Sieve Estimation with Bivariate Interval Censored Data

Bivariate interval censored data arises in many applications. However, both theoretical and computational investigations for this type of data are limited because of the complicated structure of bivariate censoring. In this paper, we propose a two-stage spline-based sieve estimator for the association between two event times with bivariate case 2 interval censored data. A smooth and explicit estimator for the joint distribution function is also available. The proposed estimators are shown to be asymptotically consistent and computationally efficient. We demonstrate the finite sample performances of the spline-based sieve estimators using both simulation studies and real data analysis from an AIDS clinical trial study. YUAN WU and XIAOLI GAO 38