Marginal regression of recurrent gap times based on semiparametric transformation cure model

Abstract

Recurrent event data frequently arise in longitudinal studies. In this article, we consider a marginal semiparametric transformation cure model for gap times between recurrent events. We assume that each individual has the possibility of being cured after each event. We proposed generalized estimating equations for parameter estimation of the regression parameters in this model. The asymptotic properties of the proposed estimators are established under independent censoring. A simulation study is conducted to investigate finite sample properties of the proposed estimators. The proposed method is illustrated using an example of the tumor recurrent data for bladder cancer patients.