Trauma, 7th Ed.

CHAPTER 19. Injury to the Brain

Alexander F. Post, Thomas Boro, and James M. Ecklund


Traumatic brain injury (TBI) is a disruption or alteration of brain function due to external forces. The disruption of function may be transient or long lasting and may vary in severity. The external forces creating the injury may be the result of a variety of insults including acceleration or deceleration, direct compression, penetrating objects, combined effects, and complex mechanisms such as in blast. It may produce subtle effects not discernible on radiological imaging, focal injuries such as fractures, contusion, subarachnoid hemorrhage (SAH), subdural hemorrhage (SDH), epidural hemorrhage (EDH), or intraparenchymal hemorrhage (IPH), or more widespread damage such as diffuse axonal injury (DAI). All injuries and symptoms, even if apparently minor on initial presentation, should be taken seriously since injuries may rapidly progress and become life-threatening.


The exact number of people suffering TBI is unknown since many individuals suffering mild or moderate TBI do not seek medical attention, and some who suffer severe traumatic injuries do not survive to receive medical attention.

Of those who do receive medical attention in an emergency department, approximately 1.4 million people per year suffer TBI. Of these patients, approximately 1.1 million are treated and released, 240,000 are hospitalized, and 50,000 die.1

Common causes for TBI are falls (28%), motor vehicle accidents (20%), pedestrian impact (19%), and assault (11%). TBI has a bimodal age distribution with the greatest risk in 0–4 and 15- to 19-year-olds. Males have 1.5 times the risk of females. The younger group is often the victim of abuse and cannot protect itself. The older group practices greater risk-taking behavior, and includes the population of new drivers and teenagers exposed to drugs and alcohol. Military personnel comprise a statistically small number of the overall TBI injuries per year, but have a higher incidence of penetrating and blast injuries resulting from combat operations.


TBI is a dynamic process and management must be tailored throughout the patient’s course. Primary injuries of the brain result from the forces imparted at the time of the accident. This includes disruption of scalp (lacerations), bone (cranial vault, skull base, facial bones), vasculature (SDH/EDH/IPH/intraventricular hemorrhage [IVH], traumatic aneurysm), or brain parenchyma (contusion, DAI).

Secondary injuries occur after the initial impact and may be more insidious and more difficult to control. They are often due to failure of autoregulation and loss of normal homeostasis. These injuries include hypoxemia, ischemia, initial hyperemia, cerebral edema, and expansion of hemorrhages leading to increased intracranial pressure (ICP), seizures, metabolic abnormalities, and systemic insults.


image Systemic Evaluation and Resuscitation

Assessment and treatment of head-injured patients often begins in the prehospital setting with family, bystanders, and off-duty medical personnel. Care continues with the primary care physician or emergency medical technician (EMT), transfers to the physician in the emergency department, and eventually involves the trauma team, neurologist, neurosurgeon, and neurointensivist. Treatments may be started at any point along the patient’s journey based on recognition of neurological signs and availability of appropriate medication, equipment, and personnel.

The nervous system does not exist in a vacuum and patients with TBI may have additional injuries. Treatments specific to TBI are often complementary or adjunctive to the treatment of the trauma patient without neurological injury. The basic principles of trauma resuscitation should be adhered to and include rapid assessment and maintenance of an airway, breathing, and circulation.

A detailed medical and surgical history should be obtained including the events preceding a trauma, a description of the accident scene, accurate description of the patient’s neurological baseline, and any subsequent changes to the neurological status. Chronic medications and medications given in the prehospital setting should be determined. Special attention should be paid to medications with the ability to alter the neurological examination including sedatives or psychopharmacologics (to restrain the altered patient), paralytics (for intubation or transportation), atropine (for cardiac resuscitation), and other mydriatics (for evaluation of ocular trauma).

Primary and secondary surveys should evaluate for systemic injuries including discrete injuries to the head and cervical spine. Open lacerations and vigorous scalp hemorrhage may lead to hypovolemia. In the newborn or premature infant, cephalohematoma may allow enough displacement of blood to produce hemodynamic instability. Raccoon’s eyes (periorbital ecchymosis), Battle’s sign (postauricular ecchymosis), and otorrhea/rhinorrhea suggest a basilar skull fracture. Palpable fractures or depressions may indicate bony injury with a higher likelihood of underlying hemorrhage or parenchymal injury. Periorbital edema or proptosis may suggest local ocular or orbital trauma. Puncture wounds may indicate a more serious, penetrating injury to the brain, spinal cord, sympathetic plexus, or vasculature. Bruits of the carotid artery or globe of the eye may represent carotid dissection or carotid-cavernous fistula, respectively. Multiple areas of swelling or bruising may indicate prior seizure activity.

image Neurological Examination

An accurate neurological examination is essential to determine diagnosis, treatment strategies, and prognosis in TBI patients. The exam may be limited due to the patient’s age, level of education, native language, presence of sedative or paralytic medication, illicit drugs, hypotension, hypoxia, hypothermia, or hypoglycemia. Examination of the pediatric patient may include further limitations due to overall neural development and degrees of myelination, inability to visualize the pupils or fundi of the premature or newborn infant, and limited cooperation.

It is critical to monitor the overall trend of the neurological examination over time. It must be understood that these examinations can and will fluctuate based on the patient’s improving or declining condition, the evolution of disease processes, and the ability of medical personnel to minimize or eliminate factors that confound an accurate neurological assessment.

In the uncooperative patient or unconscious patient with severe TBI, the exam may be limited to the Glasgow Coma Scale (GCS), pupillary reactivity, and testing of various reflex actions (Table 19-1). As the patient becomes more alert and cooperative, a more complete neurological examination will provide greater sensitivity for assessment of neurological change. The extent of the examination must be tailored to each patient’s neurological ability.

TABLE 19-1 Neurological Examination for Trauma


image Pupillary Response

The parasympathetic, pupilloconstrictor, and light reflex (pupillary reflex) can be easily and rapidly assessed in the unconscious patient. Damage to the Edinger–Westphal nucleus or uncal compression of CN III at the tentorial notch will result in pupillary dilatation (≥4 mm). If severe enough (i.e., cerebral herniation), the pupil will be fixed in this dilated position and is unresponsive to a light stimulus. Direct orbital trauma can also result in pupillary dilation/fixation in the absence of temporal lobe herniation or intracranial hypertension (IC-HTN). It should be considered (quickly) before assuming that a dilated pupil is due to brainstem compression.

In cases of orbital or periorbital trauma, facial fractures, or abnormal eye movement, ophthalmology may examine the patient and wish to instill mydriatics. To avoid confusion, this should only be permitted after concerns regarding elevated ICP have been addressed. Clear notation should be made, in the chart and at the bedside, as to when mydriatics have been used and for how long they will last.

image Glasgow Coma Scale

The GCS has become the standard for objective measurement of TBI severity.2 The patient is assessed in three parameters (best motor function [M], best verbalization [V], and best eye opening [E]) and the summation of these individual scores represents their overall GCS score (Table 19-2). A neurologically intact patient can achieve a maximum of 15 points and the most severely injured patient achieves a total of 3 points. If the patient is intubated, he or she receives a score of 1 for the verbal component and the overall scored is annotated with a “T.” For example, an intubated patient with eye opening to pain and extremity withdrawal to pain would have a GCS score of image. The same GCS score can be derived from different values of the motor, verbal, and eye components. Therefore, it is beneficial for practitioners to denote each of the subscores and not just the overall number.

TABLE 19-2 Glasgow Coma Scale (Recommended for Age ≥4)


The GCS allows practitioners to communicate quickly and reliably regarding a patient’s general condition. The postresuscitation GCS is also effective in stratifying patients into groups for definition of injury severity and overall prognosis. Patients with image are defined as having mild TBI, are usually awake, and have no focal deficits. Patients with image have moderate TBI and usually have altered sensorium and may have focal deficits. image have severe TBI, and usually will not follow commands and meet the generally accepted definition of patients in a comatose state.

As with all neurological assessments, confounding factors must be investigated and eliminated. Paralytics are used for patient transport, restraint, or cardiopulmonary stabilization and will limit the motor exam. Similarly, unrecognized spinal cord injury can produce a lower motor component score of the GCS and should be sought if suggested by the mechanism of injury. Hearing deficits, lack of hearing aids, or impairment of language function may limit the verbal component score. Finally, the pediatric population represents a special subset where modifications to the verbal component score better reflect the limited language skills of the young child3 (Table 19-3).

TABLE 19-3 Glasgow Coma Scale for Children (Recommended for Age <4)



image Plain X-Rays

Plain x-rays are most useful in the trauma setting for evaluating and clearing the cervical spine. The spine is imaged from the occiput to T1 and a C-collar is maintained until instability has been ruled out. In the C-spine AP, lateral and odontoid views are the most useful; T- and L-spine films are obtained based on mechanism of injury, degree of neurological deficits, and pain. Additionally, if CT or MRI scans are unavailable, plain x-rays can be used to determine pneumocephalus, skull fractures, and the tract of penetrating objects. It must be remembered that intracranial ricochets of projectiles may occur, and CT is the imaging modality of choice.

image CT Scan

Overwhelmingly, the CT scan has become the initial study of patients presenting after head trauma, or with a new neurological deficit. In a single, rapid pass, without patient repositioning, scans of the head, neck, chest, abdomen, and pelvis can be performed. From these, cervical, thoracic, and lumbar spine images can be reconstructed without additional radiation. Administration of contrast allows for CT angiogram reconstruction to evaluate vasculature of the head and neck and be used for diagnosis and operative planning.

CT scan findings after trauma include SDH, EDH, SAH, IPH, and IVH, contusions, hydrocephalus, cerebral edema or anoxia, skull fractures, ischemic infarction (if >12 hours old), mass effect, or midline shift.

Indications for an initial post-traumatic CT scan include image, unresponsiveness, focal deficit, amnesia for the injury, altered mental status, and signs of basilar skull fracture.

image MRI

MRI scans have better parenchymal resolution and can evaluate infarction, ischemia, edema, and DAI. MRI is also helpful to determine ligamentous injury of the spine or traumatic cord injury. It is generally performed after the initial trauma evaluation and resuscitation have been completed. MRIs have limited availability, slower image acquisition time, and increased cost. Their use in the initial assessment of trauma is not routinely recommended since intracranial surgical lesions seen on MRI are also identified on CT scan.4

image Angiography

In penetrating trauma when the tract of injury is near a known vessel distribution or when a delayed intracerebral hemorrhage occurs, angiograms are used to look for direct vessel injury or pseudoaneurysms. When CT or MRI scans are unavailable, angiograms may be used to look for mild vessel shift indicative of compressive mass lesions (e.g., SDH, EDH, IPH). More significant vessel shift may be indicative of transtentorial or subfalcine herniation and suggest the need for more rapid treatment.


image Skull Fractures

Skull fractures can be described by the state of the overlying scalp (closed or open), the number of bone fragments (simple or compound), the relationship of bone fragments to each other (depressed or nondepressed), whether the fracture enters or widens an existing cranial suture (diastatic, more common in children), and whether it involves the cranial vault or skull base. In general, lower force impacts (falls from standing) will create fractures that are more linear, closed, and without dural laceration. Higher force impacts (MVA, falls from heights, penetrating trauma) will produce compound, open fractures with a greater likelihood of underlying dural or cerebral injury. “Ping-pong” fractures are greenstick-type fractures usually seen in newborns due to the plasticity of the skull (Fig. 19-1). They show a local concavity of the skull, without sharp edges, and usually do not require intervention as the skull remodels during growth and smooths out the cosmetic deformity.5


FIGURE 19-1 CT bone windows showing ping-pong skull fracture. The multiple nondisplaced linear lucencies are normal sutures.

