A total of 2570 hospitalized adult TBI patients were included in the analyses. The mean age of the patients studied was 48.54 ± 16.38 years, of which 1934 patients (73.5%) were male. The main cause of TBI was traffic accidents (n = 1465, 57.0%), followed by stumbles (n = 45217.6%) and falls (n = 378, 14.7%). Intraparenchymal hemorrhage and brain contusion were the most common head CT findings (n = 1840, 70.2%). Multiple injuries were seen in 147 (5.7%) patients. The most common site of multiple injuries was the thorax (n = 103, 4.0%), followed by extremities (n = 37, 1.4%). The mean ISS of all patients was 15.35 ± 7.24. The mean fibrinogen level at admission was 2.54 ± 1.14 g/L, with 77 (3.0%) patients exhibiting fibrinogen levels < 1.0 g/L, 326 patients (12.7%) with levels between 1.0–1.5 g/L, 589 patients (22.9%) between 1.5–2.0 g/L, 918 patients (35.7%) between 2.0–3.0 g/L, and 660 patients (25.7%) > 3.0 g/L. A total of 198 patients died in this series. The main cause of death was brain death caused by malignant intracranial hypertension (n = 128, 64.6%), followed by multiple organ failure (n =56, 28.3%), septic shock (n = 11, 5.6%) and intracranial infection (n = 3, 1.5%). Among them, 41.9% died within 24 hours after admission, 24.2% died between 24 and 48 hours, 25.8% died between 48 and 72 hours, and 8.1% died after 72 hours.
Univariate analyses revealed that age, pupillary reactions, GCS at admission, head CT findings, and ISS were significantly related to 3-month mortality. Meanwhile, patients who died exhibited significantly higher INR, PT, APPT, and D-dimer levels, combined with lower FIB, PLT, Hb, and HCT levels relative to those who survived (Table 1). By contrast, age, pupillary reactions, GCS at admission, head CT findings, and ISS were also significantly related to 3-month favorable outcomes (Table 2). Multiple injuries were not significantly related to 3-month mortality, although the combination of TBI with either chest or abdominal injuries was significantly associated with 3-month unfavorable outcomes. Patients with an unfavorable outcome exhibited significantly higher INR, PT, APPT and D-dimer levels, combined with lower FIB, PLT, Hb, and HCT levels compared to those with a favorable outcome.
Table 1. Summary of patient characteristics and coagulation tests of the dead and surviving patients
|
Died (%)
|
Survived (%)
|
Total (%)
|
P Valuea
|
N
|
198 (100.0%)
|
2372 (100.0%)
|
2570 (100.0%)
|
|
Age (mean ± SD)
|
56.25 ± 17.81
|
47.90 ± 16.10
|
48.54 ± 16.38
|
< .001
|
Sex
|
|
|
|
.391
|
Male
|
154 (77.8%)
|
1780 (75.0%)
|
1934 (75.3%)
|
|
Female
|
44 (22.2%)
|
592 (25.0%)
|
636 (24.7%)
|
|
Mechanism of injury
|
|
|
|
|
Traffic accident
|
119 (60.1%)
|
1346 (56.7%)
|
1465 (57.0%)
|
.090
|
Fall
|
20 (10.1%)
|
358 (15.1%)
|
378 (14.7%)
|
|
Stumble
|
44 (22.2%)
|
408 (17.2%)
|
452 (17.6%)
|
|
Blow to head
|
6 (3.0%)
|
104 (4.4%)
|
110 (4.3%)
|
|
Others
|
9 (4.5%)
|
156 (6.6%)
|
165 (6.4%)
|
|
Pupillary reactions
|
|
|
|
|
Both reacting
|
103 (52.0%)
|
2145 (90.4%)
|
2248 (87.5%)
|
< .001
|
One reacting
|
50 (25.3%)
|
183 (7.7%)
|
