Clinical Question: Should warfarin (Coumadin) be added to aspirin in patients who have had a myocardial infarction or acute coronary syndrome?

Setting: Various (meta-analysis)

Study Design: Meta-analysis (randomized controlled trials)

Synopsis: This analysis tries to answer the question of whether it causes more harm than good to add warfarin to aspirin in patients who have experienced acute coronary syndrome or acute myocardial infarction. The authors did not follow standard practices for identifying studies: they searched only one database, limited their research to English-language articles, and did not look for unpublished data. These are not major limitations, however, because the studies they identified were large and published in major journals. Two reviewers independently evaluated the articles for inclusion, and two reviewers independently abstracted the data. The 10 randomized studies evaluated more than 5,900 patients. The typical dose of aspirin was between 80 and 325 mg per day. The target International Normalized Ratio was 2.0 to 2.5, and the studies compared aspirin alone with the combination of aspirin and warfarin.

Overall, percentage of deaths was not significantly different with the combination therapy (2.6 versus 2.7 percent). The annualized rates of myocardial infarction were significantly lower when the combination was used (2.2 versus 4.1 percent; number needed to treat [NNT] = 56; 95% confidence interval [CI], 47 to 79). The risk of ischemic stroke also is lower with the combination (NNT = 221; 95% CI, 163 to 518). Major bleeding episodes were more common in the combination group (1.5 versus 0.56 percent; number needed to treat to harm = 117; 95% CI, 65 to 258).

Bottom Line: Adding warfarin to aspirin prophylaxis does not affect overall death rates, although the combination decreases subsequent myocardial infarction risk and, to a lesser degree, ischemic stroke risk. As might be expected, major bleeding episodes occur more often with the addition of warfarin, though only in a small number of patients. (Level of Evidence: 1a)