I've performed an univariate binomial regression with OR (95% CI) and I obtain this result:
> full.model <- glm(Re_OUT ~ t0_FIO1 , data = db,family=binomial())
> logistic.display(full.model)
Logistic regression predicting Re_OUT : 1 vs 0
OR(95%CI) P(Wald's test) P(LR-test)
t0_FIO1 (cont. var.) 1.03 (1.03,1.04) < 0.001 < 0.001
Log-likelihood = -212.7811
No. of observations = 383
AIC value = 429.5621
Then I performed a multivariate binomial regression. I should find that the t0_FI01 has a crude OR equal to the OR of the univariate binomial regression, but this isn't so.
> full.model <- glm(Re_OUT ~ t0_FIO1 + t0_IOT , data = db,
family=binomial())
> logistic.display(full.model)
Logistic regression predicting Re_OUT : 1 vs 0
crude OR(95%CI) adj. OR(95%CI) P(Wald's test) P(LR-test)
t0_FIO1 (cont. var.) 1.03 (1.03,1.04) 1.03 (1.02,1.03) < 0.001 < 0.001
t0_IOT: 1 vs 0 63.06 (15.17,262.1) 27.86 (6.53,118.82) < 0.001 < 0.001
Log-likelihood = -189.8483
No. of observations = 383
AIC value = 385.6967
Why?
