Quantile regression model for interval-censored data with competing risks.

Our interest is to provide the methodology for estimating quantile regression model for interval-censored competing risk data. Lee and Kim [Analysis of interval censored competing risk data via nonparametric multiple imputation. Stat. Biopharm. Res. 13 (2020), pp. 367-374.] applied a censoring complete data concept suggested by Ruan and Gray [Analyses of cumulative incidence function via non-parametric multiple imputation. Sta. Med. 27 (2008), pp. 5709-5724.] to recover a missing information related with competing events. In this paper, we also applied it to a quantile regression model. The simulated censoring times of the competing events are generated with a multiple imputation technique and the survival function of right censoring times. The performance of suggested methods is evaluated by comparing with the result of a simple imputation method under several distributions and sample sizes. The AIDS dataset is analyzed to estimate the effect of several covariates on the quantiles of cause-specific CIF as a real data analysis.