Joint Modelling for Longitudinal Data and Survival Data
July 26, 2018 14:30-15:30
W303 School of Mathematics, Sichuan University
Jianxin Pan (School of Mathematics, University of Manchester, UK)
In medical practices, it is not uncommon that multiple events from each patient are observed simultaneously over time. Such events may be longitudinal observations in different categories or time to events with longitudinal biomarkers, among many others. Modelling such complex data requires joint statistical modelling strategy, which accounts for inherent associations between multiple events due to data coming from the same individual. Ignoring the inherent associations may cause very inaccurate statistical inferences and bias statistical conclusions. In the talk, I will use several case studies including Osteoarthritis trial and Hypertension trial to show the important role of inherent associations of multiple events played in statistical modelling. I will also introduce how to jointly model longitudinally multiple events through analysing the real data sets arising in the case studies.