Dilemma on graphical presentation

[Tara Boyle]

Hello all. I wonder if the experts on here can perhaps point me in the right direction. I am comparing two treatments (one medical vs one surgical) on different dependent variables, one is continous (model 1), second is ordinal categorical (model 2) and the last one is a binary categorical (model 3).

Ideally, I would like to plot them all in one forest plot, but the difficulty is they don't all produce odds ratio apart from Model 3, making it difficult to put them all in one Plot. I would like them all in one plot so the reader can look at it an quickly understand the advantage of the medical treatment over the surgical treatment.

Questions: 1. Is it possible to convert Model 1 and Model 2 into odds ratios ?
Question 2: If question 1 doesn't make sense, what would you advise? Or perhaps there no solution at all, but one has to describe in text.

MODELS

- continuousvar - is a continous variable eg age
- treatment var - is a categorical variable 0 or 1
-pt comorbidity is a categorical variable from 1-3, 3 is the sickest
-catvar is a binary a

Model 1: **Co efficient produced

reg continousvar treatmentvar ptcomordbitiy

//Here I obtain co-efficients


Model 2: **THIS PRODUCES predicted probabilities

ologit ptcomorbidity treatmentvar continousvar 

//this produces an ODDS ratio but it is difficult to interpret if the ODDS ratio is for pt comordbity 2 or 3, therefore I proceeded with 

margins, at(treatmentvar=(0/1)) predict(outcome1)) at means //SICK1
margins, at(treatmentvar=(0/1)) predict(outcome2)) at means //SICK2
margins, at(treatmentvar=(0/1)) predict(outcome3)) at means //SICK3

Model 3: **odds ratios produced

logit catvar treatmentvar continousvar