A REGRESSION SURVIVAL MODEL FOR TESTING THE PROPORTIONAL HAZARDS HYPOTHESIS

A semi-parametric generalization of the proportional hazards regressio n model is defined, whereby the hazard functions can cross for differe nt values of the covariates. In the two-sample comparison, it includes in particular the case of two Weibull distributions differing in scal e and shape parameters. A global test of the proportional hazards assu mption is proposed against such defined alternatives. Its power in the two-sample case is compared to that of previously described tests by using simulation experiments. Survival data of patients with breast ca rcinoma, including several prognostic factors, are presented as an ill ustration.