Closed Head Injury 

Updated: Feb 24, 2016
Author: Leonardo Rangel-Castilla, MD; Chief Editor: Brian H Kopell, MD 



In the United States, the incidence of closed head injury is estimated to be approximately 200 cases per 100,000 persons per year. In a population of 291.6 million people, this rate equates to more than 570,000 patients annually. Approximately 15% of these patients succumb to the injury upon arrival to the emergency department.[1, 2, 3, 4, 5, 6, 7, 8]

Traumatic injuries remain the leading cause of death in children and in adults aged 45 years or younger. Head injuries cause immediate death in 25% of acute traumatic injuries.[9] Traumatic brain injury (TBI) results in more deaths than does trauma to other specific body regions.[10] Penetrating intracranial injuries have worse outcomes than closed head injuries.[11] Motor vehicle collisions (MVCs) are the most common cause of closed head injuries for teenagers and young adults.[12] Alcohol or drug use contributes to as many of 38% of cases of severe head trauma in younger patients.[13] A recent development has been the apparent increase in brain injuries among the elderly; this increase is thought to be related to the use of anticoagulant and antiplatelet drugs.[14]

A CT scan of left frontal acute epidural hematoma is shown below.

CT scan of left frontal acute epidural hematoma (b CT scan of left frontal acute epidural hematoma (black arrow) with midline shift (white arrow). Note the left posterior falx subdural hematoma and left frontoparietal cortical contusion.

Head injury significantly contributes to deaths from trauma.[7] Annual mortality from closed head injuries is approximately 100,000 patients or 0%, 7%, and 36% of mild, moderate, and severe head injuries, respectively.[7, 9] Patients with severe head injury have a 30-50% mortality rate, and those who survive are often left with severe neurological deficits that may include a persistent vegetative state.[9] Permanent disability in survivors ranges from 10-100%, depending on the severity of the injuries. This produces more than 90,000 newly disabled patients annually, including 2500 who are in a persistent vegetative state.[7, 9]

The financial burden of head injuries in the United States is estimated to be $75-100 billion annually.[15] Injuries to the central nervous system tend to be the most costly on a per-patient basis because they often result in debilitating physical, psychological, and psychosocial deficits that, in turn, require extensive long-term rehabilitation and care.

The last 3 decades have been alternately exhilarating and frustrating for clinicians and researchers interested in TBI. Laboratory and bedside research has greatly improved our understanding of posttraumatic cerebral pathophysiology. These new insights have failed to make the transition to clinically used therapies. Many of the major clinical trials of the last decades have been negative studies that have shown us what does not work. Demonstrating the efficacy of new treatments has been extraordinarily difficult.[16]


Closed head injuries are classified as either primary or secondary. A primary injury results from the initial anatomical and physiological insult, which is usually direct trauma to the head, regardless of cause. A secondary injury results from hypotension, hypoxia, acidosis, edema, or other subsequent factors that can secondarily damage brain tissue (see Secondary injuries). Free radicals are thought to contribute to these secondary insults, especially during ischemia.

Primary injuries

The primary injury usually causes structural changes, such as epidural hematoma, subdural hematoma, subarachnoid hemorrhage, intraventricular hemorrhage, or cerebral contusion.


Cerebral concussion is defined as an altered mental state that may or may not include loss of consciousness that occurs as a result of head trauma. Concussion is also known as mild traumatic brain injury (MTBI). The American Academy of Neurology grading scale is widely used to categorize the degree of concussions.

Table 1. American Academy of Neurology Concussion Grading Scale (Open Table in a new window)

Grade 1

Grade 2

Grade 3

Transient confusion

Transient confusion


No loss of consciousness

No loss of consciousness

Brief or prolonged loss of consciousness

Concussion symptoms or mental status change resolves in 15 min or less

Concussion symptoms or mental status change resolves in more than 15 min



Sport-related concussions are frequent, with 300,000 cases reported each year.[17] Football players and boxers are particularly exposed to repetitive concussions, leading to the condition now known as chronic traumatic encephalopathy syndrome. Repetitive concussions may result in chronic subclinical motor dysfunctions linked to intracortical inhibitory system abnormalities.[18] Parkinsonian cognitive decline due to strionigral degeneration is now a well-known consequence of repetitive concussions; cumulative diffuse axonal injury effects in the midbrain are due to increased vulnerability to shear forces in that region. Increasing a player’s neck strength may be an effective way to minimize the risk of future concussions, as studies with Hybrid III dummies seem to indicate.[19]

Cerebral contusion

Cerebral contusions are commonly seen in the frontal and temporal lobes. They may accompany skull fracture, the so-called fracture contusion. The most worrisome trait of these contusions is their tendency to expand. This usually occurs from 24 hours to as long as 7-10 days after the initial injury. For this reason, cerebral contusions are often followed with a repeat head CT scan within 24 hours after injury.

Coup injuries (contusions) are caused by direct transmission of impact energy through the skull into the underlying brain and occur directly below the site of injury. Contrecoup injuries are caused by rotational shear and other indirect forces that occur contralateral to the primary injury. Rotational force causes the basal frontal and temporal cortices to impact or sweep across rigid aspects of the skull, the sphenoid wing, and petrous ridges. Delayed enlargement of traumatic intraparenchymal contusions and hematomas is the most common cause of clinical deterioration and death. However, progression of contusion is highly variable, and although most remain unchanged for days, a few enlarge, some quite rapidly.