233 (9.1%)
|
|
None reacting
|
45 (22.7%)
|
44 (1.9%)
|
89 (3.5%)
|
|
GCS at admission
|
6.73 ± 3.31
|
11.34 ± 3.49
|
10.98 ± 3.69
|
< .001
|
≤8
|
145 (73.3%)
|
586 (24.7%)
|
731 (28.5%)
|
< .001
|
9-12
|
36 (18.2%)
|
653 (27.5%)
|
689 (26.8%)
|
|
13-15
|
17 (8.6%)
|
1133 (47.8%)
|
1150 (44.7%)
|
|
CT findings
|
|
|
|
|
IPH or brain contusion
|
162 (81.8%)
|
1642 (69.2%)
|
1804 (70.2%)
|
< .001
|
SDH
|
104 (52.5%)
|
684 (28.8%)
|
788 (30.7%)
|
< .001
|
EDH
|
32 (16.2%)
|
662 (27.9%)
|
694 (27.0%)
|
< .001
|
DAI
|
19 (9.2%)
|
52 (2.2%)
|
71 (2.8%)
|
< .001
|
Multiple body injuries
|
13 (6.6%)
|
134 (5.6%)
|
147 (5.7%)
|
.594
|
Face
|
1 (0.5%)
|
14 (0.6%)
|
15 (0.6%)
|
1.000
|
Thorax
|
11 (5.6%)
|
92 (3.9%)
|
103 (4.0%)
|
.333
|
Abdomen
|
2 (1.0%)
|
8 (0.3%)
|
10 (0.4%)
|
.385
|
Extremities
|
2 (1%)
|
35 (1.5%)
|
37 (1.4%)
|
.828
|
External
|
1 (0.5%)
|
3 (0.1%)
|
4 (0.2%)
|
.719
|
ISSb
|
22.33 ± 6.12
|
14.77 ± 7.02
|
15.35 ± 7.24
|
< .001
|
INR
|
1.16 ± 0.24
|
1.06 ± 0.13
|
1.07 ± 0.14
|
< .001
|
PT(s)
|
13.44 ± 2.71
|
12.16 ± 1.55
|
12.26 ± 1.70
|
< .001
|
APTT(s)
|
31.19 ± 14.65
|
27.30 ± 7.20
|
27.60 ± 8.09
|
< .001
|
FIB(g/L)
|
2.06 ± 1.16
|
2.58 ± 1.13
|
2.54 ± 1.14
|
< .001
|
D-dimer (mg/L)
|
10.52 ± 14.67
|
6.90 ± 10.73
|
7.18 ± 11.12
|
< .001
|
PLT(×109/L)
|
156 ± 61
|
177 ± 67
|
176 ± 67
|
< .001
|
Hb(g/L)
|
124 ± 25
|
131 ± 21
|
131 ± 21
|
< .001
|
HCT(%)
|
36.44 ± 6.92
|
38.36 ± 5.69
|
38.21 ± 5.82
|
< .001
|
Abbreviations: GCS, Glasgow Coma Scale; ISS, Injury Severity Score; PLT, platelet; INR, international normalized ratio; PT, prothrombin; APTT, activated partial thromboplastin time; FIB, fibrinogen; EDH, epidural hematoma; SDH, subdural hematoma; IPH, intraparenchymal hemorrhage; DAI, diffuse axonal injury.
a P values not adjusted for multiple comparisons. Statistical significance determined by chi-square test, Student’s t-test or Mann-Whitney tests.
b ISS=AIS12+AIS22+AIS32, it is the sum of squares for the highest values in each of the three most severely injured body regions.
Table 2. Summary of patient characteristics and coagulation tests of the favorable and unfavorable outcome patients
|
Unfavorable outcome (%)
|
Favorable outcome (%)
|
Total
(%)
|
P Valuea
|
N
|
802 (100.0%)
|
1768 (100.0%)
|
2570 (100.0%)
|
|
Age (mean± SD)
|
52.65 ± 16.20
|
46.68 ± 16.13
|
48.54 ± 16.38
|
< .001
|
Sex
|
|
|
|
.885
|
Male
|
605 (75.4%)
|
1329 (75.2%)
|
1934 (75.3%)
|
|
Female
|
197 (24.6%)
|
439 (24.8%)
|
636 (24.7%)
|
|
Mechanism of injury
|
|
|
|
|
Traffic accident
|
469 (58.5%)
|
996 (56.3%)
|
1465 (57.0%)
|
.060
|
Fall
|
136 (17.0%)
|
242 (13.7%)
|
378 (14.7%)
|
|
Stumble
|
132 (16.5%)
|
320 (18.1%)
|
452 (17.6%)
|
|
Blow to head
|
21 (2.6%)
|
89 (5.0%)
|
110 (4.3%)
|
|
Others
|
44 (5.5%)
|
121 (6.8%)
|
165 (6.4%)
|
|
Pupillary reactions
|
|
|
|
|
Both reacting
|
565 (70.4%)
|
1683 (95.2%)
|
2248 (87.5%)
|
< .001
|
One reacting
|
160 (20.0%)
|
73 (4.1%)
|
233 (9.1%)
|
|
None reacting
|
77 (9.6%)
|
12 (0.7%)
|
89 (3.5%)
|
|
GCS at admission
|
7.55 ± 3.01
|
12.54 ± 2.82
|
10.98 ± 3.69
|
< .001
|
≤8
|
523 (65.2%)
|
208 (11.8%)
|
731 (28.5%)
|