Associated clinical signs suggestive of calvarial skull fractures include gross deformity and palpable skull fracture in patients with open scalp lacerations. Basilar skull fractures may show postauricular or periorbital ecchymosis, hemotympanum or laceration of the external auditory canal, and CSF rhinorrhea or otorrhea. Cranial nerve injuries may be seen with fractures of the cribriform plate (CN I, anosmia), optic canal (CN II, visual deficit), and temporal bone (CN VII, facial weakness; or CN VIII, hearing loss). Severe basilar skull fractures may result in pituitary gland injury and resultant endocrinopathies. Direct injury to vasculature that penetrates the skull base may result in arterial dissection, traumatic aneurysm formation, or traumatic carotid-cavernous sinus fistula with symptoms of cranial neuropathies, chemosis, bruits, and strokes. Fractures of air sinuses or mastoid air cells may rarely present with meningitis, even years after the initial event.

Radiographic Diagnosis

Skull fractures can be discovered on isolated plain x-rays or as part of a skeletal survey for abuse. They can be differentiated from vascular grooves or normal cranial sutures by characteristics listed in Table 19-4.6

TABLE 19-4 Differential Diagnosis of Fractures on Skull X-Rays


Most skull fractures are discovered by CT scan due to its overwhelming use in the initial evaluation of trauma patients. Plain films may be superior to CT scan in discovering linear calvarial fractures parallel to the skull base (in the plane of CT slice acquisition). CT scans provide better visualization of facial and orbital fractures, temporal bone fractures, and pneumocephalus, and better evaluation of air sinuses and mastoid air cells containing air fluid levels and varying degrees of opacification. Thin cut bone windows can be reconstructed in coronal or sagittal planes or 3D surface modeling to aid in fracture identification and surgical planning. CT angiograms/venograms are useful for assessing fractures involving skull base foramen containing vasculature (e.g., carotid canal, foramen magnum) or fractures that cross major venous sinuses (superior sagittal or transverse sinuses, jugular foramen).

All fractures must be assessed and treated in concert with the underlying brain. The following discussion of skull fracture treatment assumes that the underlying brain has been evaluated for subdural or epidural hematomas, parenchymal hemorrhages, or contusions and cerebral edema and that clinical criteria do not separately mandate operative decompression.

Closed, nondisplaced fractures do not require immediate intervention. Open skull fractures should be debrided and carefully inspected and all should receive antibiotics. Those with obvious underlying dural laceration, CSF leakage, or visible brain should be surgically repaired in layers to reduce the risks of meningitis or brain herniation through a dural defect. In the pediatric population laceration of the underlying dura can rarely lead to a growing skull fracture (leptomeningeal cyst) seen in 0.05–0.6% of skull fractures.7 Pulsations of the underlying rapidly growing brain widen the dural laceration and fracture line over time (Fig. 19-2). These are most common in children under 1 year and over 90% occur in children under 3 years old.8 Surgical repair includes wide bony exposure to repair the dural edges that retract beyond the limits of the visible fracture.


FIGURE 19-2 CT revealing growing skull fracture from leptomeningeal cyst in child. Note displaced bone and expansion of CSF-filled soft tissue.

Relative indications for surgical elevation of a depressed skull fracture include depression of more than 8–10 mm or more than the thickness of the skull (Fig. 19-3), a focal neurological deficit clearly attributable to compressed underlying brain, significant intraparenchymal bone fragments (implying dural laceration), and persistent cosmetic deformity after all swelling has subsided. There is no evidence that post-traumatic seizure (PTS) risk is improved by elevation of a simple, depressed skull fracture.9 For children, the growing brain induces remodeling of the overlying skull. If there is no dural violation, there is no difference in seizure risk, neurological outcome, or cosmesis afforded by elevation of a simple depressed skull fracture.10 Fractures that cross a major dural venous sinus may warrant a more conservative approach given the increased risk of bleeding and air embolus that may be incurred during surgical repair.


FIGURE 19-3 CT bone windows showing a depressed skull fracture that required surgical elevation and dural repair. The patient also had an underlying brain contusion and presented with a receptive aphasia.

Current recommendations support surgical repair of open fractures depressed greater than the thickness of the cranium and nonoperative management of open depressed cranial fractures if there is no evidence of dural penetration. Surgery is also supported in cases of significant intracranial hematoma, depression >1 cm, frontal sinus involvement, gross cosmetic deformity, wound infection, pneumocephalus, or gross wound contamination. If there is no gross wound contamination, primary bone fragments may be replaced without excessive infection risk.11

image Focal Cerebral Injuries

Trauma patients, and their injuries, are hardly uniform. During a single traumatic event, patients may be subjected to forces of different magnitude, direction, and duration. The following injuries are presented as separate discussions for the sake of clarity. It must be remembered that more than one injury type may, and often does, occur in the same patient.

Cerebral Contusion

Cerebral contusions are injuries to the superficial gray matter of the brain caused by a focal force. External forces striking the head cause acceleration of the intact skull or fractured skull fragments toward the brain surface. Conversely, during a motor vehicle accident, the brain continues to move toward the rapidly decelerating skull and dural folds of the falx or tentorium. “Coup” lesions are ipsilateral to the impact site and can be associated with adjacent calvarial fractures. “Contrecoup” lesions are opposite the coup lesion and result from gyral crests of the rebounding brain striking the inner table of the skull.

Fifty percent of contusions involve the temporal lobes with the temporal pole striking the sphenoid wing. Thirty-three percent involve the frontal lobes with impact of the frontal pole or abrasions of the inferior frontal lobes on the rough floor of the anterior cranial fossa. Twenty-five percent are parasagittal, “gliding” contusions caused by abrasions of the cerebral hemispheres along the fixed falx or tentorium cerebelli. Less likely locations include parietal and occipital lobes, cerebellar vermis, brainstem, and cerebellar tonsils. Ninety percent of cases show multiple or bilateral contusions.12

On CT scans, contusions are patchy, hyperdense lesions with a hypodense background. They are best appreciated on MRI where FLAIR imaging shows the hyperintense edematous background and associated SAH and gradient echo (GRE) series show “blooming” of hemorrhagic foci. Contusions may coalesce or enlarge within the first 12 hours and the associated edema will often worsen over the first several days. Vigilant monitoring of the patient with a contusion is essential, and repeat CT scanning is frequently required.

Intraparenchymal Hemorrhage

IPH or traumatic intracerebral hemorrhage (TICH) is seen in up to 8.2% of all TBI and up to 35% of severe TBI cases. Similar to contusions, TICH and associated edema may increase over time and produce increasing mass effect and neurological deterioration. Delayed traumatic intracerebral hemorrhage (DTICH) will appear in approximately 20% of cases and most occur within 72 hours of the initial trauma.13

If patients develop neurological decline referable to the TICH lesion such as IC-HTN refractory to medical treatment or increasing mass effect with impending herniation, surgical decompression is warranted. Investigation of patient subtypes has shown that surgical decompression is often necessary in patients with image, or patients with image who have frontal or temporal contusions >20 cm3 with midline shift ≥5 mm and/or cisternal compression on CT scan.11 Surgical procedures range from localized frontal or temporal craniotomy with resection of underlying focal clot to more extensive craniectomies with duraplasty, evacuation of severely contused brain, or temporal lobectomy.

Epidural Hemorrhage

EDH occurs when blood collects in the potential space between the dura and inner table of the skull. It is seen in 1% of all head trauma admissions and in 5–15% of patients with fatal head injuries. It is more common in males image, usually occurs in young adults, and is rarely seen in ages <2 or >60 since the dura is more adherent to the inner table of the skull in these groups. Ninety percent of EDHs are due to arterial bleeding that is often due to a fracture at the middle meningeal artery groove, and 10% are due to venous bleeding, usually associated with violation of a venous sinus by an occipital, parietal, or sphenoid wing fracture. EDHs are usually located at the site of impact over the lateral convexity of a cerebral hemisphere (70%), frontal (5–10%), parieto-occipital (5–10%), or posterior fossa locations (5–10%).

On CT scan, EDHs usually appear as a hyperdense, biconvex (lenticular) mass adjacent to the inner table of the skull (Fig. 19-4). This classic description occurs 84% of the time with the medial edge being straight 11% of the time and crescentic (resembling an SDH) 5% of the time.14 Unless there is sutural diastasis, the EDH is externally bounded by cranial sutures, and may cross the falx or tentorium. Additional associated findings include SDHs and cerebral contusions. Of those managed nonsurgically, 23% showed an increase in size, usually in the first 36 hours, with a mean enlargement of 7 mm. Up to 10% of EDHs are not seen on the initial CT scan and present in delayed fashion.15


FIGURE 19-4 CT showing epidural hemorrhage. Note the biconvex- or lenticular-shaped hemorrhage. On the bone windows this was adjacent to a diastatic left lambdoid suture.

The classic clinical presentation of an EDH is a brief post-traumatic loss of consciousness (LOC) followed by a lucid interval, of varying duration, proceeding to obtundation, contralateral hemiparesis, and ipsilateral pupillary dilatation. Interestingly, this only occurs in 27–50% of cases.16 LOC is seen in only 40% of cases, a lucid interval is seen in 80% of cases, a dilated pupil is seen in 60% of cases, and only 85% of these dilated pupils are ipsilateral. Kernohan’s phenomenon (a false localizing sign) occurs when some cases of EDH produce local hemispheric mass effect with compression of the contralateral brainstem against the tentorial notch and ipsilateral hemiparesis.

Overall mortality is 5–12%17 with a unilateral EDH but is increased in cases of bilateral EDH (15–20% mortality), no lucid interval (20% mortality), posterior fossa location (25% mortality), and concurrent acute SDH (25–90% mortality, seen at autopsy in 20% of patients with EDH).

Rapid diagnosis and intervention when indicated is paramount to optimize the outcome. Surgical guidelines suggest that EDH of >30 cm3 should be evacuated regardless of GCS score. EDH of <30 cm3 and <15 mm of thickness and >5 mm midline shift may be treated conservatively at a neurosurgical center with frequent neurological examinations and serial CT scanning.11 Relative indications exist for resection of EDHs that are neurologically symptomatic or have a maximal thickness >1 cm. Patients with acute EDH in coma image and anisocoria should undergo surgical evacuation as soon as possible.11 Most surgeons favor craniotomy for complete clot evacuation with meticulous hemostasis and use of tack-up sutures to decrease the potential epidural space.

Posterior fossa injury is rare, comprising <3% of head injuries; however, the majority of these lesions are EDHs. The limited space of the posterior fossa and the potential compromise of the brainstem and CSF pathways underscore the importance of rapid evacuation via suboccipital craniectomy.11 Patients without signs of mass effect or neurological deterioration may be watched conservatively with serial neurological examinations, CT scans, and a low threshold for surgical intervention.

Subdural Hemorrhage

SDH occurs when blood collects between the arachnoid and inner dural layer and is usually divided into hyperacute (<6 hours), acute (6 hours to 3 days), subacute (3 days to 3 weeks), and chronic (3 weeks to 3 months) variants. It is usually due to traumatic stretching and tearing of cortical bridging veins that cross the subdural space and drain into a dural sinus. The force may be direct (impact) or indirect (nonimpact) and may involve linear or rotational motion. Less common etiologies include coagulopathy, subdural dissection of IPH, and rupture of a vascular anomaly (AVM, aneurysm, cavernoma, dural AV fistula) into the subdural space. Patients with cerebral atrophy, cranial CSF shunts, and large arachnoid cysts (usually in the middle fossa) are predisposed to SDH given the increased traction on cortical veins. Following craniocerebral trauma, SDHs are found in 10–20% of all imaged patients and 30% of autopsies. Over 70% of patients with acute SDHs have other significant associated lesions.

