Semiparametric Methods to Contrast Restricted Mean Gap Times

• Activity Number: 128
• Type: Contributed
• Date/Time: Monday, August 5, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Xu Shu*+ and Douglas Earl Schaubel
• Companies: University of Michigan and University of Michigan
• Keywords: Gap time ; Cox regression ; Multiple imputation ; Restricted mean lifetime

Abstract

Times between successive events (i.e., gap times) are of great importance in survival analysis. Motivated by the comparison of primary- and repeat-transplant patients, our interest is in contrasting the gap time survival functions. Two major challenges in gap time analysis are non-identifiability of the marginal distributions of gap times (except for the first one), and the existence of dependent censoring. To address these issues, we propose two semiparametric methods to compare the first and second gap times with respect to restricted mean lifetime. In both methods, Cox regression is used to estimate the (conditional) survival distributions of each gap time (given the previous gap times), with multiple imputation applied to censored gap times. These methods differ essentially by their treatment of the imputed data. Large-sample properties are derived, with simulation studies carried out to evaluate finite-sample properties. We apply the proposed methods to liver transplant data to contrast survival between the first and repeat transplant patients.