In a retrospective study, well-known prognostic factors were found to predict contusion enlargement. The strongest prognostic factor is the presence of traumatic subarachnoid hemorrhage. The size of the intraparenchymal hemorrhage means that large lesions are probably in an active phase of progression at the time of the initial CT scan. The concurrent presence of a subdural hematoma was also predictive. Clinical features, such as the initial Glasgow Coma Score (GCS; see the Glasgow Coma Scale calculator) and intracranial pressure (ICP), were not predictive of progression. The ideal time for a rescan is unclear, although most of the growth seems to occur within the first 24 hours of injury.[20]

Epidural hematoma

The incidence of epidural hematomas is 1% of all head trauma admissions, as depicted in the image below.[21] Epidural hematomas most commonly (85%) result from bleeding in the middle meningeal artery. Epidural hematomas, however, may occur in locations other than in the distribution of the middle meningeal artery. Such hematomas may develop from bleeding from diploic vessels injured by overlying skull fractures.[9] Epidural hematomas are often associated with a "lucid interval," a period of consciousness between states of unconsciousness. The lucent period is presumed to end when the hematoma expands to the point that the brainstem is compromised.

CT scan of left frontal acute epidural hematoma (b CT scan of left frontal acute epidural hematoma (black arrow) with midline shift (white arrow). Note the left posterior falx subdural hematoma and left frontoparietal cortical contusion.

Subdural hematoma

The most common surgical intracranial lesion is a subdural hematoma. These occur in approximately 20-40% of patients with severe injuries, as depicted in the image below.[13, 22] A surface or bridging vessel (venous) can be torn because the brain parenchyma moves during violent head motion. The resulting bleeding causes a hematoma to form in the potential space between the dural and arachnoid. A lucid interval is less likely to develop in this type of injury than in epidural hematomas. Subdural hematomas can be the result of an arterial rupture as well; these hematomas have the peculiar location in the temporoparietal region and differ in form from those caused by the bridging vein rupture, which typically rupture in the frontoparietal parasagittal region. Hematoma thickness and the midline shift of the brain are often analyzed; when the midline shift exceeds the hematoma thickness (positive displacement factor), a poorer prognosis has been found.[23]

CT scan of left frontoparietal acute subdural hema CT scan of left frontoparietal acute subdural hematoma (black arrow). Note the moderate amount of midline shift.

Intraventricular hemorrhage

An intraventricular hemorrhage is another intracerebral lesion that often accompanies other intracranial hemorrhages, as depicted in the image below. Intraventricular blood is an indicator of more severe head trauma. Intraventricular blood also predisposes the patient to posttraumatic hydrocephalus and intracranial hypertension, which may warrant placement of an intraventricular catheter (if emergent drainage needed) or ventriculoperitoneal shunt for chronic hydrocephalus.

CT scan of bilateral acute intraventricular hemorr CT scan of bilateral acute intraventricular hemorrhages (black arrow). Note the comminuted skull fractures that involve bilateral frontal, temporal, and parietal bones (white arrow). Note the ischemic changes in both frontal lobes, subarachnoid hemorrhages in the intrahemispheric fissure and left frontal lobe, and multiple intraparenchymal hemorrhages in both frontal poles.

Diffuse axonal injury

Despite the absence of any intracranial mass lesion or history of hypoxia, some patients remain unconscious after a TBI. Brain MRI studies have demonstrated a clear correlation between white matter lesions and impairment of consciousness after injury. The deeper the white matter lesion, the more profound and persistent the impairment of consciousness.[24]

The usual cause for persistent impairment of consciousness is the condition referred to as diffuse axonal injury, as depicted in the image below. Approximately 30-40% of individuals who die from TBI reveal postmortem evidence of DAI and ischemia.[25] This type of injury commonly results from traumatic rotation of the head, with mechanical forces that act on the long axons, leading to axonal structural failure. DAI is caused by an acceleration injury and not by contact injury alone. The brain is relatively incompressible and does not tolerate tensile or shear strains well. Slow application of strain is better tolerated than rapid strain. The brain is most susceptible to lateral rotation and tolerates sagittal movements best.[26]

MRI of the brain that shows diffuse axonal injury MRI of the brain that shows diffuse axonal injury (DAI) and hyperintense signal in the corpus callosum (splenium), septum pellucidum, and right external capsule.

Recent studies suggest that the magnitude of rotational acceleration needed to produce DAI requires the head to strike an object or surface. These factors also increase the likelihood that DAI will be accompanied by other intracranial lesions.[26] These mechanical forces physically dissect these axons into proximal and distal segments. If a sufficient number of axons are involved, profound neurologic deficits and unconsciousness may ensue.

These same forces may act on the cerebral circulation, causing disruption of vessels and various forms of micro–intracerebral hemorrhages and macro–intracerebral hemorrhages, including Duret hemorrhages, which are commonly lethal when they occur in the brainstem. Duret hemorrhages of the midbrain and pons are small punctate hemorrhages that are often caused by arteriole stretching during the primary injury, as depicted in the image below. They may also result during transtentorial herniation as a secondary injury when arterial perforators are compressed or stretched.

MRI of the brain (sagittal view) that shows a Dure MRI of the brain (sagittal view) that shows a Duret hemorrhage in the splenium of the corpus callosum.

A recent study indicates that DAI and younger age may contribute to an increased risk of developing dysautonomia.[27]

Secondary injuries

Secondary insults can take many forms and can be summarized as follows:

  • Secondary intracranial insults to the brain

    • Hemorrhage

    • Ischemia

    • Edema

    • Raised intracranial pressure (ICP)

    • Vasospasm

    • Infection

    • Epilepsy[28]

    • Hydrocephalus

  • Secondary systemic insults to the brain

    • Hypoxia

    • Hypercapnia

    • Hyperglycemia

    • Hypotension

    • Severe hypocapnia

    • Fever

    • Anemia

    • Hyponatremia

The major focus in the management of acute closed head injury is the prevention of secondary injuries and the preservation of neurological functions that are not damaged by the primary injury.

