Cause specific rate functions for panel count data with multiple modes of recurrence

Abstract: Panel count data arise from longitudinal studies on recurrent events where each subject is observed only at discrete time points. If recurrent events of several types are possible, we obtain panel count data with multiple modes of recurrence. Such data is commonly encountered in medical studies, reliability experiments as well as in sociological studies. In this article, we present cause specific rate functions for the analysis of panel count data with multiple modes of recurrence and develop nonparametric estimation procedures for the same. We derive empirical estimators for the cause specific rate functions and also propose a smoothed version of the same estimators using kernel estimation method. Asymptotic properties of the proposed estimators are studied. A simulation study is conducted to assess the performance of the proposed estimators in finite samples. The practical utility of the proposed method is demonstrated using a real life data arising from skin cancer chemoprevention trial given in Sun and Zhao (2013).