Order of Death as a Test of Therapeutic Effect

Abstract

IN assessing the effect of a treatment on the action of a poison it is usual to measure either the effect on the LD₅₀ or on the time to death. In a recent investigation by one of us (D. F. H.) on the effect of treatment with dimethylformamide on the poisoning of male and female rats with dimethylnitrosamine it was found that changes in LD₅₀ provided a rather insensitive test of effect, which required large numbers of rats to establish, and that mean times to death in treated and untreated groups were not greatly different after doses which killed all the rats. Dimethylformamide had, however, a very marked effect on the order in which rats died at doses such that a few survived. Times to death cannot be calculated in these circumstances without making arbitrary decisions about “times to death” of the survivors. Orders of death, however, proved susceptible to a more rigid statistical treatment. As this may be of general use it is reported here.