Prediction of relevant exposure sources to aggregate chemical exposures from general and occupational environments: exploration of a decision tree approach

Objectives: To understand the total exposure of a human population to a chemical, it is necessary to aggregate exposures from different exposure routes (ingestion, inhalation, dermal uptake) and exposure sources (eg food, air, consumer products) from different environments (ie general, occupational, consumer use). Preventive actions or regulatory decisions require decisions to be taken on priority exposure routes and sources. This study explores the development of a quantitative decision tree to identify relevant exposure sources in the context of aggregate exposure. As a case study for spray applications, it focuses on joint exposure to a specific chemical in a consumer product through domestic use of hairspray, and exposure at the workplace involving surface spraying, such as spray application of paints. Methods: Determinant of the exposure models ART (for workers) and ConsExpo (for the general population) were used to generate a wide range of realistic exposure scenarios. The dominance of one source over another was analyzed through pairwise random comparisons. Exposure estimates from one source containing a specific determinant are compared with those from the other source, scaled by a dominance ratio that defines how much higher one source's exposure must be to be considered dominant. For each comparison, the number of times one source exceeds the other by at least a dominance ratio is counted, resulting in the occurrence. The occurrence is compared with a predefined threshold (eg 80%). If the threshold is met or exceeded, the higher-contributing source is considered dominant and no exposure aggregation is needed; otherwise, aggregation of both sources is recommended. Results: The findings indicated that the use of high- or medium-specification glove boxes, as forms of permanent encapsulation or encasing of the emission source, results in occupational exposure that is negligible compared with the exposure from consumer product use. When these glove boxes were used, hair spray exposure was the dominant source in 89% and 82% of cases, for high and medium specifications, respectively. A spraying activity with surface liquids performed outdoors (close to buildings) showed a significant trend toward occupational exposure dominance in 81% of cases. Using these three determinants, a three-layer quantitative decision tree was built to help users quickly decide whether aggregation was relevant before performing calculations. Aggregation was suggested in 91% of cases and avoided it in 9%.