Skip to main content

Table 2 Subsequent outcome characteristics of the four phenotypes

From: Machine learning derivation of four computable 24-h pediatric sepsis phenotypes to facilitate enrollment in early personalized anti-inflammatory clinical trials

Characteristice

Total

PedSep-A

PedSep-B

PedSep-C

PedSep-D

No. of patients, N (%)

404 (100)

136 (34)

102 (25)

110 (27)

56 (14)

Development of subsequent MOF empirical phenotypes

 SMOF, N (%)

7 (1.7)

0 (0.0)

0 (0.0)

1 (0.9)

6 (10.7)a,b,c

 TAMOF, N (%)

37 (9.2)

0 (0.0)

6 (5.9)a

3 (2.7)

28 (50.0)a,b,c

 IPMOF, N (%)

85 (21.0)

12 (8.8)

29 (28.4)a

22 (20)

22 (39.3)a

 MAS, N (%)

24 (5.5)

0 (0.0)

3 (2.9)

2 (1.8)

19 (33.9)a,b,c

 NPMOF, N (%)

117 (29.0)

28 (20.6)

25 (24.5)

32 (29.1)

32 (57.1)a,b,c

Infections

 Bacterial infection, N (%)

141 (34.9)

43 (31.6)

33 (32.4)

45 (40.9)

20 (35.7)

 Viral infection, N (%)

114 (28.2)

60 (44.1)b,c,d

21 (20.6)

24 (21.8)

9 (16.1)

 Fungal infection, N (%)

4 (1.0)

0 (0.0)

1 (1.0)

0 (0.0)

3 (5.4)

 Culture negative, N (%)

177 (43.8)

47 (34.6)

52 (51.0)

50 (45.5)

28 (50.0)

Sites of infectionsf

 Blood, N (%)

51 (12.6)

10 (7.4)

6 (5.9)

22 (20.0)a,b

13 (23.2)a,b

 Lung, N (%)

76 (18.8)

28 (20.6)

29 (28.4) a,c,d

12 (10.9)

7 (12.5)

 Urine, N (%)

16 (4.0)

4 (2.9)

5 (4.9)

6 (5.5)

1 (1.8)

Organ support

 MechVent, N (%)

366 (90.6)

134 (98.5)c

101 (99.0)c

79 (71.8)

52 (92.9)c

 ECMO, N (%)

30 (7.4)

5 (3.7)

9 (8.8)

6 (5.5)

10 (17.9)a

 CRRT, N (%)

52 (12.9)

1 (0.7)

7 (6.9)

7 (6.4)

37 (66.1)a,b,c

Anti-inflammatory therapies of interest

 Decadron, N (%)

94 (23.3)

50 (36.8)c,d

22 (21.6)

14 (12.7)

8 (14.3)

 Methylprednisolone, N (%)

117 (29.0)

54 (39.7)b

23 (22.5)

24 (21.8)

16 (28.6)

 IVIG, N (%)

51 (12.6)

6 (4.4)

10 (9.8)

19 (17.3)a

16 (28.6)a

 IVIG + Methylprednisolone

23 (5.7)

3 (2.2)

4 (3.9)

9 (8.2)a

7 (12.5)a

 Plasma exchange, N (%)

25 (6.2)

5 (3.7)

4 (3.9)

4 (3.6)

12 (21.4)a,b,c

 Plasma exchange + ECMO

6 (1.5)

1 (0.7)

1 (1.0)

1 (0.9)

3 (5.4)

Outcome

 Length of stay, median (IQR), d

9.0 (5.0–17.)

9.0 (5.8–15)c

10.5 (5.3–17)c

6 (2.3–15)

12.5 (7–26.5)c

 Mortality, N (%)

45 (11.1)

3 (2.2)

12 (11.7)a

11 (10.0)a

19 (33.9)a,b,c

 PICU free days, median (IQR), d

20.0 (8.0–25.0)

21.0 (14.8–24.0)d

19.0 (9.8–24.0)d

24.0 (13.3–27)a,b,d

4.5 (0.0–21.0)

  1. SMOF sequential liver failure-associated multiple organ failure, TAMOF thrombocytopenia-associated multiple organ failure, IPMOF immunoparalysis-associated multiple organ failure, MAS Macrophage Activation Syndrome, NPMOF new or progressive multiple organ failure, IQR interquartile range, MechVent mechanical ventilation, ECMO extracorporeal membrane oxygenation, CRRT continuous renal replacement therapies, IVIG intravenous gamma globulin
  2. aThe outcome characteristic of this computable phenotype is significantly higher than PedSep-A (p value < 0.05)
  3. bThe outcome characteristic of this computable phenotype is significantly higher than PedSep-B (p value < 0.05)
  4. cThe outcome characteristic of this computable phenotype is significantly higher than PedSep-C (p value < 0.05)
  5. dThe outcome characteristic of this computable phenotype is significantly higher than PedSep-D (p value < 0.05)
  6. eComparisons across all 4 computable phenotypes were performed using the Kruskal–Wallis test, the χ2 test, or the Fisher’s exact test (Additional file 1: Table S3, p < .05 for all comparisons after adjustment)
  7. fObtained at the first 3 days
Back to article page

Critical Care

ISSN: 1364-8535