Subgroup analyses in randomized clinical trials: value and limitations. Review #3 on important aspects of randomized clinical trials in cardiovascular pharmacotherapy

Two review articles previously published from our working group were dedicated to the selection of endpoints as well as to reasons for premature stopping of randomized clinical trials (RCTs).1,2 We there first discussed the importance of mortality and morbidity endpoints vs. softer endpoints like revascularization rates and the issue of endpoint adjudication. Second, we have shed light on the statistical methods and requirements to stop RCTs prematurely due to safety, futility, or overwhelming efficacy (vs. the control arm). The main objective of this article is now to provide the clinical cardiologist with information how to judge and interpret published subgroup analyses. The next section will summarize the situation regarding subgroup analysis and put the current article in context. Modern RCTs typically look at outcomes of large populations. Based on the selection of the primary endpoint(s), and of size and duration of trials, their...