Analysis of two-sample censored data using a semiparametric mixture model

In this article we study a semiparametric mixture model for the two-sample problem with right censored data. The model implies that the densities for the continuous outcomes are related by a parametric tilt but otherwise unspecified. It provides a useful alternative to the Cox (1972) proportional hazards model for the comparison of treatments based on right censored survival data. We propose an iterative algorithm for the semiparametric maximum likelihood estimates of the parametric and nonparametric components of the model. The performance of the proposed method is studied using simulation. We illustrate our method in an application to melanoma. Gang Li’s research was supported in part by the U.S. National Institute of Health (No. CA016042, No. P01AT003960). Chien-Tai Lin’s research was supported in part by the National Science Council of Taiwan (No. 89-2118-M-032-021, No. 96-2628-M-032-002-MY3).