SDHs are commonly located over the hemispheric convexities and may cover part or all of a hemisphere (holoconvexity SDH). Classically, they are crescent shaped, cross suture lines, and layer along the falx or tentorium (Fig. 19-5). Patients present with symptoms of mass effect or more diffuse brain injury. Chronic SDH may present with headaches and/or focal deficits.


FIGURE 19-5 CT showing an acute subdural hemorrhage. Note that crescentic hemorrhage crosses under the right coronal suture.

The appearance of SDHs changes with time. On CT, acute SDHs are hyperdense (60%, homogenous) or of mixed density (40%, may show “swirl sign” of active bleeding), subacute SDHs are isodense with brain, and chronic SDHs become hypodense. Some SDHs may be isodense acutely if there is coagulopathy, significant anemia, or an admixture of blood and CSF. There is inward displacement of the gray/white cortical ribbon and cortical vessels on contrast-enhanced CT. Subdural membranes may appear starting at 4 days and enhance with contrast administration.18 MRI shows varying intensities based on the age of the SDH that is, in turn, based on degradation of blood products from oxyhemoglobin to deoxyhemoglobin to methemoglobin to hemosiderin (Table 19-5).

TABLE 19-5 Appearance of SDH on CT and MRI


SDHs often have worse outcomes when compared to EDHs of similar size/shape. EDHs require force directed toward the skull and epidural vessels (i.e., middle meningeal artery), are usually of arterial origin, and present with symptoms quickly allowing for rapid diagnosis and treatment. It is postulated that SDHs have greater force and impart greater damage to the underlying brain and cortical vessels, have slower onset from venous sources, and have symptoms due to primary brain injury in addition to the midline shift and brainstem compression.

Guidelines suggest that an acute SDH with thickness >1 cm or a midline shift >5 mm should be evacuated regardless of GCS score. Patients with acute SDH <1 cm thick and midline shift <5 mm and in coma (GCS ≤8) should undergo SDH evacuation if the GCS decreases by 2 points between the time of injury and hospital admission, if they present with pupils that are asymmetric or fixed/dilated, or if the ICP ≥20 mm Hg.11 A craniotomy to evacuate an acute clot should be performed as soon as possible, and may require craniectomy and duraplasty for ICP control. A larger craniotomy flap is needed for acute clot that has a texture of thick jam. As the clot breaks down, passing through subacute to chronic stages, it liquefies and may require smaller craniotomies or even burr hole access for appropriate drainage.

Mortality ranges from 50% to 90% and may be related more to the underlying injury rather than the SDH itself. It is increased in the elderly and in patients on anticoagulants.19 Outcome is improved in patients operated on in less 4 four hours with mortality improvements from 66–90% down to 30–59%.20,21

Subarachnoid Hemorrhage

SAH is blood located between the pial and arachnoid membranes. Traumatic subarachnoid hemorrhage (tSAH) results from venous tears in the subarachnoid space.

tSAH is seen in 33% of patients with moderate head injury and is found in nearly 100% of trauma patients at autopsy. It can be seen as a sulcal hyperdensity on CT scan (Fig. 19-6) and as a FLAIR hyperintensity on MRI. It may be confused with FLAIR hyperintensities from 100% FiO2, inflammation, or propofol use. It is often seen embedded in convexity sulci or adjacent to contusions, SDH, or fractures. It may mimic aneurysmal SAH, present in the interpeduncular or other basal cisterns or layers on the tentorium. Patients may complain of headache, emesis, and lethargy and treatment is largely supportive using IV fluids, anticonvulsants, and nimodipine, a Ca2+ channel blocker, to prevent vasospasm.


FIGURE 19-6 CT with arrow pointing to a small traumatic subarachnoid hemorrhage in left central sulcus.

Vasospasm involves narrowing or closure of a vessel with subsequent ischemia or infarct in the vascular territory it supplies. It occurs in 2–41% of trauma cases with tSAH, may occur as early as 2–3 days post-trauma, has heightened incidence from days 3 to 14, and may last up to 2–3 weeks after injury.

image Diffuse Cerebral Injuries


A concussion, or mild traumatic brain injury (MTBI), is defined as an alteration of consciousness resulting from nonpenetrating injury to the brain. Classic symptoms of concussion include headache, confusion, amnesia, and sometimes LOC. Additional symptoms include deficits of motor function (incoordination, stumbling), speech (slowed, slurred, incoherent), memory or processing (amnesia, short-term memory loss, difficulty concentrating or focusing, inattention, perseveration, easy distractibility), and orientation (vacant stare, “glassy eyed,” unable to orient to time/date), and presence of irritability. When present, LOC is usually brief. The two most common grading systems for concussion are from Cantu22 and the American Academy of Neurology (AAN)23 and are shown in (Table 19-6).6

TABLE 19-6 Concussion Grading


Physiological responses to concussion include a transient increase in cerebral blood volume due to loss of vascular autoregulation. In mild cases, this may result in mild cerebral swelling, or hyperemia. In more severe cases, malignant cerebral edema24 may occur with elevated ICPs refractory to nearly all measures and 50–100% mortality. This is the presumed cause of “second impact syndrome”25,26 seen in child and teenage athletes who suffer a second concussion before fully recovering from their first one. Classically, the athlete walks off the field, suffers LOC in 1–5 minutes, and develops vascular engorgement with further neurological deterioration, cerebral herniation, and death.

CT findings may be subtle or nonexistent and include mild diffuse swelling secondary to hyperemia. Being more sensitive, MRI may show positive findings in up to 25% of cases where CT scans are normal.27 Pathologic specimens show no gross or microscopic parenchymal abnormalities in patients who have suffered a single concussion.

Most important for the treatment of concussion is the recognition of injury. Many concussions occur during athletic endeavors and more subtle signs of concussion might not prompt the layperson to seek medical attention. It is the responsibility of the athlete, coach, team doctor, teacher, EMT, or other medical professional to understand the nature of concussions, prompt the patient to seek medical attention, and initiate restriction of further head-injury-prone activities until the patient has fully recovered. All experts agree that a symptomatic player should not return to play. According to one publication contraindications for return to contact sports28 include persistent postconcussive symptoms, permanent neurological symptoms, hydrocephalus, spontaneous SAH, and symptomatic abnormalities at the foramen magnum (e.g., Chiari malformation).

If none of the above contraindications apply, then various suggestions exist regarding the timing of return to play after evaluation by a medical professional.

AAN management options for a single sports-related concussion23 are listed in (Table 19-7)6 and recommendations for multiple sports-related concussions in the same season29 are listed in (Table 19-8).6

TABLE 19-7 Recommendations for a Single Sports-Related Concussion



TABLE 19-8 Recommendations for Multiple Sports-Related Concussions in the Same Season


Diffuse Axonal Injury

DAI is a traumatic axonal stretch injury caused by overlying cerebral cortex and underlying deep brain structures moving at different relative speeds. Mild cases result in axonal stretching and transient neuronal dysfunction while more severe cases cause axonal shearing and permanent neuronal damage. DAI does not require head impact and may be caused by rapid acceleration or deceleration in a linear or rotational fashion.30

Eighty percent of DAI are microscopic and nonhemorrhagic with impaired axonal transport and delayed axonal swelling. When hemorrhagic DAI, the most severe form, is visible on CT/MRI, it must be assumed that it is the tip of the iceberg and that more widespread DAI exists. It should be readily investigated in cases where the degree of neurological injury exceeds what initial imaging would suggest. DAI represents 50% of all primary intra-axial TBI lesions and is found in 80–100% of autopsy patients in fatal injuries.

CT scans may be normal (50–80%) or may show hyperdense petechial hemorrhage (20–50%). MRI scans show multifocal hyperintense T2 signal at gray matter/white matter interfaces, especially in frontal lobes (67%), the corpus callosum (20%), and brainstem (10%).

Injury severity parallels the amount of force required to create DAI lesions and prognosis worsens with increasing number of lesions or as lesion depth progresses from the cortex to corpus callosum to brainstem. Prognosis is variable but generally related to the patient’s age, presenting neurological status, and trajectory of neurological improvement.

Penetrating Nonmissile Injury

Penetrating injuries to the brain and spinal cord may be caused by lower-velocity objects (knives, arrows, lawn darts, ice pick) and by higher-velocity missile-type projectiles. If the object is still embedded, and protruding, great care should be taken to stabilize the object during transport and evaluation. External examinations should describe entry and exit wounds if present. Scalp shaving may be required as well as control of vigorous scalp hemorrhage.

CT should be performed to localize the precise location of the foreign body or injury. Catheter angiography should be performed if the foreign body passes through the territory of any major vessels.

All radiographic evaluation, planning, and operative setup should be performed with the foreign body still embedded and removal should only proceed in the OR (Fig. 19-7). To help plan the removal, a similar or identical object may be useful to have as a reference. Broad-spectrum antibiotics should be administered and cultures may be taken at the time of surgery.


FIGURE 19-7 Intraoperative picture of a penetrating nail-gun injury into the brain. The bone flap has been removed and the dura is open showing the nail entering the brain parenchyma.

Penetrating Missile Injury

Gunshot wounds to the head (GSWH) represent the majority of penetrating cranial injuries and account for 35% of brain injury deaths in patients <45 years old. In civilian GSWH, approximately 66% of people die at the accident scene and the GSWH is the cause of death in 90% of victims.31

Traditionally, GSW have been divided into those from high muzzle velocity rounds (approximately 750–1,000 m/s, hunting rifles, military weapons) and those from low muzzle velocity rounds (approximately 200–500 m/s, most handguns). With escalating civilian firepower, however, these distinctions are becoming more blurred. The above velocities are taken as the projectile leaves the barrel of the weapon. As the bullet covers aerial distance, ricochets off objects, or passes through objects, it loses velocity and kinetic energy.

Primary injuries from gunshot wounds include direct injuries to face and scalp, pressure waves of gas combustion from the weapon (if touching or fired at close range), coup/contrecoup contusion injuries from missile impact, and destruction of brain or bone along the primary path of the original bullet or paths created by fragments of bullets and bone. High-velocity projectiles also create secondary cavitation that pushes tissue away from the bullet in a cone of injury many times wider than the projectile itself. The vacuum that follows the cavitation may pull surface debris into the wound and may serve as a nidus for infection. Some specially designed bullets mushroom, fragment, or tumble in order to increase the width of the destructive path. Projectiles can also careen around or ricochet off the inner table of the skull.

Secondary injuries from GSWs mimic those seen in other types of head trauma and include edema, enlarging contusions, DAI, loss of autoregulation, DIC and hemorrhage from vessel disruption, ischemia, infarction, and herniation. Late complications include abscess or traumatic aneurysm formation, seizures, and migration of embedded fragments of bone or debris.

Surgery must be quickly, and carefully, contemplated in patients with GSWs to the head since many of these patients have devastating and irrecoverable injuries that will not benefit from surgery. The most critical information to accurately triage a GSW patient is the postresuscitative GCS and the CT findings. Patients with GCS 3–5 without a large intracerebral hemorrhage to explain the exam generally have a poor outcome, and consideration for limited treatment is reasonable. Suicide, bilateral fixed and dilated pupils, and coagulopathy are poor prognostic factors. Poor prognosis and high mortality is also associated when bullets cross midline, pass through the geographic center of the brain, traverse the ventricles, or pass through more than one cerebral lobe. One study showed a 94% mortality among patients with GSWH who had image and flaccid or decorticate/decerebrate posturing, and half of the survivors were severely disabled.32

When surgery is performed for GSWs, all attempts should be made to evacuate hematomas causing mass effect (SDH/EDH/IPH), obtain meticulous hemostasis and reduce infection through debridement of devitalized tissue and foreign debris, and obtain a watertight closure to prevent CSF leaks. A “chain of evidence” (for forensics) should be maintained when bullet fragments are removed. When the trajectory is near a known vascular territory, an angiogram should be performed.

