[R-sig-Epi] Log-Rank Test on Relative Survival

[Josh Rosen]

Hello All,

I have another question regarding using the relsurv package. I am trying to
run a log-rank test to compare across the strata in my relative survival
plots. So far I am able to compute relative and non-relative survival:

relative.surv <-
rs.surv(Surv(time,tens)~treatment+ratetable(age=age*365.24, sex=sex,
year=year), data=patient_data_all_frame, ratetable=life_table_03_15)

nonrel.surv <- survfit(Surv(time,cens) ~ treatment, data =
patient_data_frame)

For a log-rank on the non-relative survival, I simply use the survdiff
function in the survival package:
survdiff(Surv(time,cens) ~ treatment, data=patient_data_all_frame)

I am trying to figure out how to run a log-rank on the relative survival
data as well, but can't seem to find a way to pass the survival object I
created (relative.surv) into the survdiff function.

Any help that you can provide would be greatly appreciated.

Thank you
-Josh

	[[alternative HTML version deleted]]