The art of occupational exposure modelling - development and evaluation of generic inhalation exposure models

Abstract

When working with chemical substances, workers might be exposed to chemical contaminants. In a risk assessment the exposure is compared with a toxicological limit value. In a risk assessment the toxicological effect of a chemical substance is compared with the exposure to the chemical in order to estimate the risk for adverse health effects. The preferred method to estimate exposure is measurement of actual exposure concentrations in the breathing zone of workers at the workplace. However, as the exposure assessment community is not able to collect sufficient number of exposure measurements to perform accurate exposure assessments for all relevant exposure scenarios, the availability of reliable and accurate exposure models is considered to be critical in the regulatory process. To deal with the enormous burden of exposure assessments a tiered approach has been proposed. In the first tier of this approach, relatively simple exposure models are used to efficiently assess workers’ exposure. In a second tier, more advanced modelling approaches are needed to produce more realistic exposure estimates. In the third and last tier, the exposure estimate will be based on a comprehensive exposure measurement survey. As the existing modelling approaches were evaluated and argued not to be sufficient, the new European regulation referred to as the Registration, Evaluation, Authorization and restriction of Chemicals (REACH) created a momentum for the development of new occupational exposure modelling approaches. This thesis describes the development and evaluation of two generic exposure assessment models used in regulatory risk assessment. The Stoffenmanager is a first tier exposure assessment model and should be used as a screenings tool for discrimination between scenarios of concern and scenarios of no concern. The Advanced REACH Tool (ART) is aimed to be a higher tier exposure assessment model providing more realistic and precise exposure estimates. The ART framework includes a mechanistic model and a Bayesian module. The Bayesian model combines the mechanistic model estimate with available exposure measurement data in order to arrive at an updated exposure estimate based on both sources of information. The exposure estimates provided by both models were validated using actual exposure measurements. Moreover, the reproducibility of the ART mechanistic model was evaluated. Also the uncertainty of the ART model estimates from the mechanistic model as well as the estimates based on the ART Bayesian model were compared with the uncertainty of exposure estimates based on exposure datasets collected with different sampling designs. The validation of the two exposure assessment tools with measured exposure measurements showed that the tools are useful instruments to assess exposure levels in the regulatory process. However, the lack of reproducible results when the ART mechanistic model is used by different experts showed substantial deviation in exposure estimates from occupational hygiene professionals for the same scenario. Although the reproducibility of the Stoffenmanager and the Bayesian module of ART was not explicitly evaluated it was hypothesized that more effort is needed to make these tools useful for occupational hygiene professionals. Training of the professional exposure assessment community and improving the guidance material is needed in order to let them come up with accurate estimates.