ABSTRACT
The methods described in the previous chapter are useful in the analysis of a single sample of survival times or when there are survival times for the categories of a categorical variable that the investigator wishes to compare. But in most survival investigations in medicine, there will be many other explanatory variables (covariates) that have been recorded and whose effects on survival time will be of interest. We need a suitable model that links the survival times to the covariates, but because of the special features of such data, something different from the multiple regression model described in Chapter 8 is necessary. In this chapter, we describe the most widely used approach to modelling survival data: Cox’s proportional hazards models.
