1
$\begingroup$

When I try to find 95% CI of quartile estimates of survival (including median), I get NAs for upper limits of CI. I use survfit(Surv(time = SURV_DURATION, event = CENSOR) ~ RESPONSE, data = data, conf.type = "log-log") to fit my model and as I understood it uses uses a generalization of the Brookmeyer and Crowley (1982) sign test to find the CI. But why I get NAs for upper limits ?

$\endgroup$

1 Answer 1

6
$\begingroup$

You get NA because not enough people have died yet to estimate the upper confidence.

Leaving out the precise details of how it's done in R, conceptually the way to get a confidence interval for median survival is to get a confidence interval for the survival curve and cut it at 50%. If the upper limit of the survival curve hasn't reached 50%, there is no estimate for the upper confidence limit on the median.

Using other techniques to get an upper limit (such as inverting a score test) doesn't change the basic idea: you can't get an upper confidence limit on the median until you can be confident the true survival curve is down below 50%

$\endgroup$
1
  • 1
    $\begingroup$ Thanks for explanation $\endgroup$ Jul 15, 2020 at 14:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.