Abstract

The establishment of toxicity thresholds is one of the major hurdles in industrial pathology. This is hardly surprising, as attempts to establish thresholds must take into account variations in individual susceptibility, exposure to multiple toxins which may exhibit interaction effects, and the effect of physical, chemical, biological, and social factors. Traditional methods such as animal experimentation and epidemiological research have inherent limits.

In an attempt to identify statistical methods which will improve the accuracy with which threshold doses of toxic substances can be determined in case-control studies, researchers from the GRABIT (Groupe de recherche-action en biologie du travail: Research-action group in occupational biology) will evaluate the sensitivity of different mathematical models and statistical methods, using data from a study on occupational exposure to manganese.