10/23/2015

Post date: Oct 26, 2015 1:55:06 PM

Title: Bayesian Network Meta-Analysis of Randomized Clinical Trials

Speaker: Jing Zhang, Department of Epidemiology and Biostatistics, University of Maryland, College Park

Abstract: Decision making bodies for national health care providers have a need to consider all available treatments when making recommendations for clinical practice. Network meta-analysis (NMA) expands the scope of a conventional pairwise meta-analysis to simultaneously handle multiple treatment comparisons, synthesizing both direct and indirect information and thus strengthening inference. The most popular methods to date for NMA are contrast-based (CB), which focus on reporting relative measures and may have serious issues. In this talk I will present our proposed novel arm-based (AB) method in the missing data framework by considering unobserved treatment arms as missing values to be imputed. This method provides more comprehensive reporting summaries and is more user-friendly comparing with the CB method. I will also talk about the extensions of the proposed method to handle the thorny situations: nonignorable missingness and outlyingness. We incorporated nonignorable missingness with selection models method and proposed several Bayesian detection measures to diagnose outlying trials. I will illustrate our methods with real data analyses and simulation studies.