Abstract

A hybrid test statistics was proposed in the literature to analyze matched studies with non-normally distributed outcomes. In this article, we investigated and compared the hybrid statistic with the metaanalysis t test under commonly used interim analysis settings in clinical trials. We estimated the empirical powers and the empirical type I errors among the 10,000 simulated datasets with different sample sizes, different effect sizes, different correlation coefficients for matched pairs, and different data distributions, respectively, in the interim and final analysis with 4 different group sequential methods. Results from our simulation study show that, compared to the meta-analysis t-test commonly used for data with normally distributed observations, the hybrid statistic almost keeps the powerfor data observed from normally distributed random variables and generally achieves greater power for log-normally, and multinomially distributed random variables with matched and unmatched subjects as well as with outliers. Powers rose with the increase in sample size, effect size, and correlation coefficient for the matched pairs. In addition, lower type I errors were observed by using the hybrid statistic in most of the cases studied, which indicates that this test is also conservative for data with outliers in the interim analysis of clinical trials.

Keywords: Clinical Trials, Empirical Power, Empirical Size,Interim Analysis