Close encounters with errors of the second kind: evaluating risks and benefits of long-term dual antiplatelet therapy.

This editorial refers to ‘Impact of clinical presentation on ischaemic and bleeding outcomes in patients receiving 6 or 24 month duration of dual antiplatelet therapy after stent implantation: a pre-specified anaylsis from the Prolonging Dual Antiplatelet Treatment After Grading Stent-Induced Intimal Hyperplasia (PRODIGY) trial’, by F. Costa et al. , on page doi:10.1093/eurheartj/ehv038. In a letter published in 1935 in the journal Nature , Ronald Fisher, one of the fathers of modern statistical theory, described the phenomenon that had become known as ‘errors of the second kind’—that is, ‘the logical fallacy of believing that a hypothesis has been proved to be true, merely because it is not contradicted by the available facts … .’.1 He would go on to write, ‘It would … add greatly to the clarity with which the tests of significance are regarded if it were generally understood that tests of significance, when used accurately, are capable of rejecting or invalidating hypotheses, in so far as these are contradicted by the data; but that they are never capable of establishing them as certainly true.’ In the realm of randomized clinical trials, errors of the second kind, more commonly referred to as Type II error, are typically addressed during the assessment of statistical power for a given sample size. However, Type II error remains a critical consideration in the interpretation of trial results, particularly in subgroup assessments, which invariably are less highly powered than parent studies. Eighty years after Fisher's description of Type II error, Costa et al . provide a new opportunity to examine critically what constitutes sufficient evidence to declare that a treatment has no beneficial effect, this time in the …