Hypothesis Testing of Treatment Policies in Two-Stage Randomization Designs

• Activity Number: 212
• Type: Invited
• Date/Time: Monday, August 2, 2010 : 2:00 PM to 3:50 PM
• Sponsor: WNAR
• Author(s): Xiang Guo*+ and Anastasios Tsiatis
• Companies: sanofi-aventis (China) and North Carolina State University
• Keywords: Counting process ; Intent-to-treat ; Inverse Weight ; Risk sets ; Potential outomes ; Survival analysis

Abstract

In many clinical trials related to diseases such as cancer and HIV, patients are treated by different combinations of therapies. This leads to the two-stage randomization designs, where patients are initially randomized to a primary therapy and then depending on the disease remission and patients'consent, a maintenance therapy will be randomly assigned. This design can be considered as the simplest form of the dynamic treatment regimes. We used the concepts of counting process and risk sets to find weighted estimating equations whose solution gives an estimate for the cumulative hazard function which, in turn, is used to derive an estimator for the survival distributions with right- censored data. we also proposed a method using an inverse weighted version of log-rank test for the hypothesis testing to compare treatment combinations. The result will be demonstrated through simulations