Bayesian analysis of pair-matched case-control studies subject to outcome misclassification

Area of Science:

• Epidemiology
• Biostatistics
• Medical Research

Background:

• Case-control studies are susceptible to misclassification bias, particularly with outcome assessment.
• Nondifferential outcome misclassification can distort odds ratio estimates.
• Accurate estimation is crucial for understanding disease associations.

Purpose of the Study:

• To evaluate the impact of nondifferential outcome misclassification on odds ratios in pair-matched case-control studies.
• To propose and validate a Bayesian model for adjusting odds ratio estimates affected by misclassification bias.

Main Methods:

• Developed a Bayesian statistical model to correct for outcome misclassification bias.
• Utilized a validation subgroup with confirmed outcome status for all case-control pairs.
• Incorporated prior knowledge of the classification mechanism's predictive values.

Main Results:

• The proposed Bayesian model effectively adjusts odds ratio estimates compromised by nondifferential outcome misclassification.
• Simulated data demonstrated the model's performance in correcting bias.
• Application to a multiple sclerosis prodromal phase study showed improved estimation accuracy.

Conclusions:

• Bayesian adjustment models offer a robust solution for misclassification bias in case-control studies.
• Accurate outcome classification and bias correction are vital for reliable epidemiological findings.
• This methodology enhances the validity of odds ratio estimates in observational research.