Table 1 Summary characteristics of the study population at ICU admission stratified by ordinal 6-month outcomes.

From: Mining the contribution of intensive care clinical course to outcome after traumatic brain injury

Summary characteristics

Overall

Glasgow Outcome Scale-Extended (GOSE) at 6 months post-injury

P value

  

(1) Death

(2 or 3) Vegetative or lower severe disability

(4) Upper severe disability

(5) Lower moderate disability

(6) Upper moderate disability

(7) Lower good recovery

(8) Upper good recovery

 

n

1550

318 (20.5%)

262 (16.9%)

120 (7.7%)

227 (14.6%)

200 (12.9%)

206 (13.3%)

217 (14.0%)

 

Age (years)

51 (31–66)

66 (50–76)

55 (36–68)

48 (29–61)

44 (31–56)

41 (27–53)

48 (31–65)

41 (24–61)

<0.0001

Sex

        

0.94

  Female

409 (26.4%)

78 (24.5%)

71 (27.1%)

43 (35.8%)

64 (28.2%)

49 (24.5%)

59 (28.6%)

45 (20.7%)

 

Baseline GCS (n = 1465)

8 (4–14)

5 (3–10)

6 (3–10)

8 (4–13)

8 (5–13)

9 (6–14)

13 (7–15)

13 (8–15)

<0.0001

  Mild (13–15)

390 (26.6%)

30 (10.3%)

38 (15.3%)

26 (23.4%)

42 (19.5%)

66 (34.9%)

91 (45.3%)

97 (46.4%)

 

  Moderate (9–12)

331 (22.6%)

65 (22.3%)

41 (16.5%)

28 (25.2%)

65 (30.2%)

36 (19.0%)

40 (19.9%)

56 (26.8%)

 

  Severe (3–8)

744 (50.8%)

196 (67.4%)

170 (68.3%)

57 (51.4%)

108 (50.2%)

87 (46.0%)

70 (34.8%)

56 (26.8%)

 

Marshall CT (n = 1255)

VI (II–VI)

III (II–VI)

II (II–VI)

II (II–VI)

II (II–II)

II (II–III)

II (II–II)

VI (II–VI)

0.02

  No visible pathology (I)

118 (9.4%)

8 (3.3%)

11 (5.3%)

5 (5.2%)

17 (8.7%)

25 (15.2%)

24 (13.6%)

28 (16.5%)

 

  Diffuse injury II

592 (47.2%)

56 (22.8%)

84 (40.6%)

54 (56.2%)

92 (47.2%)

100 (60.6%)

103 (58.5%)

103 (60.6%)

 

  Diffuse injury III

108 (8.6%)

42 (17.1%)

17 (8.2%)

10 (10.4%)

14 (7.2%)

9 (5.5%)

6 (3.4%)

10 (5.9%)

 

  Diffuse injury IV

16 (1.3%)

7 (2.8%)

1 (0.5%)

1 (1.0%)

4 (2.1%)

1 (0.6%)

1 (0.6%)

1 (0.6%)

 

  Mass lesion (V & VI)

421 (33.5%)

133 (54.0%)

94 (45.4%)

26 (27.1%)

68 (34.9%)

30 (18.2%)

42 (23.9%)

28 (16.5%)

 

tSAH (n = 1254)

957 (76.3%)

221 (90.2%)

176 (84.2%)

73 (76.0%)

150 (76.9%)

106 (63.9%)

125 (71.4%)

106 (63.1%)

0.12

EDH (n = 1257)

244 (19.4%)

31 (12.7%)

32 (15.3%)

21 (21.9%)

46 (23.6%)

32 (19.3%)

42 (23.9%)

40 (23.5%)

0.01

Retired (n = 1312)

353 (26.9%)

136 (61.3%)

74 (33.6%)

23 (22.1%)

12 (5.9%)

13 (7.3%)

52 (28.1%)

43 (21.8%)

0.02

Length of ICU stay (days)

8.3 (3.0–16.9)

7.0 (2.9–14.2)

17.5 (10.1–24.9)

13.1 (4.6–21.7)

9.1 (3.7–16.3)

6.9 (2.8–14.9)

3.7 (2.1–9.7)

3.9 (1.8–9.1)

<0.0001

  1. Data are counts (% of total study) for sample size (n), median (IQR) for continuous characteristics, and n (% of column group) for categorical characteristics. Units or numerical definitions of characteristics are provided in square brackets. If a characteristic had missing values for some patients in the population, the non-missing sample size was provided in parentheses—e.g., Marshall CT (n = 1255). Conventionally, TBI severity is categorically defined by baseline GCS scores as indicated in square brackets. Incidence of epidural haematoma (EDH) or traumatic subarachnoid haemorrhage (tSAH) was assessed from CT scans at ICU admission. P values are determined from proportional odds logistic regression (POLR) coefficient analysis trained on all summary characteristics concurrently. For categorical variables with k > 2 categories (e.g., Baseline GCS), P values were calculated with a likelihood ratio test (with k-1 degrees of freedom) on POLR.
Back to article page