Nonparametric tests for panel count data with unequal observation processes

Nonparametric comparison for panel count data is discussed. For the situation, most available approaches require that all subjects have the same observation process. However, such an assumption may not hold in reality. To address this, a new class of test procedures are proposed that allow unequal observation processes for the subjects from different treatment groups. The method applies to both univariate and multivariate panel count data. In addition, the asymptotic normality of the proposed test statistics is established and a simulation study is conducted to evaluate the finite sample properties of the proposed approach. The simulation results show that the proposed procedures work well for practical situations and in particular for sparsely distributed data. They are applied to a set of panel count data arising from a skin cancer study.