Posttraumatic vasospasm can be a cause of ischemic damage after severe traumatic brain injury, with parenchymal contusions and fever being risk factors. Diffuse mechanical injury and activation of inflammatory pathways may be secondary mechanisms for this vasospasm. Patients with parenchymal contusions and fever may benefit from additional screening.[29]

Cerebral ischemia

Cerebral ischemia is inadequate oxygen perfusion to the brain as a result of hypoxia or hypoperfusion. The undamaged brain tolerates low PaO2 levels better than the severely injured brain. Traumatized brain tissues are very sensitive to even moderate hypoxia (90 mm Hg). Gordon and Ponten proposed 2 explanations for this phenomenon: (1) Respiratory alkalosis may shift the oxygen-hemoglobin curve to the left, thereby increasing the affinity of the hemoglobin to the oxygen and decreasing the ease of oxygen release, and (2) uneven cerebral blood flow (CBF) may result from focal vasospasm with loss of focal autoregulation in the area of injured brain tissue.[26] Approximately one third of patients with severe head injuries have been demonstrated to experience ischemic levels of CBF.[30]

CBF is normally kept constant over a range (about 50-150 mm Hg) of cerebral perfusion pressure, as depicted in the image below. This is made possible by adjustments in vascular tone known as autoregulation (solid line). In patients with brain trauma, this autoregulation may malfunction, and CBF may become dependent on the CPP (dashed lines). Autoregulation is absent, diminished, or delayed in 50% of patients with severe head injuries.[30] The lowest CBF values occur within the first 6-12 hours after injury.[31, 32, 33, 34] The overall outcome of patients who experience ischemia is much worse than that of initially nonischemic patients.[30, 35, 36] The initial ischemia is thought to cause permanent irreversible damage even if CBF is eventually optimized. The use of Xenon CT scan to measure CBF is now part of the armamentarium to diagnose and treat abnormalities in the CBF.

Cerebral blood flow/cerebral perfusion pressure ch Cerebral blood flow/cerebral perfusion pressure chart.

Brain edema

Brain edema is another form of secondary injury that may lead to elevated ICP and frequently results in increased mortality. Brain edema is categorized into 2 major types: vasogenic and cellular (or cytotoxic) edema.

Vasogenic edema occurs when a breach in the blood-brain barrier allows water and solutes to diffuse into the brain. Most of this fluid accumulates in the white matter and can be observed on head CT scans as hypodense white matter (on T1-weighted images) or as a bright signal area on the T2-weighted MRI. The mechanism of cellular (cytotoxic) edema is less clear. Theories include the increased uptake of extracellular potassium by the injured brain cells or the transport of HCO3- and H+ for Cl- and Na+ by the injured brain tissue as the mechanism of insult.[26]

In one study, diffusion-weighted MR imaging was used to evaluate the apparent diffusion coefficient (ADC) in 44 patients with TBI (GCS < 8) and in 8 healthy volunteers. Higher ADC values have been associated with vasogenic edema, and lower ADC values have been associated with a predominantly cellular form of edema. Regional measurements of ADC in patients with focal and diffuse injury were computed. The final conclusion was that the brain swelling observed in patients with TBI appears to be predominantly cellular, as signaled by low ADC values in brain tissue with high levels of water content.[37]


The incidence of closed head injury is estimated to be approximately 200 cases per 100,000 persons per year.[2, 3, 4, 5, 6, 7]

Traumatic injuries remain the leading cause of death in children and in adults aged 45 years or younger.

The incidence varies by age, but children and young people experience closed head trauma more often than older populations.




The Glasgow Coma Scale (GCS), first introduced by Teasdale and Jennett in 1974, has been the standard for objectively assessing individuals with traumatic head injuries (see Table 2).[38] It comprises motor, verbal, and eye scores. The overall score generally refers to the best response/examination obtained within the first 6-8 hours after injury and following resuscitation and is considered to be a predictor of the patient's overall outcome.[39, 40, 41, 13, 42]

The GCS has 2 main advantages in that it provides a reproducible, objective evaluation of neurological status, and it is a relatively simple way to monitor a patient's neurological condition over time. The GCS has shortcomings because its reliability depends on the absence of confounding factors (eg, sedation, paralytics, hypothermia, hypotension, hypoxia). Additionally, it cannot compensate for lack of eye opening in patients with periorbital trauma or loss of verbal response in intubated patients, and it omits brainstem reflex assessment.[43, 44]

Most clinicians assign a verbal score of 1 and apply the modifier "T" to intubated individuals. This may not be an accurate assessment of the patient's true verbal score.[45] The motor component of the GCS score is most predictive of the severity of the brain injury and correlates most strongly with overall outcome.[45]

The GCS score is often used to categorize the severity of head injury into mild (15-13), moderate (12-9), or severe (8 or less). In general, mild head injury does not usually involve significant primary brain injury, is not associated with neurological deficits, and may or may not include loss of consciousness. Approximately 75% of head injuries are categorized as mild to moderate in nature.[46] Most authorities agree that a patient with a severe head injury is one who is unconscious and unable to follow simple commands.

Table 2. Glasgow Coma Scale (Open Table in a new window)

Best Motor Response

Obeys commands




Withdraws (abnormal flexion)


Flexion (decorticate posturing)


Extension (decerebrate posturing)


No response


Verbal Response





Inappropriate words


Incomprehensible sounds


No response


Eye Opening



To command


To pain


No response






Patients should also be evaluated for basic brainstem reflexes. This assessment includes the evaluation of pupillary reflex, corneal reflex, gag/cough reflex, oculocephalic reflex, vestibuloocular reflex, and spontaneous breathing. Pupillary asymmetry or anisocoria of more than 1 mm (up to 1 mm may be physiological) must be attributed to an intracranial lesion until proven otherwise.[45]

Careful attention must be given to the evaluation of spontaneous breathing in the ventilated patient. A common pitfall is to mistakenly assess a patient as taking spontaneous breaths when the sensitivity on the ventilator is set to a level that triggers a mechanical breath at the slightest effort by the patient.

