I computed a multivariable logistic regression model for a pre-op drug's relation to operative mortality. My covariates included the drug, age, gender, bmi, operation time, and comorbidities. I then computed covariate-adjusted mortality rates. Can I determine if the adjusted rates are significantly different? i.e get a p-value or chi square? Or is this not meaningful since it is already adjusted?
How I'm presenting my data:
univariable: 7.9% mortality w/ drug vs 2.1% mortality w/o drug, p = 0.0013
adjusted: 6.6% mortality w/ drug vs 3.3% mortality w/o, p = ????
Thanks for your help.