Use odds ratio procedure within proc logistic on imported parameter estimates?

[richart]

This may be a shot in the dark, but I have a csv of a past model I have run with coefficients, terms, SE, etc. But as far as I can tell, proc logistic only works on the cohort and variables you are calling. Any way it can act on already obtained model estimates? If not, maybe there is a macro someone knows about?

Thanks,

[mkeintz]

I'm not following.  It seems you have results in hand that are analogous to what proc logistic produces.  And you don't have the data from which those results were obtained.

So what do you want to use the procedure for?

[richart]

no, I do not have the data to the results. Effectively I want statistical software to exp(beta) so I can obtain OR. However, I have interaction terms which I know SAS can handle quite well (i.e. proc logistic oddsratio/ at, slice, etc.) and of course exponentiating my main effects in presence of interaction terms is not appropriate.