Modeling Nested Multivariate Longitudinal Count Data with Application to HIV Behavioral Intervention

• Activity Number: 586
• Type: Contributed
• Date/Time: Wednesday, August 4, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Yuda Zhu*+ and Robert E. Weiss
• Companies: University of California, Los Angeles and University of California, Los Angeles School of Public Health
• Keywords: Multivariate Longitudinal ; Discrete ; Count ; HIV Prevention ; Behavioral

Abstract

Modeling nested multivariate longitudinal count data presents a challenging statistical problem in that the correlation structure is not explicitly presented as in the multivariate normal distribution. The model we present is motivated by data collected from the Health Living Project (HLP) to evaluate the effects of a randomized counseling intervention. Primary outcomes of interest are the number of protected and unprotected sex acts the subjects engage in with HIV positive and negative partners. Sex acts are naturally nested within partner. We model the hierarchical structure of the primary outcomes and longitudinally over time.