Regression Modeling of Competing Risks Survival Data in the Presence of Covariates Based on a Generalized Weibull Distribution: A Simulation Study

Abstract

In survival analysis or medical studies each person can be exposed to more than one type of outcomes which occurrence of one of them prevents the other outcomes' occurrence; this situation is called the competing risks. Assessing the effect of covariates on the survival time (or failure time) is one of the purposes in competing risks analysis. In this paper, we study a competing risks model in the presence of covariates when the causes of failures follow generalized Weibull distributions. Covariates are entered to the model through the scale parameter of this distribution. Also in this study the competing risks are considered to be independent. Parameter estimation has been done by the maximum likelihood approach, in a real data set and a simulation study has shown the advantages of proposed model.