Preadmission use of nonaspirin nonsteroidal anti-inflammatory drugs and 30-day stroke mortality

Objectives: To examine whether preadmission use of nonaspirin nonsteroidal anti-inflammatory drugs (NSAIDs) influenced 30-day stroke mortality.

Methods: We conducted a nationwide population-based cohort study. Using medical databases, we identified all first-time stroke hospitalizations in Denmark between 2004 and 2012 (n = 100,043) and subsequent mortality. We categorized NSAID use as current (prescription redemption within 60 days before hospital admission), former, and nonuse. Current use was further classified as new or long-term use. Cox regression was used to compute hazard ratios (HRs) of death within 30 days, controlling for potential confounding through multivariable adjustment and propensity score matching.

Results: The adjusted HR of death for ischemic stroke was 1.19 (95% confidence interval [CI]: 1.02–1.38) for current users of selective cyclooxygenase (COX)-2 inhibitors compared with nonusers, driven by the effect among new users (1.42, 95% CI: 1.14–1.77). Comparing the different COX-2 inhibitors, the HR was driven by new use of older traditional COX-2 inhibitors (1.42, 95% CI: 1.14–1.78) among which it was 1.53 (95% CI: 1.02–2.28) for etodolac and 1.28 (95% CI: 0.98–1.68) for diclofenac. The propensity score–matched analysis supported the association between older COX-2 inhibitors and ischemic stroke mortality. There was no association for former users. Mortality from intracerebral hemorrhage was not associated with use of nonselective NSAIDs or COX-2 inhibitors.

Conclusions: Preadmission use of COX-2 inhibitors was associated with increased 30-day mortality after ischemic stroke, but not hemorrhagic stroke. Use of nonselective NSAIDs at time of admission was not associated with mortality from ischemic stroke or intracerebral hemorrhage.