Moment-Based Covariate Adjustment Method for Treatment Effect Estimation in Randomized Clinical Trials

• Activity Number: 177
• Type: Contributed
• Date/Time: Monday, August 1, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Stephen George*+ and Junling Ma and Xiaofei Wang
• Companies: Duke University and Duke University and Duke University
• Keywords: Covariate adjustment ; Efficiency ; Empirical likelihood ; Randomized clinical trial ; Semiparametric method

Abstract

In the analysis of randomized clinical trials, the covariates that correlate with the primary outcome are often adjusted in the estimation of treatment effect in order to improve efficiency and to compensate for any lack of baseline covariate balance between treatment arms. There are many different techniques for adjusting for baseline covariates, one of which commonly used is the analysis of covariance if the primary outcome is quantitative. Our research focuses on the approach advocated by Koch et al. (1998) and Tsiatis et al. (2007), which has the advantage of incorporating the covariates in treatment effect estimation while requiring no joint regression modeling of treatment and covariates. We propose a moment-based covariate adjustment method that constrain all higher order moments of the covariate distribution. The proposed method requires no specification of the parametric form of any regression model and the estimates of its regression coefficients. Asymptotic properties of the proposed method are established. Simulation studies show that the proposed method has nice finite sample properties and perform well compared to existing methods. Data example will be provided.