|| Checking for direct PDF access through Ovid
Evaluation of data relevance, reliability and contribution to uncertainty is crucial in regulatory health risk assessment if robust conclusions are to be drawn. Whether a specific study is used as key study, as additional information or not accepted depends in part on the criteria according to which its relevance and reliability are judged. In addition to GLP-compliant regulatory studies following OECD Test Guidelines, data from peer-reviewed scientific literature have to be evaluated in regulatory risk assessment of pesticide active substances. Publications should be taken into account if they are of acceptable relevance and reliability. Their contribution to the overall weight of evidence is influenced by factors including test organism, study design and statistical methods, as well as test item identification, documentation and reporting of results. Various reports make recommendations for improving the quality of risk assessments and different criteria catalogues have been published to support evaluation of data relevance and reliability. Their intention was to guide transparent decision making on the integration of the respective information into the regulatory process. This article describes an approach to assess the relevance and reliability of experimental data from guideline-compliant studies as well as from non-guideline studies published in the scientific literature in the specific context of uncertainty and risk assessment of pesticides.Relevant (appropriate) and reliable (trustworthy) data are needed for regulatory decision making.Selection of data used in risk assessment and weight of evidence can directly influence the decision making process.Transparent evaluation criteria are essential to understand why and how certain data are relied on.Specific requirements for experimental data from different sources for health risk assessment of pesticides are described.Adherence to good scientific and reporting practice is key to facilitate regulatory use of published research results.