SEA-MAKE score as a tool for predicting major adverse kidney events in critically ill patients with acute kidney injury: results from the SEA-AKI study

Annals of Intensive Care

Table 2 Creating the simplified SEA-MAKE scoring system using the multivariable regression method^a

Predictors	B	S.E.	Adjusted odds ratio^b (95% CI)	P	Simplified scores^c
Glasgow coma scale < 10	1.00	0.11	2.72 (2.17, 3.40)	< 0.001	3
Respiratory rate > 24 breaths/min	0.44	0.09	1.55 (1.29, 1.86)	< 0.001	1
Vasopressor use	0.45	0.09	1.56 (1.30, 1.88)	< 0.001	1
On mechanical ventilation	0.60	0.11	1.81 (1.45, 2.27)	< 0.001	2
Urine output < 0.5 L/day	0.58	0.09	1.78 (1.48, 2.14)	< 0.001	2
Serum creatinine rising ≥ 3 times	1.56	0.18	4.77 (3.38, 6.73)	< 0.001	5
Blood urea nitrogen ≥ 40 mg/dL [14.28 mmol/L]	0.84	0.11	2.33 (1.89, 2.86)	< 0.001	3
Hematocrit < 30%	0.59	0.10	1.81 (1.50, 2.19)	< 0.001	2
Platelet counts < 150 × 10³ cell/µL	0.34	0.10	1.40 (1.15, 1.70)	0.001	1

^aPooled data from 10 imputations after Forward Wald selection method in binary logistic regression model
^bAdjusted for co-variates
^cSchneeweiss’s scoring system
All of the variance inflation factors (VIFs) coefficients in collinearity statistics were less than 1.3 and tolerance coefficients exceeded 0.79 in multicollinearity test among all co-variates