Help with RPSFT model implementation in SAS

#1
Hi! I am doing an analysis on overall survival endpoint. The patients in the placebo group will be switched to treatment group in the open-label phase on voluntary base. I plan to use Rank-preserving structural failure time method (RPSFT) model. Is anyone familiar with this method and know how to implement it in SAS?
 

hlsmith

Less is more. Stay pure. Stay poor.
#2
I have not used that procedure before. I went to do a quick Google search and nothing apparent popped up within the first couple of pages. It seems to be related/relevant to Jamie Robins Time-Varying Exposures work and perhaps Marginal Structural Models.


I will note that I saw an R program article that is only about a month old on it. You could try that program or use its source documentation to try and reproduce the approach in SAS. I bet you there is a SAS user group article on it, or perhaps one of Miguel Hernan (Harvard) programs on his University Causality page may cover the topic. Though I believe many of those approaches are related to observational data and use of propensity scoring.


Is your data observational or experimental? Also, please update your post, in that I would be interested to see where you landed in your pursuits in regards to this question.


P.S., the Analysis of Observational Health Care Data Using SAS book addresses time-vary treatments as well I believe.