Title: Diagnostics and predictions for joint models of survival and multivariate longitudinal data
Authors: Marcella Mazzoleni - University of Milano Bicocca (Italy) [presenting]
Mariangela Zenga - University of Milano-Bicocca (Italy)
Abstract: The joint models analyse the effect of longitudinal covariates onto the risk of an event. The longitudinal and the survival sub-models compose the joint models. The survival sub-model is a proportional hazard model, while the longitudinal sub-model is a linear multivariate mixed model. An Expectation-Maximisation (EM) algorithm which maximises the joint likelihood function is implemented. For testing the goodness of fit of the algorithm some diagnostics elements will be presented, such as residuals for both sub-model and the estimated survival function. Moreover, the dynamic predictions for the survival part of the model based on the longitudinal covariates will be shown.