Home > OS >  How can I evaluate the Kaplan Meier estimator at specific time points?
How can I evaluate the Kaplan Meier estimator at specific time points?

Time:07-13

I have a Kaplan Meier object from the R package survival. I can plot it and etcetera, but I would like to evaluate it at specific time points.

How can I create a function that evaluates the KM estimator at specific time points using the KM object?

library(survival)


KM <- survfit(Surv(time, status) ~ 1,data = lung)

> KM(1)
Error in KM(1) : could not find function "KM"

CodePudding user response:

Use summary() with the times argument:

library(survival)

KM <- survfit(Surv(time, status) ~ 1,data = lung)
summary(KM, times = seq(0, 5) * 90)
#> Call: survfit(formula = Surv(time, status) ~ 1, data = lung)
#> 
#>  time n.risk n.event survival std.err lower 95% CI upper 95% CI
#>     0    228       0    1.000  0.0000        1.000        1.000
#>    90    201      27    0.882  0.0214        0.841        0.925
#>   180    160      36    0.722  0.0298        0.666        0.783
#>   270    108      30    0.575  0.0338        0.513        0.645
#>   360     70      24    0.434  0.0358        0.369        0.510
#>   450     48      16    0.329  0.0356        0.266        0.406
summary(KM, times = seq(0, 5) * 90)$surv
#> [1] 1.0000000 0.8815789 0.7216707 0.5753102 0.4340441 0.3287158
  • Related