Chapter 22. The Cox Proportional Hazards Model
22.1 Introduction 413
22.2 Synopsis 414
22.3 Examples 415
22.4 Details & Notes 424
Like the log-rank test discussed in the previous chapter, the Cox proportional hazards model (sometimes referred to as 'Cox regression') is used to analyze event or 'survival' times. This is a modeling procedure that adjusts for censoring, may include numeric covariates, one or more of which may be time-dependent, and requires no assumption about the distribution of event times. Categorical factors may also be included as covariates in the model.
Examples of where the Cox proportional hazards model might be useful in clinical trials is a comparison of time to death among treatment groups in a cancer trial ...