Blast Traumatic Brain Injury

Although it may be seen in the civilian population, the military combatant is especially at risk for explosive blast traumatic brain injury (bTBI). Head injury accounts for approximately 20% of all combat-related injuries in recent modern wars33,34 including Operation Enduring Freedom in Afghanistan and Operation Iraqi Freedom. In these two conflicts, explosive bTBI frequently results from the use of improvised explosive devices (IEDs). Modern helmets, body armor, rapid transport of injured personnel, and forward-based field hospitals have provided unprecedented rates of warfighter survival and have allowed for a better understanding of the effects of bTBI.

bTBI may be composed of four distinct types of injuries that can occur separately, or in concert, to varying degrees. Primary blast injury occurs from overpressure. This has long been known to contribute to injuries in air-filled organs such as the lungs, GI tract, and middle ear. Possible contribution of overpressure to brain injury is currently the focus of much study. Secondary blast injury results from penetrating objects that are energized by the explosive. Tertiary blast injury results from the patient being thrown and striking the ground or other object. Quaternary blast injury results from additional factors not included above (e.g., thermal, toxic, hypoxia). Blast injuries in an enclosed space produce an enhanced and complex wave pattern as forces reflect off walls and various objects to impact the head at multiples angles and to multiple degrees.

A blast-induced TBI often includes closed and penetrating TBI components, and many of these patients also have additional serious injuries such as traumatic limb amputations or hemorrhagic shock. The combination of these factors makes it difficult to assess the true contribution of primary or quaternary blast effects on patients with TBI after a blast-induced injury. In milder cases combatants may not recognize that they suffer delayed effects of bTBI or knowingly hide their deficits to remain with their combat unit.

Mild and moderate bTBI can cause concussion symptoms including headache, confusion, amnesia, and altered mental status. Postconcussive symptoms (difficulty concentrating, sleep disturbance, and mood alteration) also can be difficult to differentiate from post-traumatic stress disorder that may also be involved.

Patients with more severe bTBI may have hyperemia and severe cerebral edema early in their course requiring decompressive craniectomy. Higher rates of traumatic pseudoaneurysm and vasospasm are also seen, versus similar civilian closed and penetrating TBI patients, and may require more frequent endovascular or open vascular repair.35

Extracranial Vascular Injury

Post-traumatic intracranial strokes may result from injuries to the internal carotid artery (ICA), common carotid artery (CCA), or vertebral artery (VA) in the neck. Rarely, the injury is due to penetrating trauma with direct ICA, CCA, or VA injury. More often, vessel dissection is sustained in motor vehicle accidents, falls, neck rotation, spine fracture, or iatrogenic injury (surgery, chiropractic maneuvers).

Traumatic dissection is more common in the ICA, occurs in 0.08–0.4% of blunt trauma patients, and runs from a few centimeters above the carotid bifurcation to the skull base. Less common is VA dissection that is usually at the C1–C2 level. Vessel dissection allows blood to collect between the adventitia and media (pseudoaneurysm formation) or between the intima and media of the vessel wall (luminal stenosis). This intramural hematoma may expand or may propagate with distal propagation being more common.

Spontaneous dissections occur in younger to middle-aged adults with 70% between 35 and 50 years old. Seven percent occur in adolescents, and spontaneous dissection is rare in children. Sixty to 90% of patients present with headache and neck pain that is often unrelenting. Symptoms and diagnosis may occur hours to weeks after the initial trauma and include TIAs or strokes, Horner’s syndrome, and, less often, carotid bruits, pulsatile tinnitus, and lower cranial nerve palsies (CN 9–12).

On CT, injuries are seen as a linear lucency within an enhancing vessel and represent the flap separating the true and false vessel lumens. MRI shows a crescentic band surrounding the native flow void and ultrasound shows an echogenic intimal flap.

Twenty percent of cases have an associated injury such as cervical spine injury or silent dissection of another vessel. Treatment consists of heparin followed by Coumadin anticoagulation with balloon angioplasty in select cases.

Child Abuse/Nonaccidental Trauma

Nonaccidental trauma (NAT) is traumatic injury deliberately inflicted on infants and children. The concept was first described as an injury triad in infants consisting of long bone metaphyseal fractures, SDHs, and retinal hemorrhages36 and has become known in common parlance as “whiplash shaken infant syndrome” or “shaken baby syndrome.”37 It is almost certainly underreported and represents the primary etiology of brain injury death in children <2 years old.

Some medical professionals find the cases difficult due to awkward discussions with the child’s parents, an emotional attachment to the child, frequent lack of accurate information, rare confessions from perpetrators, and the medicolegal implications of child abuse accusations. A heightened level of suspicion must be maintained since missed recognition of NAT returns the child to a harmful environment, almost always results in continuation or escalation of the abuse, and may result in the patient’s death.

The two most common histories are no trauma and trivial blunt trauma such as a short-height fall from bed/low surface. Except for the rare EDH with middle meningeal arterial bleeding, low-height falls (household falls, head to impact distance <3 ft) do not result in life-threatening brain injuries.38,39 With no trauma history, the only indicators of NAT may be feeding difficulty, emesis, lethargy, irritability, abnormal movements, seizures, unresponsiveness, or apnea.

Radiologically, NAT shows multiple brain injuries that are more severe than expected given the reported history. Impact injuries include skull fractures, superficial scalp lacerations or swelling, and injuries to the underlying brain and have a high association with other organ injuries. Historically, some fracture patterns have been incorrectly considered more suspicious for child abuse. Fractures that are multiple, compound, diastatic, midline, or nonparietal or that cross suture lines (multiple bones) may denote a greater degree of imparted force but are not pathognomonic for NAT.

A shaking mechanism can result in injuries of differing ages and diffusely distributed SDH. The most frequent hemorrhagic finding is a combination of convexity and interhemispheric SDH (often posterior). Some experts believe interhemispheric SDH has highest specificity for abuse of any intracranial injury. SDH, SAH, and retinal hemorrhages are far more commonly seen in abused children than in nonabused children. EDHs can occur but are much more commonly accidental. Retinal hemorrhages are seen in 65–95% of children with inflicted head injuries and may be unilateral or bilateral. However, severe bilateral retinal hemorrhages are occasionally seen in accidental trauma and usually have a well-defined mechanism of action with major application of force (e.g., MVA).

Given the right circumstances, practically any pattern of hemorrhage or fracture can result from either accidental or inflicted trauma. However, inflicted injury is the only known illness or condition with the combination of acute SDH, skeletal fractures, and severe bilateral retinal hemorrhages. Finally, the severity of injury is staggering with 15–38% overall mortality and 60% mortality if the patient is comatose on presentation. Survivors face a 60–70% likelihood of significant neurological handicap.

image Medical Management of Traumatic Brain Injury

There are multiple treatments for TBI that may be used in series or parallel. Although a dedicated discussion of prehospital care is beyond the scope of this chapter, some treatments can be started in the field and others may be added as more advanced equipment or qualified personnel are available in the emergency department, neuroscience floor, intensive care unit, or operating room.

Basic measures should be implemented in all patients undergoing monitoring and management of TBI. They should be in an ICU setting with frequent monitoring of vital signs, fluid intake, and output and neurological examinations (as permitted). Appropriate monitoring may require multiple invasive lines for blood pressure (arterial line), volume assessment (Swan–Ganz), administration of fluids, medication, or nutrition (central venous catheter), urine output or temperature (Foley catheter), ICP, cerebral tissue oxygenation, or cerebral blood flow (CBF). Patients should be kept normothermic and euvolemic with isotonic fluids (i.e., NS + 20 mEq KCl/L). They should have GI prophylaxis against Cushing’s (stress) ulcers that are frequently seen in severe head injury and cases of elevated ICP.

The head of bed should be elevated to 30–45°, the neck should be kept midline, and the fit of the patient’s cervical collar and endotracheal tube stabilizer should be assessed to prevent compression of the jugular veins and promote venous outflow from the head. Most important is frequent assessment of the patient’s condition, determination of response to therapies implemented, and willingness to adjust care strategies in a fluid manner for optimal outcome.

Over the past two decades, it has become clear that systematic treatment of TBI patients by dedicated neurologists, neurosurgeons, neurointensivists, and surgical trauma and critical care teams has resulted in marked improvement in patient survival and outcome.40 Central to this effort has been the creation, implementation, and refinement of Guidelines for the Management of Severe Traumatic Brain Injury41 with companion guidelines for prehospital management of TBI,42 pediatric TBI,43 surgical management of TBI,11 penetrating TBI,44 and field management of combat-related head trauma.45 These guidelines have undertaken a critical evaluation of the available literature and we readily refer the reader to these excellent works for further detail into the creation of current recommendations.

Blood Pressure and Oxygenation

TBI results from primary injury at the time of impact followed by secondary injury in the minutes, hours, and days that follow. While the medical practitioner cannot “take back” the events of the primary injury, every effort must be made to mitigate or eliminate the secondary injuries that are often more severe. Prior to, or during, transport to a hospital setting, a significant portion of patients may experience periods of hypoxemia or hypotension.46 A single episode of hypoxemia (apnea, cyanosis or O2 saturation <90% in the field, or image mm Hg) or hypotension image is an independent predictor of worse outcome in TBI.4750

Oxygen saturation and blood pressure monitoring should start in the field and continue in the hospital setting with the goal of identifying, avoiding, and rapidly correcting hypoxemia or hypotension, as described above. Empiric oxygen administration should start as early as possible and endotracheal intubation may be required. Similarly, isotonic or hypotonic saline, plasma, colloid, blood, or intravenous pressors may be required to avoid hypotension.41

image Intracranial Pressure Assessment

To understand the rationale behind ICP management, one must start with understanding how pressure in the intracranial space differs from that in other body compartments. If a patient sustains an injury to his or her arm or leg, the surrounding soft tissue has a significant ability to expand outwards from the humerus or femur. By contrast, in cases of TBI, the brain is unable to expand because of the rigid skull.

A useful concept is the modified Monro–Kellie hypothesis, first proposed by Monro51 and verified by Kellie.52 Assuming that the skull is completely inelastic, that the ventricular space is confluent, and that pressures are equally and readily transmitted throughout the intracranial space, the hypothesis states that there is a balance between the brain, blood volume, and CSF contained in the intracranial space. Increases in the volume of one constituent (e.g., cerebral edema, hyperemia) or addition of new components (e.g., tumor, hemorrhage) mandate compensatory decreases in other constituents to maintain the same ICP.

Mildly increased, localized pressure in the brain causes neurological dysfunction of the immediate area. More severe pressure increases cause local tissue compression, shift of intracranial structures, subfalcine and transtentorial herniation, and both local and distant neurological dysfunctions. In the most severe cases, herniation causes compression at the level of the brainstem with direct tissue damage to the pons and medulla, occlusion of brainstem vasculature, infarction, and death.

ICP Monitoring

Normal ICPs vary by age and are considered to be <10–15 mm Hg in adults and older children, 3–7 mm Hg in children, 1.5–6 mm Hg in infants, and may be subatmospheric in the neonates. IC-HTN is seen in 13% of trauma patients with a normal head CT, 60% of patients with an abnormal head CT (hemorrhage, contusion, edema, herniation, or compressed basal cisterns), and ∼60% of patients with a normal head CT plus two or more of the following select criteria on motor examination: age >40, SBP <90 mm Hg, and unilateral or bilateral abnormal posturing (decorticate or decerebrate).53 Therefore, ICP monitoring is recommended in patients with severe TBI image and an abnormal CT scan or with severe TBI, a normal CT scan, and two or more of the select criteria listed above. ICP monitoring may also be considered in patients without an accurate neurological examination due to sedatives, paralytics, or general anesthesia required for other reasons (e.g., difficult ventilator management, agitation, need for additional non-neurological surgery).

