A method for sensitivity analysis to assess the effects of measurement error in multiple exposure variables using external validation data

George O. Agogo*, Hilko van der Voet, Pieter van 't Veer, Pietro Ferrari, David C. Muller, Emilio Sánchez-Cantalejo, Christina Bamia, Tonje Braaten, Sven Knüppel, Ingegerd Johansson, Fred A. van Eeuwijk, Hendriek C. Boshuizen

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

1 Citation (Scopus)

Abstract

Background: Measurement error in self-reported dietary intakes is known to bias the association between dietary intake and a health outcome of interest such as risk of a disease. The association can be distorted further by mismeasured confounders, leading to invalid results and conclusions. It is, however, difficult to adjust for the bias in the association when there is no internal validation data. Methods: We proposed a method to adjust for the bias in the diet-disease association (hereafter, association), due to measurement error in dietary intake and a mismeasured confounder, when there is no internal validation data. The method combines prior information on the validity of the self-report instrument with the observed data to adjust for the bias in the association. We compared the proposed method with the method that ignores the confounder effect, and with the method that ignores measurement errors completely. We assessed the sensitivity of the estimates to various magnitudes of measurement error, error correlations and uncertainty in the literature-reported validation data. We applied the methods to fruits and vegetables (FV) intakes, cigarette smoking (confounder) and all-cause mortality data from the European Prospective Investigation into Cancer and Nutrition study. Results: Using the proposed method resulted in about four times increase in the strength of association between FV intake and mortality. For weakly correlated errors, measurement error in the confounder minimally affected the hazard ratio estimate for FV intake. The effect was more pronounced for strong error correlations. Conclusions: The proposed method permits sensitivity analysis on measurement error structures and accounts for uncertainties in the reported validity coefficients. The method is useful in assessing the direction and quantifying the magnitude of bias in the association due to measurement errors in the confounders.

Original languageEnglish
Article number139
JournalBMC Medical Research Methodology
Volume16
Issue number1
DOIs
Publication statusPublished - 2016

Fingerprint

Vegetables
Fruit
Uncertainty
Mortality
Self Report
Smoking
Diet
Health
Neoplasms
Direction compound

Keywords

  • Attenuation-contamination matrix
  • Bayesian MCMC
  • EPIC study
  • Measurement error
  • Validation study

Cite this

Agogo, George O. ; van der Voet, Hilko ; van 't Veer, Pieter ; Ferrari, Pietro ; Muller, David C. ; Sánchez-Cantalejo, Emilio ; Bamia, Christina ; Braaten, Tonje ; Knüppel, Sven ; Johansson, Ingegerd ; van Eeuwijk, Fred A. ; Boshuizen, Hendriek C. / A method for sensitivity analysis to assess the effects of measurement error in multiple exposure variables using external validation data. In: BMC Medical Research Methodology. 2016 ; Vol. 16, No. 1.
@article{aabdcb089afa4c82a5a66f4e70007249,
title = "A method for sensitivity analysis to assess the effects of measurement error in multiple exposure variables using external validation data",
abstract = "Background: Measurement error in self-reported dietary intakes is known to bias the association between dietary intake and a health outcome of interest such as risk of a disease. The association can be distorted further by mismeasured confounders, leading to invalid results and conclusions. It is, however, difficult to adjust for the bias in the association when there is no internal validation data. Methods: We proposed a method to adjust for the bias in the diet-disease association (hereafter, association), due to measurement error in dietary intake and a mismeasured confounder, when there is no internal validation data. The method combines prior information on the validity of the self-report instrument with the observed data to adjust for the bias in the association. We compared the proposed method with the method that ignores the confounder effect, and with the method that ignores measurement errors completely. We assessed the sensitivity of the estimates to various magnitudes of measurement error, error correlations and uncertainty in the literature-reported validation data. We applied the methods to fruits and vegetables (FV) intakes, cigarette smoking (confounder) and all-cause mortality data from the European Prospective Investigation into Cancer and Nutrition study. Results: Using the proposed method resulted in about four times increase in the strength of association between FV intake and mortality. For weakly correlated errors, measurement error in the confounder minimally affected the hazard ratio estimate for FV intake. The effect was more pronounced for strong error correlations. Conclusions: The proposed method permits sensitivity analysis on measurement error structures and accounts for uncertainties in the reported validity coefficients. The method is useful in assessing the direction and quantifying the magnitude of bias in the association due to measurement errors in the confounders.",
keywords = "Attenuation-contamination matrix, Bayesian MCMC, EPIC study, Measurement error, Validation study",
author = "Agogo, {George O.} and {van der Voet}, Hilko and {van 't Veer}, Pieter and Pietro Ferrari and Muller, {David C.} and Emilio S{\'a}nchez-Cantalejo and Christina Bamia and Tonje Braaten and Sven Kn{\"u}ppel and Ingegerd Johansson and {van Eeuwijk}, {Fred A.} and Boshuizen, {Hendriek C.}",
year = "2016",
doi = "10.1186/s12874-016-0240-1",
language = "English",
volume = "16",
journal = "BMC Medical Research Methodology",
issn = "1471-2288",
publisher = "Springer Verlag",
number = "1",