< .001
|
9-12
|
226 (28.2%)
|
463 (26.2%)
|
689 (26.8%)
|
|
13-15
|
53 (6.6%)
|
1097 (62.0%)
|
1150 (44.7%)
|
|
CT findings
|
|
|
|
|
IPH or brain contusion
|
687 (85.7%)
|
1117 (63.2%)
|
1804 (60.2%)
|
< .001
|
SDH
|
361 (45.0%)
|
427 (24.2%)
|
788 (30.7%)
|
< .001
|
EDH
|
173 (21.6%)
|
521 (29.5%)
|
694 (27.0%)
|
< .001
|
DAI
|
50 (6.2%)
|
20 (1.2%)
|
71 (2.8%)
|
< .001
|
Multiple body injures
|
70 (8.7%)
|
77 (4.4%)
|
147 (5.7%)
|
< .001
|
Face
|
3 (0.4%)
|
2 (0.7%)
|
15 (0.6%)
|
.509
|
Thorax
|
57 (7.1%)
|
46 (2.6%)
|
103 (4.0%)
|
< .001
|
Abdomen
|
6 (0.7%)
|
4 (0.2%)
|
10 (0.4%)
|
.014
|
Extremities
|
16 (2%)
|
21 (1.2%)
|
37 (1.4%)
|
.158
|
External
|
3 (0.4%)
|
1 (0.1%)
|
4 (0.2%)
|
.176
|
ISSb
|
21.5 ± 6.4
|
12.57 ± 5.73
|
15.35 ± 7.24
|
< .001
|
INR
|
1.10 ± 0.16
|
1.06 ± 0.13
|
1.07 ± 0.14
|
< .001
|
PT(s)
|
12.59 ± 1.92
|
12.11 ± 1.57
|
12.26 ± 1.70
|
< .001
|
APTT(s)
|
28.50 ±9.62
|
27.19 ± 7.26
|
27.60 ± 8.09
|
< .001
|
FIB(g/L)
|
2.45 ± 1.39
|
2.58 ± 1.01
|
2.54 ± 1.14
|
.016
|
D-dimer (mg/L)
|
10.97 ± 15.44
|
5.46 ± 7.89
|
7.18 ± 11.12
|
< .001
|
PLT(×109/L)
|
165 ± 70
|
181 ± 65
|
176 ± 67
|
< .001
|
Hb(g/L)
|
125 ± 24
|
133 ± 20
|
131 ± 21
|
< .001
|
HCT(%)
|
36.83 ± 6.62
|
38.83 ± 5.29
|
38.21 ± 5.82
|
< .001
|
Abbreviations: GCS, Glasgow Coma Scale; ISS, Injury Severity Score; PLT, platelet; INR, international normalized ratio; PT, prothrombin; APTT, activated partial thromboplastin time; FIB, fibrinogen; EDH, epidural hematoma; SDH, subdural hematoma; IPH, intraparenchymal hemorrhage; DAI, diffuse axonal injury.
a P values not adjusted for multiple comparisons. Statistical significance determined by chi-square test, Student’s t-test or Mann-Whitney tests.
b ISS=AIS12+AIS22+AIS32, it is the sum of squares for the highest values in each of the three most severely injured body regions.
The shape of the relationship between fibrinogen level and clinical outcomes was examined by univariate analyses with linear and cubic spline functions. The relationship between fibrinogen level and the probability of favorable outcome and mortality was curvilinear, and the fitting degree of cubic spline functions was higher than that of a linear function. According to the curve correlation between fibrinogen levels and the probability of mortality, we found that when fibrinogen levels were <2.0 g/L, mortality was inversely correlated with those levels. However, when they were >2.0 g/L, the association between mortality and fibrinogen level was lost. Therefore, fibrinogen levels were divided into two subgroups of <2.0 g/L and ≥2.0g/L (Fig. 1). For patients with fibrinogen levels < 2.0 g/L, the levels were an independent prognostic factor of 3-month mortality in multivariate analyses (odds ratio [OR], 0.91; 95% confidence interval [CI], 0.89–0.93; P < .001). Age, GCS, pupillary reactions, EDH, and INR were also independent prognostic factors for mortality in this group. The risk of death in patients with fibrinogen levels < 2.0 g/L was 0.622 times higher (95% CI, 0.439–0.882; P = .008) than that in patients with fibrinogen levels ≥ 2.0 g/L (Table 3).
