Differential associations of beta-blockers with hemorrhagic events for chronic heart failure patients on warfarin

Purpose Beta-blockers have many different physiologic effects that could potentially influence the risk of hemorrhagic events in chronic heart failure patients (CHF) on warfarin. We examined how different beta-blockers vary in their associated risk of a hemorrhagic event. Methods We used databases from the Department of Veterans Affairs (VA) that contain information on medications prescribed, diagnoses, and hospitalizations. We identified patients with CHF on warfarin and either metoprolol, carvedilol, atenolol, or no beta-blocker during 1999-2001. We modeled time to first hemorrhagic event using a Cox proportional hazards model, adjusting for age, ethnicity, comorbidities, and other factors. INR levels were examined in a subsample of 3546 patients. Results We identified 66988 CHF patients on warfarin. Hemorrhagic events occurred in 15.3% of the sample and, in 3.8% of the sample, the hemorrhage was considered severe. Compared to patients on carvedilol, the hazards ratio for a new hemorrhagic event was 1.25 (1.17, 1.34) for no beta-blocker, 1.27 (1.18, 1.38) for atenolol, and 1.38 (1.28, 1.48) for rnetoprolol. No differences in INR levels were evident among the four groups. Conclusions The risk for a hemorrhagic event among CHF patients on warfarin may be affected by beta-blocker use and varies depending on which beta-blocker is prescribed. Copyright (c) 2006 John Wiley & Sons, Ltd.