}

A method for sensitivity analysis to assess the effects of measurement error in multiple exposure variables using external validation data. / Agogo, George O.; van der Voet, Hilko; van 't Veer, Pieter; Ferrari, Pietro; Muller, David C.; Sánchez-Cantalejo, Emilio; Bamia, Christina; Braaten, Tonje; Knüppel, Sven; Johansson, Ingegerd; van Eeuwijk, Fred A.; Boshuizen, Hendriek C.

In: BMC Medical Research Methodology, Vol. 16, No. 1, 139, 2016.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - A method for sensitivity analysis to assess the effects of measurement error in multiple exposure variables using external validation data

AU - Agogo, George O.

AU - van der Voet, Hilko

AU - van 't Veer, Pieter

AU - Ferrari, Pietro

AU - Muller, David C.

AU - Sánchez-Cantalejo, Emilio

AU - Bamia, Christina

AU - Braaten, Tonje

AU - Knüppel, Sven

AU - Johansson, Ingegerd

AU - van Eeuwijk, Fred A.

AU - Boshuizen, Hendriek C.

PY - 2016

Y1 - 2016

N2 - Background: Measurement error in self-reported dietary intakes is known to bias the association between dietary intake and a health outcome of interest such as risk of a disease. The association can be distorted further by mismeasured confounders, leading to invalid results and conclusions. It is, however, difficult to adjust for the bias in the association when there is no internal validation data. Methods: We proposed a method to adjust for the bias in the diet-disease association (hereafter, association), due to measurement error in dietary intake and a mismeasured confounder, when there is no internal validation data. The method combines prior information on the validity of the self-report instrument with the observed data to adjust for the bias in the association. We compared the proposed method with the method that ignores the confounder effect, and with the method that ignores measurement errors completely. We assessed the sensitivity of the estimates to various magnitudes of measurement error, error correlations and uncertainty in the literature-reported validation data. We applied the methods to fruits and vegetables (FV) intakes, cigarette smoking (confounder) and all-cause mortality data from the European Prospective Investigation into Cancer and Nutrition study. Results: Using the proposed method resulted in about four times increase in the strength of association between FV intake and mortality. For weakly correlated errors, measurement error in the confounder minimally affected the hazard ratio estimate for FV intake. The effect was more pronounced for strong error correlations. Conclusions: The proposed method permits sensitivity analysis on measurement error structures and accounts for uncertainties in the reported validity coefficients. The method is useful in assessing the direction and quantifying the magnitude of bias in the association due to measurement errors in the confounders.

AB - Background: Measurement error in self-reported dietary intakes is known to bias the association between dietary intake and a health outcome of interest such as risk of a disease. The association can be distorted further by mismeasured confounders, leading to invalid results and conclusions. It is, however, difficult to adjust for the bias in the association when there is no internal validation data. Methods: We proposed a method to adjust for the bias in the diet-disease association (hereafter, association), due to measurement error in dietary intake and a mismeasured confounder, when there is no internal validation data. The method combines prior information on the validity of the self-report instrument with the observed data to adjust for the bias in the association. We compared the proposed method with the method that ignores the confounder effect, and with the method that ignores measurement errors completely. We assessed the sensitivity of the estimates to various magnitudes of measurement error, error correlations and uncertainty in the literature-reported validation data. We applied the methods to fruits and vegetables (FV) intakes, cigarette smoking (confounder) and all-cause mortality data from the European Prospective Investigation into Cancer and Nutrition study. Results: Using the proposed method resulted in about four times increase in the strength of association between FV intake and mortality. For weakly correlated errors, measurement error in the confounder minimally affected the hazard ratio estimate for FV intake. The effect was more pronounced for strong error correlations. Conclusions: The proposed method permits sensitivity analysis on measurement error structures and accounts for uncertainties in the reported validity coefficients. The method is useful in assessing the direction and quantifying the magnitude of bias in the association due to measurement errors in the confounders.

KW - Attenuation-contamination matrix

KW - Bayesian MCMC

KW - EPIC study

KW - Measurement error

KW - Validation study

U2 - 10.1186/s12874-016-0240-1

DO - 10.1186/s12874-016-0240-1

M3 - Article

VL - 16

JO - BMC Medical Research Methodology

JF - BMC Medical Research Methodology

SN - 1471-2288

IS - 1

M1 - 139

ER -