Table 2.
Variable | Type | Unit | Univariate logistic regression analysisa | Random forest | |||||
|
|
|
OR (95% CI) | P value | Rank | Decrease of impurity, % | Nodes using that attribute, n | Rank | |
Age (years) | Continuous | Years | 1.1 (1.03-1.07) | <.001 | 1b | 43 | 3238 | 1 | |
Mechanical ventilation >24 hours | Categorical | Yes/no | 5.8 (3.9-8.6) | <.001 | 3 | 21 | 297 | 21 | |
Preoperative creatinine clearance | Continuous | μmol/L | 0.97 (0.96-0.98) | <.001 | 1b | 39 | 2544 | 4 | |
Length of stay in the ICUc | Ordinal | —d | — | — | 2 | 26 | 590 | 20 | |
|
>72 hours | — | — | 7.6 (4.9-11.9) | <.001 | — | — | — | — |
|
24-72 hours | — | — | 1.7 (0.9-2.8) | <.001 | — | — | — | — |
Procedure other than isolated CABGe | Categorical | Yes/no | 2.9 (1.8-2.5) | <.001 | 6 | 28 | 370 | 15 | |
Blood product within 48 hours | Categorical | Yes/no | 2.9 (2.0-4.2) | <.001 | 5 | 28 | 452 | 14 | |
Intraoperative TEEf | Categorical | Yes/No | 2.0 (1.3-3.1) | .002 | 10 | 27 | 568 | 18 | |
EURO IIg score | Continuous | Percent | 1.07 (1.05-1.09) | <.001 | 17 | 41 | 2716 | 2 | |
Preoperative hemoglobin | Continuous | gm/dL | 0.98 (0.97-0.99) | <.001 | 1b | 40 | 2766 | 3 | |
Preoperative A-Fibh | Categorical | Yes/no | 2.3 (1.4-3.6) | <.001 | 7 | 35 | 486 | 6 | |
Timing of IABPi | Ordinal | — | — | — | 4 | 29 | 329 | 12 | |
|
Preoperative | — | — | 1.4 (0.6-2.9) | .42 | — | — | — | — |
|
Intraoperative | — | — | 6.8 (1.9-23.1) | .002 | — | — | — | — |
Intraoperative inotropes | Categorical | Yes/no | 2.1 (1.4-3.0) | <.001 | 8 | 27 | 514 | 17 | |
COPDj | Categorical | Yes/no | 1.7 (1.1-2.7) | .02 | 14 | 33 | 689 | 9 | |
CVDk | Categorical | Yes/no | 1.8 (1.1-2.9) | .01 | 13 | 29 | 516 | 13 | |
DMl | Categorical | Yes/no | 0.9 (0.6-1.4) | .79 | 16 | 39 | 995 | 5 | |
Frail | Categorical | Yes/no | 2.0 (1.1-3.5) | .03 | 12 | 30 | 381 | 11 | |
History of turn down | Categorical | Yes/no | 8.2 (2.8-24.3) | <.001 | 1b | 21 | 93 | 22 | |
EFm categories | Ordinal | — | — | — | 9 | 33 | 89 | 10 | |
|
30%-50% | — | — | 1.4 (0.9-2.1) | .18 | — | — | — | — |
|
<30% | — | — | 2.1 (1-4.2) | .04 | — | — | — | — |
Gender | Categorical | Yes/no | 1.2 (0.8-1.9) | .47 | 16 | 35 | 752 | 7 | |
Aortic stenosis | Ordinal | — | — | — | 14 | 26 | 899 | 16 | |
|
Moderate | — | — | 1.4 (0.6-2.8) | .43 | — | — | — | — |
|
Severe | — | — | 1.6 (1.1-2.5) | .01 | — | — | — | |
Mitral insufficiency | Ordinal | — | — | — | 15 | 26 | 899 | 19 | |
|
Moderate | — | — | 1.4 (0.9-2.1) | .07 | — | — | — | — |
|
Severe | — | — | 2.3 (1.02-4.6) | .03 | — | — | — | — |
Postoperative arrhythmias | Categorical | Yes/no | 1.8 (1.3-2.8) | .002 | 11 | 34 | 746 | 8 |
aAnalysis was done using univariate logistic regression with a P value of <.05 considered to be statistically significant.
bThese variables were all equally ranked as 1st because they had almost equal odds ratios and a P value of <.001.
cICU: intensive care unit.
dNot applicable.
eCABG: coronary artery bypass graft.
fTEE: transesophageal echo.
gEURO II: European System for Cardiac Operative Risk Evaluation II.
hA-Fib: atrial fibrillation.
iIABP: intra-aortic balloon pump.
jCOPD: chronic obstructive pulmonary disease.
kCVD: cerebrovascular disease.
lDM: diabetes mellitus.
mEF: ejection fraction.