OBJECTIVES: Healthcare decision-makers need more head-to-head drug comparison trials. Industry rarely sponsors such studies, preferring placebo comparators. We present an example of using meta-analysis for an indirect comparison of 2 drugs, with the results subsequently validated in a direct comparison trial. METHODS: The drugs for comparison were abciximab and tirofiban, both GPIIb/IIIa inhibitors used in patients with acute coronary syndromes undergoing percutaneous coronary interventions (PCI). A class effect has been assumed for these agents, although they differ in molecular structure and duration of action. We identified all placebo-controlled trials of each drug in PCI patients, and meta-analysed the odds ratios (OR) for death or MI (D/MI) at 30 days for each set of studies. These ORs were then compared using a general linear model. RESULTS: For D/MI at 30 days, the results of the meta-analysis of the 6 abciximab studies [OR=0.52 (0.43, 0.63)] appeared superior to the results of the 3 tirofiban studies [OR=0.73 (0.55, 0.96)], although the differences did not reach significance. In the model, a non-significant (p=0.10) abciximab advantage was observed. The ratio of the ORs of tirofiban/placebo and abciximab/placebo meta-analyses was 1.4, suggesting a higher risk of D/MI at 30 days for tirofiban relative to abciximab. After completion of these analyses, the results of a randomized comparison trial of tirofiban vs. abciximab in PCI patients were announced (for D/MI at 30days OR=1.26, p=0.04). The results demonstrated an advantage for abciximab, in keeping with our meta-analysis result. The magnitude of the efficacy difference was similar to that we had predicted. CONCLUSIONS: This is the first time an indirect comparison of drugs using meta-analytic techniques has been validated with a contemporaneous RCT. This method should be used to predict results of direct comparisons of drugs anytime such a trial is contemplated.