Skip to main content

Table 3 Baseline patient characteristics between MIMIC-III and eICU

From: A machine learning-based risk stratification tool for in-hospital mortality of intensive care unit patients with heart failure

Variables

MIMIC-III (n = 5676)

eICU (n = 1349)

Age, mean (years, SD)

70.7 (12.6)

67.8 (13.7)

Men, n (%)

3161 (55.7)

802 (59.5)

BMI, mean (kg/m2, SD)

29.2 (7.5)

31.5 (12.9)

SBP, mean (mmHg, SD)

117.4 (16.5)

119.1 (19.8)

PTT, mean (s, SD)

41.9 (21.9)

42.0 (18.7)

Temperature, mean (°C, SD)

36.8 (0.6)

36.7 (0.6)

GCS, mean (SD)

11.9 (3.4)

13.4 (2.7)

Respiratory rate, mean (/min, SD)

19.1 (3.9)

20.3 (4.1)

Phosphate, mean (mg/dL, SD)

3.8 (1.3)

4.1 (1.4)

Calcium, mean (mmol/L, SD)

8.4 (0.8)

8.5 (0.7)

Glucose, mean (mg/dL, SD)

145.8 (50.9)

148.6 (58.3)

Urine output, mean (mL, SD)

1908.6 (1167.4)

1837.4 (1408.6)

Arterial base excess, mean (mmol/L, SD)

0.03 (4.3)

NA

Mortality (%)

10.5

12.8

  1. SD standard deviation, BMI body mass index, SBP systolic blood pressure, PTT partial thromboplastin time, GCS Glasgow Coma Scale, bp blood pressure, NA not available