Demetriades D, Kuncir E, Velmahos GC, Rhee P, Alo K, Chan LS. Outcome and Prognostic Factors in Head Injuries With an Admission Glasgow Coma Scale Score of 3. Arch Surg. 2004;139(10):1066-1068. doi:10.1001/archsurg.139.10.1066
Copyright 2004 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2004
To identify significant risk factors associated with mortality in patients with a Glasgow Coma Scale score of 3.
Trauma registry study.
Level I urban trauma center.
A total of 760 patients with head injury with an admission Glasgow Coma Scale score of 3. Analysis was performed in all patients and in only patients who reached the hospital alive and had no major extracranial injuries (exclusion of patients with a chest or abdominal Abbreviated Injury Score [AIS] >3).
Main Outcome Measures
Stepwise logistic regression analysis was used to identify independent risk factors associated with mortality.
Blunt trauma accounted for 477 (63%) and penetrating trauma for 283 (37%) of the 760 head injuries. Penetrating trauma was significantly more likely to be associated with a lack of vital signs on admission (15% vs 9%; P = .03). Overall mortality was 76% (94% for penetrating injuries and 65% for blunt injuries; P<.001). Overall, 79% of patients had a head AIS of 4 or greater. Mortality in the subgroup was 64% (320/497) and was significantly higher in penetrating vs blunt trauma (89% vs 52%; P<.001). Penetrating trauma, high head AIS, hypotension on admission, and age older than 55 years were independent significant risk factors associated with mortality. Only 10% of the 177 survivors had good functional outcome at hospital discharge. Eighty-six patients (17% of those with vital signs on admission) became organ donors.
Patients with head injury with an admission Glasgow Coma Scale score of 3 have a poor prognosis. Mechanism of injury, head AIS, hypotension on admission, and age play a critical role in outcome. These patients are an important source of organ donation and should be evaluated and resuscitated aggressively.
Patients with head injury with low Glasgow Coma Scale (GCS) scores on hospital admission have a poor prognosis. A GCS score of 3 is the lowest possible score and is associated with an extremely high mortality rate, with some researchers suggesting that there is no chance of survival. The aims of this study are to evaluate mortality in a large group of patients with head injury with admission GCS scores of 3 and to identify significant factors that are associated with outcome.
This trauma registry–based study included all patients with head injuries and admission GCS scores of 3 admitted to the Los Angeles County and University of Southern California Trauma Center during a 10-year period (January 1, 1993, to December 31, 2002). Los Angeles paramedics are not allowed to administer narcotics, sedatives, or paralytic agents to trauma patients. The trauma registry has 130 data fields and is maintained by a trained team of 7 full-time nurses. The following data were retrieved from the trauma registry for this study: age, race, sex, mechanism of injury, hospital admission vital signs, Abbreviated Injury Score (AIS) (1990 version) of each body area, Injury Severity Score, prehospital airway management, survival status, functional results on hospital discharge (preinjury capacity, temporary disability, or permanent disability), and intensive care unit and hospital stay. The head AIS was calculated on the basis of computed tomographic, operative, or autopsy findings.
Analysis was performed in 2 study populations: (1) the whole population of patients with GCS scores of 3 (population A; N = 760) and (2) the subgroup of patients who were alive on hospital admission and had no other major extracranial injuries (population B; n = 497). For each study population, bivariate analysis was first performed to examine the relationship of age, sex, mechanism of injury, Injury Severity Score, head AIS, and admission systolic blood pressure with mortality (χ2 test). Stepwise logistic regression analysis was then used to identify the independent risk factors associated with mortality. The effects of the risk factors on mortality were measured using adjusted odds ratios and 95% confidence intervals. Statistical software was used for the analysis (SAS release 8.2; SAS Inc, Cary, NC).
During the 10-year study, there were 760 patients with head injury with an admission GCS score of 3. Blunt trauma accounted for 63% of cases (n = 477) and penetrating trauma for 37% (n = 283). The mean ± SD age of patients with blunt trauma was 39 ± 19 years and of those with penetrating trauma was 29 ± 14 years (P<.001). The mean ± SD overall Injury Severity Score was 27 ± 15 (26 ± 14 in those with blunt trauma and 27 ± 17 in those with penetrating trauma; P = .18). Penetrating trauma was significantly more likely than blunt trauma to be associated with a head AIS of 4 or more (83% vs 70%; P<.001) and no vital signs on hospital admission (15% vs 9%; P = .03). Overall mortality was 76% (577 patients), and it was significantly higher in penetrating injuries (94% vs 65%; P<.001). The odds ratio of death in penetrating trauma was 7.8 (95% confidence interval, 4.6-13.5; P<.001). Mortality correlated with head AIS in blunt trauma (P<.001) but remained uniformly high in all head AIS categories in penetrating trauma (P = .65) (Table 1).
