Online Program Home
My Program

Abstract Details

Activity Number: 250
Type: Contributed
Date/Time: Monday, August 1, 2016 : 2:00 PM to 3:50 PM
Sponsor: Biometrics Section
Abstract #320071
Title: Induced Smoothing and Efficient Variance Estimation for the Accelerated Gap Times Model with Recurrent Events Data
Author(s): Tianmeng Lyu* and Gongjun Xu and Chiung-Yu Huang and Xianghua Luo
Companies: University of Minnesota and University of Minnesota and The Johns Hopkins University and University of Minnesota
Keywords: accelerated failure time model ; induced smoothing ; log-rank weight ; recurrent event ; variance estimation

A broad class of semiparametric regression models has recently been proposed for the analysis of gap times from a recurrent event process. Among them, the accelerated failure time model (AFTM) is specifically attractive due to its direct physical interpretation. However, the rank-based estimating functions routinely used for AFTMs are non-smooth step functions, and hence the corresponding estimators may not be well defined. Moreover, the commonly used resampling or perturbation based variance estimation for AFTMs require solving rank-based estimating equations repeatedly and hence could become unstable. In this paper, we extend the induced smoothing method to the AFTM for recurrent gap time data and propose a smooth estimating function which permits the application of standard numerical methods. For variance estimation, we propose to use an efficient resampling procedure that does not require solving the estimating equations repeatedly, and hence the computing time is significantly reduced. Simulation studies and a data application are presented to compare the performance of the non-smooth and the proposed smooth estimating functions with various variance estimation methods.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

Copyright © American Statistical Association