1. Standards
   Class I data are insufficient to support a treatment standard for this topic.
2. Guidelines
   Class II data are insufficient to support a treatment standard for this topic.
3. Options
   
   1. Class III data support the assertion that civilian regions having organized trauma care systems have better outcomes. This, combined with Class III data from military studies, would suggest that continuing to improve on the military's existing organized trauma care system is appropriate.
   2. Class III civilian data supports the recommendation that patients with GCS score 9-13 should be transported to a trauma center for evaluation.
   3. Patients with Glasgow Coma Scale (GCS) score 14 should not return to duty until disorientation resolves. GCS data obtained in the hyperacute setting, particularly concerning decisions for expectant management, should be used cautiously as it may overestimate the severity of intracranial injury. Pupillary examination may have limited usefulness due to the frequency of blast injury and the potential for traumatic iridoplegia resulting in fixed, dilated pupils which are not indicative of severe brain injury. Both GCS score and pupillary examination should be obtained, documented and repeated throughout the transport as frequently as is practical in order to follow and report the patient's clinical course.

Triage and transport decisions in combat scenarios may be complicated by tactical conditions. The essence of the decision making process involves making an assessment and categorizing patient status as return to duty, requires evacuation, or not likely to survive. In addition, those being evacuation must be further categorized as to the level of care required.

It is important to make a neurologic assessment and determine whether the patient has a nonsurvivable injury. It must be kept in mind that horrific appearing injuries involving the face and head may be survivable. If damage is limited to a single hemisphere of the brain, a tremendous amount of brain loss - coupled with massive facial tissue loss and scalp bleeding - may have the inaccurate initial appearance of being nonsurvivable. Patient responsiveness should be carefully assessed in a serial fashion. Any purposeful or repetitive movement, any ability to follow commands, and the presence of smooth, spontaneous respirations can be indicators of a survivable injury. Alternatively, absence of spontaneous respirations and/or heartbeat is a uniformly poor prognostic indicator. Massive bilateral skull and brain tissue loss are not survivable. Here again it should be emphasized that bilateral fixed, dilated pupils may be the result of direct trauma to the globe from blast injury or blunt trauma and no assessment should rely solely on pupillary examination. Overall neurologic status should be determined from the best reproducible segment of the neurologic examination: motor, verbal, eye-opening, or pupillary examination.

Return to duty decisions must be made based on a combination of medical and tactical factors. A head injury which is potentially life-threatening for the patient or which affects the patient's ability to make appropriate life-and-death decisions will demand that the medic, in most cases, recommend removal from active engagement. In instances that are not clear-cut, the chain of command can be used to assist in the decision. Patients with GCS ≤ 14 should not return to duty until they are oriented to person, place, time, and situation.

From a neurologic standpoint, the decision to evacuate must be made based upon the immediate condition of the patient and the likelihood for short-term improvement, the threat that the injury poses to the patient, the threat that the patient may pose to the unit or mission, and the availability of evacuation assets. Patients with GCS 3-8 should be evacuated to a facility with neurosurgical capability, potentially bypassing a closer facility in order to insure the level of care necessary is made available in the most expeditious fashion. Finally, the urgency of evacuation must be considered. The possible danger to evacuation personnel, vehicles, and/or aircraft must be weighed against the immediate needs of the patient. It is incumbent upon the medical provider to base recommendations on the medical needs of the patient first, and the chains of command of both the tactical unit and evacuation unit will determine whether the evacuation asset is dispatched. However, the requesting medic must be aware of the implications of making a recommendation for a priority evacuation from a hot landing zone and carefully consider whether the patient can be stabilized on-site without increased morbidity or mortality.

In tactical environments where explosions are common, it should be kept in mind that the GCS score may be artificially low for a period of time due to the patient being rendered unconscious from a blast. Additionally, pupillary examination may demonstrate fixed, dilated pupils which are due to globe trauma and not brain injury. Dehydration, combat stress, and traumatic brain injury (TBI) all may result in global neurologic dysfunction. In mild cases of disorientation the decision to evacuate should be made based upon serial examinations over whatever time is available. If the neurologic status is deteriorating, the decision to evacuate becomes clear. If the patient rapidly improves to normal, there may be an opportunity to return to duty. Appropriate frequency of reexamination has not been established.

