Predicting one-year mortality of critically ill patients with early acute kidney injury: data from the prospective multicenter FINNAKI study

Table 4 Day 3 model for one-year mortality by multivariate logistic regression analysis

Predictor	Odds ratio (95% CI)	P	Independent contribution %^a
The highest bilirubin within D1 to D3 (per μmol/l)^b	1.02 (1.01-1.03)	<0.001	22.6
Age, years (per year)	1.04 (1.02-1.06)	<0.001	15.6
Mechanical ventilation on D3	2.73 (1.62-4.61)	<0.001	15.0
SAPS II score without points given for age, renal components, bilirubin, and type of admission (per point)	1.03 (1.00-1.05)	0.02	12.3
The lowest BE value on D3 (per mmol/l)	0.92 (0.87-0.97)	0.001	11.8
Number of co-morbidities	1.26 (1.05-1.52)	0.015	8.0
Dependence of assistance in premorbid functional performance preceding the acute illness^c	1.76 (0.91-3.42)	0.094	5.9
Admission type according to SAPS II^d			8.9
Unscheduled surgical	3.97 (0.46-33.98)	0.208
Medical	6.59 (0.79-55.05)	0.08

^aPresents the independent contribution percentage of the variable to the predictive performance of the model; ^bthe proportion of missing values of the bilirubin concentration within the first three days was 6.0%. ^ccompared to normal or unable to work; ^dcompared to scheduled surgical. Non-significant predictors for one-year mortality included to the analysis: APACHE II diagnostic group, gender, daily performance, admission type according to SAPS II, severe sepsis between admission and D3, use of norepinephrine on D3, the lowest platelet value on the D3, the highest Kidney Disease: Improving Global Outcomes (KDIGO) stage during D1 to D3, and difference in number of organ failures, including renal failure, on D3 and D1 (ΔOF)). CI, confidence interval; D1/D3 the first or third day on the ICU; SAPS, Simplified Acute Physiology Score; BE, base excess; ICU, intensive care unit.

ISSN: 1364-8535