Bad reporting or bad science? Systematic data evaluation as a means to improve the use of peer-reviewed studies in risk assessments of chemicals
In this study we assess the applicability of a set of reliability criteria proposed by Ågerstrand et al. This was done by evaluating the reliability of 12 non-standard peer-reviewed ecotoxicity and toxicity studies for Bisphenol A. There was an overall agreement between the evaluator and the authors of the papers regarding the result of the evaluations. This suggests that the criteria offer enough guidance to be a useful and consistent evaluation tool. It provides a transparent and structured approach, and ensures that a minimum and similar set of criteria is used. The evaluation of the peer-reviewed ecotoxicity and toxicity studies concludes that important information is sometimes missing, and therefore the studies do not always meet common regulatory requirements regarding reporting. Whether this is due to insufficient reporting or due to poorly performed studies is not known. To improve the reporting, and thereby promote reliability and reproducibility, researchers, reviewers, and editors are recommended to use the suggested criteria as a guideline. In conclusion, in order to improve the reliability of peer-reviewed studies, and to increase their use in regulatory risk assessments of chemicals, the dialog between regulators, researchers, and editors regarding how to evaluate and report studies needs to be strengthened.