topic Re: Linear Cox Regression model for repeated measurements with the same ID in Statistical Procedures

Linear Cox Regression model for repeated measurements with the same ID

Strs1309 — Wed, 26 Jun 2024 09:56:50 GMT

Hello! I have got a question about fitting Cox regression model to obs with different measurements with the same ID. I’m new to SAS so I don’t know exactly how to fix this problem.
I have got a dataset of 471 obs. 53 out of them have got new measurements but the id of obs are the same. I tried to use this code
proc phreg data = imottkm;
class x1;
model tgvhd*imt(0)= x115;
random x1/noclprint;
hazardratio 'Relapse' x115/cl=both;
run;
but the results are a bit weird for me. How could I make this thing work? Is there any way to use random statement in Cox model?

Re: Linear Cox Regression model for repeated measurements with the same ID

ballardw — Wed, 26 Jun 2024 13:48:21 GMT

Please discuss things in terms of the variables you show not some underlying use that is not defined. For instance, your are asking about "same ID". Since there is no variable named ID shown in the code we don't know if you even have "ID" in the model anywhere or how it may be used.

Re: Linear Cox Regression model for repeated measurements with the same ID

Strs1309 — Wed, 26 Jun 2024 14:21:20 GMT

X1 is an ID of a patient. X115 is a relapse of disease. Tgvhd-time to event (censoring death, last visit or getting an infection). Event here is imt (the fact of being infected). I have 471 patients in general on the first transplantation (39 of them got infection). After first transplantation only 53 went to further transplantations (and 8 of them got infected there). The problem is that we need to consider all the events of infections but these 53 people before further transplantations had body checks again so they have different measurements (for example, on the transplantation 1 id 222 had donor haplo, but on the further transplantation id 222 had donor mud). Is there a way to do such analysis ?

Re: Linear Cox Regression model for repeated measurements with the same ID

sbxkoenk — Wed, 26 Jun 2024 14:23:30 GMT

Mixed effects cox regression models are used to model survival data when there are repeated measures on an individual, individuals nested within some other hierarchy, or some other reason to have both fixed and random effects.

Think about some good Google search terms, like:

mixed effects cox regression
Cox Regression repeated measures
Cox Regression time-varying covariates

Then enter these terms in a Google search bar with the "site:" parameter

site:documentation.sas.com
site:support.sas.com
site:blogs.sas.com
site:communities.sas.com
site:lexjansen.com

SAS Tip: Learn lexjansen.com
https://communities.sas.com/t5/SAS-Tips-from-the-Community/SAS-Tip-Learn-lexjansen-com/td-p/436336

I am sure you will get some good hits!

BR, Koen