[R] Survival Rate Estimates

[Terry Therneau]

---begin included message --
Is there an automated way to use the survival package to generate
survival
rate estimates and their standard errors?  To be clear, *not *the
survivorship estimates (which are cumulative), but the survival *rate *
estimates...
--- end --

The classic method in epidemiology is simple hazard rates = total
events / total time.  This is done in more general form by the pyears
(person-years) routine.

  pfit <- pyears(Surv(time, status) ~ agegrp + sex, data=mydata)

It's advantage over simple tables is the ability to do time-dependent
categories.  For instance if I want death rates by decade of age:
  acut <- tcut(mydata$age, 0:12 *10)
  pfit2 <- pyears(time, status) ~ acut + sex, data=mydata)
The first table was based on age at the start, the second on current
age.

For non-parametric estimates of the hazard function look at the
"survival" task view on CRAN; there are several choices.

Terry Therneau