| he objective of this study was to use animal based oral toxicity data to predict the occurrence of harmful effects of accidental oral exposures of children to pesticides. The source of childhood poisoning data was 35,792 ingestions of pesticides in children age five and under in 1989 provided by the American Association of Poison Control Centers (AAPCC). Acute oral toxicity values were obtained or calculated for 477 pesticide products identified in the AAPCC database based on studies reviewed by toxicologists at the U.S. Environmental Protection Agency. Two other factors, age of child and state (solid or liquid), were considered in the analysis.;A total of 11,375 cases of oral child exposures in the AAPCC data were identified where both the medical outcome and toxicity of the pesticide could be determined. Even with three predictors, product oral toxicity, state (solid or liquid), and age of child, the probit model was not favored based on the likelihood ratio chi-square test. Analysis of chemical types of pesticides did reveal one group of pesticides, cholinesterase-inhibiting insecticides, which fit the probit model well. Any of the three models tested including the probit, logistic, and gompertz, provided a good explanation of the data for cholinesterase inhibitors. Adding age to the model did not improve the fit significantly. The best overall model was:;Based on a cost estimation model provided by the U.S. Consumer Product Safety Commission and data from AAPCC, the total estimated medical and social costs of childhood pesticide poisoning are... |
