A hierarchical Bayesian framework has been developed for exposure assessment that makes use of statistical sampling-based techniques to estimate the posterior probability of the 95th percentile or arithmetic mean of the exposure distribution being located in one of several exposure categories. The framework can synthesize professional judgment and monitoring data to yield an updated posterior exposure assignment for routine exposure management. The framework is versatile enough that it can be modified for use in epidemiological studies for classifying the arithmetic mean instead of the 95th percentile into several exposure categories. Various physico-chemical exposure models have also been incorporated in the hierarchical framework. The use of the framework in three settings has been illustrated. First, subjective judgments about exposure magnitude obtained from industrial hygienists for five tasks were treated as priors in the Bayesian framework. Monitoring data for each task were used to create a likelihood function in the hierarchical framework and the posterior was predicted in terms of the 95th percentile being located in each of the four AIHA exposure categories. The accuracy of the exposure judgments was then evaluated. Second, we illustrate the use of exposure models to develop priors in this framework and compare with monitoring data in an iron foundry. Finally, we illustrate the use of this approach for retrospective exposure assessment in a chemical manufacturing facility, to categorize exposures based on arithmetic mean instead of 95th percentile.
- decision making
- exposure model
- judgment accuracy
- occupational exposure judgment