B1424
Title: Joint models of longitudinal and survival data with censoring and outliers
Authors: Lang Wu - University of British Columbia (Canada) [presenting]
Abstract: In a survival model with a time-dependent covariate, the covariate may be left-censored due to a lower detection limit, and its observed values may contain outliers. Motivated by an HIV vaccine study, a robust method is proposed for joint models of longitudinal and survival data, where the outliers in longitudinal data are addressed using a multivariate t-distribution for b-outliers and an M-estimator for e-outliers. A computationally efficient method is also proposed for approximate likelihood inference. The proposed method is evaluated by simulation studies. Based on the proposed models and method, the HIV vaccine data is analyzed and a strong association is found between longitudinal biomarkers and the risk of HIV infection.
