Multi-State Model with Missing Continuous Covariate

• Activity Number: 554
• Type: Contributed
• Date/Time: Wednesday, August 12, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Wenjie Lou* and Richard J. Kryscio
• Companies: University of Kentucky and University of Kentucky
• Keywords: multi-state model ; missing covariate ; maximum simulated likelihood ; longitudinal data

Abstract

Multi-state models are very useful tools to model chronic disease process in which patients might go through several difference states. For example, in the dementia study patients might experience an intermittent state called mild cognitive impairment (MCI) before they become demented or dead. The common way to account the personal difference in the disease process in the model is to add covariates in the transition intensity functions. In most applications, covariates have to be fully observed. In this paper, we proposed a maximum simulated likelihood method to handle the missing continuous covariates in Multi-state models. Our simulation study shows that the proposed method works quite well in most MAR cases. We also apply the method to a real dataset, a longitudinal dementia study cohort of 732 subjects. Keywords: multi-state model, missing covariate, maximum simulated likelihood