Statistical criteria for evaluating chemicals as positive or negative in the hepatocyte/DNA repair assay.

The usefulness of the hepatocyte/DNA repair assay was enhanced when statistical techniques were applied to the evaluation of a chemical as positive or negative. Using our test procedure, we are 95% confident that the false positive rate is less than 1% if the net nuclear grain count for the test chemical is 3 counts higher than the solvent control for the same animal. Additionally, quantitation of the proportion of cells in repair should be used to corroborate the results of the net nuclear grain count and is recommended for inclusion in the criteria for evaluating a chemical.