On the Usefulness and Reliability of Existing QSBRs for Risk Assessment and Priority Setting

Abstract

Abstract The results of a survey of quantitative structure-biodegradability relationships (QSBRs) are presented. The survey was made to evaluate the use of QSBRs for risk assessment and priority setting. It appears that few models can actually be used for legislative purposes. The major reason for this is that many models are either based on biodegradation end-points that are inconvenient for legislative purposes, or are not reproducible, or are based on too few data. It is concluded that single descriptor models cannot accurately predict the biodegradability of a broad range of chemicals. Most of these models therefore apply to narrow subclasses of chemicals only. The development of multiple descriptor models is still rather modest. More generally applicable models are based on molecular fragment contributions to biodegradability. However, for most of these models the chemical domain is uncertain. Despite the fact that none of the models in the survey meet the criteria set for judging a model to be useful for legislative purposes, three models are given the benefit of the doubt and are qualified as “best available”. Finally, some new validation results on the models selected are presented and recommendations for model development are given.