Estimation of HC5 Taking Into Account Uncertainties of Individual Dose Response Curves and Species Sensitivity Distribution. J. S. Jaworska, Procter & Gamble; and T. Aldenberg, RIVM
Species Sensitivity Distributions (SSD) are traditionally based on NOECs from chronic toxicity tests. Currently, the use of NOEC values has been critisized because of its poor statistical properties, including lack of confidence intervals (CI). For this reason the uncertainty of the SSD itself (primary uncertainty) has only been evaluated while the points used to build it (the NOECs) were treated as without error (for a Bayesian approach see Aldenberg & Jaworska 2000, Ecotox. Env. Saf. 46, 1-18). A current tendency expressed by academics and regulatory agencies is to shift from NOEC to EC10-25 (Effective Concentration) in evaluation of chemical effects on individual species. This development should be followed by similar modification of methods to construct SSDs and eventually to calculate Hazardous Concentrations that affect 5% of species (HC5) and their CI. This paper discusses a novel method that takes into account uncertainties of individual ECx values to construct the SSD. The uncertainty of the dose response curve and therefore uncertainty of ECx is calculated with Bayesian regression. Next, the ECx uncertainty distributions are simulated (e.g. by Monte Carlo techniques) to build a family of SSDs (form of secondary uncertainty), each of them being uncertain due to the limited number of species used to construct it (primary uncertainty). Primary and secondary uncertainty will be expressed as CIs. In combination, both types of uncertainties will greatly affect the uncertainty of the HC5. Consequences to probabilistic risk assessment will be discussed.
Go to . . .
2000 SRA Annual Meeting Table of Contents
2000 SRA Annual Meeting Author Index
Main Abstracts Menu Page
RiskWorld Home Page