Abstract
Cyclooxygenase-2 inhibitors relieve pain from inflammatory conditions by decreasing the gastrointestinal side effects from traditional nonsteroidal anti-inflammatory drugs. Basic research provided plausible mechanisms and some observational epidemiological studies, case-control and cohort, indicated that patients prescribed with cyclooxygenase-2 inhibitors and nonsteroidal anti-inflammatory drugs had increased risks for myocardial infarction and stroke. Because patients prescribed with cyclooxygenase-2 inhibitors were systematically different, uncontrolled and uncontrollable confounding by indication was as large as the observed risks. Thus, epidemiological studies or their meta-analyses could not discern whether, and if so, how much, the risks were real. A comprehensive meta-analysis of randomized trials indicated that cyclooxygenase-2 inhibitors increased the risk of vascular events by 42%, almost exclusively myocardial infarction, as did high-dose regimens of ibuprofen and diclofenac, but not naproxen. Individual clinical judgments and policy decisions should include cardiovascular disease and noncardiovascular disease risks including gastrointestinal side effects and clinical benefits including improved quality of life from less pain and disability.