Semiparametric AUC Model with Placement Values

• Activity Number: 256
• Type: Contributed
• Date/Time: Monday, August 2, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Lin Zhang*+ and Jack Tubbs
• Companies: Quintiles and Baylor University
• Keywords: semi-parametric ; AUC ; ROC curve

Abstract

In this paper, we present a semi-parametric procedure for the analysis of the Area Under the ROC Curve (AUC) using placement values of the treated population with respect to the placebo population. Since these placement scores are quantiles from the placebo population, we will use beta regression to estimate the AUC. This approach will allow for both discrete and continuous covariates. When the covariates are discrete, the model effectively reduces to the binary regression model for AUC given by Dodd and Pepe (2003). When the covariates are continuous, the results are easily interpreted as they are similar to results given in standard generalized linear models software. Simulation result will be provided.