Proper evaluation entails setting the ventilator sensitivity to zero, then reevaluating the patient after a few seconds, which does not allow time for hypoxia or hypercapnia to develop. The presence of any of the above reflexes confirms that the patient has at least basic brainstem reflexes. A complete absence of brainstem reflexes is an ominous sign.


Causes include the following:

  • Motor vehicle collisions

  • Motorcycle collisions

  • Assaults

  • Falls

  • Other



Laboratory Studies

Lab studies include the following:

  • CBC count including platelet count

  • Blood chemistries

  • Prothrombin time (PT) or international normalized ratio (INR)

  • Activated partial thromboplastin time (aPTT)

  • Anticonvulsant (eg, phenytoin) level - For patients who have been previously loaded or who were previously on anticonvulsant medications to ensure therapeutic levels

Imaging Studies

CT scanning of the head is the criterion standard for patients with acute closed head injuries.[16] A head CT scan is warranted, except for patients with only minor head trauma who are neurologically intact and not intoxicated with drugs or alcohol. Advantages and disadvantages of head CT scan are summarized in the following table. Table 3. Advantages and Disadvantages of CT scanning in the Head Trauma Evaluation

Table. (Open Table in a new window)



Noninvasive and rapid

Traumatic vascular lesions may be missed.

Very sensitive for acute hemorrhage

DAI is likely to be missed.

Defines nature of ICH (ie, SDH†, SAH‡)

Motion artifact may limit study.

Defines anatomical location of lesion

Posterior fossa lesions are poorly depicted.

Identifies fractures of the cranium

Depressed skull fractures at the vertex (or along the plane of an axial scan) are poorly depicted.

Sensitive to detecting intracranial air

The scanner has a weight limit, and a patient may be too heavy.

Sensitive in identifying foreign objects

A patient may decompensate while in the scanner.

*Intracranial hemorrhage

†Subdural hemorrhage

‡Subarachnoid hemorrhage


CT scans are helpful in assessing the degree of intracranial injury, in predicting outcome, and, if findings are normal, in avoiding unnecessary hospitalization.[47, 48] CT scans are very sensitive to acute hemorrhage or skull fractures. CT scans aid in evaluating (1) intracranial hemorrhage, (2) skull fractures, (3) mass effect and midline shift, (4) obliteration of the basal cisterns, and (5) evidence of herniation (subfalcine, tonsillar, or uncal).

CT scans cannot diagnose a concussion (which is a clinical diagnosis) and are poor for diagnosing DAI. If DAI has occurred, CT scans may show small hemorrhages in the corpus callosum and cerebral peduncles. In this case, an MRI of the brain should be obtained on an elective basis when the patient is clinically stable because no effective treatment of DAI is currently available. MRI is more sensitive for detecting brainstem injuries, posterior fossa lesions, and brain edema. For advantages and disadvantages of CT scanning in patients with closed head injuries, see Table 3.

As a general rule, a repeat head CT scan is recommended within 4-8 hours of the initial scan in patients with intracranial hemorrhages and/or coagulopathies.[49, 50, 51] A repeat head CT scan is recommended sooner in patients who are deteriorating neurologically.

Spinal cord injuries should be considered possible in patients with closed head injuries. Spinal cord injuries are present in up to 10% of these patients.[45] Accordingly, the cervical spine should be evaluated (with 3 views) during the initial evaluation. C1-C2 should be evaluated with a thin-cut CT scan in intubated patients. If any abnormalities are noted on the initial cervical plain radiographs, this area should be further evaluated with a CT scan. An MRI may be necessary to image a spinal cord injury. A rigid cervical collar (Philly) should remain on at all times while the patient is being evaluated.

The results of one study found that computed tomographic angiography (CTA) findings used in addition to other screening criteria may help identify injuries not captured using conventional screening guidelines alone.[52]

In a systematic review of the clinical utility of single photon emission CT (SPECT) for TBI, SPECT was shown to have some advantages over CT and MRI in the detection of mild TBI and to have excellent negative predictive value. The authors suggest it may be an important second test in settings where CT or MRI are negative after a closed head injury with post-injury neurologic or psychiatric symptoms. The most commonly abnormal regions revealed by SPECT in cross-sectional studies were frontal (94%) and temporal (77%) lobes. SPECT was found to outperform both CT and MRI in both acute and chronic imaging of TBI, particularly mild TBI. It was also found to have a near 100% negative predictive value.[53]

The findings of one study strongly suggest that diffusion tensor imaging (DTI), but not "classic" MRI sequences, is a more precise and accurate measurement to assess a degree of brain injury after blunt trauma. DTI is a valuable research tool to further the understanding of pathophysiological mechanism(s) evoked by blast injury and may become a prognostic tool.[54]

Henninger et al found that in 136 patients 50 years or older who were admitted to a neurologic/trauma ICU, preexisting leukoaraiosis (white matter hyperintensities) was significantly associated with a poor outcome at 3 and 12 months. According to the study findings, the independent association between leukoaraiosis and poor outcome remained when the analysis was restricted to patients who survived up to 3 months, had moderate-to-severe TBI [enrollment Glasgow Coma Scale (GCS) ≤12; P = 0.001], or had mild TBI (GCS 13-15; P = 0.002), respectively.[55]

According to one study, meningeal enhancement on contrast-enhanced fluid-attenuated inversion recovery (FLAIR) images can help detect traumatic brain lesions and other abnormalities that are not identified on routine unenhanced MRI in symptomatic patients with mild traumatic brain injury. The authors therefore recommended contrast-enhanced FLAIR MR imaging when a contrast MR study is indicated in patients with symptomatic prior closed mild head injury. In a study of 25 patients, 3 additional cases of brain abnormality were detected with the contrast-enhanced FLAIR images. Meningeal enhancement was identified on contrast-enhanced FLAIR images in 9 cases, while the other routine image sequences showed no findings of traumatic brain injury. Overall, the additional contrast-enhanced FLAIR images revealed more extensive abnormalities than routine imaging in 37 cases.[56]

Patients who arrive with a decreased GCS and normal findings on head CT scans may have another condition that needs to be considered. These include the following:

  • Acute ischemic stroke (within 24 h) that is not seen on head CT scan

  • Postictal state

  • Spinal cord injury

  • Intoxication or effects of illicit drug use

  • Prior medical conditions (Speaking with family members may help in differentiating acute from chronic conditions.)

