Please tell me what is the right proc for one categorical (three levels) vs continuous (non normal)

[khageswar]

I have dataset which has one categorical varaible with three levels 

i.e. Group is the variable under which 1=Attended medication less than a year, 2= one - two years and 3= more than two years

and growth_index value is continuous but not normal.

please suggest the suitable proc...

[PaigeMiller]

First you have to tell us what analysis you want to perform on this data.

[khageswar]

Trying to find out any difference exist among the groups?

[PaigeMiller]

---

@khageswar wrote:
Trying to find out any difference exist among the groups?

---

Mean difference? Variability difference? Distribution difference? Other?

[Ksharp]

You also could do PROC GLM for ANOVA, it is robust for non-normal data.
Or you could try Non-parameter method SAVAGE test,check the last example 
in PROC NPAR1WAY documentation.