Scott et al. conducted a systematic review to evaluate the risk of myocardial infarction (MI) associated with cyclo-oxygenase-2 (COX2) inhibitors and traditional NSAIDs. The review found a small increased risk of MI associated with COX2 inhibitors, particularly rofecoxib. Although a fixed-effects model of 14 case-control studies suggested a slightly increased risk of MI with NSAID use, a random-effects model of the same data and analysis of 6 cohort studies found no such link. An increased risk of MI was found in four RCTs of NSAID use in colonic adenoma. In an analysis of 14 RCTs that compared COX2 inhibitors with traditional NSAIDs in patients with arthritis, the odds ratio for MI with COX2 inhibitors was 1.6 (95% CI 1.1-2.4), although most of this risk was accounted for by rofecoxib. COX2 inhibitors were, however, associated with a reduced risk of serious gastrointestinal events. An analysis of previous systematic reviews showed increased risks of MI associated with rofecoxib and celecoxib.