Higher mortality and worse outcomes are seen in patients with ICP persistently above 20 mm Hg.54 Therefore, most centers consider IC-HTN to be defined as image. ICP reduction measures are recommended when ICP thresholds exceed 20 mm Hg.41

The most accurate, low-cost, and reliable ICP technology is the fluid-coupled ventriculostomy catheter, or external ventriculostomy drain (EVD), connected to an external strain gauge.41 Another advantage to ventriculostomy placement is that CSF drainage can be performed as a therapeutic measure to control ICP. Other ventricular catheters using fiber-optic or microstrain gauge transduction are more costly and roughly as accurate. Parenchymal ICP monitors require less tissue penetration and do not require the ability to localize the ventricle. However, they cannot be recalibrated in situ and may be subject to measurement drift. Parenchymal monitors are diagnostic (able to measure ICP) while ventricular catheters have the added benefit of being therapeutic (able to drain CSF). Subarachnoid, subdural, and epidural monitors tend to be less accurate.

Cerebral Perfusion Pressure

The post-traumatic brain is at risk for local ischemia in the region of defined traumatic lesions, as well as global ischemia from a more diffuse loss of cerebral autoregulation. Neurological dysfunction may come from direct tissue injury or may come from impaired function of structurally intact neural tissue. For neural tissue to function, it must have adequate CBF to meet the metabolic demand. CBF depends on cerebral perfusion pressure (CPP), which is MAP – ICP. Studies have shown that the injured adult brain is more susceptible to ischemia if the CPP trends below 50 mm Hg.55 Similar studies in children have shown improved survival in those patients who sustain imagemm Hg.56 Studies in the adult population have shown that keeping the image results in unacceptably higher rates of adult respiratory distress syndrome (ARDS) without significantly improved outcome.57,58 There is likely an age-dependent continuum of optimal CPP measurements. Current recommendations support avoidance of image in children, <50 in adults, and >70 in either population.41

Cerebral Blood Flow and Metabolism

CT perfusion can provide measure of CBF at a single point in time. It measures the relative cerebral blood volume, CBF, and mean transit time after injection of iodinated contrast. It has been used extensively in stroke patients and has been investigated in TBI patients to determine the potential viability of contusional and pericontusional tissue, and to help guide other therapeutic strategies such as optimized hyperventilation (HPV).

Intermittent measurements of the jugular venous oxygen saturation (SjvO2) in the bulb of the jugular vein can also be used to assess cerebral perfusion. Normal venous saturation of oxygen is approximately 50–69% and studies have shown that multiple episodes of venous desaturation (<50%) or sustained and profound desaturations are associated with poor outcome.59 In addition, excessively high SjvO2 (>75%) is associated with poor outcome and may indicate hyperemia or significant areas of infarction that will not extract oxygen. The arterial–jugular venous oxygen content difference (AJdO2) may also be calculated.

More focal measurements of CBF include transcranial Doppler (TCD) ultrasonography and parenchymal CBF probes. Thermal diffusion probes provide local CBF measurements based on the thermal temperature difference between microprobes and the relative conductive properties of cerebral tissue and convective properties of blood flow. Probes are often placed in penumbral tissue that is thought to be “at risk” but still salvageable.

Cerebral microdialysis involves placement of a microprobe into penumbral tissue, and measurement of neurochemicals that diffuse into a dialysate through a semipermeable membrane. Neurochemical changes indicative of primary and secondary brain injury are seen in penumbral tissue and TBI patients with poor clinical outcome have been shown to have elevated levels of neurotransmitters, elevated lactate/pyruvate ratios, and abnormal lactate and glutamate levels.60

Brain tissue oxygen tension (PbtO2) monitoring allows direct measurement of focal tissue oxygen tension in a specific region of the brain. Probes are placed in a penumbral area of white matter and allow measurement of local oxygen content or delivery. Normal PbtO2 is approximately 32 mm Hg and studies have shown that patients with multiple or prolonged episodes of image mm Hg have increased morbidity and mortality.61

Current recommendations suggest that SjvO2 and PbtO2 may be monitored as adjuncts to ICP and CPP and therapies should be targeted to keep image and image mm Hg.41

image Intracranial Pressure Management

A ventricular catheter, as mentioned previously, is an excellent method to measure ICP because it facilitates CSF drainage, which is a powerful tool to control ICP. Additional methods follow in this section.

Analgesics and Sedatives

TBI patients, by definition, have suffered trauma and will have increased levels of stress, agitation, and, possibly, discomfort. Patients may suffer discomfort or anxiety from their initial traumatic injury, invasive monitoring, the ICU environment, and procedures. They may be disoriented and/or agitated due to neurological injury or prescribed medication. Pain and agitation can cause increased sympathetic tone, increased temperature, and hypertension. Left unopposed, they can lead to increased venous and ICP, increased metabolic demand, and resistance to controlled ventilation.

Patients may require sedatives or psychotropic medication to prevent self-injurious behavior and dislodgement of airway, vascular lines, or monitoring equipment. Patients on ventilators may require sedatives or paralytics to allow appropriate lung excursion or timing of breath patterns. The medications used to treat pain and agitation, and the doses used, must be carefully monitored and administered so that a balance is achieved between their beneficial effect in reducing pain and anxiety and their side effects of hypotension, alteration or obliteration of the neurological examination, and rebound ICP elevation. Initially, haloperidol may be useful for agitation given its relatively nonsedating quality. If further sedatives or paralytics are needed and a neurological examination becomes unobtainable, an ICP monitor should be placed.

Short-acting agents are preferred in order to facilitate frequent, intermittent neurological examination. Continuous infusion administration may be preferable to bolus administering as this avoids the potential for transient ICP increases seen with some analgesics. Increasingly, fentanyl and its related derivatives (remifentanil, sufentanil) are becoming the agents of choice for acute and longer-term analgesia. They are short acting, reversible, and conducive to administration by continuous infusion. Midazolam and propofol are two commonly used agents for sedation. Midazolam is a short-acting benzodiazepine that is also effective for sedation of the ventilated TBI patient.

Propofol is a hypnotic anesthetic commonly used for treatment of TBI patients with rapid onset and a very short half-life that facilitates rapid neurological assessment. It also reduces cerebral metabolism and oxygen consumption and exerts a neuroprotective effect. Propofol use should be limited in both concentration and duration to avoid propofol infusion syndrome.62 First described in children, and later in adults, excessively high doses or extensive durations of propofol use can result in hyperkalemia, hepatomegaly, metabolic acidosis, rhabdomyolysis, renal failure, and death. Caution should be used if doses exceed 5 mg/(kg h) or 48 hours of therapy in adults.

Although the use of analgesics or sedatives has not shown an independent improvement on neurological outcome, their effectiveness in ICP reduction ensures that they will be used for the foreseeable future.

Hyperosmolar Therapy

While the exact mechanism by which mannitol provides beneficial outcome is unclear, two primary methods are postulated. In the first few minutes, it produces immediate plasma expansion with reduced hematocrit and blood viscosity, improved rheology, and increased CBF and O2 delivery. This reduces ICP and is most notable in patients with image mm Hg.63,64 Over the next 15–30 minutes, and lasting 1.5–6 hours, mannitol produces an osmotic effect with increased serum tonicity and withdrawal of edema fluid from the cerebral parenchyma.

When given as a bolus, the ICP reduction is evident at 1–5 minutes and peaks at 20–60 minutes. The initial bolus of mannitol, for acute ICP reduction in cases of neurological worsening or herniation, should be dosed at 1 g/kg with subsequent administration at smaller doses and longer intervals (i.e., 0.25–0.5 g/kg Q 6 hours). Mannitol opens the blood–brain barrier (BBB) and may cross the BBB itself, drawing water into the brain and transiently exacerbating vasogenic cerebral edema. Furosemide may also be used synergistically with mannitol65 to reduce cerebral edema through increased serum tonicity and reduced CSF production.

There has been concern that continuous mannitol infusions lead to elevated serum levels of mannitol, sequestering of mannitol within brain tissue, rebound shifts of water back into the brain, and worsening outcomes. It was thought that bolus administration reduced this effect and was more effective than mannitol infusions for ICP reduction66 with an added benefit of maximized rheologic increase in CBF. More recent data suggest that there are no significant data to support this.67 The significant water shifts employed by the use of mannitol mandate accurate measurement of urine output and fluid replacement to maintain euvolemia. Accurate diagnosis of diabetes insipidus (DI) may be precluded in the presence of mannitol.

Acute tubular necrosis (renal failure) may be seen when mannitol is used in high doses, in patients with preexisting renal disease, or with other nephrotoxic drugs. Serum osmolality should be monitored and use of mannitol should be restricted when serum osmolality is >320 mOsm/L.68 It is imperative to follow urine output to allow replacement of urinary electrolyte loss and continued avoidance of hypotension and hypovolemia.

Although TBI patients usually get mannitol in conjunction with ICP monitoring, some patients may benefit from high-level empiric dosing.42 No strong evidence supports empiric prehospital administration of mannitol to TBI patients69 but mannitol may be of benefit in patients with acute mass lesions and may be used as a bridge toward definitive therapy such as operative evacuation of mass lesions. Comatose patients acutely presenting with operative subdural hematomas or with operative intraparenchymal temporal lobe hemorrhages and abnormal pupillary dilatation demonstrated improved clinical outcomes when treated preoperatively with large doses of mannitol, approximately 1.4 g/kg.70,71

Hypertonic Saline

As with mannitol, hypertonic saline (HS) is thought to lower ICP through two mechanisms. First, an oncotic pressure gradient, across the BBB, results in mobilization of water from brain tissue and hypernatremia. Second, rapid plasma dilution and volume expansion, endothelial cell and erythrocyte dehydration, and increased erythrocyte deformability lead to improvements in rheology, CBF, and oxygen delivery. HS is often administered as a continuous infusion of 25–50 mL/h of 3% saline (replacing the patient’s isotonic IV fluid) or bolus infusions of 10–30 mL of 7.2%, 10%, or 23.4% saline solution. Onset of clinical response can be within minutes and may last for hours making HS a candidate for use in cases of severe ICP elevation or acute herniation syndrome.

A serum sodium goal of 145–160 mEq/L is frequently used although higher serum sodium levels may be necessary to achieve clinical goals. Serum sodium and osmolality levels should be aggressively followed as excessively rapid increases in sodium, seen during HS administration, may result in central pontine myelinolysis. This occurrence is most often seen in patients with preexisting, chronic hyponatremia and is rarely seen in the chronically normonatremic patient treated with HS. HS may also induce or exacerbate pulmonary edema in patients with underlying cardiac or pulmonary deficits.


HPV lowers PCO2 with subsequent vasoconstriction, reduction of cerebral volume, and reduction in ICP. Time of onset ranges from 30 seconds to 1 hour; peak effect may be seen at 8 minutes and may last up to 15–20 minutes. This rapid onset of action makes HPV particularly effective in the treatment of an IC-HTN crisis and as a bridge to more definitive therapy (i.e., surgical decompression) or while other ICP reduction measures take effect.

Although HPV was once used as a first-line therapy, concern has grown regarding prophylactic or prolonged use in TBI patients. During the first 24 hours after injury, patients with severe TBI show CBF reduction of at least 50%.72Forced vasoconstriction, through HPV, can lower this further. Depending on the degree of functional autoregulation, there may be increases in the oxygen extraction fraction or shunting of blood to ischemic areas with widening of the total ischemic territory.

