Correcting for extraneous death in the evaluation of morbidity or mortality from tumor.

In an earlier paper (2) a method of evaluating tumor morbidity was presented and its usefulness discussed. Although the usefulness of the method is considered to be as discussed in that paper, and the interpretations applied to the data are still valid, the method used to correct for extraneous deaths was in error. The purpose of this note is to correct that error by presenting a simple and valid method for correcting for extraneous death. The author would also like to call attention to an excel lent discussion by Spicer (4), of methods of evalu ating tumor mortality, which discusses other closely related methods. The percentage of tumors in a population of ani mals is subject to the vicissitudes of fate. The acci dental death of some animals in one experimental group can lower the tumor percentage and will often make comparisons with other groups dif ficult. Correction for accidental death is laborious and frequently of questionable validity. The same information as is gleaned from comparisons of tu mor incidence can be obtained from a considera tion of the probability of a single animal or small group of animals developing tumors within a given time period. This probability is a permanency within the conditions of the experiment and does not change with the accidental demise of some of the animals, provided the accidental deaths are random. The proposed method is concerned with