B1600
Title: Jointly modeling lung function decline, nutritional evolution and pulmonary exacerbation onset
Authors: Pedro Miranda Afonso - Erasmus University Medical Center (Netherlands) [presenting]
Rhonda Szczesniak - Cincinnati Children Hospital Medical Center (United States)
Dimitris Rizopoulos - Erasmus University Rotterdam (Netherlands)
Grace Zhou - University of Cincinnati (United States)
John Clancy - Cystic Fibrosis Foundation (United States)
Anushka Palipana - Cincinnati Children's Hospital Medical Center (United States)
Erika Rasnick - Cincinnati Children's Hospital Medical Center (United States)
Cole Brokamp - University of Cincinnati (United States)
Patrick Ryan - University of Cincinnati (United States)
Ruth Keogh - London School of Hygiene and Tropical Medicine (United Kingdom)
Eleni-Rosalina Andrinopoulou - Erasmus Medical Center (Netherlands)
Abstract: Cystic fibrosis (CF) is an inherited disease primarily affecting the lungs and gastrointestinal tract. It is of clinical interest to simultaneously investigate the association between the risk of recurrent pulmonary exacerbations (PEx), lung function measured as percent-predicted of forced expiratory volume in 1 second (FEV1) decline, nutritional status (BMI) evolution, and the risk of lung transplant/death. Previous work has been limited to continuous longitudinal markers and time-to-first PEx, and ignored the spatial variability among individuals. This was mainly due to the unavailability of appropriate and robust statistical methods and software. We propose a Bayesian hierarchical model for jointly modeling multiple longitudinal markers, and recurrent and terminal event processes. We account for the individual geographical location and explore different forms of association between the markers and the events of interest. The developed model is available in the R statistical package JMbayes2. Full MCMC algorithm implementation in C++ enables model fit in a timely fashion, despite its complexity. The proposed multivariate joint model allows more efficient use of all available data. It thereby brings new insights into CF disease progression across different geographical regions and enhances our understanding of risks posed by PEx.