Other Tests

Order serum sodium, urine specific gravity, urine osmolarity, and serum osmolarity tests for individuals with urine output of 250 mL/h or more for 3 or more consecutive hours (pediatric patients, >3 mL/kg/h) and for patients who are thought to have diabetes insipidus. Large doses of mannitol can mask diabetes insipidus by producing a high urine output.


Patients with a severe brain injury (GCS score < 8), those who have labile blood pressure, those who require intensive care monitoring, and those who need surgical intervention are likely to require the placement of an indwelling urinary catheter (Foley), placement of a central venous access catheter, and invasive blood pressure monitoring via an arterial line.

ICP monitoring in patients with closed head injuries is a matter of controversy; however, most authors agree that invasive ICP monitoring is warranted in patients with a GCS score of 8 or less and an abnormal CT scan finding, in patients with suspected severe brain edema, or in any situation in which the ICP is suspected to be significantly elevated. The ICP offers data that supplement a reliable neurological examination and can be crucial in patients whose examination findings are affected by sedatives, paralytics, and other factors.

  • Patients with an abnormal head CT scan finding, a GCS score of less than 8, or both who require emergent surgery on another organ system should also be considered for some form of ICP monitoring before going to the operating room (or perhaps in the operating room) because frequent neurological examinations are not possible in this setting.

  • ICP monitoring can take 1 of 2 forms, either as an intraventricular catheter or an intracranial fiberoptic monitor (Camino).

    • The intraventricular catheter is preferred in closed head injuries when the ventricles are large enough to accommodate a catheter. The advantage of the catheter is the ability to drain CSF if the ICP is elevated (>20 mm Hg), although ventricles compromised by mass effect make draining much CSF difficult. An accurate pressure reading can be lost if the ventricle collapses around the catheter tip during drainage.

    • The advantage of the fiberoptic catheter is that ICP can be monitored in patients who have very small ventricles, in whom ventriculostomy catheters cannot be inserted. The pressure measurements are not prone to fluctuations in ventricular size. Either procedure provides adequate ICP monitoring.



Medical Care

Intracranial hypertension is a common neurologic complication in patients who are critically ill. Intracranial hypertension is the common pathway in the presentation of traumatic head injury. The underlying pathophysiology of increased intracranial pressure (ICP) is the subject of intense basic and clinical research, which has led to advances in the understanding of the physiology related to ICP. Few specific treatment options for intracranial hypertension have been subjected to randomized trials, however, and most management recommendations are based on clinical experience.

Normal values of intracranial pressure

In healthy individuals with closed cranial fontanelles, central nervous systems contents, including brain, spinal cord, blood, and cerebrospinal fluid (CSF), are encased in a noncompliant skull and vertebral canal, constituting a nearly incompressible system. In the average adult, the skull encloses a total volume of 1450 mL: 1300 mL of brain, 65 mL of CSF, and 110 mL of blood. The Monroe-Kellie hypothesis states that the sum of the intracranial volumes of blood, brain, CSF, and other components is constant, and that an increase in any one of these must be offset by an equal decrease in another.

The reference range of ICP varies with age. Values for pediatric subjects are not as well established. Normal values are less than 10-15 mm Hg for adults and older children, 3-7 mm Hg for young children, and 1.5-6 mm Hg for term infants. ICP can be subatmospheric in newborns. ICP values greater than 20-25 mm Hg require treatment in most circumstances. Sustained ICP values of greater than 40 mm Hg indicate severe, life-threatening intracranial hypertension.

Cerebral dynamics overview

Cerebral perfusion pressure (CPP) depends on the mean systemic arterial pressure (MAP) and ICP, which is determined by the following relationship:

CPP = MAP – ICP where MAP = (1/3 systolic BP) + (2/3 diastolic BP)

As a result, CPP can be reduced from an increase in ICP, a decrease in blood pressure, or a combination of both factors. Through the normal regulatory process called pressure autoregulation, the brain is able to maintain a normal cerebral blood flow (CBF) with a CPP that ranges from 50-150 mm Hg. At CPP values less than 50 mm Hg, the brain may not be able to compensate adequately, and CBF falls passively with CPP.

After injury, the ability of the brain to autoregulate may be absent or impaired. When CPP is within the normal autoregulatory range (50-150 mm Hg), this ability of the brain to pressure autoregulate also affects the response of ICP to a change in CPP. When pressure autoregulation is intact, decreasing CPP results in vasodilation, which allows the CBF to remain unchanged. This vasodilation can result in an increase in ICP. Likewise, an increase in CPP results in the vasoconstriction of cerebral vessels and may reduce ICP. When pressure autoregulation is impaired or absent, ICP decreases and increases with changes in CPP.[57]

Perhaps the least invasive method of lowering ICP is to elevate the head of the patient to 30°. Some researchers have demonstrated improved ICP control with elevation (head of bed) to 45°, but recent evidence from using multimodality monitoring has suggested a 30° head elevation for maximum benefits.[26] Note that head elevation may reduce cerebral perfusion, even as it lowers ICP.

