Derivation of Expedited Health Benchmarks Using Empirical Distributional Estimates of Potential Toxi. R. G. Hetes, US Environmental Protection Agency, RTP, NC 27709; and D. Hattis, Clark University, Worcester, MA. 01610
There are a large number of chemicals which do not have current health benchmarks to support rule-making or risk decision-making (e.g., setting site cleanup standards). Using current EPA methods of RfD development for these large number of chemicals is prohibitive. Therefore, methods must be developed to expedite the derivation such benchmarks. While expedited benchmarks may have greater uncertainty, the social value of using expedited benchmarks has been documented. Even for chemicals for which insufficient data are available to support the development of an RfD (or RfC or CSF) under current EPA procedures, there may still be a wealth of information on the potential toxicity of that chemical. While this information may be inadequate to support an RfD this information may be sufficient to support the prediction of a probabilistic estimation of what an RfD may be should additional data become available. Relevant information on potential toxicity may include chemical properties, chemical structure, acute toxicity test results, predicted or observed target organs, or predicted or observed mechanisms of toxicity. These data may be used to estimate an RfD, which may have greater uncertainty than that found in a typical RfD may still provide a useful estimate for use in regulatory development. Empirical distributional estimates of potential toxicity can be used to derive estimates of protective health benchmarks. This method is based on a hierarchy of methods or surrogates based on available data and generates a probabilistic estimate of an RfD (or similar using expedited benchmarks has been documented. Even for chemicals for which insufficient data are ining the available to support the development of an RfD (or RfC or CSF) under current EPA procedures, there may still be a wealth of information on the potential toxicity of that chemical. While this information may be inadequate to support an RfD this information may be sufficient to support the prediction of a probabilistic estimation of what . . . . . . . [RiskWorld Note: Submitted abstract incomplete]