topic Re: Comparing survival curves at multiple timepoints in Statistical Procedures

Comparing survival curves at multiple timepoints

Melk — Tue, 30 May 2017 17:35:36 GMT

I have run a PROC LIFEREG and have obtained a LR p-value comaring 2 strata. I would additionally like to compare my survival curves and get p-values at specific timepoints. How can I go about doing this?

Here is my code for the overall survival analysis:

proc lifetest data=dat maxtime=60 plots=survival(cl atrisk(maxlen=13 outside(0.15)) test);
   time fu*fail(0);
   strata group / test=logrank;
   symbol1 v=none color=blue line=1;
   symbol2 v=none color=red line=2;
   label fu = "Follow-up, mo";
run;

I would like to compare my curves at 24 and 36 months.

Re: Comparing survival curves at multiple timepoints

KevinViel — Wed, 31 May 2017 16:20:37 GMT

Melk,

It may depend what you mean by "compare". Forgive the brevity and potentially tangential citation, but you may find:

Caution: Hazards Crossing! Using the Renyi Test Statistic in Survival Analysis

https://www.pharmasug.org/proceedings/2011/SP/PharmaSUG-2011-SP06.pdf

of interest.

Good luck,

Kevin

Re: Comparing survival curves at multiple timepoints

Melk — Fri, 02 Jun 2017 00:05:19 GMT

Thank you - to answer my own question, I used the proportion surviving and standard error to perform a 2 sample z test.