A Model for Selecting Short-Term Tests of Carcinogenicity

A theoretical model is presented that selects among available short term tests of carcinogenicity and then classifies chemicals on the basis of the results. The model is simplified in ignoring the potency of various carcinogens but does take account of the social cost of falsely classifying a chemical as a carcinogen as well as the social cost of falsely classifying a chemical as a noncarcinogen. If some of the currently available short-term tests, separately or in sets, are capable of classifying chemicals with even moderate accuracy, this framework would serve to improve the regulation of suspect carcinogens.