Abstract 3271: The Risk Of Non-selective NSAIDs And Selective Cyclooxygenase-2 Inhibitors Is Not Modified By Use Of Evidence-based Pharmacotherapy In Chronic Heart Failure
Introduction: It is generally perceived that excess cardiovascular risk of NSAIDs is primarily driven by selective COX-2 inhibitors. We studied if evidence-based pharmacotherapy in heart failure (HF) modified the cardiovascular risk of NSAIDs (non-selective NSAIDs and selective COX-2 inhibitors) in a large unselected cohort of patients with chronic HF.
Methods: The study population comprised 107,092 patients surviving first HF hospitalization in Denmark 1995–2004, identified from the nationwide Danish Patient Registry. Analyses were by multivariable time-dependent Cox proportional-hazard models, adjusted in a propensity-score based analysis.
Results: The mean observation period was 2.6 years (SD 2.4); 60,974 (56.9%) patients died, and 8,970 (8.4%) and 39,984 (37.5%) were re-hospitalized for MI and HF, respectively. NSAID treatment was initiated in 37,574 (35.1%) patients at some point after discharge. Selective COX-2 inhibitors as well as non-selective NSAIDs were associated with increased risk of death, with a dose-dependent response (Fig⇓), and increased risk of re-hospitalization for MI or HF, although partial effects of unmeasured confounders cannot be excluded. The risk was unaffected by the use of ACE inhibitors, beta-blockers, spironolactone or statins.
Conclusion: Contrary to general beliefs, this study indicates that both non-selective NSAIDs and selective COX-2 inhibitors are associated with increased mortality and cardiovascular risk in chronic HF. This increased risk was not modified by the use of evidence-based pharmacotherapy and NSAIDs should be used with caution in all patients with chronic HF.