Hyperventilation became a popular means of reducing ICP in the 1970s. However, recent studies raise concern that aggressive hyperventilation exacerbates cerebral ischemia. One study shows that patients who were hyperventilated to a PCO2 level of 25 mm Hg had worse outcomes than patients who were kept at a nearly normal PCO2 level.[50] In addition, in most patients, hyperventilation is not necessary to control ICP.[58, 59, 33, 60] Currently, hyperventilation (PCO2 of 30-35 mm Hg) is recommended to reduce ICP for only a short period, as a temporizing measure while other methods of ICP control are initiated. Hyperventilation reduces ICP only temporarily, progressively losing effectiveness after 16 hours of continuous use.[61]

Mannitol probably has several mechanisms of action. One obvious mechanism is through osmotic diuresis via drawing edema from the cerebral parenchyma. This usually takes 15-30 minutes, and the effect usually lasts 1.5-6 hours. Another mechanism is by immediate plasma expansion and decreased blood viscosity, thereby improving blood flow and eventually resulting in intracranial vasoconstriction in an attempt to maintain constant blood flow. This vasoconstriction ultimately leads to decreased intracranial volume (Monroe-Kelly hypothesis) and decreased ICP.[62, 63] Mannitol is also considered a free radical scavenger.[64] Administration of this drug in severe traumatic brain injury patients studied both with jugular bulb oxygen saturation[65] as well as with multimodal brain monitoring[66] suggests a potential change in the internal milieu that would improve cerebral oxygenation.

Serial serum osmolarity levels must be checked to maintain an osmolarity of no greater than 315-320 mOsm/kg H2 O to avoid acute renal failure. Some studies have raised concern that the early use of mannitol can lead to hypotension, with an associated worse outcome.[67] For this reason, patients treated with mannitol must be kept euvolemic with isotonic fluid resuscitation as required.

Although some evidence suggests that barbiturates may be effective in lowering refractory ICP, such administration often causes depressed myocardial function and CPP.[45] These drugs often have an associated morbidity and do not significantly change outcome.[68, 11]

A barbiturate-induced coma with EEG burst suppression is often a "last ditch effort" to reduce the ICP and should be reserved only for patients with refractory ICP who are unresponsive to other measures. One may even consider decompressive craniectomy prior to the use of barbiturates. Barbiturate serum levels are a poor estimate of therapeutic effect and should not be followed for treatment purposes. For this reason, all patients should have EEG monitoring to monitor for induced burst suppression. A loading dose of pentobarbital can be administered as a 10 mg/kg bolus (over 30 min), followed by 5 mg/kg/h for 3 doses, titrated to a low level of bursts per minute (2-5). Barbiturates are contraindicated in hypotensive patients.

One must give special attention to preventing hypotension.[25, 35] Data from the Traumatic Coma Data Bank (TCDB) reveal that hypotension in patients with severe TBI increases the mortality rate from 27% to 50%.[26] Traditional management has included fluid restriction in order to minimize cerebral edema, but this practice may be dangerous in patients who already have intravascular volume depletion.

Cerebral edema may occur regardless of the amount of intravenous fluid administered, and hypervolemia, per se, does not cause brain edema if the serum sodium level and osmolarity are within normal limits.[26] However, managing patients who have closed head injuries with liberal amounts of hypotonic intravascular fluid may cause intracerebral hemorrhages to blossom. Smaller amounts of hypertonic solutions may be equally effective without the risk of fluid overload.[69, 70] The ultimate goal of the management of patients with closed head injuries is to maintain a state of euvolemia. In a euvolemic patient who is hemodynamically stable, two-thirds maintenance of isotonic solution is recommended. Avoid hypotonic fluids because they may decrease serum osmolarity and increase brain swelling.

In addition, patients with closed head injuries are prone to acute coagulopathies. These coagulopathies are often the result of release of thromboplastin and tissue-activating protein from injured brain tissue. The release of these proteins leads to abnormal intravascular clotting, which consumes clotting factors, platelets, and fibrinogen and ultimately results in elevated PT and aPTT. In patients with acute intracranial hemorrhages, these coagulopathies must be addressed and corrected promptly.

Fresh frozen plasma (FFP) transfusions until the coagulopathy is corrected is the preferred method. This is especially true for individuals who are taking anticoagulants (eg, warfarin) and who are at high risk of continued bleeding. Winter and colleagues have shown that prophylactic FFP administration in individuals with closed head injuries is of no benefit.[71] Vitamin K plays an important role in correcting the coagulopathy; however, it usually takes 24-48 hours to be activated. During this interval, the patient's intracranial hemorrhage is likely to worsen. Recombinant activated factor VII (rFVIIa) is a relatively new pharmaceutical agent developed for use in patients with hemophilia in whom inhibitors to clotting factors VIII or IX have developed.

The use of rFVIIa in neurosurgery has lagged behind its use in other fields, although the body of literature on such uses is growing. Various uses are pertinent to the neurosurgeon, including the treatment of patients with coagulation disorders, those patients who have experienced trauma, and those patients with perioperative hemorrhage, intracerebral hemorrhage, or subarachnoid hemorrhage. rFVIIa is a safe and effective agent with the potential to revolutionize the treatment of neurosurgical patients with hemorrhage. Cost is a major impediment to the widespread use of rFVIIa, and some evidence suggests that its use in the neurosurgical population may be subject to higher risk than in other populations studied thus far. Although further study is needed to better delineate the safety and efficacy of the drug, rFVIIa is clearly an agent with tremendous promise.[72] In placebo-controlled trials, the off-label use of rFVIIa in high doses increased the risk of arterial events but not venous thromboembolicevents, particularly among older patients.[73] Until more clinically significant data emerge, caution should be exercised when using rFVIIa in off-label settings.[74]

Pyrexia commonly occurs in patients with head injuries, possibly because of posttraumatic inflammation, direct damage to the hypothalamus, or secondary infection. The most common cause is fever secondary to an underlying infection. Less common is an unexplained fever or "neurogenic" fever estimated to occur in approximately 8% of patients who have head injuries with pyrexia.[75] Regardless of the cause of the elevated temperature, pyrexia alone increases metabolic expenditure, glutamate release, and neutrophil activity, while causing blood-brain barrier breakdown.

