Validation Testing in Thorough QT/QTc Clinical Trials∗

In order to validate the results of a thorough QT/QTc (the duration of depolarization and repolarization of ventricles or the duration corrected for heart rate) clinical trial, ICH E14 recommended to include a concurrent positive control treatment in the trial. It further recommended that validation is achieved if the positive control has an effect on the mean QT/QTc interval of about 5 ms. Zhang (2008) discussed the intersection-union test approach for testing the validation hypotheses and an alternative global average test approach. In this article, we further discuss the difference and relationship of the two sets of hypotheses and the difference in the efficiencies of the two approaches. We conclude that validation can be achieved if either one test rejects the null hypotheses without inflating the family-wise Type I error rate. However, using both approaches may improve the efficiency in validation assessment.