Hi all,

I'm not quite sure why I'm getting confused about this, but I am! So I have 3 different areas and my study period is over 17 years with the DV being hospitalizations. In SAS, I did negative binomial regression and am mainly interested in the interaction term of area*period because I want to know if there are differences in n between the three areas over time. I believe I can calculate the incidence rate ratio (relative risk) by exponentiating the beta estimate or by adding an lsmeans statement in SAS.

However, what I'm confused on is getting the IRR over time, not just for area 1 vs area 2, area 1 vs area 3, etc. Because I need to get the IRR for the 3 areas over that 17 year period (for which I have n for each area for each calendar year). Thanks!