Predictive models (short - term and long term animal studies; biological monitoring) are useful in occupational toxicology. For instance, validated in vivo predictive tests with experimental animals are useful in assessing skin irritation, allergic contact dermatitis, fibrosis, etc. You may want to confer with this NIOSH abstract:

http://www.cdc.gov/niosh/nioshtic-2/20039047.html

Ali and Chithan, thank you very much!

Hi Mihaela

Use of biomarkers in occupational toxicology and epidemiological research supplies information on various stages of the multistep process of putative disease. It allows the identification of biological signs, or biomarkers, that may indicate an increased risk for disease. Markers of exposure have an established position in environmental and occupational health; some reflect progression along the causal pathway to disease, and others reflect innate or acquired susceptibility to the effects of etiological agents. If unequivocal markers of effects and markers of susceptibility to disease could be developed, the identification of individuals/groups at increased risk would then be helpful to the field of preventive medicine.

http://www.sciencedirect.com/science/article/pii/S0378427498002525

In Nanotoxicology and Nanomedicine - quantitative nanostructure activity relationships is used to predict the pharmacokinetics and biodistribution of nanomaterials in the body

https://www.crcpress.com/product/isbn/9781482203875

Predictive models, especially for cumulative toxic hazards with possible long term effects could contribute to legal lowering thresholds of certain biomarkers of occupational exposure previously considered safe. Once again, thank you all for providing useful information.