Online Program Home
  My Program

Abstract Details

Activity Number: 57 - Some Recent Developments and Applications of the Empirical Likelihood Method
Type: Topic Contributed
Date/Time: Sunday, July 30, 2017 : 4:00 PM to 5:50 PM
Sponsor: International Chinese Statistical Association
Abstract #323613
Title: Proportional Likelihood Ratio Models for Competing Risks
Author(s): Chiung-Yu Huang* and Tony Sit and Gongjun Xu
Companies: Johns Hopkins University and Chinese University of Hong Kong and University of Michigan
Keywords: density ratio model ; semicompeting risks ; mixture problem ; empirical likelihood

The comparison of overall survival in cancer patients with different metastasis sites can provide important insights into developing new diagnostic and management strategies. However, metastasis is an intermediate event preceding death and is subject to right censoring. As a result, naive comparisons among patients who develop metastatic disease during follow-up yield invalid inferential results because of the bias induced by censoring. On the other hand, treating metastasis as a time-dependent covariate in regression models does not compare subpopulations directly. This research aims to develop statistical methods for comparing failure times of individuals with different types of intermediate events. Our idea is to embed the problem into the competing risk framework by defining the type of competing event according to the intermediate event, and model the conditional density of failure time given the type of failure event using the semiparametric likelihood ratio model. Model estimation procedures based on conditional likelihoods and full likelihoods are developed. Hypothesis testing procedures are discussed. The proposed methods are applied to a pancreatic cancer study.

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

Back to the full JSM 2017 program

Copyright © American Statistical Association