Abstract
In many countries, animal experiments can only be performed when their necessity has been demonstrated in a legal document. As the usefulness of animals in research is also a significant societal and political issue, criteria to structure debates and evaluations are needed. Here, background information is given on laboratory animal studies. Moreover, parameters that may be considered in judging their usefulness are suggested. The discussion is strictly focused on animals used as tools/test systems/models to provide information on humans. In this context, general features and performance characteristics of models are discussed. Examples are given for well-recognized criteria (e.g., robustness, relevance, predictivity) to judge the usefulness of predictive models. The main hypothesis put forward here is that a benefits evaluation (usefulness metrics) is only possible within sharply circumscribed “use domains”. Examples are given for the research fields of drug and vaccine research, toxicology, disease pathogenesis, and basic biological research. Efficacy, safety, and quality studies are highlighted as “use domains” within the field of drug discovery and production. A further separation into individual diseases, drug targets or symptoms is suggested for, e.g., efficacy studies or pathophysiology. Finally, an outlook is given on the evaluation of model advantages and disadvantages to arrive at their “net benefit”. Moreover, the need to compare the net benefits of animal models versus that of their alternatives is highlighted.