CHAPTER 23 Survival Regression 327

Chapter 23

Survival Regression

S

urvival regression is one of the most commonly used techniques in

biostatistics. It overcomes the limitations of the log-rank test (see

Chapter 22) and allows you to analyze how survival time is influenced by one

or more predictors (the X variables), which can be categorical or numerical. In this

chapter, we introduce survival regression. We specify when to use it, describe its

basic concepts, and show you how to run survival regressions in statistical

software and interpret the output. We also explain how to build prognosis curves

and estimate the sample size you need to support a survival regression.

Note: Because time-to-event data so often describe actual survival, when the

event we are talking about is death, we use the terms death and survival time. But

everything we say about death applies to the first occurrence of any event, like

pre-diabetes patients restoring their blood sugar to normal levels, or cancer sur-

vivors suffering a recurrence of cancer.

IN THIS CHAPTER»

» ^{Knowing when to use survival}

regression»

» ^{Grasping the concepts behind}

survival regression»

» ^{Running and interpreting the}

outcome of survival regression»

» ^{Peeking at prognosis curves»}

» ^{Estimating sample size for survival}

regression