An analysis of the use of historical control data in the assessment of regulatory pesticide toxicity studies.

The concurrent control group is the most important reference for the interpretation of toxicity studies. However, pooled information on control animals from independent studies, i.e., historical control data (HCD), is also used for the interpretation of results. Currently, an overview on actual HCD use in regulatory toxicology is lacking. Therefore, we evaluated the HCD use of the Joint FAO/WHO Meeting on Pesticide Residues from 2004 to 2021 and compared it with recommendations in regulatory guidelines and in the literature. We found that HCD was used routinely and exclusively to avoid potential false positive decisions regarding the treatment-relatedness of effects, mostly using the HCD range, i.e., the most extreme values, as a benchmark. HCD were not used to avoid potential false negative decisions or for quality control of the index study. The central assumption of the HCD use, namely that the HCD and control group of the index study follow the same underlying distribution because they are samples of the same data generation process, was not investigated, although numerous factors potentially contribute to effect variation between the different control groups pooled in the HCD. We recommend that the existing guidelines be revised to improve the robustness and transparency of toxicological assessments.