POWER OF LOGRANK TEST AND COX REGRESSION-MODEL IN CLINICAL-TRIALS WITH HETEROGENEOUS SAMPLES

This paper evaluates the loss of power of the simple and stratified lo grank tests due to heterogeneity of patients in clinical trials and pr oposes a flexible and efficient method of estimating treatment effects adjusting for prognostic factors. The results of the paper are based on the analyses of survival data from a large clinical trial which inc ludes more than 6000 cancer patients. Major findings from the simulati on study on power are: (i) for a heterogeneous sample, such as advance d cancer patients, a simple logrank test can yield misleading results and should not be used; (ii) the stratified logrank test may suffer so me power loss when many prognostic factors need to be considered and t he number of patients within stratum is small. To address the problems due to heterogeneity, the Cox regression method with a special hazard model is recommended. We illustrate the method using data from a gast ric cancer clinical trial. (C) 1997 by John Whey & Sons, Ltd.