A multiple imputation approach for the Cox-Aalen cure model with interval-censored data

Interval censored survival data, where the exact event time is only known to lie in an interval, is commonly encountered in practice. Furthermore, medical advancements have made it possible for a fraction of patients to be cured. In this article, we analyze interval-censored data using the Cox-Aalen model with a cure fraction, where the probability of being uncured is determined by a logistic regression model and the failure times of the uncured subjects are modelled by the Cox-Aalen model with fixed covariates. We propose a multiple imputation (MI) scheme for obtaining parameter and variance estimates for both the cure probability and survival distribution of the uncured subjects. One major advantage of the proposed MI scheme is its simplicity since it avoids computational complexity resulting from interval censoring and presence of a cure fraction. The presented approach can be implemented by using the existing software for the Cox-Aalen model with right censored data. Simulation studies indicate that the approach performs well for practical situation. We apply the proposed method to the analysis of the data from hypobaric decompression sickness (HDS) study.