Severe TBI patients should aim to be normocarbic image. HPV should be avoided during the first 24 hours postinjury when CBF is most reduced. If HPV is necessary after the first 24 hours, short-term, mild HPV image can be effective for ICP control necessary to implement other treatment strategies. Further, moderate HPV produces even further reductions in CBF,73 and should be avoided except for very brief periods while other therapies are prepared. Prophylactic HPV image is contraindicated as it is associated with increased ischemia and worse outcomes.74

When HPV is used, further monitoring should be considered to monitor oxygen delivery and may include jugular venous saturation or PbtO2 monitors.

Decompressive Craniectomy

When the above therapies fail to provide adequate control of ICP, other second-tier therapies can be considered. These include decompressive craniectomy, temporal lobectomy, optimized HPV, barbiturate coma, and hypothermia.

Some cases of TBI require acute craniotomies to address focal lesions (e.g., SDH, EDH, IPH). The bone is removed, the lesion is resected, and the dura and bone are replaced. More severe cases of TBI may develop diffuse cerebral edema, contusions of large size in eloquent areas, or multiple, coalesced contusions. In these cases, it may be preferable to leave the bone flap off. Additionally, when ICP is refractory to the previously mentioned techniques, a decompressive craniectomy effectively expands the intracranial space to lower the ICP.

The most common decompressive craniectomy is unilateral hemispheric. Bifrontal and bilateral hemispheric craniectomies (Fig. 19-8) have also been described and are based on the location and severity of the underlying lesion(s). In decompressive craniectomies the dura is opened widely and areas of noneloquent contused and devitalized brain can be removed if required. In the hemispheric technique at least a 12-cm cranial flap is removed. The brain is then contained only by the augmented dural covering and the more compliant scalp.


FIGURE 19-8 CT of a bilateral hemispheric decompressive craniectomy performed in a patient with severe edema from a likely second impact syndrome.

While decompressive craniectomy has been shown to be effective in reducing ICPs, many studies are observational or case series, lack appropriate control subjects, or do not achieve statistical significance with regard to all end points.7578 As a result this procedure has not yet been definitively proven to improve outcomes in the trauma patient. Two international trials are currently ongoing (DECRA and RESCUEicp) to address this question.

Decompressive craniectomy is most commonly used as a second-tier option for patients with IC-HTN refractory to maximal medical management and success is dependent on patient selection.79 Ideally, operative intervention should occur within 48 hours of the initial injury and before ICP has surpassed 40 mm Hg for sustained periods of time. Outcomes tend to be more favorable in younger patients with diffuse injury and limited secondary injury. Regardless of the preoperative indications or patient profile, continuing postoperative IC-HTN greater than 35 mm Hg has been associated with 100% mortality.80

Early decompressive craniectomy, as a primary treatment, has the advantage of rapid control of IC-HTN11,81; however, there are a number of potential complications. These include infection, subdural hygromas, hydrocephalus, syndrome of the trephined, perfusion breakthrough, and cerebral infarction. Early surgical intervention may be an option for patients presenting with severe unilateral or bilateral cerebral edema, parenchymal lesions resistant to initial medical management of ICP, or other injuries whose management conflicts with standard ICP control measures (e.g., patients with acute respiratory distress syndrome requiring elevated ventilatory pressures).


Barbiturates benefit TBI patients by decreasing metabolic demand for oxygen (CMRO2), decreasing free radicals and intracellular calcium, and lowering ICPs. Side effects such as immunosuppression and hypotension from reduced sympathetic tone and mild cardiodepression often limit their use. TBI patients in coma (GCS ≤8) receiving barbiturate therapy have infection and respiratory complication rates in excess of 50%80 and significant systemic hypotension is present in 25% of patients82 despite adequate intravascular volume and pressor therapy.

Patients with hemodynamic instability, sepsis, respiratory infection, or cardiac risk factors are excluded from this therapy and those receiving barbiturates should be closely monitored for signs of cardiac compromise or infection with cessation of therapy if systemic effects of the treatment become significant and unmanageable. A pretherapy echocardiogram and intratherapy use of a Swan–Ganz catheter should be considered.

Barbiturates clearly reduce ICP but studies have shown both improved and worsened outcomes for TBI patients receiving barbiturate therapy. There is no role for prophylactic barbiturate therapy in TBI patients, as it increases hypotension without significantly improving outcome,83,84 and it is to be used only as a second-line therapy when other treatment measures have failed.

A typical pentobarbital regimen is a loading dose of 10 mg/kg over 30 minutes followed by a 5 mg/(kg h) infusion for 3 hours. A maintenance dose of 1 mg/(kg h) should then be started.41 Serum barbiturate levels of 3–4 mg% should be maintained, although poor correlation exists between serum level, therapeutic benefit, and systemic complications. Continuous electroencephalographic evaluation is more reliable and dosing to the point of EEG burst suppression produces near-maximal reductions of CMRO2 and CBF.


Induced prophylactic hypothermia attempts to improve outcome in patients with severe TBI through reduction of cerebral metabolism, ICP, inflammation, lipid peroxidation, excitotoxicity, cell death, and seizures. Side effects of hypothermia include decreased cardiac function, thrombocytopenia, elevated creatinine clearance, pancreatitis, and shivering with associated elevations in ICP. Initial interest for induced hypothermia stemmed from anecdotes (e.g., a child trapped in a frozen lake), single-center clinical trials, and four meta-analyses.8589 Although its use has been adopted by some trauma centers and there is level 1 evidence for its use in V-fib or V-tach MI, initial literature had not shown statistically significant improvements in mortality directly attributable to induced hypothermia for the trauma patient.

Meta-analyses of more recent data and subsequent guidelines from the Brain Trauma Foundation41 note a nonstatistically significant trend toward mortality reduction (compared to normothermic controls) when target temperatures were maintained for greater than 48 hours. Hypothermia-treated patients also had significantly higher Glasgow Outcome Scale scores. Additionally, it was found that patients who were hypothermic on admission had improved outcomes when hypothermia was maintained. Results are limited, however, by the small sample sizes and potential confounding factors of each study in the meta-analyses.

Hypothermic therapy is an option in the patient with severe TBI. Selected patients should be cooled relatively early in their care or maintained in a cooled state if hypothermic on arrival. A target temperature of 32–33°C should be achieved and, if possible, maintained for greater than 48 hours. These patients should be closely monitored for the untoward effects of hypothermia such as electrolyte abnormalities, hypocoagulability, and cardiac rhythm alterations.90 Rewarming of these patients should be very slow, generally not exceeding more than 1° per 24 hours.


Glucocorticoids are not recommended for improving outcome or reducing ICP in TBI.41,91,92 Side effects of steroid use include coagulopathies, hyperglycemia, and increased infection and are reflected in poor outcomes.

Antiseizure Prophylaxis

PTSs are deleterious in the TBI patient for many reasons including elevated metabolic demand that exacerbates ischemia and increased ICP. TBI patients at increased risk for PTS include those with image, depressed skull fractures, cortical contusions or hemorrhage (SDH, EDH, IPH), and penetrating hemorrhage or seizure within 24 hours of head injury.93 Anticonvulsants have been shown to effectively reduce the risk of early PTS (<7 days postinjury) but not late PTS (>7 days postinjury)94 and the prophylactic administration of phenytoin or carbamazepine is indicated for the prevention of early PTS only41 (i.e., 7 days after injury). Despite no controlled study showing an equivalent efficacy for early PTS prevention, many centers are now using levetiracetam (Keppra®) because of reduced side effects and ease of administration.

image Specific System Considerations


At rest, all patients have basal energy expenditure (BEE) dependent on their sex, age, height, and weight. All injured patients show an increase in BEE regardless of neurological course. Patients who are sedated and paralyzed may show BEE increases to 120–130% of baseline.97 Comatose patients (GCS ≤8) with isolated head injury have BEE approximately 140% (range 120–250%).96,97 Mortality is reduced in patients who receive full caloric replacement by 1 week postinjury98 and at least 15% of calories should be supplied as protein. Since it may take 2–3 days to ramp up feedings, nutritional replacement should start by 72 hours postinjury.

Enteral feeding is preferred over parenteral nutrition as it provides enhanced immunocompetence and a reduced risk profile.99 If the patient has diminished gastric motility, a jejunal feeding tube can be placed since patients with severe TBI can tolerate early jejunal feeding even in the presence of gastric dysfunction and absent small bowel activity.100 Total parenteral nutrition should be started if enteral feeding is not possible or if higher nitrogen intake is required.


Trauma patients may incur infection as part of their initial injury from gross wound contamination or immunosuppression, or iatrogenically from open surgical procedures, intubation for mechanical ventilation, and invasive monitoring equipment. In general, antibiotic coverage should be targeted toward specific organisms and removed as soon as possible to decrease drug-resistant strains of bacteria or alterations in normal floral patterns and bacterial overgrowth (i.e., C. difficile colitis).

Perioperative antibiotics are generally only recommended for the first 24 hours. Routine flushing or exchange of ventricular catheters is not recommended.41 Conflicting evidence precludes recommendations for periprocedural antibiotics during EVD placement, although one study has shown that use of rifampin-impregnated ventriculostomy catheters resulted in an overall decrease in infection rates with a concomitant increase in rifampin-resistant organisms.101

Coagulopathy and DVT Prophylaxis

Patients suffering head trauma often develop or present with coagulopathy even without medical comorbidities or medications. Most critically ill trauma patients will have decreased levels of plasma antithrombin (AT) activity. Head-injured patients, however, tend to have increased rates of coagulopathy with supranormal AT activity that can progress to disseminated intravascular coagulation and fibrinolysis (DICF), and expansion of existing contusions and delayed development of additional hemorrhages.102,103 Coagulopathy is especially prevalent in penetrating brain injury.

For all trauma patients, the patients’ medical history and review of systems should specifically address coagulopathic medical disorders, prior episodes of trauma, bleeding or clot formation, use of specific antiplatelet or anticoagulant medication (aspirin, warfarin, low-molecular-weight heparin), and medications that have antiplatelet compounds as a component. Laboratory studies should be performed including measurements of prothrombin time (PT), activated partial thromboplastin time (aPTT), international normalized ratio (INR), platelet count (plt), and bleeding time or platelet function assay (when necessary).

Coagulopathies should be rapidly and aggressively treated until the patient achieves a normal coagulation profile. Effects of warfarin anticoagulation may be reversed by administration of vitamin K, fresh frozen plasma (FFP), or prothrombin complex concentrate; effects of heparin may be reversed with protamine sulfate; and thrombocytopenia or platelet deactivation may be treated with donor platelet transfusion. Unfortunately, patients may have life-threatening intracranial hemorrhage or developing cerebral edema with rising ICPs and these treatments require time for transfusion.

