Maximum Likelihood Estimation for Cox's Regression Model Under Case-Cohort Sampling

From MaRDI portal





In the Prentice case-cohort design right censored survival times and censoring indicators are observed for all samples (cohorts) whence the covariate values are known only in cases for which the survival time was not censored and for a random sub-sample (sub-cohort) from cases with censored survival times. A typical example is an epidemiological study at which the covariates are observed for all ill persons and for a control group of healthy persons. Cox's regression model is considered for the effect of covariates. A maximum likelihood estimator for the regression coefficients is described and an EM-type algorithm is proposed for the computation of this estimator. The obtained estimator is compared with the estimators proposed by \textit{S. G. Self} and \textit{R. L. Prentice} [Ann. Stat. 16, No. 1, 64--81 (1988; Zbl 0666.62108)] and \textit{K. Chen} and \textit{S.-H. Lo} [Biometrika 86, No. 4, 755--764 (1999; Zbl 0940.62108)]. Simulation results are presented.




Cited in
(44)








This page was built for publication: Maximum Likelihood Estimation for Cox's Regression Model Under Case-Cohort Sampling

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4677096)