Interpreting extremely high HZ while accounting for interaction with time

EGR

New Member
#1
Hi all,
I'm trying to figure out how to interpret extremely high HZ when the proportional assumption does not hold, and i insert to the model the term + interaction term with time.
for example:
Wald Pv Hazard ratio B Variables​
0.187 1.913 0.649 high LDL at baseline​
>0.001 18690020.05 16.744 EvYear03CVACount​
>0.001 0.081 -2.518 Lntime_EvYear03CVACount​
As you can see both term and interaction term are statistically significant, but the HZ of the term is extremely high and i don't know how to correctly interpret it.
Any help will be much appreciated.
Elad
 

EGR

New Member
#3
Sorry! by writing HZ i meant Hazard ratio.

And what exactly do you mean "to play around with distributions"?