Joint Modeling of Longitudinal and Primary Endpoint Data -- An Introduction

Marie Davidian

Joint modeling of longitudinal measures and primary endpoints in order to understand the association between the time trajectory of variables such as measures of immunological or virological activity in HIV-infected patients and an outcome of interest such as death has generated considerable recent interest. From a formal statistical modeling persepective, this work involves unifying models and methods for longitudinial analysis and primary endpoint analysis in a single framework. Thus, joint modeling requires familiarity both with techniques for representing longitudinal data and for analysis of the primary endpoint. I will review standard methods for longitudinal analysis, analysis of endpoints such as time to an event, and then indicate how these may be combined in various ways to provide a basis for joint analysis.

Return to Biostatistics Working Group