Skip to main content

Table 3 The “traditional” risk model of AKI derived through logistic regression with automated backward variable selection

From: Prediction of acute kidney injury risk after cardiac surgery: using a hybrid machine learning algorithm

Characteristics AORs, 95% CI β coefficient Risk score
CARE score
1 Ref Ref 0
2 2.32 (1.95–2.76)  − 0.3139  − 4
3 3.83 (3.02–4.85) 0.2756 3
4 9.64 (6.77–13.73) 1.2220 14
NYHA Functional Class
0 Ref Ref 0
1 0.85 (0.66–1.10)  − 0.3184  − 4
2 1.10 (0.91–1.34)  − 0.0781  − 1
3 1.31 (1.08–1.59) 0.0889 1
4 1.76 (1.28–2.43) 0.4427 5
Anemia
No Ref Ref 0
Yes 1.60 (1.38–1.86) 0.2124 2
Redo sternotomy
No Ref Ref 0
Yes 1.45 (1.13–1.86) 0.2427 3
Smoking status
Never Ref Ref 0
Current 1.21 (0.98–1.51) 0.1007 1
Former 1.42 (1.21–1.66) 0.1146 1
BMI 1.04 (1.03–1.06) 0.0434  
 ≤ 18   0.4774  − 5
19–24   Ref 0
25–29   0.2387 3
 ≥ 30   0.7595 8
  1. CARE score, Cardiac Anesthesia Risk Evaluation (CARE) mortality risk score; NYHA class, New York Heart Association Classification; BMI, body mass index