Multi-state Mixture Cure Model

In modern medical research,the time data of multiple recurrence event has attracted extensive attention because of its important data structure characteristics,and we usually use multistate model to analyze the time data of recurrent events and terminal events.In this paper,a multi-state mixture cure model is proposed to analyze the time data of multiple recurrence events and terminal events,and we add frailty to explain the heterogeneity between subjects.Considering the existence of cured subjects,we use EM algorithm to fit the multi-state mixture cure model to obtain the estimation of parameters and baseline risk function.In the numerical simulation,we verify the validity of the multiple recurrence model estimation process under different scenarios,and compare the fitting results of the multiple recurrence model and the primary recurrence model.Finally,we use the proposed multiple recurrence model to the simulated data data Multiv in the R package.