Dear all:
PROC LIFETEST in SAS gives estimated quartiles and the median in it's output together with associated CIs (but unfortunately it can't be written in a data set). These seem to be obtained via "inverse regression" from the CIs of the Kaplan-Meier estimate. It may happen, however, that still more than 50% of the subjects are at risk at the end of the time course and the median can't be estimated, yet, in some cases, the lower limit of the 95% CI can. The question is: Does anybody know how exactly(!!) this is done in SAS in this case? Although several things appear plausible, I did not quite succeed to exactly get the same value in all example I looked at. I'd be very grateful for some ideas on how exactly it is done.
Best regards
Markus
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|