An industrial perspective on a quantitative estimation of risk associated with low level exposures of humans, with...

Having been presented with a set of data on acrylonitrile to be utilized as a case study, an assessment was performed of the potential risk of carcinogenesis associated with low level exposures to a material with the characteristics in the supplied data package. This study attempts to present the risk assessment process in an open manner, clearly identifying the multiple complex steps involved in the process and the uncertainty associated with each of the steps. The approach used was truly pragmatic, as the author believes that, in many cases, those performing risk assessments stand on dogma, failing to recognize that what is scientifically best in addressing uncertainty is not that which is most conservative (i.e., uses a worst case decision mode), but rather that which reduces the uncertainty the most. All available data of suitable quality should be utilized in either performing the risk assessment or in checking the result. All the mathematical steps involved in the risk assessment process have limited biological basis. As a result, any risk assessment process that results in predictions which are refuted by real data should be rejected.