Table 3. Correlation between fibrinogen levels and mortality
3-month mortality for patients with fibrinogen level <2.0g/L
|
Predictors
|
OR (95% CI)
|
P Value a
|
Age
|
1.035 (1.019-1.051)
|
< .001
|
GCS
|
0.811(0.739-0.891)
|
< .001
|
Pupillary reactions
|
2.240 (1.467-3.421)
|
< .001
|
EDH
|
0.280 (0.149-0.526)
|
< .001
|
INR
|
4.316 (1.556-11.975)
|
.005
|
FIB
|
0.522 (0.273-0.995)
|
.048
|
3-month mortality for patients with fibrinogen level < 2.0 g/L vs. ≥2.0 g/L groups
Predictors
|
OR (95% CI)
|
P Value a
|
AGE
|
1.035 (1.024-1.047)
|
< .001
|
GCS
|
0.760 (0.716-0.807)
|
< .001
|
Pupillary reactions
|
1.874 (1.404-2.500)
|
< .001
|
EDH
|
0.399 (0.253-0.630)
|
< .001
|
PT
|
1.215 (1.125-1.312)
|
< .001
|
FIBGb
|
0.622 (0.439-0.882)
|
.008
|
Abbreviations: GCS, Glasgow Coma Scale; INR, international normalized ratio; PT, prothrombin; FIB, fibrinogen; EDH, epidural hematoma; SDH, subdural hematoma.
a P values adjusted for multiple comparisons. Statistical significance determined by Multiple logistic regression analysis.
b FIBG means patients with fibrinogen level < 2.0 g/L or ≥2.0 g/L groups.
Further analyses of the correlation between fibrinogen levels and the probability of favorable outcomes showed that when the fibrinogen levels were <2.5 g/L, the likelihood of favorable outcomes increased in association with fibrinogen level. This positive association was not seen in patients with fibrinogen levels between 2.5 and 3.0 g/L. Further elevation of fibrinogen to levels > 3.0g/L revealed a decrease in favorable outcomes in association with increases in the levels (Fig. 2). For patients with fibrinogen levels < 2.5g/L, the levels were an independent prognostic factor for 3-month favorable outcomes (OR, 1.654; 95% CI, 1.186–2.306; P = .003). Age, GCS score, D-dimer, SDH, EDH, IPH and brain contusion, and thorax injury were also independent prognostic factors for favorable outcomes in this group. Similarly, for patients with fibrinogen levels > 3.0 g/L, multivariate analyses showed that the levels were also an independent prognostic factor for 3-month favorable outcomes (OR, 0.771; 95% CI, 0.607–0.979; P = 0.033), with GCS, D-dimer, IPH, and brain contusions also independent prognostic factors for favorable outcomes in this group (Table 4).
Table 4. Correlation between fibrinogen levels and the probability of favorable outcomes.
3-month favorable outcome for patients with fibrinogen level <2.5g/L
|
Predictors
|
OR (95% CI)
|
P Value a
|
AGE
|
0.965 (0.955-0.975)
|
< .001
|
GCS
|
1.624 (1.539-1.714)
|
< .001
|
SDH
|
0.601 (0.440-0.821)
|
.001
|
EDH
|
1.615 (1.124-2.321)
|
.010
|
IPH or brain contusion
|
0.448 (0.302-0.665)
|
< .001
|
D-dimer
|
0.975 (0.964-0.986)
|
< .001
|
FIB
|
1.654 (1.186-2.306)
|
.003
|
Thorax injury
|
0.509 (0.277-0.934)
|
.029
|
3-month favorable outcome for patients with fibrinogen level >3.0 g/L.
Predictors
|
OR (95% CI)
|
P Value a
|
GCS
|
1.555 (1.450-1.668)
|
< .001
|
IPH or brain contusion
|
0.446 (0.246-0.808)
|
.008
|
D-dimer
|
0.932 (0.899-0.966)
|
< .001
|
FIB
|
0.771 (0.607-0.979)
|
.033
|
Abbreviations: GCS, Glasgow Coma Scale; FIB, fibrinogen; EDH, epidural hematoma; SDH, subdural hematoma; IPH, intraparenchymal hemorrhage.
a P values adjusted for multiple comparisons. Statistical significance determined by Multiple logistic regression analysis.