A Bayesian Missing Data Framework for Generalized Multiple Outcome Mixed Treatment Comparisons

• Activity Number: 236
• Type: Topic Contributed
• Date/Time: Monday, August 5, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Health Policy Statistics Section
• Author(s): Hwanhee Hong*+ and Haitao Chu and Jing Zhang and Bradley P. Carlin
• Companies: Division of Biostatistics, University of Minnesota and University of Minnesota School of Public Health and University of Minnesota School of Public Health and University of Minnesota
• Keywords: Bayesian hierarchical model ; Markov chain Monte Carlo ; missingness mechanism ; network meta-analysis

Abstract

Bayesian statistical approaches to mixed treatment comparisons (MTCs) are becoming more popular due to their flexibility and interpretability. Many randomized clinical trials report multiple outcomes with possible inherent correlations. MTC data are typically sparse and researchers often choose study arms based on previous trials. In this paper, we summarize existing hierarchical Bayesian methods for MTCs with a single outcome, and we introduce novel Bayesian approaches for multiple outcomes. We do this by incorporating missing data and correlation structure between outcomes through contrast- and arm-based parameterizations that consider any unobserved treatment arms as missing data to be imputed. We also extend the model to apply to all types of generalized linear model outcomes, such as count or continuous. We develop a new measure of inconsistency under our missing data framework, having more straightforward interpretation and implementation. We offer a simulation study under various missingness mechanisms (e.g., MCAR, MAR, and MNAR) providing evidence that our models outperform existing models in terms of bias and MSE, then illustrate our methods with two real MTC datasets.