Evaluation of Predictive Capacities of Biomarkers Based on Research Synthesis

• Activity Number: 253
• Type: Contributed
• Date/Time: Monday, August 10, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Satoshi Hattori* and Xiao-Hua Zhou
• Companies: Kurume University and University of Washington
• Keywords: diagnostic study ; meta-analysis ; predictive value curve ; prognostic study ; receiver operating characteristics ; research synthesis

Abstract

The objective of diagnostic studies or prognostic studies is to evaluate and compare predictive capacities of biomarkers. Suppose we are interested in evaluation and comparison of predictive capacities for a binary outcome of continuous biomarkers based on research synthesis. In analysis of each study, subjects are often classified into two groups of the high- and low-expression groups according to a cut-off value and statistical analysis is based on a 2x2 tables defined by the response and the high- or low-expression of the biomarker. Since the cut-off is study-specific, it is difficult to interpret a combined summary measure such as a odds ratio based on the standard meta-analysis techniques. We develop a method to estimate positive or negative predictive curves, which are an alternative to the receiver operating characteristic curve, based on research synthesis. Our method is useful to summarize positive or negative predictive value reported in published studies and to compare biomarkers accounting for heterogeneity of cut-off values among studies.