Decision-Dependent Choice of Models. J. D. Wilson, Resources for the Future, DC; and R. M. Putzrath, Georgetown Risk Group, DC
When data are abundant, complex models are usually necessary to incorporate all of the appropriate information into an accurate risk assessment. The resulting risk estimate, with its attendant uncertainty, can be used in multifactorial decision analyses. In contrast with risk assessment, safety assessment informs a dichotomous decision process. Thus, the model(s) used to analyze the data can often be less complex. RfDs and ADIs are safety assessments, i.e., determinations of a level below which no action will be taken. As such, they do not provide information about exposures above this level. Determining generically whether higher exposures are likely to pose a significant risk can become a complex enterprise. For a specific chemical and exposure, however, a series of simpler analysis may provide sufficient information for an informed decision. One critical factor is how much larger the exposure is than the RfD or ADI. The NOAEL and the response at the next higher dose can then provide a crude estimate of the dose-response curve at low doses. Similarly, a comparison of NOAELs of all known toxic endpoints can provide information on whether they need to be considered by the decision-maker. A decision matrix will be presented that screens those situations where an exposure above the RfD or ADI is likely to be acceptable or unacceptable. This screen would allow resources for a full risk assessment to be reserved for those cases where the benefits of an exposure initially screened as "unacceptable" are sufficiently large to require a more accurate analysis.
Go to . . .
2001 SRA Annual Meeting Table of Contents
2001 SRA Annual Meeting Author Index
Main Abstracts Menu Page
RiskWorld Home Page