An imputation-based solution to using mismeasured covariates in propensity score analysis

Yenny Webb-Vargas, Kara Rudolph, David Lenis, Peter Murakami, Elizabeth A. Stuart

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

Although covariate measurement error is likely the norm rather than the exception, methods for handling covariate measurement error in propensity score methods have not been widely investigated. We consider a multiple imputation-based approach that uses an external calibration sample with information on the true and mismeasured covariates, multiple imputation for external calibration, to correct for the measurement error, and investigate its performance using simulation studies. As expected, using the covariate measured with error leads to bias in the treatment effect estimate. In contrast, the multiple imputation for external calibration method can eliminate almost all the bias. We confirm that the outcome must be used in the imputation process to obtain good results, a finding related to the idea of congenial imputation and analysis in the broader multiple imputation literature. We illustrate the multiple imputation for external calibration approach using a motivating example estimating the effects of living in a disadvantaged neighborhood on mental health and substance use outcomes among adolescents. These results show that estimating the propensity score using covariates measured with error leads to biased estimates of treatment effects, but when a calibration data set is available, multiple imputation for external calibration can be used to help correct for such bias.

Original languageEnglish (US)
Pages (from-to)1824-1837
Number of pages14
JournalStatistical Methods in Medical Research
Volume26
Issue number4
DOIs
StatePublished - Aug 1 2017
Externally publishedYes

Fingerprint

Propensity Score
Multiple Imputation
Imputation
Calibration
Covariates
Measurement Error
Treatment Effects
Vulnerable Populations
Estimate
Exception
Biased
Mental Health
Health
Eliminate
Likely
Simulation Study
Norm
Therapeutics

Keywords

  • Causal inference
  • measurement error
  • multiple imputation
  • nonexperimental study

ASJC Scopus subject areas

  • Epidemiology
  • Statistics and Probability
  • Health Information Management

Cite this

An imputation-based solution to using mismeasured covariates in propensity score analysis. / Webb-Vargas, Yenny; Rudolph, Kara; Lenis, David; Murakami, Peter; Stuart, Elizabeth A.

In: Statistical Methods in Medical Research, Vol. 26, No. 4, 01.08.2017, p. 1824-1837.

Research output: Contribution to journalArticle

Webb-Vargas, Yenny ; Rudolph, Kara ; Lenis, David ; Murakami, Peter ; Stuart, Elizabeth A. / An imputation-based solution to using mismeasured covariates in propensity score analysis. In: Statistical Methods in Medical Research. 2017 ; Vol. 26, No. 4. pp. 1824-1837.
@article{db862903d54741daa74c8046f0b50d27,
title = "An imputation-based solution to using mismeasured covariates in propensity score analysis",
abstract = "Although covariate measurement error is likely the norm rather than the exception, methods for handling covariate measurement error in propensity score methods have not been widely investigated. We consider a multiple imputation-based approach that uses an external calibration sample with information on the true and mismeasured covariates, multiple imputation for external calibration, to correct for the measurement error, and investigate its performance using simulation studies. As expected, using the covariate measured with error leads to bias in the treatment effect estimate. In contrast, the multiple imputation for external calibration method can eliminate almost all the bias. We confirm that the outcome must be used in the imputation process to obtain good results, a finding related to the idea of congenial imputation and analysis in the broader multiple imputation literature. We illustrate the multiple imputation for external calibration approach using a motivating example estimating the effects of living in a disadvantaged neighborhood on mental health and substance use outcomes among adolescents. These results show that estimating the propensity score using covariates measured with error leads to biased estimates of treatment effects, but when a calibration data set is available, multiple imputation for external calibration can be used to help correct for such bias.",
keywords = "Causal inference, measurement error, multiple imputation, nonexperimental study",
author = "Yenny Webb-Vargas and Kara Rudolph and David Lenis and Peter Murakami and Stuart, {Elizabeth A.}",
year = "2017",
month = "8",
day = "1",
doi = "10.1177/0962280215588771",
language = "English (US)",
volume = "26",
pages = "1824--1837",
journal = "Statistical Methods in Medical Research",
issn = "0962-2802",
publisher = "SAGE Publications Ltd",
number = "4",

}

TY - JOUR

T1 - An imputation-based solution to using mismeasured covariates in propensity score analysis

AU - Webb-Vargas, Yenny

AU - Rudolph, Kara

AU - Lenis, David

AU - Murakami, Peter

AU - Stuart, Elizabeth A.

PY - 2017/8/1

Y1 - 2017/8/1

N2 - Although covariate measurement error is likely the norm rather than the exception, methods for handling covariate measurement error in propensity score methods have not been widely investigated. We consider a multiple imputation-based approach that uses an external calibration sample with information on the true and mismeasured covariates, multiple imputation for external calibration, to correct for the measurement error, and investigate its performance using simulation studies. As expected, using the covariate measured with error leads to bias in the treatment effect estimate. In contrast, the multiple imputation for external calibration method can eliminate almost all the bias. We confirm that the outcome must be used in the imputation process to obtain good results, a finding related to the idea of congenial imputation and analysis in the broader multiple imputation literature. We illustrate the multiple imputation for external calibration approach using a motivating example estimating the effects of living in a disadvantaged neighborhood on mental health and substance use outcomes among adolescents. These results show that estimating the propensity score using covariates measured with error leads to biased estimates of treatment effects, but when a calibration data set is available, multiple imputation for external calibration can be used to help correct for such bias.

AB - Although covariate measurement error is likely the norm rather than the exception, methods for handling covariate measurement error in propensity score methods have not been widely investigated. We consider a multiple imputation-based approach that uses an external calibration sample with information on the true and mismeasured covariates, multiple imputation for external calibration, to correct for the measurement error, and investigate its performance using simulation studies. As expected, using the covariate measured with error leads to bias in the treatment effect estimate. In contrast, the multiple imputation for external calibration method can eliminate almost all the bias. We confirm that the outcome must be used in the imputation process to obtain good results, a finding related to the idea of congenial imputation and analysis in the broader multiple imputation literature. We illustrate the multiple imputation for external calibration approach using a motivating example estimating the effects of living in a disadvantaged neighborhood on mental health and substance use outcomes among adolescents. These results show that estimating the propensity score using covariates measured with error leads to biased estimates of treatment effects, but when a calibration data set is available, multiple imputation for external calibration can be used to help correct for such bias.

KW - Causal inference

KW - measurement error

KW - multiple imputation

KW - nonexperimental study

UR - http://www.scopus.com/inward/record.url?scp=85027702416&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85027702416&partnerID=8YFLogxK

U2 - 10.1177/0962280215588771

DO - 10.1177/0962280215588771

M3 - Article

C2 - 26037527

AN - SCOPUS:85027702416

VL - 26

SP - 1824

EP - 1837

JO - Statistical Methods in Medical Research

JF - Statistical Methods in Medical Research

SN - 0962-2802

IS - 4

ER -