Review and validation of exposure assessment methods

Abstract

The purpose of this research is twofold, to standardize and to validate exposure assessment methods. First, the attempt is made to standardize the manner in which exposure assessment methods are developed. Literature on the subject is reviewed and seven common elements discovered to be common are discussed. The seven elements are causative agents, exposure groups, exposure-modifying parameters, industrial hygiene measurement data, misclassification issues, validation issues, and reliability issues. It is believed that thinking in terms of these elements will yield more consistent and complete exposure assessment models. Three types of exposure estimation methods are reviewed in this form. These methods are selected because they are the most thorough and represent the most frequently used and referenced types of estimation strategies: the statistical model, the deterministic model, and the multiplicative model. Second, the paper reports on an attempt to validate a semiquantitative exposure assessment model against industrial hygiene data collected from employees of one firm in the maritime industry. The set of data contains 440 samples with 75 percent of them censored by the method limit of detection. Methods to calculate an average concentration with nondetectable data are discussed. It is concluded that (1) the model does not predict the data well, (2) the industrial hygiene data does not properly fit the tails of a lognormal distribution, and (3) that average exposure to benzene in the (un)loading of petrochemicals from tankers is decidedly below exposure limits.