Detecting Outlying Trials in Network Meta-Analysis

• Activity Number: 345
• Type: Topic Contributed
• Date/Time: Tuesday, August 11, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Jing Zhang* and Haoda Fu and Bradley P. Carlin
• Companies: University of Maryland and Eli Lilly and Company and University of Minnesota
• Keywords: Network meta-analysis ; Trial-level outliers ; Detection measures

Abstract

Network meta-analysis (NMA) expands the scope of a conventional pairwise meta-analysis to simultaneously handle multiple treatment comparisons. However, some trials may appear to deviate markedly from the others, and thus be inappropriate to be synthesized in the NMA. In addition, the inclusion of these trials in evidence synthesis may lead to bias in estimation. We call such trials trial-level outliers. To the best of our knowledge, while heterogeneity and inconsistency in NMA have been extensively discussed and well addressed, few previous papers have considered the proper detection and handling of trial-level outliers. In this talk, I will introduce our proposed Bayesian outlier detection measures, which are then applied to a diabetes data set. Simulation studies comparing our approaches in both the arm-based and contrast-based model settings will also be discussed in this talk.