Similar analyses were performed for a subset of population A that excludes patients with no vital signs on hospital admission and patients with major extracranial injuries (chest or abdomen AIS >3). This subpopulation included 497 patients (65% of all patients). The epidemiologic and injury characteristics of this group are given in Table 2. Penetrating trauma dominated in patients younger than 30 years, and blunt trauma dominated in the older age groups (Table 2). Male sex was associated with a significantly higher incidence of penetrating trauma than female sex (37% vs 18%; P = .001).
The mean ± SD Injury Severity Score was 23 ± 11, and there was no significant difference between blunt and penetrating trauma (22 ± 11 vs 24 ± 11; P = .92). Overall, 50 patients (11%) were hypotensive on hospital admission (systolic blood pressure, <90 mm Hg), and there was no significant difference between blunt and penetrating trauma (10% vs 12%; P = .37). Overall, 79% of patients had a head AIS of 4 or more. It was significantly more likely that patients with a GCS score of 3 had a head AIS of 4 or more in penetrating vs blunt trauma (92% vs 69%; P<.001).
Overall mortality in population B was 64% (320/497). Mortality was significantly higher in penetrating vs blunt trauma (89% vs 52%; P<.001). The odds ratio of death in penetrating trauma was 7.8 (95% confidence interval, 4.4-13.8; P<.001). The mortality rate is uniformly high for all head AIS categories in penetrating trauma, and it increases with increased head AIS in blunt trauma, the same findings as in population A (Table 3).
Stepwise logistic regression analysis, with mechanism of injury, age, head AIS, and hypotension as risk factors, identified penetrating trauma, high head AIS, hypotension on hospital admission, and age older than 55 years as significant risk factors for mortality (Table 4).
Overall, 28 (16%) of 177 survivors in population B were discharged to a rehabilitation facility with the classification of permanent disability. This was similar in both mechanisms of injury (blunt trauma: 16% [26/160]; penetrating trauma: 12% [2/17]; P = .69). Ten percent (18/177) of the survivors were classified at discharge as having good functional outcome with preinjury functional capacity (blunt trauma: 10% [16/160]; penetrating trauma: 12% [2/17]; P>.99). No long-term follow-up was available for functional outcome evaluation.
The mean ± SD overall hospital stay was 7.4 ± 10.0 days, including 3.6 ± 5.8 intensive care unit days. The hospital stay according to mechanism of injury and survival status is given in Table 5.
A GCS score of 3 due to head trauma is associated with a high mortality rate, and some researchers1 suggested that there are no chances of survival. Predicting which of these patients are likely to survive or die may be useful in the allocation of limited hospital resources or in the facilitation of possible organ donation. There are few studies investigating this subject, and the existing studies include only a few patients.1- 3 Kotwica and Jakubowski1 reported the only study specifically addressing patients with head trauma with a GCS score of 3. In a review of 111 patients with blunt head trauma with a GCS score of 3 and excluding those with major extracranial injuries, the authors reported a mortality rate of 89% within the first 30 days, 7% survival in a vegetative state, and 4% survival with satisfactory results. In a study of 190 patients with penetrating craniocerebral injuries with GCS scores of 3, 4, or 5, Levy et al2 reported that 106 had a GCS score of 3. Overall mortality was 95%. All 5 survivors in the study had severe or moderate disability.
The present study identified 4 significant risk factors for death in patients with a GCS score of 3: penetrating trauma, high head AIS, age older than 55 years, and hypotension on hospital admission. To our knowledge, no study has attempted to compare blunt with penetrating trauma in head injuries with GCS scores of 3. The results of this study show that the mechanism of injury is critical in determining outcome. Overall mortality in blunt trauma was 65% and in penetrating trauma was 94%. When patients with no vital signs on admission and those with severe extracranial trauma were excluded, mortality was 52% and 89%, respectively. A possible explanation for this statistically significant difference might be the higher head AIS in penetrating trauma. Many patients with blunt trauma might have a diffuse axonal injury and a fairly low head AIS, a condition that is often associated with a better outcome than cerebral tissue destruction or hemorrhage, which occurs in penetrating trauma. However, as seen in Table 3, even after adjusting for head AIS, the mortality rate in penetrating trauma was dramatically higher than in blunt trauma. Another interesting finding is the low incidence of hypotension in patients with severe head injuries, even in the absence of other major extracranial injuries. Only 11% of patients who reached the hospital alive had hypotension on admission. The presence of hypotension is a statistically significant risk factor for mortality.
Despite the poor outcome and the financial costs of caring for this group of patients, every effort should be made to provide care similar to that in survivable patients. Besides the few survivors with good functional results, organ donation is another major benefit of intensive care of these hopeless patients. Optimization of the physiologic status reduces cardiac arrest before organ donation and preserves the organs in good functional condition.
Correspondence: Demetrios Demetriades, MD, PhD, Los Angeles County and University of Southern California Healthcare Network, 1200 N State St, Room 1105, Los Angeles, CA 90033 (firstname.lastname@example.org).
Accepted for publication January 26, 2004.