B1744
Title: Bayesian joint model for time-to-event and longitudinal markers with association based on within-individual variability
Authors: Marco Palma - University of Cambridge (United Kingdom) [presenting]
Ruth Keogh - London School of Hygiene and Tropical Medicine (United Kingdom)
Angela Wood - University of Cambridge (United Kingdom)
Graciela Muniz Terrera - University of Edinburgh (United Kingdom)
Jessica Barrett - University of Cambridge (United Kingdom)
Abstract: In multiple biomedical fields, there is an increasing interest in quantifying within-individual variability of health indicators measured over time, e.g. blood pressure, to inform about disease progression. Simple summary statistics (e.g. the standard deviation for each individual) are often used, not accounting for the longitudinal nature of the data. In addition, when these summary statistics are used as covariates in a regression model for time-to-event outcomes, the estimates of the hazard ratios are subject to regression dilution. To overcome these issues, a joint model is built where the association between the time-to-event outcome and multivariate longitudinal markers is specified in terms of the within-individual variability of the latter. A mixed-effect location-scale model is used to analyse the longitudinal markers, their within-individual variability and their correlation. For the time-to-event outcome, a proportional hazard regression model is considered with a flexible specification of the baseline hazard. A shared parameter structure is assumed for the joint model. The model can be used to quantify within-individual variability for the longitudinal markers and their association with the time-to-event outcome. The model is illustrated on the primary biliary Cirrhosis dataset available in R.
