Semiparametric analysis of recurrent discrete time data with competing risks

Abstract

The regression analysis of the cumulative incidence function on recurrent discrete time data with competing risks has not been widely investigated. We propose semiparametric analysis for regression modelling of the cumulative incidence function for recurrent discrete time competing risks data. The maximum likelihood inferences are developed for the estimation of the model parameters in a discrete time competing risks model, which are based on a working independence likelihood. We derive the sandwich variance estimator to account for correlations between multiple discrete times within each subject. Simulation studies show that the procedures perform well. The proposed methods are illustrated with two applications, a study of contraceptive use in Indonesia and data on unemployment in Germany.

Keywords:

• Competing risks
• cumulative incidence function
• maximum likelihood inferences
• recurrent discrete times

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

This work was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education [NRF-2018R1D1A1B07041070].