Recombinant activated coagulation factor VII (rFVIIa) rapidly forms a complex with tissue factor to produce thrombin and, separately, converts factor X to its active form, factor Xa, resulting in a “thrombin burst” at the site of tissue damage.104 It is FDA approved for use in hemophiliacs and patients with antibodies to factor VIII or IX, and has been studied off-label in cases of ICH and trauma patients requiring rapid craniotomy in the face of coagulopathy. Its effects on neurological outcome and mortality, as well as its cost burden (∼$10,000/dose), are currently under investigation and have not been fully defined.105,106

Trauma patients, in general, and head-injured patients, in particular, are at risk for venous thromboembolism (VTE) such as deep venous thrombosis (DVT) and pulmonary embolus (PE). Neurological risk factors for DVT and PE include stroke or spinal cord injury, need for prolonged surgery or prolonged bed rest, SAH or head injury causing altered coagulation or dehydration, and increased blood viscosity from cerebral salt wasting and treatment of cerebral edema.107,108 The incidence of DVTs in neurosurgical patients ranges from 19% to 50%. Low-risk, prophylactic measures against DVTs include passive range of motion, early ambulation, rotating beds, and electrical stimulation of calf muscles. If DVTs are not already present, pneumatic compression boots (PCBs) and sequential compression devices may be safely used and can reduce the incidence of DVTs to 1.7–2.3% and PEs to 1.5–1.8%.109

Active, pharmacologic anticoagulation can increase the effectiveness of DVT prophylaxis with the risk of additional hemorrhagic complications. Low-molecular-weight heparins have a higher ratio of anti–factor Xa to anti–factor IIa activity, versus unfractionated heparin, have greater bioavailability after subcutaneous injection, and have more predictable plasma levels. They can be added to PCBs without significantly increased risk of hemorrhage,110 and their use is recommended in postoperative neurosurgery patients.41 There are no universally accepted recommendations for the method and timing of postoperative anticoagulation and this should be tailored to each patient. One study has shown no increased incidence of hemorrhagic complications once full anticoagulation was resumed 3 days after craniotomy.19


image Prognosis

To interpret and compare the effectiveness of various treatments, common end points are necessary for communication between practitioners or comparison of studies. The Glasgow Outcome Scale111 (Table 19-9) is a widely used outcome grading scale with many studies separating patients into those with good outcome (image or 5), those with poor outcome (image or 3), and those who are dead image. Although its separation of patient categories is relatively coarse and may not identify the subtleties of recovering TBI patients, it remains a useful tool for describing patient outcome just as the GCS is a useful tool for measuring a patient’s neurological examination.

TABLE 19-9 Glasgow Outcome Score


The medical practitioner is often called upon to make predictions of outcome based on limited information early on in the patient’s course. The patient’s ultimate neurological outcome may not be fully evident until weeks or months of treatment have taken place in hospitals, rehabilitation centers, and at home.

Various studies and meta-analyses112114 show that worse prognosis is seen in patients with bilaterally dilated (> 4 mm) or absent pupillary light reflexes, absent oculocephalic or oculovestibular reflexes, increased injury severity scale (> 40), advanced age (> 60 and possibly < 2), hypotension (image, worse with concomitant hypoxemia), abnormal CT scan (extensive tSAH, compression or obliteration of basal cisterns), persistent image mm Hg, elevated ICP during the first 24 hours, or presence of apolipoprotein E allele. There is a stepwise, increasing probability of poor outcome with worsening initial total GCS scores (especially image) and some studies show worse prognosis based on lower GCS subscores (motor ≤3, eye opening ≤2, verbal response ≤2).

image Brain Death Determination and Organ Donation

“Brain death” denotes the absence of any observable neurological activity in the brain and the irreversibility of cessation of the cardiopulmonary system or the entire brain. It must be explained to patients’ families that brain death is a legally binding death and a true clinical death. The requirements of a brain death examination may vary slightly between states or medical facilities but retain similar core elements.

There must be no complicating conditions (hypothermia <32.2°C, hypotension [SBP <90], exogenous sedatives, paralytics, drug/alcohol, hepatic encephalopathy, hyperosmolar coma, atropine, recent CPR/shock/anoxia) to confuse the neurological exam. Patients have fixed, dilated pupils and no observable corneal, oculocephalic, oculovestibular, gag, or cough reflexes. There is no movement to deep central or peripheral pain and no spontaneous breathing is seen on disconnection from the ventilator with PaCO2 >60 mm Hg (i.e., apnea test).

If the patient is unable to tolerate an apnea test or if parts of the brain death protocol are equivocal, secondary tests may be used to confirm or augment the above information. Most commonly used are cerebral angiography to show absence of intracranial flow and cerebral radionuclide angiogram to show absent uptake in brain parenchyma.

Head-injured patients who progress to brain death may be candidates for organ donation. Specialized organ procurement organizations are present in most states and represent a party separate from the treating team and with no conflict of interest regarding the patient’s care. Although the patient’s death is unfortunate, organ donation can provide family members with a slightly more positive conclusion to a series of unfortunate events.


1. Langlois JA, Rutland-Brown W, Thomas KE. Traumatic Brain Injury in the United States: Emergency Department Visits, Hospitalizations, and Deaths. Atlanta, GA: Centers for Disease Control and Prevention, National Center for Injury Prevention and Control; 2006.

2. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A practical scale. Lancet. 1974;2(7872):81–84.

3. Hahn YS, Chyung C, Barthel MJ, et al. Head injuries in children under 36 months of age. Demography and outcome. Childs Nerv Syst. 1988;4(1):34–40.

4. Wilberger JE, Deeb Z, Rothfus W. Magnetic resonance imaging after closed head injury. Neurosurgery. 1987;20:571–576.

5. Loesser JD, Kilburn HL, Jolley T. Management of depressed skull fracture in the newborn. J Neurosurg. 1976;44:62–64.

6. Greenberg MS. Handbook of Neurosurgery. 6th ed. New York, NY: Thieme; 2006.

7. Arseni CS. Growing skull fractures of children. A particular form of post-traumatic encephalopathy. Acta Neurochir. 1966;15:159–172.

8. Lende R, Erickson T. Growing skull fractures of childhood. J Neurosurg. 1961;18:479–489.

9. Jennett B. Epilepsy after Non-Missile Head Injuries. 2nd ed. London: Heinemann; 1975:179.

10. Steinbok P, Flodmark O, Martens D, et al. Management of simple depressed skull fractures in children. J Neurosurg. 1987;66;506–510.

11. Bullock MR, Chesnut R, Ghajar J, et al. Surgical management of depressed cranial fractures. Neurosurgery. 2006;58(3):S2-1–S2-106.

12. Katzman GL. Cerebral contusion. In: Osborn AG, Blaser SI, Salzman KL, et al., eds. Diagnostic Imaging: Brain. 1st ed. Salt Lake City, UT: Amirsys; 2004:26–29.

13. Gudeman SK, Kishore PR, Miller JD, et al. The genesis and significance of delayed traumatic intracerebral hematoma. Neurosurgery. 1979;5: 309–313.

14. Tsai FY, Teal JS, Hieshima GB. Neuroradiology of Head Trauma. Baltimore: University Park Press; 1984.

15. Borovich B, Braun J, Guilburd JN, et al. Delayed onset of extradural hematoma. J Neurosurg. 1985;63:30–34.

16. Mckissok W, Taylor JC, Bloom WH, et al. Extradural hematoma: observations on 125 cases. Lancet. 1960;2:167–172.

17. Rivas JJ, Lobato RD, Sarabia R, et al. Extradural hematoma: analysis of factors influencing the courses of 161 patients. Neurosurgery. 1988;23: 44–51.

18. Munro D, Merritt HH. Surgical pathology of subdural hematomas: based on a study of one hundred and five cases. Arch Neurol Psychiatry. 1936;35:64–78.

19. Kawamata T, Takeshita M, Kubo O, et al. Management of intracranial hemorrhage associated with anticoagulant therapy. Surg Neurol. 1995;44:438–443.

20. Seelig JM, Becker DP, Miller JD, et al. Traumatic acute subdural hematoma: major mortality reduction in comatose patients treated within four hours. N Engl J Med. 1981;304:1511–1518.

21. Wilberger JE, Harris M, Diamond DL. Acute subdural hematoma: morbidity, mortality and operative timing. J Neurosurg. 1991;74: 212–218.

22. Cantu RC. Head injuries in sports. Br J Sports Med. 1996;30:289–296.

23. Quality Standards Subcommittee of the American Academy of Neurology. Practice parameter: the management of concussion in sports [summary statement]. Neurology. 1997;48:581–585.

24. Bruce DA, Alavi A, Bilanium L, et al. Diffuse cerebral swelling following head injuries in children: the syndrome of “malignant brain edema”. J Neurosurg. 1981;54:170–178.

25. Schneider RC. Head and Neck Injuries in Football. Baltimore: Williams & Wilkins; 1973.

26. Saunders RL, Harbaugh RE. Second impact in catastrophic contact-sports head trauma. JAMA. 1984;252:538–539.

27. Levin HS, Williams DH, Eisenberg HM, et al. Serial MRI and neurobehavioral findings after mild to moderate closed head injury. J Neurol Neurosurg Psychiatry. 1992;55:255–262.

28. Cantu RC. Return to play guidelines after a head injury. Clin Sports Med. 1998;17:45–60.

29. Flannagan PP, Bailes JF. Neurological injury in athletes. Contemp Neurosurg. 1998;20(17):1–7.

30. Gennarelli TA, Thibault LE, Adams JH, et al. Diffuse axonal injury and traumatic coma in the primate. Ann Neurol. 1982;12:564–574.

31. Kaufman HH. Civilian gunshot wounds to the head. Neurosurgery. 1993;32:962–964.

32. Benzel EC, Day WT, Kesterson L, et al. Civilian craniocerebral gunshot wounds. Neurosurgery. 1991;29:67–72.

33. Carey ME, Joseph AS, Morris WJ, et al. Brain wounds and their treatment in VII Corps during Operation Desert Storm, February 20 to April 15, 1991. Mil Med. 1998;163:581–586.

34. Bellamy RF, Maningas PA, Vayer JS. Epidemiology of trauma: military experience. Ann Emerg Med. 1986;15:1384–1388.

35. Ling G, Bandak F, Armonda R, et al. Explosive blast neurotrauma. J Neurotrauma. 2009;26:815–825.

36. Kempe CH, Silverman FN, Steele BF, et al. The battered-child syndrome. JAMA. 1962;7(181):17–24.

37. Caffey J. The whiplash shaken infant syndrome: manual shaking by the extremities with whiplash-induced intracranial and intraocular bleedings, linked with residual permanent brain damage and mental retardation. Pediatrics. 1974;54(4):396–403.

38. Lyons TJ, Oates RK. Falling out of bed: a relatively benign occurrence. Pediatrics. 1993;92(1):125–127.

39. Reiber GD. Fatal falls in childhood. How far must children fall to sustain fatal head injury? Report of cases and review of the literature. Am J Forensic Med Pathol. 1993;14(3):201–207.

40. Vukic L, Negovetic D, Kovac D, et al. The effect of implementation of guidelines for the management of severe head injury on patient treatment and outcomes. Acta Neurochir. 1999;141(11):1203–1208.

41. Bratton SL, Chestnut RM, Ghajar J, et al. Guidelines for the management of severe traumatic brain injury, 3rd ed. J Neurotrauma. 2007;24(S1): S1–S106.

42. Gabriel EJ, Ghajar J, Jagoda A, et al. Guidelines for Pre-Hospital Management of Traumatic Brain Injury. New York, NY: Brain Trauma Foundation; 2000.

43. Adelson PD, Bratton SL, Carney NA, et al. Guidelines for the acute medical management of severe traumatic brain injury in infants, children and adolescents. Pediatr Crit Care Med. 2003;4:S417–S491.

44. Guidelines for the management of penetrating brain injury. J Trauma. 2001;51:S3–S6.

45. Knuth T, Letarte PB, Ling G, et al. Guidelines for the Field Management of Combat-Related Head Trauma. New York, NY: Brain Trauma Foundation; 2005.

46. Stocchetti N, Furlan A, Volta F. Hypoxemia and arterial hypotension at the accident scene in head injury. J Trauma. 1996;40:764–767.

47. Chesnut RM, Marshall LF, Klauber MR, et al. The role of secondary brain injury in determining outcome from severe head injury. J Trauma. 1993;34:216–222.

48. Fearnside MR, Cook RJ, McDougall P, et al. The Weatmead Head Injury Project outcome in severe head injury. A comparative analysis or pre-hospital, clinical and CT variables. Br J Neurosurg. 1993;7: 267–279.

