Correlation of Animal Eye Test Data with Human Experience for Household Products: An Update

Abstract Accidental consumer eye exposures to household products provide useful data for determining whether animal tests adequately predict human response. Reports of accidental consumer exposures from mid-1983 through 1984 were compared to two animal methods and to the results of a similar analysis for the period 1979-1980. The new comparison confirmed the earlier analysis in showing that the low-volume rabbit eye test is more predictive of human eye response than the Federal Health and Safety Administration (modified Draize) test. Furthermore, the data show that rigorous follow-up after consumer accidents can provide useful information for assessing the validity of animal studies used to estimate human safety. Finally, because the low-volume method correlates better with human eye effects than the Draize method does for household products, it is the method that should be used to validate any new in vitro eye testing procedures. In fact, any laboratory test developed to assess a health or safety effect must be validated for its ability to predict this effect in humans.