A MEDLINE search without date limits was performed using combinations of the keywords "combat," "triage," "evacuation," "tactical," "casualty," and "head injury." The 286 articles listed were reviewed in abstract form and 25 were selected for full review. None contained pertinent data specifically related to the evacuation and triage of neurologically injured patients. A MEDLINE search from 1970 to 1999 using the keywords "trauma systems," "trauma centers," "emergency medical services," "prehospital care," and "ambulance transports" identified 147 articles. Careful review and analysis of all 147 articles permitted an assessment of trauma systems and the role of EMS in managing patients with severe TBI.

Since the late 1970s, several investigators have tried to demonstrate the efficacy of EMS systems and trauma systems. Studies performed in the late 1970s and early 1980s attempted to show that excessive "preventable" trauma deaths occurred in regions without organized EMS or trauma care.1 The investigators' methodology relied on physician panels who reviewed patient care case by case and then used various consensus methods to determine the appropriateness of the treatment. This technique has been criticized as being too subjective because blinding of the panel participants to the treatment setting is often extremely difficult and the various means used to reach consensus produce different results.2 Later studies relied on series of patients treated at one or more trauma centers and compared them with those patients treated in a non-trauma center within a region3 or across the United States,4 using prospectively collected, standardized data on severity and outcome. In all comparisons between organized and nonorganized EMS and trauma systems, patient outcome was worse without organization.3,5 A number of studies and their methodologies have been summarized in publications.4,6 To deliver the best possible trauma care, it is crucial that trauma victims first receive competent on-scene prehospital EMS care before being removed directly to a hospital. In addition, because victims of severe trauma usually have a life-threatening condition, the receiving hospital must be sufficiently equipped and qualified to take care of their injuries.

Recent literature suggests that the outcome of trauma patients clearly improve when prehospital care, triage, and admission to designated trauma centers are coordinated within regional trauma systems. It should be noted, however, that nearly all of these studies refer to the general trauma patient, and only a few primarily address the patient with TBI. There are no published data suggesting that the lack of a trauma care system is superior to organized systems. There is a retrospective study that compared head trauma outcome before and after the implementation of a trauma system in Oregon, which reported that an odds ratio of 0.80 for mortality after system implementation.7

A report of preventable deaths in San Diego County compared non-TBI and TBI deaths before and after instituting a regional trauma care system.8 Reviewers were blinded to the facility where care was rendered. Preventable deaths for non-TBI cases decreased from 16/83 (20%) to 2/211 (1%) (p < 0.005), and for TBI cases, preventable deaths decreased from 4/94 (5%) to 1/149 (0.7%) (p < 0.10), respectively, before and after the trauma system was put in place.

Another before and after study compared outcome of injured patients in a rural hospital before it chose to meet American College of Surgeons Committee on Trauma guidelines for a level II trauma center with outcome after it became a level II trauma center.9 Survival for all patients who had a calculated probability of survival of 25% was 13% before and 30% after meeting trauma center criteria. For patients with closed head trauma, the survival was 15.4% before and 32% after meeting the criteria.

Several articles studied the EMS system's impact within the overall trauma system. One study in New Delhi, India and in Charlottesville, Virginia, compared mortality rates after head injury using the motor score portion of the GCS to stratify patients.10 While outcome was not statistically different in those patients with the lowest motor scores, mortality in patients with a motor score of 5 was notably different. Patients in Charlottesville had a mortality of 4.8%, whereas those in New Delhi had a mortality of 12.5% (p = 0.001). The authors postulated that one reason for this difference may be that only 0.5% of patients in New Delhi arrived to the hospital by ambulance, versus 84% in Charlottesville. In addition, only 7% of patients in New Delhi arrived at the hospital within 1 hour and an additional 33% in 2-3 hours, compared with 50% within 1 hour and an additional 39% within 3 hours in Virginia. Thus, the lack of an EMS system and delay in presentation were thought to be important factors that account for the difference in outcome between the two cities.

