Cox regression with covariate measurement error

Chengcheng Hu, D. Y. Lin

Research output: Contribution to journalArticle

31 Citations (Scopus)

Abstract

This article deals with parameter estimation in the Cox proportional hazards model when covariates are measured with error. We consider both the classical additive measurement error model and a more general model which represents the mis-measured version of the covariate as an arbitrary linear function of the true covariate plus random noise. Only moment conditions are imposed on the distributions of the covariates and measurement error. Under the assumption that the covariates are measured precisely for a validation set, we develop a class of estimating equations for the vector-valued regression parameter by correcting the partial likelihood score function. The resultant estimators are proven to be consistent and asymptotically normal with easily estimated variances. Furthermore, a corrected version of the Breslow estimator for the cumulative hazard function is developed, which is shown to be uniformly consistent and, upon proper normalization, converges weakly to a zero-mean Gaussian process. Simulation studies indicate that the asymptotic approximations work well for practical sample sizes. The situation in which replicate measurements (instead of a validation set) are available is also studied.

Original languageEnglish (US)
Pages (from-to)637-655
Number of pages19
JournalScandinavian Journal of Statistics
Volume29
Issue number4
StatePublished - Dec 2002
Externally publishedYes

Fingerprint

Cox Regression
Measurement Error
Covariates
Cumulative Hazard Function
Estimator
Measurement Error Model
Partial Likelihood
Cox Proportional Hazards Model
Score Function
Moment Conditions
Random Noise
Estimating Equation
Asymptotic Approximation
Likelihood Function
Gaussian Process
Linear Function
Normalization
Parameter Estimation
Sample Size
Regression

Keywords

  • Censoring
  • Corrected score
  • Mismeasured covariates
  • Partial likelihood
  • Proportional hazards
  • Survival data

ASJC Scopus subject areas

  • Mathematics(all)
  • Statistics and Probability

Cite this

Cox regression with covariate measurement error. / Hu, Chengcheng; Lin, D. Y.

In: Scandinavian Journal of Statistics, Vol. 29, No. 4, 12.2002, p. 637-655.

Research output: Contribution to journalArticle

@article{11bda626d23c4052995c6df864ceb666,
title = "Cox regression with covariate measurement error",
abstract = "This article deals with parameter estimation in the Cox proportional hazards model when covariates are measured with error. We consider both the classical additive measurement error model and a more general model which represents the mis-measured version of the covariate as an arbitrary linear function of the true covariate plus random noise. Only moment conditions are imposed on the distributions of the covariates and measurement error. Under the assumption that the covariates are measured precisely for a validation set, we develop a class of estimating equations for the vector-valued regression parameter by correcting the partial likelihood score function. The resultant estimators are proven to be consistent and asymptotically normal with easily estimated variances. Furthermore, a corrected version of the Breslow estimator for the cumulative hazard function is developed, which is shown to be uniformly consistent and, upon proper normalization, converges weakly to a zero-mean Gaussian process. Simulation studies indicate that the asymptotic approximations work well for practical sample sizes. The situation in which replicate measurements (instead of a validation set) are available is also studied.",
keywords = "Censoring, Corrected score, Mismeasured covariates, Partial likelihood, Proportional hazards, Survival data",
author = "Chengcheng Hu and Lin, {D. Y.}",
year = "2002",
month = "12",
language = "English (US)",
volume = "29",
pages = "637--655",
journal = "Scandinavian Journal of Statistics",
issn = "0303-6898",
publisher = "Wiley-Blackwell",
number = "4",

}

TY - JOUR

T1 - Cox regression with covariate measurement error

AU - Hu, Chengcheng

AU - Lin, D. Y.

PY - 2002/12

Y1 - 2002/12

N2 - This article deals with parameter estimation in the Cox proportional hazards model when covariates are measured with error. We consider both the classical additive measurement error model and a more general model which represents the mis-measured version of the covariate as an arbitrary linear function of the true covariate plus random noise. Only moment conditions are imposed on the distributions of the covariates and measurement error. Under the assumption that the covariates are measured precisely for a validation set, we develop a class of estimating equations for the vector-valued regression parameter by correcting the partial likelihood score function. The resultant estimators are proven to be consistent and asymptotically normal with easily estimated variances. Furthermore, a corrected version of the Breslow estimator for the cumulative hazard function is developed, which is shown to be uniformly consistent and, upon proper normalization, converges weakly to a zero-mean Gaussian process. Simulation studies indicate that the asymptotic approximations work well for practical sample sizes. The situation in which replicate measurements (instead of a validation set) are available is also studied.

AB - This article deals with parameter estimation in the Cox proportional hazards model when covariates are measured with error. We consider both the classical additive measurement error model and a more general model which represents the mis-measured version of the covariate as an arbitrary linear function of the true covariate plus random noise. Only moment conditions are imposed on the distributions of the covariates and measurement error. Under the assumption that the covariates are measured precisely for a validation set, we develop a class of estimating equations for the vector-valued regression parameter by correcting the partial likelihood score function. The resultant estimators are proven to be consistent and asymptotically normal with easily estimated variances. Furthermore, a corrected version of the Breslow estimator for the cumulative hazard function is developed, which is shown to be uniformly consistent and, upon proper normalization, converges weakly to a zero-mean Gaussian process. Simulation studies indicate that the asymptotic approximations work well for practical sample sizes. The situation in which replicate measurements (instead of a validation set) are available is also studied.

KW - Censoring

KW - Corrected score

KW - Mismeasured covariates

KW - Partial likelihood

KW - Proportional hazards

KW - Survival data

UR - http://www.scopus.com/inward/record.url?scp=0036901081&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0036901081&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0036901081

VL - 29

SP - 637

EP - 655

JO - Scandinavian Journal of Statistics

JF - Scandinavian Journal of Statistics

SN - 0303-6898

IS - 4

ER -