Major bleeding risk associated with warfarin and co-medications in the elderly population

Purpose: Warfarin management in the elderly population is complex as medicines prescribed for concomitant diseases may further increase the risk of major bleeding associated with warfarin use. We aimed to quantify the excess risk of bleeding-related hospitalisation when warfarin was co-dispensed with potentially interacting medicines. Methods: A retrospective cohort study was undertaken over a 4-year period from July 2002 to June 2006 to examine bleeding risk associated with medications co-administered in patients taking warfarin using an administrative claims database from the Australian Department of Veterans' Affairs. All veterans aged 65years and over who were new users of warfarin were followed until death or study end. Risk of bleeding was assessed using a Poisson GEE model adjusting for age, gender, socioeconomic status, co-morbidity index, previous bleeding related hospitalisations and indicators of health service use. Results: Overall, 17661 veterans who used warfarin at any time during the study period were included. The overall incidence rate of bleeding-related hospitalisations was 4.1 (95% CI 3.7-4.6) per 100 person-years in veterans who were not receiving potentially interacting medicines. Bleeding-related hospitalisation rates were significantly increased when warfarin was co-prescribed with low-dose aspirin (Adjusted rate ratio (AdjRR) 1.44, 95% CI 1.00-2.07), clopidogrel (AdjRR 2.23, 95% CI 1.48-3.36), clopidogrel with aspirin (AdjRR 3.44, 95% CI 1.28-9.23), amiodarone (AdjRR 3.33, 95% CI 1.38-8.00) and antibiotics (AdjRR 2.34, 95% CI 1.55-3.54). Conclusions: Models assessing bleeding risk with warfarin should take account of the range of potentially harmful medicine combinations used in elderly people with comorbid conditions.