Several regression methods are also available, including nonlinear regression and maximum likelihood estimation. These have the advantage of producing not only the parameters but also a confidence interval for the dose-response curve predicted. The disadvantage of all the foregoing methods is that they may produce poor results if the data are not normally distributed or if the data do not cover the range of toxic effects well. That is, the data must include at least one dosage with essentially zero response and one with about 100% response. For example, if a bioassay on a wastewater produces 30% mortality at 100% concentration, these methods will not work.
TABLE 19.2 Qualitative Classification of Toxic Compounds
LD50 Range (mg/kg)
LD5o of Example
Supertoxic Extremely toxic Highly toxic Moderately toxic Slightly toxic Practically nontoxic
5 or less 5-50
Strychnine in rat Ethanol in mouse
Parathion in goats DDT in rat
Dioxin in guinea pig
Was this article helpful?