A Bayesian approach for estimating the partial potential impact fraction with exposure measurement error under a main...

Area of Science:

• Epidemiology
• Biostatistics
• Public Health

Background:

• The partial potential impact fraction is crucial for assessing disease burden and potential prevention strategies in populations.
• Exposure measurement error can lead to biased estimates of the partial potential impact fraction, hindering accurate public health interventions.
• Existing methods often struggle to adequately correct for these measurement errors.

Purpose of the Study:

• To develop and validate a Bayesian approach for adjusting partial potential impact fraction estimates when exposure data is measured with error.
• To leverage the strengths of a main study/internal validation study design for robust exposure error correction.
• To apply the novel method to real-world data for estimating the impact of dietary changes on colorectal cancer.

Main Methods:

• A Bayesian statistical framework was employed to adjust for exposure measurement error.
• The reclassification approach was utilized within a main study/internal validation study design.
• Extensive simulations were conducted to evaluate the performance of the proposed estimators.

Main Results:

• The developed Bayesian method effectively corrects for exposure measurement error in partial potential impact fraction estimation.
• Simulation studies demonstrated the accuracy and reliability of both point and credible interval estimators.
• The approach was successfully applied to the Health Professionals Follow-up Study.

Conclusions:

• The proposed Bayesian approach provides a valid and robust method for estimating the partial potential impact fraction in the presence of exposure measurement error.
• This method enhances the accuracy of public health assessments and informs evidence-based interventions.
• Accurate estimation is vital for understanding the impact of modifiable risk factors on disease incidence, such as colorectal cancer.