An overview of six clinical trials of aspirin in coronary heart disease

Abstract

Data from six randomized, placebo‐controlled clinical trials of aspirin, involving a total of 10,703 post‐myocardial infarction patients, have been compared and combined. An apparent heterogeneity of findings among the six trials with respect to all‐cause mortality is noted (P = 0.08). The reasons for this heterogeneity are explored. Two common methods of combining the results of several trials – the logrank and the log odds ratio methods – yield two‐sided P‐values of 0.112 and 0.110, respectively, for the aspirin‐placebo effect on all‐cause mortality. Alternative statistical methods of combining the results of the six trials give two‐sided P‐values ranging from 0.017 to 0.487.