Dunnett's Multiple Comparison Tests for Survival Data

This document is a technical report that details the methodology for conducting Dunnett's multiple comparison tests of treatments against a control group using survival data. The report outlines the computation of power and sample size necessary for such studies, specifically employing the one-sided logrank statistic to compare multiple treatment groups with a common control. It discusses the family-wise error rate (FWER) adjustment to address the multiplicity issue inherent in multiple testing scenarios. The document explains the one-way design approach, emphasizing the correlation among tests due to the shared control group. It also provides technical details regarding the logrank test, including the formulation of null and alternative hypotheses and the statistical properties of the test statistic. Furthermore, the report includes a power calculation example, illustrating the process of determining the required sample size to achieve a specified power level while controlling the FWER. The assumptions regarding the exponential survival distribution and the parameters involved in the calculations are also described.