Title: Joint models for longitudinal and survival data: An orthodox best linear unbiased predictor approach
Authors: Renjun Ma - University of New Brunswick (Canada) [presenting]
Xingde Duan - Guizhou University of Finance and Economics (China)
Abstract: In medical studies, longitudinal and survival outcomes are frequently collected over time on each of many subjects. As a random effects Cox survival model can be characterized as an auxiliary Poisson random effects model, we can employ our techniques on joint modelling for different types of longitudinal data to handle joint modelling of longitudinal and survival outcomes. An optimal estimation of our model has been developed using orthodox best linear unbiased predictor of random effects. The analysis results do not rely on any distributional assumption of random effects. The approach will be illustrated with real data examples.