No Prediction Model, No Validation Study

There has recently been increasing interest in the development and validation of alternative methods that could. be used in the place of in vivo toxicity tests. The goal is that a toxicologist will be able to test a substance by an alternative method, convert the results obtained into correct predictions of toxic hazard and, ultimately, use the predictions for making decisions about the safety of a test substance. If a toxicologist can be assured that the predictions obtained from an alternative method will lead to correct risk assessment decisions, the method may replace tlw in vivo test. Theoretical discussions on the assessment of alternative method validity have emphasised the importance of confirming the technical performance of an alternative method, but have been less clear about how to confirm that a method correctly predicts a toxic endpoint. 13 The discussions commonly recommend that the data obtained from a validation study be thoroughly searched after the study has been completed, in order to determine whether the alternative method might be useful for making predictions. In this situation, the data from the validation study are used to construct the models to be used for making toxicity