|
Abstract:
|
Joint frailty models have been proposed for analysis of correlated survival data (e.g., recurrent or clustered failure times) in the presence of informative dropout or a dependent terminal event like death. The estimation in such models often involves MCEM algorithms, which is slow in convergence and difficult to implement, limiting the further application of such models. In this paper we apply Gauss quadrature techniques for estimation. The unspecified baseline hazard is approximated by a piecewise constant one. The resulting parametric model can be fitted easily by Gauss quadrature tools in standard software, e.g., SAS Proc NLMIXED. Simulation studies show such approximation yields satisfactory results for both normal and Gamma frailty models. Our method greatly facilitates the application of the joint frailty models in practice. Two real examples are given for illustration purpose.
|
