Standard Error Estimation for the Joint Analysis of Survival and Longitudinal Data

• Activity Number: 660
• Type: Invited
• Date/Time: Thursday, August 8, 2013 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Paul David Baines*+ and Jane-Ling Wang and Cong Xu
• Companies: UC Davis and UC Davis and UC Davis
• Keywords: Joint modeling ; Survival analysis ; Longitudinal data ; EM Algorithm

Abstract

Joint modeling of survival and longitudinal data has been studied extensively in recent literature. The likelihood approach is one of the most popular estimation methods employed within the joint modeling framework. Typically the parameters are estimated using maximum likelihood, with computation performed by the EM algorithm. However, one drawback of this approach is that standard error (SE) estimates are not automatically produced when using the EM algorithm. Many different procedures have been proposed to obtain the asymptotic variance-covariance matrix for the parameters when the number of parameters is typically small. In the joint modeling context, however, there is an infinite dimensional parameter, the cumulative baseline hazard function, which makes the problem much more complicated. We study the performance of existing methods for EM SE estimation when applied to this semi-parametric setting, evaluate their precision and computational speed and compare them with results from the profile likelihood and bootstrap methods.