Use of tnm as continuous variable in univariate or multivariate analysis

#1
Hello,

I hope I put this post in the correct group.

May I use TNM stage (1,2,3....), N stage (0,1,2,3.....) or grade (1,2,3,4....) as continuous variable when performing logistic regression or cox regression for survival?

For example can I say that T stage (It was not trasformed it in a categorical variable (HR 3.2 CI 95% 2.3-10.1, p=0.001) is a predictor of poor survival?

If it is allowed do I have any advantages/disadvantages in doing that?

Thank you very much for your help in advance.
 
#2
Yes, you can. But you can be better off by using dummy variables for all categories of TNM except for the reference category. Ultimately, a model selection criterion (like cross-validation or AIC) will tell you which model is better.
 
#3
Thank you very much.

I do not understand why most people in their paper convert T or N to categorical variables.....

I will study those modell then.