Compare Meta-Analysis Methods in Diagnostic Test Accuracy Studies

• Activity Number: 550
• Type: Contributed
• Date/Time: Wednesday, August 7, 2013 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Jin-Hua Chen*+ and Chun-Shu Chen and Tzeng Sheng-Li
• Companies: Graduate Institute of Biostatistics, China Medical University and National Changhua University of Education and Department of Public Health, China Medical University
• Keywords: meta-analysis ; diagnostic test ; sROC ; bivariant model

Abstract

Meta-analysis is a very import step in EBM (Evidence-Based Medicine) and popular method to try to answer the question in uncertainty clinical problem. We would introduce briefly the difference within systematic review and meta-analysis in general steps for medical research. They would pool the statistics with different trials in same sampling design for interesting topics. Today, we focus on the method for diagnostic test accuracy study in pooling the sensitivity, specificity, diagnostic odds ratio (DOR) in clinical medical studies. There have some generally applicable methods: univariate, summary ROC curve (sROC) and bivariate model. Our main approach will compare and explain the data-analytic results of these methods. We would show some simulation the results in some scenarios.