A Bartlett type correction for Wald test in Cox regression model

In clinical trials with time to an event endpoint, the Wald test in a Cox regression model may be used to assess the difference between two treatments. In this paper, a Bartlett type correction to the Wald test is derived and applied to the analysis of data from a clinical trial on actuate myelogenous leukemia (AML). Monte Carlo simulations are also performed to assess the actual type I error and power of the proposed Bartlett corrected test.