Pyrexia is also thought to exacerbate oxygen radical production and cytoskeletal proteolysis.[76, 77] These changes may further compromise the injured brain and worsen neuronal damage. For this reason, the source of the fever must be identified and corrected.

Although the source of the infection is sought, maintain body temperature in a normothermic range with acetaminophen. However, despite sound physiological justification for treating fever in brain-injured patients, no evidence indicates that doing so improves outcome.

Hyperglycemia has also been shown to have a detrimental effect on induced brain ischemia. Clinical trials support the correlation between hyperglycemia and poor overall outcome in patients with head injuries and recommend that euglycemia be maintained at all times.[78]

Some patients with severe head injuries may develop hypertension, either from an exacerbation of a chronic process or as a result of the head injury. Keep systolic blood pressure less than 180 mm Hg, particularly in patients who have an intracranial hemorrhage. This value requires adjustment for patients with a history of uncontrolled hypertension. If possible, avoid nitroprusside because it is a cerebral vasodilator and may actually increase ICP. A nicardipine grip is preferred in patients whose blood pressure is difficult to control. Corticosteroids have occasionally been used but have no proven benefit for patients with severe head injuries.[11]

Effective treatment of intracranial hypertension involves the meticulous avoidance of factors that precipitate or aggravate increased ICP. When ICP becomes elevated, ruling out new mass lesions that should be surgically evacuated is important. Medical management of increased ICP should include sedation and paralysis, drainage of CSF, and osmotherapy with either mannitol or hypertonic saline. For intracranial hypertension refractory to initial medical management, barbiturate coma, hypothermia, or decompressive craniectomy should be considered. Steroids are not indicated and may be harmful in the treatment of intracranial hypertension that results from TBI.[57]

Surgical Care

As a general rule, indications for surgery include any intracranial mass lesion that causes significant or progressive neurological compromise, particularly a decreased level of consciousness. The overall outcome of individuals with an intracranial lesion that causes significant mass effect is improved with rapid decompression; therefore, operating on these patients as soon as possible is advisable.

Before operating, one must always consider the patient's condition and refrain from relying solely on radiographic evidence. For example, some patients with severe cerebral atrophy (eg, elderly patients) may accommodate a large intracranial hemorrhage, whereas most young individuals may experience neurological deficits with relatively smaller intracranial hemorrhages. Note that some intracranial hemorrhages may be actively bleeding during the initial head CT scan; what may appear as relatively small on the initial scan may actually become quite significant in a short period of time. In this case, the patient's physical examination findings are more valuable in evaluating his or her intracranial status than the initial head CT scan findings.

Some authors have suggested a decompressive craniotomy (ie, removal of a bone flap with or without dural opening) to provide more space for the brain to expand, for the treatment of uncontrollable ICPs before irreversible ischemic brain damage has occurred. The role of decompressive craniotomy in the absence of compressive pathology (such as subdural hematoma) in patients with closed head injuries has not been well documented. Most authors, however, agree that children benefit more from decompressive craniotomies than adults, and some authors are advocates of very early decompressive craniotomies for uncontrollable ICP in children.[79] It seems clear that older individuals, particularly those older than 50 years, do less well with elective decompressive craniotomies.[80, 81]

One study investigated complications associated with the use of a dural substitute, the Neuro-Patch, during decompressive craniectomy. Results suggested that it has not been found to increase the incidence of neurosurgical site infection and hydrodynamic complications, including subdural hygroma and cerebrospinal fluid leakage, following decompressive craniectomy or cranioplasty for severe TBI. However, patients with the Neuro-Patch more often encounter extra-axial hematoma at the site of craniectomy, which forms a compressive lesion on the adjacent brain.[82]

A metaanalysis of 2 randomized controlled trials, despite a relatively small number of patients, convincingly and strongly suggested that the early induction of hypothermia to and below 35°C for 48 h before or soon after craniotomy improves outcomes in patients with intracranial hematomas after severe traumatic brain injury. The study is trendsetting rather than a recommendation and requires confirmation by a prospective clinical trial.[83]

Despite the poor overall prognosis of patients with closed head injury and bilateral fixed and dilated pupils, one study suggested that a good recovery may be possible if an aggressive surgical approach is taken, particularly those with extradural hematoma. Of 82 patients who underwent surgery for extradural or subdural hematoma, in those with extradural haematoma, the mortality rate was 29.7%, with a favorable outcome seen in 54.3%. In patients with acute subdural haematoma, the mortality rate was 66.4%, with a favorable outcome seen in 6.6%.[84]


Obtain consultations as necessary for other accompanying injuries (eg, plastic surgeons for facial lacerations), realizing that the patient's closed head injury takes precedence over all other non—life-threatening injuries once the patient is stabilized.


Before, physicians thought that patients with closed head trauma should be on NPO status. Now, the new modality is to provide nutrition as soon as possible. The consequences of hypermetabolism, hypercatabolism, and an altered immune function are part of the response to traumatic head injury. Once a person with acute traumatic brain injury develops this hyperdynamic state, the resultant excessive protein breakdown ensues. This can lead to malnutrition. Lack of nutrient supplementation in these patients is associated with increased morbidity and mortality. Enteral nutrition is the preferred mode of feeding but is often not tolerated in the patient with head injury. Parenteral nutritional support can be given to these patients without worsening cerebral edema.[85]


The patient should be prescribed bed rest.



