Environmental Teratogenesis, Statistics for

AbstractTeratology is the study of adverse effects of environmental agents on the developing system. The major source of information for studying potential teratogenic (developmental) effects of chemicals on humans is generally the animal dose–response experiments. Quantitative risk assessment involves an assessment of the relationship between exposure and adverse health effects to derive an allowable exposure level. Acceptable levels of human exposure are derived by dividing the lower confidence limit benchmark dose by the product of several uncertainty factors. This article describes the statistical models and benchmark approach for quantitative risk assessment of developmental effects. The benchmark dose method involves selection of a dose–response function to fit animal experimental data. The benchmark dose and its lower confidence limit are then estimated from the fitted dose–response model. The dose–response function for continuous response data is modeled using the litter‐based analysis. The dose–response function for dichotomous response data is modeled using beta binomial.