Christopher’s research interests related to his thesis center around the development of non-parametric methods for missing data structures. In his collaboration with his advisor, Mark van der Laan, they have developed inferential methods for the Nonparametric Maximum Likelihood Estimator for bivariate truncated data, and introduced a new class of estimators, the Locally Efficient Estimators, for bivariate right censored data. He and collaborator Peter Bacchetti of the University of California at San Francisco have developed methods to estimate the hazard function for interval censored data as smooth functions of time dependent covariates.

Christopher’s current interests involve the development of statistical models based on historical data to improve patient care in a managed health care environment.