Statistical analysis of time to event data from preclinical safety pharmacology studies

We describe and discuss parametric models for statistical analysis of safet y pharmacology studies in which the response is time to some event. The eve nt is a qualitative change situated in time. The models include possible ce nsoring which can be caused by end of study before the event has occurred o r death of the animal before end of study, for example. The models proposed in, this article include covariates that do not depend on time. The presen tation is accompanied by case studies analyzed by the SAS procedure LifeReg .