Toxicological animal studies are essential to risk assessment. The reliability of the studies depends on the extent to which study design, conduct, and analysis have a high internal validity. Several tools exist for assessing the internal validity of human and animal studies, however, a dedicated tool for animal toxicokinetic studies is lacking. We developed a questionnaire to systematically assess toxicokinetic animal studies. The questionnaire consists of seven main domains with questions for which points are given by the assessor multiplied by fixed weighting factors (WF). The scores are then aggregated at both the domain and overall levels and expressed as a percentage of the maximum score. We established four confidence tiers based on the percentage of the achieved score. Experts were able to discriminate studies based on internal validity (overall assessment rating between 37.6% and 86.9%, n = 7 studies). The questionnaire was also a helpful tool for less experienced scientists (master students, postgraduates) helping them to assess the internal validity in a systematic way. By implementing a WF into the calculation and performing the assessment at the domain level, both innovative steps in the assessment process, the transparency of the assessment is increased. In a limited exploratory study, we found that Large Language Models (AI) were able to complete the questionnaire, however, failing to discriminate studies' quality. This newly developed questionnaire provides a transparent, granular, and effective framework for assessing the internal validity of animal toxicokinetic studies, offering reliable discrimination across quality tiers for both experienced and toxicologists in training.
Batke et al. (Wed,) studied this question.