Skip to main content

Table 2 The heterogeneous features for training set (1661 cases) according to blood tests

From: Clinical applications of machine learning in the survival prediction and classification of sepsis: coagulation and heparin usage matter

Features

Cluster one

Cluster two

Cluster three

Cluster four

P value

number of each cluster

211

1215

46

189

 

survival (%)

165 (78.2%)

1005 (82.7%)

35 (76.1%)

132 (69.8%)

0.020

Age, median (IQR), year

66 (54–76)

66 (54–76.5)

67.5 (55–80)

66 (57–77)

0.933

Male, no. (%)

102 (48.3%)

663 (54.6%)

17 (37.0%)

110 (58.2%)

0.023

Top ten blood test varies, median (IQR), unit

 PTT, sec

30.6 (26.5–36.3)

29.6 (26.2–34.3)

150 (122.9–150)

51.6 (43.5–60.6)

0.000

 Neutrophils, %

51 (33–61.7)

84 (77–89.9)

81.8 (77.2–86)

83 (76.4–89)

0.000

 PT, sec

14.3 (13.1–16.4)

14.2 (13.1–16.0)

17.7 (15.3–23.8)

30.2 (21.9–43.1)

0.000

 INR(PT), NULL %

1.3 (1.1–1.5)

1.3 (1.1–1.5)

1.9 (1.5–3.2)

3.4 (2.2–5.2)

0.000

 Lymphocytes, %

26 (15.5–34.3)

7 (4–11.3)

7 (4.6–14.0)

8 (4–13)

0.000

 White blood cells, K/μL

6.1 (3.4–10.8)

13.2 (8.9–8.1)

10.7 (7.0–8.4)

12.1(8.5–17.4)

0.000

 Platelet count, K/μL

2.3 (2.0–2.5)

2.4 (2.2–2.5)

2.3 (2.2–2.4)

2.3 (2.1–2.5)

0.000

 MCHC, %

33.3 (32.3–34.3)

33 (32–34.1)

32.5 (32.2–34)

32 (31–33.3)

0.000

 Albumin, %

2.9 (2.5–3.3)

2.9 (2.5–3.4)

2.9 (2.5–3.4)

2.9 (2.4–3.2)

0.000

 Red blood cells, K/μL

3.5 (3.1–4.0)

3.7 (3.3–4.2)

3.7 (3.2–4.0)

3.7 (3.1–4.0)

0.000