R help - May 2007 - Survival problem with two populations?

Hi all,
   i'm modeling a survival problem about the duration of a treatment.
Patient leave the treatment following a mixed weibull distribution (2
weibull). The first are patient that returns to take drugs and the
latter are patient  that ends the treatment because they don't want to
take drugs anymore. The two curves are obviously mixed as in this
figure


*
*
**
***                ****
****          **********
**********************

What survival model can i apply? I think kaplan meyer and cox are
tight to models described by only one distribution and i don't know
other R modules that let me do other type of analysis...

any suggestion?

thanks and sorry for my english,
  nelson



-- 
--
Consulenze Linux e Windows

http://nelsonenterprise.net