49. Marmarou A, Anderson RL, Ward JD, et al. Impact of ICP instability and hypotension on outcome in patients with severe head trauma. J Neurosurg. 1991;75:159–166.

50. McHugh GS, Engel DC, Butcher I, et al. Prognostic value of secondary insults in traumatic brain injury: results from the IMPACT study. J Neurotrauma. 2007;24:287–293.

51. Monro A. Observations on the Structure and Function of the Nervous System. Edinburgh: Creech & Johnson; 1823:5.

52. Kellie G. An account of the appearances observed in the dissection of two of the three individuals presumed to have perished in the storm of the 3rd, and whose bodies were discovered in the vicinity of Leith on the morning of the 4th November 1821 with some reflections on the pathology of the brain. Trans Med Chir Soc Edinb. 1824;1:84–169.

53. Narayan RK, Kishore PRS, Becker DP, et al. Intracranial pressure: to monitor or not to monitor? A review of our experience with severe head injury. J Neurosurg. 1982;56:650–659.

54. Juul N, Morris GF, Marshall SB, et al. Intracranial hypertension and cerebral perfusion pressure: influence on neurological deterioration and outcome in severe head injury. J Neurosurg. 2000;92:1–6.

55. Nordstrom CH, Reinstrup P, Xu W, et al. Assessment of the lower limit for cerebral perfusion in severe head injuries by bedside monitoring of regional energy metabolism. Anesthesiology. 2003;98:809–814.

56. Elias-Jones AC, Punt JA, Turnbull AE, et al. Management and outcome of severe head injuries in the Trent region 1985–90. Arch Dis Child. 1992;67:1430–1435.

57. Contant CF, Valadka AB, Gopinath SP, et al. Adult respiratory distress syndrome: a complication of induced hypertension after severe head injury. J Neurosurg. 2001;95:560–568.

58. Robertson CS, Valadka AB, Hannay HJ, et al. Prevention of secondary ischemic insults after severe head injury. Crit Care Med. 1999;27: 2086–2095.

59. Sheinberg M, Kanter MJ, Robertson CS, et al. Continuous monitoring of jugular venous oxygen saturation in head-injured patients. J Neurosurg. 1992;76:212–217.

60. Kett-White R, Hutchinson PJ, Al-Rawi PG, et al. Adverse cerebral events detected after subarachnoid hemorrhage using brain oxygen and microdialysis probes. Neurosurgery. 2002;50:1213–1221.

61. van den Brink WA, van Santbrink H, Steyerberg EW, et al. Brain oxygen tension in severe head injury. Neurosurgery. 2000;46:868–876.

62. Fong JJ, Sylvia L, Ruthazer R, et al. Predictors of mortality in patients with suspected propofol infusion syndrome. Crit Care Med. 2008;36: 2281–2287.

63. Barry KG, Berman AR. Mannitol infusion. Part III. The acute effect of the intravenous infusion of mannitol on blood and plasma volume. N Engl J Med. 1961;264:1085–1088.

64. McGraw CP, Howard G. The effect of mannitol on increased intracranial pressure. Neurosurgery. 1983;13:269–271.

65. Marion DW, Letarte PB. Management of intracranial hypertension. Contemp Neurosurg. 1997;19(3):1–6.

66. James HE. Methodology for the control of intracranial pressure with hypertonic mannitol. Acta Neurochir. 1980;51:161–172.

67. Roberts I, Schierhout G, Wakai A. Mannitol for acute traumatic brain injury. Cochrane Database Syst Rev. 2003;(2):CD001049.

68. Mendelow AD, Teasdale GM, Russel T, et al. Effect of mannitol on cerebral blood flow and cerebral perfusion pressure in human head injury. J Neurosurg. 1985;63:43–48.

69. Wakai A, Roberts IG, Schierhout G. Mannitol for acute traumatic brain injury. Cochrane Database Syst Rev. 2007;(1):CD001049.

70. Cruz J, Minoja G, Okuchi K. Improving clinical outcomes from acute subdural hematomas with the emergency preoperative administration of high doses of mannitol: a randomized trial. Neurosurgery. 2001;49: 864–871.

71. Cruz J, Minoja G, Okuchi K. Major clinical and physiological benefits of early high doses of mannitol for intraparenchymal temporal lobe hemorrhages with abnormal pupillary widening: a randomized trial. Neurosurgery. 2002;51:628–638.

72. Bouma GJ, Muizelaar JP, Stringer WA, et al. Ultra early evaluation of regional cerebral blood flow in severely head injured patients using xenon enhanced computed tomography. J Neurosurg. 1992;77: 360–368.

73. Obrist WD, Langfit TW, Jaggi JL, et al. Cerebral blood flow and metabolism in comatose patients with acute head injury. Relationship to intracranial hypertension. J Neurosurg. 1984;61:241–253.

74. Muizelaar JP, Marmarou A, Ward JD, et al. Adverse effects of prolonged hyperventilation in patients with severe head injury. A randomized clinical trial. J Neurosurg. 1991;75:731–739.

75. Sahuquillo J, Arikan F. Decompressive craniectomy for the treatment of refractory high intracranial pressure in traumatic brain injury. Cochrane Database Syst Rev. 2006;(1):CD003983.

76. Polin RS, Shaffrey ME, Bogaev CA, et al. Decompressive bifrontal craniectomy in the treatment of severe refractory posttraumatic cerebral edema. Neurosurgery. 1997;41:84–92.

77. Munch E, Horn P, Schurer L, et al. Management of severe traumatic brain injury by decompressive craniectomy. Neurosurgery. 2000;47:315–322.

78. Citero G, Andrews PJ. Refractory elevated intracranial pressure: intensivist’s role in solving the dilemma of decompressive craniectomy. Intensive Care Med. 2007;33:45–48.

79. Salazar AM, Cooper P, Ecklund J. Traumatic brain injury. In: Neurological Disorders: Course and Treatment. 2nd ed. Elsevier Science; 2002.

80. Valadka AB, Andrews BT, eds. Neurotrauma: Evidence Based Answers to Common Questions. New York, NY: Thieme; 2005.

81. Chesnut RM. Care of central nervous system injuries. Surg Clin North Am. 2007;87:119–156.

82. Roberts I, Sydenham E. Barbiturates for acute traumatic brain injury. Cochrane Database Syst Rev. 1999;(3):CD000033.

83. Rea GL, Rockswold GL. Barbiturate therapy in uncontrolled intracranial hypertension. Neurosurgery. 1983;12:401–404.

84. Ward J, Becker DP, Miller JD, et al. Failure of prophylactic barbiturate coma in the treatment of severe head injury. J Neurosurg. 1985;62: 383–388.

85. Sydenham E, Roberts I, Alderson P. Hypothermia for traumatic head injury. Cochrane Database Syst Rev. 2009;(2):CD001048.

86. Saxena M, Andrews PJD, Cheng A. Modest cooling therapies (35°C to 37.5°C) for traumatic brain injury. Cochrane Database Syst Rev. 2008;(3):CD006811.

87. McIntyre LA, Fergusson DA, Hebert PC, et al. Prolonged therapeutic hypothermia after traumatic brain injury in adults: a systematic review. JAMA. 2003;298:2992–2999.

88. Henderson WR, Dhingra VK, Chittock DR, et al. Hypothermia in the management of severe traumatic brain injury: a systemic review and meta-analysis. Intensive Care Med. 2003;29:1637–1644.

89. Harris OA, Colford JM, Good MC, et al. The role of hypothermia in the management of traumatic brain injury: a meta-analysis. Arch Neurol. 2002;59:1077–1083.

90. Jurkovich GJ. Environmental cold-induced injury. Surg Clin North Am. 2007;87:247–267.

91. Roberts IG. Aminosteroids for acute traumatic brain injury. Cochrane Database Syst Rev. 1999;(3).

92. Alderson P, Roberts I. Corticosteroids for acute traumatic brain injury. Cochrane Database Syst Rev. 2005;(1):CD000196.

93. Yablon SA. Posttraumatic seizures. Arch Phys Med Rehab. 1993;74: 983–1001.

94. Temkin NR, Dimken SS, Wilensky AJ, et al. A randomized double-blind study of phenytoin for the prevention of post-traumatic seizures. N Engl J Med. 1990;323:497–502.

95. Clifton GL, Robertson CS, Choi SC. Assessment of nutritional requirements of head injured patients. J Neurosurg. 1986;64:895–901.

96. Clifton GL, Robertson CS, Grossman RG, et al. The metabolic response to severe head injury. J Neurosurg. 1984;60:687–696.

97. Deutschman CS, Konstantinides FN, Raup S, et al. Physiologic and metabolic response to isolated closed head injury. J Neurosurg. 1986; 64:89–98.

98. Rapp RP, Young B, Twyman D, et al. The favorable effect of early parenteral feeding on survival in head injured patients. J Neurosurg. 1983;58:906–912.

99. Ukleja A, Romano MM. Complications of parenteral nutrition. Gastroenterol Clin North Am. 2007;36:23–46.

100. Grahm TW, Zadrozny DB, Harrington T. The benefits of early jejunal hyperalimentation in the head-injured patient. Neurosurgery. 1989; 25:729–735.

101. Zabramski JM, Whiting D, Darouiche RO, et al. Efficacy of antimicrobial-impregnated external ventricular drain catheters: a prospective, randomized, controlled trial. Neurosurgery. 2003;98:725–730.

102. Owings JT, Bagley M, Gosselin R, et al. Effect of critical injury on plasma antithrombin activity: low antithrombin levels are associated with thromboembolic complications. J Trauma. 1996;41:396–405.

103. Kaufman HH, Moake JL, Olson JD, et al. Delayed and recurrent intracranial hematomas related to disseminated intravascular clotting and fibrinolysis in head injury. Neurosurgery. 1980;7:445–449.

104. Hoffman M, Monroe DM 3rd. A cell-based model of hemostasis. Thromb Haemost. 2001;85(6):958–965.

105. McQuay N Jr, Cipolla J, Franges EZ, et al. The use of recombinant activated factor VIIa in coagulopathic traumatic brain injuries requiring emergent craniotomy: is it beneficial? J Neurosurg. 2009;111(4): 666–671.

106. Brown CV, Foulkrod KH, Lopez D, et al. Recombinant factor VIIa for the correction of coagulopathy before emergent craniotomy in blunt trauma patients. J Trauma. 2010;68(2):348–352.

107. Hamilton MG, Hull RD, Pineo GF. Venous thromboembolism in neurosurgery and neurology patients: a review. Neurosurgery. 1994;34:280–296.

108. Olson JD, Kaufman HH, Moake J, et al. The incidence and significance of hemostatic abnormalities in patients with head injuries. Neurosurgery. 1989;24:825–832.

109. Black PM, Baker MF, Snook CP. Experience with pneumatic calf compression in neurology and neurosurgery. Neurosurgery. 1986;18: 440–444.

110. Frim DM, Barker FG 2nd, Poletti CE, et al. Postoperative low-dose heparin decreases thromboembolic complications in neurosurgical patients. Neurosurgery. 1992;30:830–832.

111. Jennett B, Bond M. Assessment of outcome after severe brain damage: a practical scale. Lancet. 1975;1:480–484.

112. Zink BJ. Traumatic brain injury outcome: concepts for emergency care. Ann Emerg Med. 2001;37:318–332.

113. Miller JD, Butterworth JF, Gudeman SK, et al. Further experience in the management of severe head injury. J Neurosurg. 1981;56: 650–659.

114. Chesnut RM, Ghajar J, Maas AIR, et al. Early Indicators of Prognosis in Severe Traumatic Brain Injury. New York, NY: Brain Trauma Foundation. Available at: Accessed April 2, 2010.