Table 2 Feature selection of stepwise logistic regression model

Demographic characteristics

 Weight (with each 10 kg increment)

0.971 [0.947, 0.995]

0.017

 Admission type (unscheduled surgical as reference)

  Medical

1.195 [1.054, 1.356]

0.005

  Scheduled surgical

0.696 [0.492, 0.986]

0.041

Comorbidities

 Cerebrovascular disease

1.395 [1.198, 1.626]

 < 0.001

 Paraplegia

1.685 [1.340, 2.118]

 < 0.001

 Metastatic solid tumor

1.587 [1.241, 2.029]

 < 0.001

Vital signs

 HR (with every 10 beats/min increment)

1.047 [1.014, 1.082]

0.005

 RR

1.019 [1.007, 1.031]

0.001

 SpO2

0.943 [0.921, 0.964]

 < 0.001

GCS

 GCS eyes

0.773 [0.723, 0.826]

 < 0.001

 GCS motor

0.844 [0.811, 0.879]

 < 0.001

Blood biochemistry

 PaCO2 (with every 10 mmHg increment)

1.151 [1.076, 1.231]

 < 0.001

 Glucose

1.002 [1.001, 1.004]

 < 0.001

 Hemoglobin

0.910 [0.880, 0.940]

 < 0.001

 BUN

1.008 [1.005, 1.012]

 < 0.001

 Creatinine (with every 0.1 mg/dl increment)

0.986 [0.979, 0.994]

 < 0.001

 Sodium (with every 10 mmol/L increment)

1.397 [1.275, 1.532]

 < 0.001

 Calcium

0.883 [0.813, 0.959]

0.003

 Anion gap

1.061 [1.042, 1.081]

 < 0.001

 PTT

1.005 [1.002, 1.008]

0.002

 INR

1.182 [1.064, 1.314]

0.002

Ventilator parameters

 Ppeak

1.037 [1.027, 1.047]

 < 0.001

 PEEP

1.082 [1.057, 1.107]

 < 0.001

 Inspiratory flow rate

1.011 [1.008, 1.013]

 < 0.001

Norepinephrine (with every 0.1 mcg/kg/min increment)

1.308 [1.228, 1.394]

 < 0.001

Intravenous antibiotics

0.843 [0.743, 0.957]

0.008

CRRT

1.734 [1.319, 2.280]

 < 0.001