Time trends in the HERS secondary prevention trial: much ado about nothing?

The Heart and Estrogen/progestin Replacement Study was the first randomized, double-blind, placebo-controlled study evaluating whether hormone replacement therapy (HRT) reduces the incidence of nonfatal myocardial infarction (MI) and coronary heart disease (CHD) death. The main conclusion was that HRT provided no benefit for secondary prevention. This paper examines the statistical evidence for a subsidiary analysis suggesting that the effects of HRT varied over time; that the HRT group had more cardiac events in year 1, but fewer by the last years of the study. An accurate reading of the HERS results indicates an increased incidence of MI and CHD death in the HRT group in year 1, but no evidence that the HRT group had fewer CHD events by the end of the study.