Skip to main content

Table 2 Uni- and multivariable prediction models for ICU mortality in 5062 children admitted to intensive care units with sepsis and septic shock

From: Defining benefit threshold for extracorporeal membrane oxygenation in children with sepsis—a binational multicenter cohort study

Predictor variable

 

Univariable model

 

Multivariable model

 

Group

Predictor

OR (95% CI)

p value

OR (95% CI)

p value

Respiratory

PaO2/FiO2 ratio

2.83 (2.34–3.43)

< 0.001

1.57 (1.28–1.91)

< 0.001

Ventilation during the first hour

3.97 (3.15–4.96)

< 0.001

NS

 

Intubated

23.19 (14.58–36.89)

< 0.001

15.92 (9.76–25.96)

< 0.001

High frequency ventilation

7.59 (5.50–10.46)

< 0.001

2.26 (1.48–3.45)

< 0.001

Inhaled nitric oxide

7.15 (4.93–10.37)

< 0.001

NS

 

Circulatory

SBP < 5th percentile

3.06 (2.45–3.82)

< 0.001

1.81 (1.39–2.36)

< 0.001

Cardiac arrest pre-admission

6.50 (3.82–11.07)

< 0.001

2.36 (1.23–4.52)

0.01

Shock on presentation

4.78 (3.73–6.07)

< 0.001

2.31 (1.76–3.04)

< 0.001

Metabolic

Lactate (mmol/l)

1.26 (1.20–1.32)

< 0.001

1.12 (1.07–1.18)

< 0.001

Renal

Renal replacement therapy

8.23 (6.06–11.17)

< 0.001

2.44 (1.72–3.47)

< 0.001

Neurologic

Dilated, unresponsive pupils

48.11 (23.00–11,062)

< 0.001

35.01 (15.35–79.84)

< 0.001

Patient

Immunosuppression

1.65 (1.29–2.10)

< 0.001

2.66 (1.95–3.63)

< 0.001

  1. OR odds ratio, CI confidence interval, SBP systolic blood pressure