A0413
Title: Log-rank test with coarsened exact matching
Authors: Nakahiro Yoshida - University of Tokyo (Japan) [presenting]
Abstract: It is of special importance in the clinical trial to compare survival times between the treatment group and the control group. In practice, the distributions of the covariates differ between the two groups since the distribution of the assignments depends on the covariates of the individuals. The complex structure between covariates, treatment assignment and survival time makes it difficult to apply parametric methods, such as the propensity score by logistic regression, to correctly match individuals of the different groups to assess the treatment effect. The coarsened exact matching (CEM) is considered, which does not need any parametric models, and a weighted log-rank statistic based on CEM is proposed. Asymptotic properties of the weighted log-rank statistic are given, i.e., the asymptotic normality under the null hypothesis and the consistency of the test. Simulation studies show that the log-rank statistic with CEM is more robust than the log-rank statistic with the propensity score matching.
