Additive hazards regression with covariate measurement error

The additive hazards model specifies that the hazard function conditional o n a set of covariates is the sum of an arbitrary baseline hazard function a nd a regression function of covariates. This article deals with the analysi s of this semiparametric regression model with censored failure time data w hen covariates are subject to measurement error. We assume that the true co variate is measured on a randomly chosen validation set, whereas a surrogat e covariate (i.e., an error-prune version of the true covariate) is measure d on all study subjects. The surrogate covariate is modeled as a linear fun ction of the true covariate plus a random error. Only moment conditions are imposed on the measurement error distribution. We develop a class of estim ating functions for the regression parameters that involve weighted combina tions of the contributions from the validation and nonvalidation sets. The optimal weight can be selected by an adaptive procedure. The resulting esti mators are consistent and asymptotically normal with easily estimated varia nces. Simulation results demonstrate that the asymptotic approximations are adequate for practical use. Illustration with a real medical study is prov ided.