Analysis of genotoxicity data in a regulatory context.

Analytical methods for regulatory tests usually must be defined before testing. To take this into account and to minimise equivocal interpretations, a sequential strategy is recommended. Assay validity must be verified and results then classed as clearly negative, clearly positive, or uncertain based on historical data. Where there is uncertainty, standard parametric or non-parametric statistical methods should be used with appropriate corrections to assess the significance. The biological importance of statistically significant data should then be evaluated using historical data.