|
Activity Number:
|
86
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, July 31, 2016 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Section on Nonparametric Statistics
|
|
Abstract #321370
|
|
|
Title:
|
Predictive Accuracy of Markers or Risk Scores for Interval-Censored Survival Data
|
|
Author(s):
|
Yuan Wu* and Xiaofei Wang and Kouros Owzar
|
|
Companies:
|
Duke University and Duke University and Duke University
|
|
Keywords:
|
Interval censoring ;
sieve estimation ;
AUC
|
|
Abstract:
|
Time-dependent predictive accuracy measures of markers or risk scores for right censored survival data have been discussed extensively in the literature. Interval censored data is another type of survival data commonly encountered in cancer or other medical disease studies. The existing methods of estimating time-dependent predictive accuracy measures are not applicable for interval censored data. In this proposal we propose to develop nonparametric sieve estimators for time-dependent area under ROC curve (AUC) based on sensitivity and specificity for markers or risk scores for interval censored survival data. Simulation studies show that our proposed estimate for the AUC is consistent and the variance estimation for the estimated AUC is also provided for hypothesis testing. Additionally, the proposed method will be illustrated with data from real cancer studies subject to interval censoring.
|
Authors who are presenting talks have a * after their name.
