Assessing techniques for quantifying the impact of bias due to an unmeasured confounder: An applied example

Julie Barberio, Thomas P. Ahern, Richard F. Maclehose, Lindsay J. Collin, Deirdre P. Cronin-Fenton, Per Damkier, Henrik Toft Sørensen, Timothy L. Lash

Research output: Contribution to journalArticlepeer-review

Abstract

Purpose: To compare the magnitude of bias due to unmeasured confounding estimated from various techniques in an applied example. Patients and Methods: We examined the association between dibutyl phthalate (DBP) and incident estrogen receptor (ER)-positive breast cancer in a Danish nationwide cohort (N=1,122,042). Cox regression analyses were adjusted for age and active drug compounds contributing to DBP exposure. We estimated the hazard ratios (HRs) that would have been observed had one of the DBP sources been unmeasured and calculated the strength of confounding by comparing to the fully adjusted HR. We performed a quantitative bias analysis (QBA) of the “unmeasured” confounder, using external information to specify the bias parameters. Upper bounds on the bias were estimated and E-values were calculated. Results: The adjusted HR for incident ER-positive breast cancer among women with high-level (≥10,000 cumulative milligrams) versus no DBP exposure was 2.12 (95% confidence interval 1.12 to 4.05). Removing each DBP source in isolation resulted in negligible change in the HR. The bias estimates from the QBA ranged from 1.00 to 1.01. The estimated maximum impact of unmeasured confounding ranged from 1.01 to 1.51. E-values ranged from 3.46 to 3.68. Conclusion: The impact of bias due to simulated unmeasured confounding was negligible, in part, because the unmeasured variable was not independent of controlled variables. When a suspected confounder cannot be measured in the study data, our exercise suggests that QBA is the most informative method for assessing the impact. E-values may best be reserved for situations where uncontrolled confounding emanates from an unknown confounder.

Original languageEnglish (US)
Pages (from-to)627-635
Number of pages9
JournalClinical Epidemiology
Volume13
DOIs
StatePublished - Jul 1 2021

Bibliographical note

Funding Information:
This work was supported in part by Susan G. Komen for the Cure (CCR13264024) awarded to Thomas P. Ahern and the US National Library of Medicine (R01LM013049) awarded to Timothy L Lash. Lindsay J. Collin and Timothy L. Lash were supported in part by awards from the US National Cancer Institute (F31CA239566 and R01CA166825, respectively). Thomas P. Ahern was supported in part by an award from the US National Institute of General Medical Sciences (P20 GM103644).

Funding Information:
This work was supported in part by Susan G. Komen for the Cure (CCR13264024) awarded to Thomas P. Ahern and the US National Library of Medicine (R01LM013049) awarded to Timothy L Lash. Lindsay J. Collin and Timothy L. Lash were supported in part by awards from the US National Cancer Institute (F31CA239566 and

Funding Information:
R01CA166825, respectively). Thomas P. Ahern was supported in part by an award from the US National Institute of General Medical Sciences (P20 GM103644).

Funding Information:
Julie Barberio's doctoral stipend and tuition are supported by an award to Emory University from Amgen, Inc. The Department of Clinical Epidemiology, Aarhus University Hospital, receives funding for other studies from companies in the form of research grants to (and administered by) Aarhus University. None of these studies have any relation to the present study. Timothy L. Lash is a member of the Methods Advisory Council for Amgen, Inc, for which he is compensated as a consultant. Lindsay J. Collin reports grants from NCI and NCATS during the conduct of the study. The authors report no other conflicts of interest in this work.

Publisher Copyright:
© 2021 Barberio et al.

Keywords

  • Bias analysis
  • The E-value
  • Unmeasured confounding

PubMed: MeSH publication types

  • Journal Article

Fingerprint

Dive into the research topics of 'Assessing techniques for quantifying the impact of bias due to an unmeasured confounder: An applied example'. Together they form a unique fingerprint.

Cite this