The second study compared trauma patients with an injury severity score (ISS) of 9 or more in Seattle and Monterrey, Mexico.11 Patients were taken to an urban hospital in Monterrey and to a level I trauma center in Seattle. Overall mortality was 55% in Monterrey and 34% in Seattle (p = 0.001). Deaths in Monterrey occurred in the field (40%) and in the ED (11%) compared with Seattle where 21% died in the field and 6% in the ED (p = 0.001 and 0.003, respectively). In addition, at hospital arrival, 39% of patients in Monterrey had a systolic blood pressure less than or equal to 80 mm Hg compared with 18% (p = 0.001) in Seattle. Of those patients who were hypotensive, 5% in Monterrey and 79% in Seattle underwent endotracheal intubation in the field (p = 0.001) and 70% in Monterrey and 99% in Seattle had fluid resuscitation en route (p = 0.001).

The need for the in-house presence of the trauma surgeon 24 hours a day versus the ability of the trauma surgeon to respond quickly to the hospital has generated significant controversy. A report from one level II trauma center in Oklahoma concluded that level II trauma centers with attending trauma surgeons who are available but not "in-house" have outcomes as good as those with surgeons present in the hospital at all times.12 This study was performed internally comparing daytime hours when the attending trauma surgeon was in-hospital versus evening and night hours when call was taken from outside. Using survival as predicted by the Major Trauma Outcome Study, this study evaluated 3,689 patients with major trauma. Overall survival was 97% with a predicted survival of 96%. Subgroup analysis revealed that, for patients with a trauma score < 12, predicted survival and actual survival was 84%. In comparing whether the trauma surgeon was present, patients with severe thoracoabdominal trauma had a predicted survival of 79% and actual survival of 77% when the surgeon was in-house and a predicted and actual survival of 74% and 81% when the surgeon was called in from outside. In addition, patients with head trauma had predicted survival of 61% and actual of 63% when the surgeon was immediately available, and 57% predicted and 63% actual when the surgeon came in from home. All p-values were described as nonsignificant. Whether or not the trauma surgeon takes call from home, the important point in delivering trauma care to the patient is the physical presence of an appropriate team at the time of patient arrival in the ED.

Another issue that has also resulted in significant controversy relates to experience and patient volume criteria. Using data collected by trauma nurse coordinators, a retrospective studyevaluating volume measurements on patient outcome compared trauma centers in Chicago. The trauma centers treating larger volumes of trauma patients were found to have better patient outcomes than those with fewer admissions. Patients transported to low volume centers had a 30% greater chance of death when compared with high-volume centers.13 However, a recent report questions the impact of case volume on patient outcome. Richardson et al.14 evaluated mortality and morbidity outcomes, such as length of stay of trauma patients by case volume per attending surgeon. They found no difference based upon annual case volume or years of experience. While the optimal number of cases per trauma center and per trauma surgeon may be debated, the individual physicians on the treating team must have adequate experience to make the complex decisions often required when caring for a patient with severe multisystem or brain injury.

Another study that evaluated 1,332 patients with femoral fractures who underwent operative repair compared outcome in terms of morbidity and mortality between trauma centers and non-trauma centers.2 Morbidity was 21% in the trauma centers and 33% in the nontrauma centers (p = 0.001), and mortality was 1.0% versus 2.2% respectively.

Several studies from Quebec demonstrated similar results. Mortality for all trauma patients before implementation of a trauma system was 20%, but only 10% after the system was put in place.15 A subsequent review of trauma care in Quebec compared the outcome of 2,756 trauma patients transported directly to a trauma center with 1,608 patients who first were treated at a local hospital and subsequently transferred to the trauma center.16 Mortality was 4.8% for patients taken directly to the trauma center and 8.9% if transfer occurred later (p = 0.001).

