My dependent variables are patient satisfaction scores questions which are ordinal variable. My independent variables are continuous (age , length of stay, Case mix index, no of physicians, no of floor switches). I ran Mann Whitney test initially by making my independent variables dichotomous My data was non normally distributed so I ran spearman correlation.Case max index and no of physician variables were significantly correlated with a few of the dependent variables. I decided to run ordinal logistic regression. I want to control for case mix index and see if there is any relationship between no of physicians and patient satisfaction score questions. How do I do that ?