A risk‐neutral default for chemical risk management

Abstract

In many risk management decisions concerning industrial chemicals, including decisions on classification and labeling, lack of toxicity data is interpreted as (or has the same implications as) absence of toxicity. In other words, if the toxicity of a chemical is unknown, it is treated as having no or low toxicity. This practice is difficult to defend from a decision-theoretical point of view. We apply standard decision theory to toxicity data and investigate an alternative approach in which substances with unknown properties are treated as if they had the average toxicity among tested substances in the group to which they belong. An index of acute toxicity is proposed and then used to define a risk-neutral hazard default that can be applied to industrial chemicals for which no specific information on acute toxicity is available. It is proposed that such a risk-neutral approach is preferable to the current practice of treating substances with unknown acute toxicity in the same manner as substances that can reasonably be assumed to have no such harmful properties. The risk-neutral approach could be generalized to other toxicological endpoints.