Is a Cox regression the right test for my question?

#1
I would like to test if the value of a blood parameter is linked to death (age as a covariate)....example: higher serum creatinine = higher risk to die?
I know how to make a cox regression with two groups (e.g.two different medications), but can I also work with a question like described above by maybe forming different categories, like

0,5-0,7mg/dl
0,71-0,9mg/dl
0,91-1,1mg/dl
1,11-1.30 mg/dl.....and so on....

so I would expect that in the fist groups I will not have any deaths and the higher the crea value will get I will have a higher death rate? What values will I get?

Is there a better test to perform?

Thank you for your suggestions :)