I have the following model fit:
Call: coxph(formula = surv ~ data$Arm * x) coef exp(coef) se(coef) z p data$ArmLevel1 0.062322 1.06 0.70338 0.0886 0.93 covariate1 -0.000495 1.00 0.00710 -0.0697 0.94 data$ArmLevel1:covariate1 0.003183 1.00 0.00857 0.3714 0.71
What's the easiest way to test that covariate1 is significant controlling for Arm? Thus I want to use this model object to test that covariate1 is significant given that ArmLevel=0 and that ArmLevel =1?
If there's a way to look at this both sliced (thus both p-values) and pooled, that would be great!