A risk set calibration method for failure time regression by using a covariate reliability sample

Regression parameter estimation in the Cox failure time model is considered when regression variables are subject to measurement error. Assuming that repeat regression vector measurements adhere to a classical measurement model, we can consider an ordinary regression calibration approach in which the unobserved covariates are replaced by an estimate of their conditional expectation given available covariate measurements. However, since the rate of withdrawal from the risk set across the time axis, due to failure or censoring, will typically depend on covariates, we may improve the regression parameter estimator by recalibrating within each risk set. The asymptotic and small sample properties of such a risk set regression calibration estimator are studied. A simple estimator based on a least squares calibration in each risk set appears able to eliminate much of the bias that attends the ordinary regression calibration estimator under extreme measurement error circumstances. Corresponding asymptotic distribution theory is developed, small sample properties are studied using computer simulations and an illustration is provided.

[1]  J. Cornfield.  Joint dependence of risk of coronary heart disease on serum cholesterol and systolic blood pressure: a discriminant function analysis. , 1962, Federation proceedings.

[2]  David R. Cox,et al.  Regression models and life tables (with discussion , 1972 .

[3]  R. Prentice,et al.  Commentary on Andersen and Gill's "Cox's Regression Model for Counting Processes: A Large Sample Study" , 1982 .

[4]  R. Prentice.  Covariate measurement errors and parameter estimation in a failure time regression model , 1982 .

[5]  Raymond J. Carroll,et al.  On errors-in-variables for binary regression models , 1984 .

[6]  W. Willett,et al.  Reproducibility and validity of a semiquantitative food frequency questionnaire. , 1985, American journal of epidemiology.

[7]  John D. Kalbfleisch,et al.  Misspecified proportional hazard models , 1986 .

[8]  L. J. Wei,et al.  The Robust Inference for the Cox Proportional Hazards Model , 1989 .

[9]  T. Nakamura,et al.  Proportional hazards model with covariates subject to measurement error. , 1992, Biometrics.

[10]  M. Feinleib,et al.  Statistical Models for Longitudinal Studies of Health , 1992 .

[11]  L. Lissner,et al.  Dietary underreporting by obese individuals--is it specific or non-specific? , 1995, BMJ.

[12]  M. Pepe,et al.  Auxiliary covariate data in failure time regression , 1995 .

[13]  R. Prentice,et al.  Measurement error and results from analytic epidemiology: dietary fat and breast cancer. , 1996, Journal of the National Cancer Institute.

[14]  W. Tsai,et al.  On using the Cox proportional hazards model with missing covariates , 1997 .

[15]  Myunghee Cho Paik.  Multiple Imputation for the Cox Proportional Hazards Model with Missing Covariates , 1997, Lifetime data analysis.

[16]  R. Prentice,et al.  Regression calibration in failure time regression. , 1997, Biometrics.

[17]  M. Davidian,et al.  Estimating the parameters in the Cox model when covariate variables are measured with error. , 1998, Biometrics.

[18]  R. Little,et al.  Proportional hazards regression with missing covariates , 1999 .

[19]  Yijian Huang,et al.  Cox Regression with Accurate Covariates Unascertainable: A Nonparametric-Correction Approach , 2000 .

[20]  Haoxuan Zhou,et al.  Failure time regression with continuous covariates measured with error , 2000 .