R help - Sep 2009 - R-squared in proportional hazards regression

Hi

I'm trying to interpret the output from the coxph command. One of the
outputs is:
Rsquare= 0.01   (max possible= 0.069 )

Given that the methods for calculating R squared and its validity with
proportional hazards regression are debated in the literature - can anyone tell
me exactly how this part of the coxph output is calculated and how I derive my
R-squared from it?

Thanks
Helen