Validation of the EU environmental risk assessment for veterinary medicines

Abstract

An alarming decline of vulture populations (up to 95%) in Pakistan in the late 1990__s has recently been attributed to the use of the anti-inflammatory drug diclofenac in cattle. Several species are now threatened with extinction, a tragedy that demonstrates that an environmental risk assessment of veterinary medicines is legitimate. The fate and behaviour of pharmaceuticals in the environment has been studied incidentally since the second half of the 20th century. The possible effect of residues on the quality of the environment was a concern that had been recognised since shortly after medicines were actually found in the environment towards the end of the 1960s. About a decade or two later, these concerns about pollution and effects had reached the regulatory agenda and were addressed in the European legislation on veterinary medicines. Notably from 1994 onwards, when the EU Committee for Veterinary Medicinal Products (CVMP) initiated the drafting of a guidance document for the risk assessment of veterinary medicines, an increasing number of publications on use, emission, fate, occurrence, effect, and risk of pharmaceuticals has been published. In this thesis, the representation of the environment in the risk model and the consequences of the choices regarding variables and relationships in the natural system were addressed in light of existing legislation and policy on environmental quality. Where possible, empirical validation of individual exposure models was performed, demonstrating the accuracy of the model for a specified use. I have outlined and discussed several technical and conceptual flaws in the use of risk assessment in the registration process. The formalisation of the protection goals by the Note for Guidance has no legal status, the guidance does not contain all European environmental quality criteria and includes a soil exposure trigger for the full risk assessment that was not well founded. Particularly the variability in possible outcomes for the soil exposure trigger is troublesome, since it gives the impression that science may produce information that will not lead to better judgement. Screening level exposure models are not always protective and lack standardised scenarios for uniform assessment. Risk mitigation measures are not validated, and give a false impression that the risk to the environment is reduced to acceptable levels. Products that pose a threat to environmental quality at or after use may now be registered and be used. Authorities responsible for water and land quality cannot infer from the registration of a veterinary medicine that the use complies with the quality standards set by environmental legislation. This legislation now obliges these authorities to regulate and enforce product use and slurry use on a case by case basis through permits. All of the shortcomings identified undermine, to a greater or to a lesser extent, the legitimacy of the registration process and the use of scientific opinions. Policy makers, regulators, and scientists should engage in a reconnaissance of regulatory goals, assessment scales, model approaches, and uncertainty and variability of data associated with the assessment. Only then, a risk model (with or without triggers) can be applicable to the situation of interest and provide a solid framework for making risk-benefit decisions to the satisfaction of producers, decision-makers and consumers.