st: Using -stpm2- for Mean, Median Survival Times with CIs for subgroups

[graupel75@gmail.com]

Dear statalisters,

Building on earlier threads, I would like to get mean and median
survival times (with CIs) for three exposure groups from a flex
parametric survival model (-stdp2-, aka "Royston-Paramar" models),
setting some co-variate categories to baseline levels. I've consulted
Royston and Lambert's 2011 Stata book, the SJ articles for -stpm- and
-stpm2- along with a range of pdfs from Royston and others.

The main issue is: -predict rmst- and -predict median- give a value
for each individual, and not the (sub)-population values. Should these
be evaluated with -at- options for baselines of the covariates, and
then summarized for each of the exposure categories? I am not
convinced this a valid approach.

Any clarification or suggestions would be greatly appreciated.

cheers-
Andrew
______________________________________________
Epidemiologist
Centre for Infectious Disease Epidemiology Research (CIDER)
Saw Swee Hock School of Public Health
National University of Singapore
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/