An editorial notes that at face value, these new IMPROVE-IT results appear to have important clinical implications. However, it cautions that subgroup analyses from trials can be unreliable, and this may be especially true when the treatment effects are small, results from the primary analyses are of marginal statistical significance, or emphasis is placed on nominally significant findings in the context of such multiple analyses. It adds that because there is unlikely to ever be enough additional randomised data to definitively test these new hypotheses, and given the various strands of compelling genetic and randomised trial evidence suggesting that relative cardiovascular benefit is likely to be similar in those with and without diabetes mellitus, it is reasonable to conclude that ezetimibe reduces atherosclerotic cardiovascular risk through lowering LDL cholesterol irrespective of the presence or absence of diabetes mellitus. Furthermore, to quantify which patients might be expected to have a smaller or larger absolute benefit from ezetimibe, it recommends classifying their risk of atherosclerotic cardiovascular disease by using a relevant and reliable risk score and then applying the relative benefits estimated by the overall IMPROVE-IT result.