Parameters:
(1) serum creatinine (CR) in mg/dL
(2) lactic acid (A) during surgery in mg/dL
(3) cardiopulmonary bypass (CPB) time (T) in minutes
(4) additive Euroscore (E)
|
Parameter
|
Finding
|
Points
|
|
serum creatinine
|
<= 0.96 mg/dL
|
18
|
|
|
0.97 to 1.26
|
25
|
|
|
1.27 to 1.49
|
30
|
|
|
>= 1.50 mg/dL
|
39
|
|
lactate
|
<= 17.8 mg/dL
|
5
|
|
|
18.0 to 23.9
|
6
|
|
|
24 to 36.8
|
9
|
|
|
36.9 to 51.0
|
13
|
|
|
>= 51.1 mg/dL
|
21
|
|
CPB time
|
<= 81 minutes
|
7
|
|
|
82 to 104
|
9
|
|
|
105 to 138
|
12
|
|
|
139 to 181
|
15
|
|
|
>= 182 minutes
|
20
|
|
Euroscore
|
<= 2
|
7
|
|
|
3 to 4
|
10
|
|
|
5 to 7
|
20
|
|
|
8 to 10
|
30
|
|
|
>= 11
|
40
|
total score =
= SUM(points for all 4 parameters)
Interpretation:
• minimum score: 37
• maximum score: 120
• The higher the score the greater the risk of acute kidney injury.
|
Total Score
|
Percent AKI
|
AKI Risk
|
|
< 50
|
4%
|
very low
|
|
50 to 70
|
27%
|
low
|
|
70 to 80
|
47%
|
moderate
|
|
80 to 90
|
64%
|
high
|
|
> 90
|
100%
|
very high
|
Alternatively:
Y =
= (0.114 * (CRATE)) – 9.24
probability of AKI =
= 1 / (1 + EXP((-1) *Y))
Alternatively a logistic regression model can be used:
X = (0.336 * (Euroscore)) + (0.01 * (CPB)) + (0.031 * (lactate)) + (2.268 * (creatinine)) – 8.422
where:
• The lactate has units of mmol/L vs mg/dL in the rest of the paper.
proability of AKI =
= 1 / (! + EXP((-1) * X))
Performance:
• The area under the ROC curve was 0.81.
