R help - Jan 2006 - Variance of expected survival based on a Cox model with frailty

Dear R users,

I am interested in the expected survival probabilities of an "example
person",
calculated on the basis of a Cox model with shared frailty. The example person
has covariate values equal to zero and a frailty term equal to its expected
value. I calculate the expected survival probabilities in the following way:

fit1=coxph(...)
base1=basehaz(fit1,centered=F)
surv1=exp(-base1$hazard).

Then I would like to calculate the variance of the expected survival
probability but
it seems to be a bit tricky.
Has anybody calculated the variance of expected survival probability and
willing to share the code?

Thank you in advance for any help.

Marjo Pyy-Martikainen