Dr Jüni and colleagues point out that comparisons of the effects of drug treatments can be confounded by indication in observational studies. In our study, we took steps to avoid such a bias in both the study design and the analysis, even though it seems unlikely that the choice of NSAIDs in patients with rheumatoid arthritis would be influenced by cardiovascular risk factors.1 As stated in the "Methods" section, we excluded patients prescribed anticoagulants or antiplatelet agents, including aspirin, during the year prior to their entry into the study. We also excluded patients who received such prescriptions 30 days or less prior to their index date. In addition, we assessed a number of cardiovascular risk factors at baseline and adjusted for them in the analysis. The adjustment had little effect on the odds ratios.
Watson DJ, Rhodes T, Cai B, Guess HA. Risk of Myocardial Infarction Associated With Selective COX-2 Inhibitors: Questions Remain—Reply. Arch Intern Med. 2002;162(22):2642. doi: