Combining expert ratings and exposure measurements: a random effect paradigm

Area of Science:

• Occupational Health
• Epidemiology
• Statistical Modeling

Background:

• Accurate estimation of exposure group (EG)-specific means is crucial for epidemiological studies.
• Combining expert judgment with exposure measurements presents statistical challenges, particularly regarding between-worker variability.
• Existing methods may not fully integrate ordinal expert ratings with quantitative exposure data.

Purpose of the Study:

• To present a novel paradigm for integrating ordinal expert ratings with exposure measurements.
• To develop a statistical model that accounts for between-worker effects in estimating EG-specific means.
• To provide a framework for combining expert judgment and exposure data for enhanced epidemiological analysis.

Main Methods:

• A nested two-way random effects model was developed, combining expert ratings with exposure measurements.
• Gibbs sampling was employed to fit the model, incorporating prior information on variance components.
• An approximate formula was derived to estimate EG-specific means, leveraging information across EGs.

Main Results:

• The proposed paradigm effectively combines expert ratings and exposure measurements, accounting for between-worker variance.
• Application to dust exposure data in a steel factory revealed differences in EG-specific means compared to estimates without ratings.
• The model allows for estimation of rating-specific means under various hypotheses, demonstrating flexibility.

Conclusions:

• The developed model offers a robust framework for integrating independent expert exposure ratings with quantitative measurements.
• This approach enhances the accuracy of exposure group mean estimation for epidemiological research.
• It is recommended that expert exposure ratings be established independently of existing exposure measurements for optimal model performance.