Table 2.
Logistic regression model for the prediction of postoperative hypoxemia
| Variables | Odds ratio | Lower limit of 95% CI | Upper limit of 95% CI | P value |
|---|---|---|---|---|
| BMI | 1.32 | 1.15 | 1.54 | < 0.001 |
| PF | 0.99 | 0.99 | 1.00 | 0.011 |
| Stanford (A as reference) | 0.22 | 0.06 | 0.66 | 0.011 |
| WBC | 1.21 | 1.06 | 1.40 | 0.008 |
| Age | 1.03 | 0.99 | 1.08 | 0.128 |
| HCT | 0.89 | 0.80 | 0.98 | 0.016 |
| CBP time | 0.99 | 0.98 | 1.00 | 0.031 |
| pH | 0.0002 | 2*10−8 | 0.74 | 0.048 |
Note: The logistic regression model was selected by using stepwise forward selection procedure, AIC was used to decide the inclusion of a variable. The odds ratio was reported for each one unit increase for each variable
Abbreviations: CI confidence interval, WBC white blood cell count, PF PaO2/FiO2, BMI body mass index, CBP cardiopulmonary bypass, HCT hematocrit