Quantile Regression Adjusting for Dependent Censoring

• Activity Number: 624
• Type: Contributed
• Date/Time: Thursday, August 5, 2010 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Ruosha Li*+ and Limin Peng+
• Companies: Emory University and Emory University
• Keywords: Depentent Censoring ; Quantile Regression ; Copula ; Estimating Function

Abstract

In many clinical studies the primary interest is to study the progression of disease. However, the observation on time to some non-terminating landmark event of disease is often subject to censoring by patient dropout, which can be informative to the unobserved disease outcome. In this work we concern performing quantile regression in such a scenario and develop modeling and inference procedures that appropriately adjust for dependent censoring. Based on the proposed unbiased estimating equations, we are able to make simultaneous inferences on the disease progression as well as the association between the disease endpoint and dependent censoring. The resulting estimators are shown to be uniformly consistent and converge weakly to a Gaussian process. We demonstrate the properties of the proposed estimators via simulation studies and a real example.