Bias correction methods for misclassified covariates in the Cox model: Comparison of five correction methods by simulation and data analysis

Heejung Bang, Ya Lin Chiu, Jay S. Kaufman, Mehul D. Patel, Gerardo Heiss, Kathryn M. Rose

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Measurement error/misclassification is commonplace in research when variable(s) cannot be measured accurately. A number of statistical methods have been developed to tackle this problem in a variety of settings and contexts. However, relatively few methods are available to handle misclassified categorical exposure variable(s) in the Cox proportional hazards regression model. In this article, we aim to review and compare different methods to handle this problem - naive methods, regression calibration, pooled estimation, multiple imputation, corrected score estimation, and MC-SIMEX - by simulation. These methods are also applied to a life course study with recalled data and historical records. In practice, the issue of measurement error/ misclassification should be accounted for in design and analysis, whenever possible. Also, in the analysis, it could be more ideal to implement more than one correction method for estimation and inference, with proper understanding of underlying assumptions.

Original languageEnglish (US)
Pages (from-to)381-400
Number of pages20
JournalJournal of Statistical Theory and Practice
Volume7
Issue number2
DOIs
StatePublished - Jan 1 2013

Fingerprint

Bias Correction
Cox Model
Model Comparison
Simulation Analysis
Covariates
Data analysis
Misclassification
Measurement Error
Corrected Score
Regression Calibration
Proportional Hazards Regression
Cox Regression
Hazard Models
Multiple Imputation
Categorical
Statistical method
Regression Model
Simulation

Keywords

  • ARIC
  • Childhood SES
  • Cox proportional hazards regression
  • Measurement error
  • Misclassification
  • Recalled error

ASJC Scopus subject areas

  • Statistics and Probability

Cite this

Bias correction methods for misclassified covariates in the Cox model : Comparison of five correction methods by simulation and data analysis. / Bang, Heejung; Chiu, Ya Lin; Kaufman, Jay S.; Patel, Mehul D.; Heiss, Gerardo; Rose, Kathryn M.

In: Journal of Statistical Theory and Practice, Vol. 7, No. 2, 01.01.2013, p. 381-400.

Research output: Contribution to journalArticle

Bang, Heejung ; Chiu, Ya Lin ; Kaufman, Jay S. ; Patel, Mehul D. ; Heiss, Gerardo ; Rose, Kathryn M. / Bias correction methods for misclassified covariates in the Cox model : Comparison of five correction methods by simulation and data analysis. In: Journal of Statistical Theory and Practice. 2013 ; Vol. 7, No. 2. pp. 381-400.
@article{0cabd518a8dc4f97a2b71d834565134b,
title = "Bias correction methods for misclassified covariates in the Cox model: Comparison of five correction methods by simulation and data analysis",
abstract = "Measurement error/misclassification is commonplace in research when variable(s) cannot be measured accurately. A number of statistical methods have been developed to tackle this problem in a variety of settings and contexts. However, relatively few methods are available to handle misclassified categorical exposure variable(s) in the Cox proportional hazards regression model. In this article, we aim to review and compare different methods to handle this problem - naive methods, regression calibration, pooled estimation, multiple imputation, corrected score estimation, and MC-SIMEX - by simulation. These methods are also applied to a life course study with recalled data and historical records. In practice, the issue of measurement error/ misclassification should be accounted for in design and analysis, whenever possible. Also, in the analysis, it could be more ideal to implement more than one correction method for estimation and inference, with proper understanding of underlying assumptions.",
keywords = "ARIC, Childhood SES, Cox proportional hazards regression, Measurement error, Misclassification, Recalled error",
author = "Heejung Bang and Chiu, {Ya Lin} and Kaufman, {Jay S.} and Patel, {Mehul D.} and Gerardo Heiss and Rose, {Kathryn M.}",
year = "2013",
month = "1",
day = "1",
doi = "10.1080/15598608.2013.772830",
language = "English (US)",
volume = "7",
pages = "381--400",
journal = "Journal of Statistical Theory and Practice",
issn = "1559-8608",
publisher = "Taylor and Francis",
number = "2",

}

TY - JOUR

T1 - Bias correction methods for misclassified covariates in the Cox model

T2 - Comparison of five correction methods by simulation and data analysis

AU - Bang, Heejung

AU - Chiu, Ya Lin

AU - Kaufman, Jay S.

AU - Patel, Mehul D.

AU - Heiss, Gerardo

AU - Rose, Kathryn M.

PY - 2013/1/1

Y1 - 2013/1/1

N2 - Measurement error/misclassification is commonplace in research when variable(s) cannot be measured accurately. A number of statistical methods have been developed to tackle this problem in a variety of settings and contexts. However, relatively few methods are available to handle misclassified categorical exposure variable(s) in the Cox proportional hazards regression model. In this article, we aim to review and compare different methods to handle this problem - naive methods, regression calibration, pooled estimation, multiple imputation, corrected score estimation, and MC-SIMEX - by simulation. These methods are also applied to a life course study with recalled data and historical records. In practice, the issue of measurement error/ misclassification should be accounted for in design and analysis, whenever possible. Also, in the analysis, it could be more ideal to implement more than one correction method for estimation and inference, with proper understanding of underlying assumptions.

AB - Measurement error/misclassification is commonplace in research when variable(s) cannot be measured accurately. A number of statistical methods have been developed to tackle this problem in a variety of settings and contexts. However, relatively few methods are available to handle misclassified categorical exposure variable(s) in the Cox proportional hazards regression model. In this article, we aim to review and compare different methods to handle this problem - naive methods, regression calibration, pooled estimation, multiple imputation, corrected score estimation, and MC-SIMEX - by simulation. These methods are also applied to a life course study with recalled data and historical records. In practice, the issue of measurement error/ misclassification should be accounted for in design and analysis, whenever possible. Also, in the analysis, it could be more ideal to implement more than one correction method for estimation and inference, with proper understanding of underlying assumptions.

KW - ARIC

KW - Childhood SES

KW - Cox proportional hazards regression

KW - Measurement error

KW - Misclassification

KW - Recalled error

UR - http://www.scopus.com/inward/record.url?scp=84876138375&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84876138375&partnerID=8YFLogxK

U2 - 10.1080/15598608.2013.772830

DO - 10.1080/15598608.2013.772830

M3 - Article

AN - SCOPUS:84876138375

VL - 7

SP - 381

EP - 400

JO - Journal of Statistical Theory and Practice

JF - Journal of Statistical Theory and Practice

SN - 1559-8608

IS - 2

ER -