Joint Modeling of Longitudinal Drug Using Pattern and Time to First Relapse in Cocaine Dependence Treatment Data

• Activity Number: 134
• Type: Contributed
• Date/Time: Monday, August 4, 2014 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistical Computing
• Author(s): Jun Ye*+ and Yehua Li and Yongtao Guan
• Companies: University of Akron and Iowa State University and University of Miami
• Keywords: Akaike information criterion ; Functional principal components ; Generalized longitudinal data ; Interval censoring ; Louis formula ; Penalized splines

Abstract

An important endpoint variable in a cocaine rehabilitation study is the time to first relapse of a patient after the treatment. We propose a joint modeling approach based on functional data analysis to study the relationship between the baseline longitudinal cocaine-use pattern and the interval censored time to first relapse. The variations within the generalized longitudinal trajectory are modeled through a latent Gaussian process, which is characterized by a few leading functional principle components. The association between the baseline longitudinal trajectories and the posttreatment first relapse is built upon the latent principal component scores. The mean and eigenfunctions of the latent longitudinal process as well as the hazard function of time to first relapse are modeled nonparametrically using penalized splines, and the parameters in the joint model are estimated by a Monte Carlo EM algorithm based on Metropolis-Hastings steps. Our analysis of the cocaine dependence treatment data will show the relapse time is related to the trend in the cocaine-use tractors rather than the average use amount.