Bayesian adjustment for measurement error in continuous exposures in an individually matched case-control study
2011

Bayesian Method for Correcting Measurement Error in Case-Control Studies

Sample size: 271 publication Evidence: moderate

Author Information

Author(s): Gabriela Espino-Hernandez, Paul Gustafson, Igor Burstyn

Primary Institution: University of British Columbia

Hypothesis

Can a Bayesian method effectively correct for measurement error in multiple continuous exposures in individually matched case-control studies?

Conclusion

The proposed Bayesian method can correct for measurement error in continuous exposures, showing little adjustment needed for the level of measurement error observed.

Supporting Evidence

  • The Bayesian method was illustrated using data from a study on perfluorinated acids and thyroid hormone levels.
  • Sensitivity analysis showed that larger measurement errors lead to more substantial adjustments.
  • The method can be implemented in WinBUGS software for practical use.

Takeaway

This study created a new way to fix mistakes in measuring things in health studies, helping to understand if certain chemicals affect pregnant women’s thyroid health.

Methodology

A Bayesian approach was used to correct for measurement error in exposures, employing conditional logistic regression and random-effect models.

Potential Biases

Potential biases may arise from the assumptions of non-differential measurement error.

Limitations

The method assumes no misclassification of case/control status and may not generalize well to other settings.

Participant Demographics

Pregnant women aged 18 or older, with 96 cases and 175 matched controls.

Statistical Information

Confidence Interval

95% credible intervals were computed for odds ratios.

Digital Object Identifier (DOI)

10.1186/1471-2288-11-67

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication