Le critère de mortalité et sa quantité d'information dans l'experimentation radiobiologique

Abstract

Mortality is a quantal response with a definite frequency distribution. Probit analysis is the appropriate method providing a maximum amount of information from the lethal quantal response. It should be used whenever a numerical assessment is required. The method using a 100 per cent lethal dose for controls is of limited value. However it is simple and economical. A frequency ratio method may be used instead of probit analysis. Mortality (or survival) proportions provide a maximum amount of information, within one third—and two third—limits, that is, in the vicinity of the median effective dose. In screening experiments, the sequential test of mortality yields early data with the fewest possible animals.