How many censored data is too many for Kaplan-Meire estimate?

#1
Hello there. I am new to this forum and want to ask your advice on survival-time calculation.
I have a data set which contains 60 patients (dogs).
During 36-month follow-up period, 6 patients died, and we have lost 7 patients for follow-up.
Other patients (n=47) were still alive. This means we have 54 censored samples (lost or alive).

In this case, is Kaplan-Meire analysis still accurate to conduct?
I understand that it doesn't provide median survival time because the number of patients who died is so small to calculate it.

Any comments and input would be greatly appreciated!
Thank you in advance for your time.
 

hlsmith

Not a robit
#4
So you are not looking to control for or test any hypotheses, just report survi val, correct. If so, I don't see the harm, make sure you slap on some confidence intervals on what ever metric you report. Ideally graphically. The confidence intervals will help the review take in censoring, events, and small sample/rate outcome.
 
#5
@hlsmith Thank you for your response.
Yes, that is correct. I just want to report a median follow-up time, the survival rate at 36 months (with CI), and survival curve with censored events.
One more question. As I said at my initial post, is it correct not to have a median survival time bc there were only 6 dogs had died during the study?
 
Last edited: