Shortcomings of LD50-values and acute toxicity testing in animals.

Abstract The author argues that the standard practice of determining an LD₅₀ with 95% confidence limits is a waste of animals since the extra statistical precision is undermined by numerous other factors. For example, the LD₅₀ of a given chemical often varies by at least 10-fold between different animal species and strains. Also, environmental factors can result in substantial differences in the LD₅₀ of a given chemical in a given animal strain. Therefore, the use of this measure as a general index of toxicity, as a guide to further pharmacological or toxicological studies, or as a guide to human toxicity involves unnecessary and needless taking of animal life. These points will be amplified and substantiated by actual examples. Finally, the author will make some suggestions as to what could constitute a satisfactory acute animal test given current scientific and regulatory needs.