A Flexible Two-Dimensional Distribution for Combining Studies in Meta-Analyses of Binary Events

• Activity Number: 233
• Type: Contributed
• Date/Time: Monday, August 4, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Jingjing Yan*+ and Eloise Kaizar and Steven N. MacEachern
• Companies: Ohio State University and Ohio State University and Ohio State University
• Keywords: meta-analysis ; binary data ; GLMM ; two-dimensional space

Abstract

Under common meta-analysis approaches, such as the Mantel-Haenszel method or the DerSimonian-Laird method, one would first choose a summary metric (e.g.,relative risk, odds ratio or risk difference) and then 'combine' across studies. The choice between different summary metrics is not straightforward given their different mathematical properties, ease of interpretation and variability across studies. In this paper, we proposed a flexible summary metric that would allow a smooth transition between the common relative measures of effect size (relative risk and odds ratio). The newly proposed metric is incorporated into a complete hierarchical model based on a GLM for a 2 by 2 table (with fixed row margins, as from a controlled clinical trial). We utilize such model to describe a more flexible probability distribution for studies across a two-dimensional space, to estimations an average effect and to predict future studies.