A0424
Title: Large sample results for frequentist multiple imputation for Cox regression with missing covariate data
Authors: Frank Eriksson - University of Copenhagen (Denmark) [presenting]
Torben Martinussen - University of Copenhagen (Denmark)
Soren Feodor Nielsen - Copenhagen Business School (Denmark)
Abstract: Incomplete information on explanatory variables is commonly encountered in studies of possibly censored event times. A popular approach to deal with partially observed covariates is multiple imputation, where a number of completed data sets that can be analyzed by standard complete data methods are obtained by imputing missing values from an appropriate distribution. Using a consistent and asymptotically linear but inefficient initial estimator, we impute missing values conditional on the observed data ensuring compatibility with a Cox regression model. We show that estimators of both the finite-dimensional regression parameter and the infinite-dimensional cumulative baseline hazard parameter by Cox regression applied to the completed data sets are consistent and weak convergence is established. We derive a consistent estimator of the covariance operator. Simulation studies and an application to a study on survival after treatment for liver cirrhosis show that the estimators perform well with moderate sample sizes and indicate that iterating the multiple-imputation estimator increases the precision.
