405 – Inference and Variance Estimation - 1
Exploring the Method for Analyzing Interval Censored Data Using Imputation in Competing Risks Model
Sponsor: Survey Research Methods Section
Keywords: Competing risks, Cumulative Incidence Function, Interval Censoring, Multiple Imputation
Tiandong Li
Westat
Ulrike Luderer
University of California, Irvine
Dean Baker
We consider the problem of analyzing interval censored data comparing cumulative incidence functions in the presence of competing risks. In this paper, we propose a method based on multiple imputation, which will impute the exact event time for interval censored data and take advantage of standard estimation methods for right censored data. Simulations are carried out to examine the performance of the proposed method. We analyze data from the National Children's Study to estimate cumulative risks of transition between Probability of Pregnancy Statuses and to examine the effect of major demographic variables.