Skip to main content

Table 2 Univariable analysis of variables from history and physical examination

From: Predictive modeling in pediatric traumatic brain injury using machine learning

 

Traumatic brain injury (Cases) N = 39

No traumatic brain injury (Controls) N = 156

p-value*

History

Loss of consciousness

25 (64%)

8 (5%)

<0.001

If Yes, then number who lost consciousness for > 1 minute (%)

24 (96%)

4 (50%)

<0.001

Difficult arousal (%)

27 (70%)

3 (2%)

<0.001

Vomiting (%)

11 (28%)

26 (17%)

0.113

Seizure activity (%)

6 (15%)

0 (0%)

<0.001

Confusion/Disorientation (%)

33 (85%)

3 (2%)

<0.001

(Preverbal) irritability (%) (N = 7 vs 22)

0 (0%)

1 (5%)

1

(Verbal) Headache (%) (N = 32 vs 130)

8 (25%)

45 (35%)

0.401

(Verbal) Amnesia (%) (N = 32 vs 131)

1 (3%)

6 (5%)

1

Physical examination

Signs of altered mental status (%)

36 (95%)

1 (1%)

<0.001

Presence of unequal pupils (%)

(26%)

(1%)

<0.001

Clinical signs of skull fracture (%)

2 (5%)

1 (1%)

0.103

Signs of base of skull fracture (%)

13 (33%)

4 (3%)

<0.001

Presence of scalp hematoma (%)

20 (51%)

36 (23%)

<0.001

Frontal (%) (N = 20 vs 36)

1 (5%)

13 (36%)

0.011

Presence of scalp laceration

7 (18%)

32 (21%)

0.825

(Preverbal) with open fontanelle (N = 7 vs 22)

4 (57%)

17 (77%)

0.357

Presence of tense fontanelle among those with open fontanelles (N = 4 vs 17)

3 (75%)

0 (0%)

0.03

  1. *Chi-Square or Fisher Test when appropriate.