These findings apply to both adults and children transported by EMS systems directly from the scene to trauma centers. For example, in a study of 1,320 children of whom 98 sustained severe head injuries, mortality for the children brought directly from the accident scene to a pediatric trauma center was 27%. However, children transported first to the nearest available hospital and subsequently transferred to the trauma center had a mortality of 50%.17

A number of studies attempted to evaluate the differences and difficulties associated with providing trauma care in rural settings compared with urban settings that have integrated trauma systems. Rogers et al.18 reviewed trauma deaths in an organized urban trauma system compared with a rural state without a formal trauma system. The authors suggest that the higher incidence of prehospital deaths may be related to delays in discovering the patient and the longer response and transport times required in the rural setting, particularly for interhospital transfers.

Young et al.19 compared the outcome of patients with an ISS > 15 who were transported directly to their level I trauma center with those who were first taken to another rural hospital and subsequently transferred. Outcome measures included mortality, total hospital days, and ICU days. When all patients were included the two groups did not differ. However, when patients who died within the first 24 hours were excluded, length of stay, both in the hospital and in the ICU, was significantly longer (p < 0.05) in the group transferred from another hospital, although there was no difference in mortality. The GCS of the patients who died within the first 24 hours should, however, be noted. The GCS for the patients taken directly to the trauma center was 5, compared with 10 for those patients transferred from an outside hospital (p < 0.05). In addition, of patients who died in the first 24 hours (probability of survival > 0.50), the observed mortality for the direct transport group was 28% (7/25) compared with 75% (12/16) in the transferred group (p < 0.05). The authors stated that although these differences were noted between the groups, the study did not identify specific subgroups that would clearly benefit from direct transport to the trauma center. However, they did recommend that whenever possible patients with major trauma should be transported from the scene directly to a trauma center.

As noted in the section on Glasgow Coma Scale, a significant percentage of patients with hospital GCS scores 9-13 have serious intracranial injury requiring neurosurgical intervention and poor outcome, but no studies were found that compared outcomes based upon choice of destination.

Severe TBI patients transported to trauma centers without prompt neurosurgical care or intracranial pressure monitoring are at risk for a poor outcome. Acute subdural hematomas in severe TBI patients are associated with 90% mortality if evaluated more than 4 hours after injury and only 30% mortality if evaluated earlier.20 If subdural evaluation is done in less than 2 hours after injury, one study reported a 70% decrease in mortality.21 To achieve this surgical timing, 24-hour availability of CT scanning is necessary. Intracranial pressure monitoring guides specific treatment to maintain cerebral perfusion and is recommended based on supporting scientific evidence for improved patient outcome given in the Guidelines for the Management of Severe Head Injury.22

A recent study of 4014 patients involved in motor vehicle collisions reported that a GCS ≤ 14 predicted the need for hospital admission after arrival at a trauma center.23 Hospital admission rates were 96% for GCS ≤ 12, 73% for GCS 13-14 and 32% for GCS 15. The authors concluded that activation of the trauma system should be strongly considered for GCS ≤ 14. However, since the resources available in combat differ, these Guidelines recommend holding patients with GCS 14 in the field for observation.

The combat management of the acutely head injured patient is complicated by tactical, logistical, and medical considerations. Ideally, this "fog of war" would clear, allowing the combat medic the luxury of being able to provide the best available care based on civilian standards practiced in the U.S. on a sunny day with no distractors. Unfortunately, this is likely to be the exception in combat, and the medics must be given the tools, training, and confidence to be able to provide optimal care under these most demanding of circumstances to the most deserving patients in the world.

Future investigations should focus on rapid evaluation of the neurologically injured patient. Examination algorithms which are rapidly administered, reliable, and feasible in a combat environment are essential. Diagnostic tools or devices that are accurate, lightweight, rugged enough for combat use, and simple to use under tactical conditions should be developed.