Statistics Colloquium

Location

Mathematics/Psychology : 103

Date & Time

October 10, 2014, 11:00 am12:00 pm

Description

Speaker
Hwanhee Hong, PhD 
Postdoctoral Fellow 
Department of Mental Health, JHU.

Title
 Hierarchical Bayesian Methods for Multiple Outcomes in Network Meta-Analysis

Abstract
Biomedical decision makers confronted with questions about the comparative effectiveness and safety of interventions often wish to combine all sources of data. Network meta-analysis (NMA) is a meta-analytic statistical technique that extends traditional meta-analysis of two treatments to simultaneously incorporate the findings from several studies on multiple treatments. In the NMA data framework, since few head-to-head comparisons are available, we must combine indirect and direct evidence. Many randomized clinical trials report multiple, possibly correlated outcomes for each treatment. Moreover, aggregated-level NMA data are typically sparse and researchers often choose study arms based on previous trials, further complicating identification of the best treatment. In this talk, we introduce novel Bayesian approaches for multiple outcomes simultaneously, rather than in separate NMA analyses. We do this by incorporating partially observed data and its correlation structure between outcomes through contrast-and arm-based parameterizations that consider any unobserved treatment arms as missing data to be imputed. Furthermore, availability of individual patient-level data (IPD) broadens the scope of NMA, and enables us to incorporate patient-level information into the analysis. As such, we propose a Bayesian IPD NMA modeling framework for bivariate continuous outcomes. We illustrate this approach using diabetes treatment, and show its practical implications. Finally, we close with a brief description of areas for future research.