Modeling Batched Gaussian Longitudinal Data Subject to Informative Dropout

• Activity Number: 111
• Type: Topic Contributed
• Date/Time: Monday, August 2, 2010 : 8:30 AM to 10:20 AM
• Sponsor: ENAR
• Author(s): Paul S. Albert*+ and Joanna H. Shih
• Companies: Eunice Kennedy Shriver National Institute of Child Health and Human Development and National Cancer Institute
• Keywords: dropout ; marginal inference ; missing data ; pooling ; repeated measures ; longitudinal data

Abstract

Modeling longitudinal data subject to informative dropout is an active area in statistical research. This article focuses on modeling such longitudinal data when the outcome at each follow-up time is collected in batches rather than individually collected. The motivating example is a study that compared the weight of mice over time between a control and a treatment group, where animal weight was measured in groups of five animals per cage. We develop both a shared parameter and a pattern mixture modeling approach for accounting for potentially informative dropout due to an animal's death. Through simulations, we show that both approaches due well under a correctly specified dropout mechanism. However, the pattern mixture modeling approach is more sensitive to informative dropout and more robust to model misspecification than the shared random parameter model.