We propose a flexible joint longitudinal-survival framework to examine the
association between longitudinally collected biomarkers and a time-to-event
endpoint. More specifically, we use our method for analyzing the survival
outcome of end-stage renal disease patients with time-varying serum albumin
measurements. Our proposed method is robust to common parametric assumptions in
that it avoids explicit distributional assumptions on longitudinal measures and
allows for subject-specific baseline hazard in the survival component. Fully
joint estimation is performed to account for the uncertainty in the estimated
longitudinal biomarkers included in the survival model.