[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Marc Busson <firstname.lastname@example.org>
Wed, 19 Mar 2008 16:17:29 +0100
I work on a series of grafts to evaluate the role of a genotype (variable: v1) on outcome of the graft: alive or dead
My problem is that all recipients who are geno A are alive, so number of recipients geno A who are dead = 0.
In kaplan Meier, I have 100% of survival for geno A and 70% for geno B at 2 years and logrank test give a p value = 0.01
I want to do a COX analysis with 2 other covariates v1(2 categories: geno A and geno B) ,v2 (Chronic GVHD) ,v3 (diagnostic)
The Cox procedure give no coefficient for v1 or a mistake due to this particular categorie (genoB, dead = 0)
Is it possible to do a COX which take into account this particular sheme of data.
Thank you for your Help
BUSSON Marc, PhD
H˘pital Saint Louis
* For searches and help try: