ABSTRACT
In addition to the estimation of covariate effects, the time-varying transformation models of Chapter 13 can be further used to predict the ranking of a subject’s health status at one or multiple time points. We describe in this chapter the methods for estimating the conditional distribution functions and some of their useful functionals based on the models and estimation methods of Chapter 13. These estimators lead to useful statistical indices to quantitatively track the changes of health status in biomedical studies.
