Interval-Censored Data Analysis of Gaps between Recurrent Events with Cure Fraction

Interval-censored data of gaps between recurrent events arise as the exact gaps can not be available but only know, such as whether the event time of interest is no later than the examination time. The recurrent event times subject to this interval censoring are often encountered in epidemiological studies and clinical trials. In this paper we propose a gaps model which is a derivative of the marginal rate function of the recurrent events processes. This model can accommodate the cure fraction as well. The proposed model can be easily extended to the general case, however, for an easy representation, parametric model is assumed in this paper. As an example of application, the bladder tumor recurrent event data are reanalyzed by using the proposed model.