Joint Analysis of Longitudinal Measurements and Competing Risks Failure Time Data

• Activity Number: 255
• Type: Contributed
• Date/Time: Tuesday, August 8, 2006 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Ning Li*+ and Robert Elashoff and Gang Li
• Companies: University of California, Los Angeles and University of California, Los Angeles and University of California, Los Angeles
• Keywords: competing risks model ; EM algorithm ; joint modeling ; longitudinal data ; mixed effects model ; multiple outcomes

Abstract

We study a joint model for longitudinal measurements and competing risks survival data. Our model consists of a linear mixed effects sub-model for the longitudinal outcome and a proportional cause-specific hazards frailty sub-model (Prentice et al, 1978) for the competing risks survival data, linked together by some latent random effects. We propose to obtain the maximum likelihood estimates of the parameters by an EM-based algorithm and estimate their standard errors using a profile likelihood method. Our joint model offers a flexible approach to handle non-ignorable missing data in the longitudinal measurements after event times. It is also an extension of previous joint models with a single failure type, providing a possible way to model informative censoring as a competing risk. Our method is evaluated using both simulated data and a clinical trial for the scleroderma lung disease.