Kaplan Meier - confidence interval for the median

["Barker, Chris" <Chris.Barker@cvt.com>]

 I'd like to find out whether or not the method, used in SPLUS,  for computing the confidence interval for the median of a Kaplan Meier estimate is the method of

Brookmeyer, Crowley, “A confidence Interval for the Median Survival Time”, Biometrics, 38, 29-41, 1982.

 Interestingly, for the simple dataset of , (10, 15, 23, 30, 35, 52, 100+) (and + indicates censoring)
 SPLUS gives, the median and CI as 30 (15, NA), (R reports the upper limit as inf)

 while SAS gives 30 (15,52)

  Chris Barker, Ph.D.
Consulting Statistician
   CV Therapeutics