@bullet @bullet @bullet Auxiliary Covariate Measurement Problem in Failure Time Regression Auxiliary Covariate Measurement Problem Failure Time Regression

SUMMARY • In • • • A semi-parametric method, Estimated Partial Likelihood(EPL) method, is proposed for the estimation of the relative risk parameters when auxiliary covariates are present. The asymptotic distribution theory is derived for the proposed estimator for the case in which the surrogate or mismeasured covariates are categorical. The asymptotic relative efficiency of the EPL estimator with respect to the fully paramet-ric maximum likelihood analysis and a partial likelihood analysis based on those with true covariates only are examined under the exponential model. The EPL analysis is found to compare favorably with more model-dependent analysis and more efficient than the partial likelihood analysis in most cases of practical importance. Small sam-pIe properties are investigated by simulation studies and a real data example is used to illustrate the EPL method. Doctoral dissertation by Haibo Zhou. The authors thank Ed Davis for the helpful discussion about SOLVD data and comments.