I am looking at a survival analysis using Cox PH models. I am interested in the effect of a continuous variable V1 on survival.

I want to look at interaction terms with my model. I believe variable V2 is also partly predictive of survival but I am not sure to what extent and if it is directly influencing V1. I was told by a supervisor to look at interactions between the two variables by creating a product variable (V1*V2) and including that in my model. I was also told that P>0.1 is considered significant due to a loss of power when creating the product term. As such I have the following output:

V1 p<0.001 expB 1.25, V1*V2 p=0.075 ExpB 0.85, V2 p<0.001 expB 1.744

My interpretation is therefore. V1 is predictive of death, but V2 more so, and that V1 and V2 interact with each other.

What do I do next? V1 is the variable of interest and V2 is a biological confounder.

  • $\begingroup$ This is not really an interpretation, you are just reading the results. Therefore your question is really unspecific, could you elaborate ? What is exactly that bother you ? $\endgroup$ – Dan Chaltiel Feb 26 at 14:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.