/* Specifying a Cox Proportional Hazards Model */

use http://www.stata-press.com/data/r9/drugtr, clear

/* st allows us to see how the data is set for survival analysis */

stset studytime, failure(died)

/* Let's have a look at our variables before we get started */

summarize

/* Let's specify a model to measure the efficacy of our drug on the cancer patients controlling for age*/

stcox drug age

/* We see that the drug results in a lower hazard-therefore a longer survival time-controlling for age */

/* Now let's report relative risk (the hazard ratio) for a more standard five-year change */

replace age = age/5
stcox drug age, bases(s) nolog

line s _t, sort c(j)