 141,201 patients.

Abstract

To evaluate the predictive value of pre-fracture medication usage on 30-day mortality following a hip fracture. Information on age, sex, fracture type, time of death and Charlson co-morbidity index (CCI) was collected from the Danish National Patient Registry on all patients above 60 years, sustaining a hip fracture during the period January 1995 to December 2013. Information on drug usage was obtained from the Danish National Prescription Database. Hazard ratios were calculated with 30-day mortality as the outcome. A univariate and 3 multivariate analyses were conducted with increasing adjustments, starting with age, sex and fracture type, adding co-morbidity and dose in the latter. 141,201 patients were included and a total of 12 drugs/drug groups were identified for analysis. Increased mortality was evident in all analyses for antiarrhythmics, beta blockers, proton pump inhibitors, loop diuretics, opioids, acetaminophen and for psycholeptics. For ACE-inhibitors, increased mortality was found in all analyses, except after adjustment for co-morbidity and dose. For thiazide diuretics, a significantly reduced mortality was evident in all but the univariate analyses while NSAIDs and statins were associated with a significantly reduced mortality in all analyses. For calcium channel blockers, an insignificant decrease was found after adjustment for dose. Further analysis showed a dose-response relationship for all drugs except ACE-inhibitors and calcium channel blockers. The study shows a correlation between pre-fracture usage of certain drugs and 30 day mortality after a hip fracture.