Uncertainty of testing methods--what do we (want to) know?

It is important to stimulate innovation for regulatory testing methods. Scrutinizing the knowledge of (un)certainty of data from actual standard in vivo methods could foster the interest in new testing approaches. Since standard in vivo data often are used as reference data for model development, improved uncertainty accountability also would support the validation of new in vitro and in silico methods, as well as the definition of acceptance criteria for the new methods. Hazard and risk estimates, transparent for their uncertainty, could further support the 3Rs, since they may help focus additional information requirements on aspects of highest uncertainty. Here we provide an overview on the various types of uncertainties in quantitative and qualitative terms and suggest improving this knowledge base. We also reference principle concepts on how to use uncertainty information for improved hazard characterization and development of new testing methods.

Effects assessment: boron compounds in the aquatic environment.

In previous studies, boron compounds were considered to be of comparatively low toxicity in the aquatic environment, with predicted no effect concentration (PNEC) values ranging around 1 mg B/L (expressed as boron equivalent). In the present study, we describe an evaluation of toxicity data for boron available for the aquatic environment by different methods. For substances with rich datasets, it is often possible to perform a species sensitivity distribution (SSD). The typical outcome of an SSD is the Hazardous Concentration 5% (HC5), the concentration at which 95% of all species are protected with a probability of 95%. The data set currently available on the toxic effects of boron compounds to aquatic organisms is comprehensive, but a careful evaluation of these data revealed that chronic data for aquatic insects and plants are missing. In the present study both the standard assessment factor approach as well as the SSD approach were applied. The standard approach led to a PNEC of 0.18 mg B/L (equivalent to 1.03 mg boric acid/L), while the SSD approach resulted in a PNEC of 0.34 mg B/L (equivalent to 1.94 mg boric acid/L). These evaluations indicate that boron compounds could be hazardous to aquatic organisms at concentrations close to the natural environmental background in some European regions. This suggests a possible high sensitivity of some ecosystems for anthropogenic input of boron compounds. Another concern is that the anthropogenic input of boron could lead to toxic effects in organisms adapted to low boron concentration.