Covariate Adjustment in Cardiovascular Randomized Controlled Trials

In randomized controlled trials, patient characteristics are expected to be well balanced between treatment groups; however, adjustment for characteristics that are prognostic can still be beneficial with a modest gain in statistical power. Nevertheless, previous reviews show that many trials use unadjusted analyses. In this article, we review current practice regarding covariate adjustment in cardiovascular trials among all 84 randomized controlled trials relating to cardiovascular disease published in the New England Journal of Medicine, The Lancet, and the Journal of the American Medical Association during 2019. We identify trials in which use of covariate adjustment led to a change in the trial conclusions. By using these trials as case studies, along with data from the CHARM trial and simulation studies, we demonstrate some of the potential benefits and pitfalls of covariate adjustment. We discuss some of the complexities of using covariate adjustment, including how many covariates to choose, how covariates should be modeled, how to handle missing data for baseline covariates, and how adjusted analyses are viewed by regulators. We conclude that contemporary cardiovascular trials do not make best use of covariate adjustment and that more frequent use could lead to improvements in the efficiency of future trials.