A Bayesian Hierarchical Model for Mixed Treatment Comparison Meta-Analysis Accommodating Missingness

• Activity Number: 294
• Type: Contributed
• Date/Time: Tuesday, August 5, 2014 : 8:30 AM to 10:20 AM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Yulun Liu*+ and Stacia M. DeSantis and Yong Chen
• Companies: University of Texas Health Science Center at Houston and University of Texas Health Science Center at Houston and University of Texas School of Public Health
• Keywords: Bayesian hierarchical model ; Missingness mechanism ; Meta-analysis ; Mixed treatment comparison

Abstract

The mixed treatment comparisons (MTCs) methodology, which incorporates both direct and indirect treatment comparisons, has been applied across disciplines to assess clinical effectiveness, but it has been under-utilized in addition research despite costly clinical trials that often result in null findings. Addition trials are challenging to assimilate into a summary measure using meta-analytic methods due to the tendency to report multiple correlated outcomes and the challenge of missing outcomes. We propose a Bayesian hierarchical MTC method for multivariate, missing (binary and continuous) outcomes to improve efficiency and reduce the impact of potentially unreported outcomes. We conduct extensive simulation studies to evaluate performance of our proposed method compared to the univariate MTC meta-analysis model under three missingness mechanisms: MCAR, MAR, and MNAR, and illustrate our method on two Cochrane database systematic reviews of naltrexone (NAL) and acamprosate (ACA).