Reproducibility of toxicity test data as a function of mouse strain, animal lot, and operator

The toxicity screening test method developed at the University of San Francisco was evaluated for reproducibility. The variables addressed were strain of mouse, lot of animals, and operator. There was a significant difference in response between Swiss Webster mice and ICR mice, with the latter exhibiting greater resistance. These two strains of mice are not interchangeable in this procedure. Variation between individual animals was significant and unavoidable. In view of this variation, between-lot and between-operator variations appear to have no practical significance. The significant variation between individual animals stresses the need for average values based on at least four animals, and preferably values based on at least two experiments and eight animals. Efforts to compare materials should be based on the evaluation of relatively simple responses using substantial numbers of animals, rather than on elaborate evaluation of single animals