Medication Summary

Neurosurgeons have commonly given prophylactic anticonvulsants for individuals with intracranial hemorrhages. The appropriate duration of treatment is not well established. Individuals who have experienced seizure activity can reasonably be treated with anticonvulsants for 6-12 months, after which reevaluation is necessary. Temkin and colleagues suggest that treatment for longer than 8 days after injury does not reduce the frequency of long-term seizure disorders. Anticonvulsants may be used to treat early (< 7 d) posttraumatic seizures. According to Greenberg, prophylactic anticonvulsants do not reduce the frequency of late (>7 d) posttraumatic seizures.[86]

The recommended anticonvulsant medication in adults is phenytoin or fos-phenytoin (18 mg/kg of loading dose), ensuring therapeutic levels of 10-20 mg/dL. Note that a "therapeutic level" does not necessarily have a direct bearing on adequate control of seizures. A relatively common adverse effect of chronic phenytoin use is gingival hyperplasia and hirsutism, which precludes chronic use in children. Phenobarbital is an acceptable alternative in children who require long-term anticonvulsive therapy (10-20 mg/kg loading dose, then 3-5 mg/kg/d divided bid/tid) to achieve a therapeutic level of 10-40 mg/dL.

The new antiepileptic drug levetiracetam is used in the setting of acute brain injury for seizures treatment or prophylaxis; it is a desirable alternative to phenytoin. It is associated with fewer complications when it is used as monotherapy. Checking for therapeutic levels is not needed. The dose is 500 mg bid IV or PO and advance to 1000 mg bid.[87]


Class Summary

These agents are indicated for short-term (1 wk) or long-term (6- to 12-mo) posttraumatic seizure control for patients who have experienced posttraumatic seizure activity. Phenobarbital may be considered in children for long-term anticonvulsive therapy.

Phenytoin (Dilantin)

Used for acute seizure prophylaxis in individuals with closed head injuries.

Phenobarbital (Barbita, Luminal, Solfoton)

Used for acute seizure prophylaxis in children with closed head injuries.



Further Outpatient Care

Most patients with moderate-to-severe head injuries likely benefit from outpatient physical and occupational therapy. Once patients' acute issues have been addressed, many patients require cognitive rehabilitation (outpatient or inpatient).

Further Inpatient Care

Most patients require physical and occupational therapy, depending on the severity of the head injury.

Inpatient & Outpatient Medications

Initially, an anticonvulsant regimen should be started for patients with moderate or severe head injuries. Cease administration if no seizure activity occurs within the first 7 days after injury. For patients who have seizure activity in this time period, or who have undergone surgical procedures, one may opt to continue anticonvulsants for 6-12 months.


Once the patient's acute issues have been addressed, seek long-term placement (eg, long-term acute care, skilled nursing facility, inpatient cognitive rehabilitation center) if the patient continues to require significant medical attention or assistance (eg, because of ventilation, need for significant nursing care).


See Secondary injuries.


The prognosis is affected by many factors, including (1) the type of injury (penetrating vs blunt), (2) severity of the injury and accompanying neurological deficit, (3) the age of the patient, (4) comorbid conditions, and (5) secondary injuries.

The Glasgow Coma Scale (GCS) has been reported to be the most predictive of neurological outcome at 1 year after severe head injury, while the 24-hour GCS score is the strongest predictor of cognitive recovery at 2 years after injury in patients with moderate-to-severe head injury.[88, 89]

In a study of mortality in 44 elderly patients (≥75 years) who underwent an operation for acute subdural hematoma, patients who were independent had a 1-year mortality of 42%, versus 69% for dependent patients (median follow-up, 4.2 yr; range, 2.5 to 6.4 yr). Patients taking antithrombotics had a 56% mortality after the first postoperative year, versus 30% for those not taking antithrombotics. All patients with an admission Glasgow coma scale score of 3-8 died within the first postoperative year if they used antithrombotics or were dependent before the injury.[90]

Patients in the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) who were treated via craniotomy or craniectomy for SDH between 2005 and 2012 were studied for mortality, other adverse outcomes, and length of hospital stay. The most common individual adverse events were death (18% of the patients died within 30 days of surgery) and intubation for more than 48 hours (19%). In total, 34% experienced a serious adverse event other than death, 8% of patients returned to the operating room, and the average hospital length of stay was 9.8 ± 9.9 days. Increased mortality was associated with gangrene, ascites, American Society of Anesthesiologists (ASA) Class 4 or higher, coma, and bleeding disorders. Reduced mortality was associated with age less than 60 years.[91]

A study of patients with chronic subdural hematoma found that neurologic status on admission was the best predictor of outcome. In addition, age, brain atrophy, thickness and density of hematoma, subdural accumulation of air, and antiplatelet and anticoagulant therapy were found to correlate significantly with prognosis.[92]

In a retrospective study of patients with epidural hematoma based on the Nationwide Inpatient Sample (NIS) from 2003 to 2010, 5189 admissions were identified in the database, and incidence was highest in the second decade (33.4%). Median length of stay in the hospital was about 4 days in each year, and in-hospital mortality and complication rate were 3.5% and 2.9%, respectively.[93]

Pupillary function before and after resuscitation has some predictive value. In patients who initially have bilateral unreactive pupils (and whose pupils do not regain function), approximately 85% die or remain in a persistent vegetative state, compared with 15% of those who regain pupillary function.[26]

Age also influences overall outcome. Infants and very young children tend to have a higher mortality rate. This is most likely due to the nature of their injuries and associated prolonged episodes of apnea. The mortality rate from closed head injury remains relatively constant until after the age of 35 years, at which time it begins to rise dramatically.[94]

The development and duration of fever is clearly associated with worse prognoses.[21]

The results from one study found that insulin deficiency due to diabetes mellitus (DM) imparts an increased risk for mortality in patients with moderate-to-severe traumatic brain injury (TBI) compared with patients without DM (14.4% versus 8.2% ).[95]

Patient Education

Educate the family about the type of injury the patient has sustained and what natural progression or recovery should be expected. The patient and the patient's family can be instructed regarding home exercises, wound care, and administration of long-term antibiotics.