Child Neurology
7th Edition

Chapter 9
Postnatal Trauma and Injuries by Physical Agents
John H. Menkes
Richard G. Ellenbogen
CRANIOCEREBRAL TRAUMA
In the Western world, accidents constitute the major cause of death of children between the ages of 5 and 19 years (Table 9.1) (1). From 1951 to 1971, the number of children with head injuries admitted to hospitals in Newcastle-upon-Tyne, United Kingdom, increased sixfold, reaching 13.9% of all admissions to pediatric wards (2). This increase probably was partially caused by a change in policy such that more apparently minor injuries were admitted for observation. In the United States, the situation is similar; cranial and major facial injuries are responsible for 3.6% of hospital admissions and 3.3% of days spent in the hospital, and they are the most common neurologic conditions requiring hospital admission for patients younger than 19 years of age (3). Head injury is now the most common cause of death and disability of children in the United States, causing death in approximately 7,000 children each year (4). It is also the cause of significant cognitive and motor-sensory dysfunction in the pediatric population, with an estimated economic burden approximating $10 billion a year in the United States alone (5).
A high proportion of head injuries cause death at the site of the accident or on the way to the hospital, so admission figures reflect only part of the incidence. In the San Diego Prospective Survey, as many as 85% of fatalities occurred at these times (6). Since the 1980s, child abuse has been recognized increasingly. The incidence of reported cases rose from 10.1 in 1,000 children in 1976 to 39 in 1,000 children in 1990 (7), with one of the more recent estimates of the incidence of fatal abuse in children younger than 4 years of age being at least 10 in 100,000 (8). As the reported frequency of maltreatment has risen, it has become an important cause of head injury. In one series, in infants younger than age 1 year, 36% of all head injuries and 95% of injuries resulting in intracranial hemorrhage or other major cerebral complications were caused by child abuse (9). In older children, the proportion of injury from abuse is lower, but still significant. In addition, children who were hospitalized as a consequence of child abuse remained in the hospital more than twice as long as other hospitalized children (10).
Although no accurate data exist, cranial trauma commonly occurs in childhood sports. Brain concussion occurs in an estimated 19 in 100 American football players per year. Bicycling is probably the second-most-common sport leading to head injury. In an Australian series published in 1987, bicycling was responsible for more than 20% of all head injuries in children (11). The remarkable effectiveness of bicycle helmets in preventing head injuries in children has been attested to in several studies (12,13).
This chapter presents only postnatal injuries; perinatal injuries are described in Chapter 6. This chapter discusses the diagnosis and nonsurgical management of head injuries and considers their pathophysiology and pathology based on the severity of the craniocerebral trauma and the complications and sequelae of head injuries.
Dynamics and Pathophysiology
The adverse effects of head injury are primary and secondary. The primary effects are the result of physical forces, mainly acceleration and deceleration, that act on the brain through shear-strain deformation (a change in shape without a change in volume) and through compression-rarefaction strain (a change in volume without a change in shape). The brain is injured through these two mechanisms acting alone, in combination, or in succession (14,15,16). Secondary effects result from the various processes that complicate the injury.
Shear-Strain Deformation
Shear strain occurs when two layers slide on each other, moving in parallel in opposite directions (16). The injury from shearing or tearing is responsible for most lesions, especially in an infant or a young child, whose skull is more easily deformed than an adult’s. Deformation
P.660

absorbs much of the energy of impact, reducing the adverse effects of acceleration and deceleration but adding to the risk of tearing of blood vessels. Because of its elasticity and ability to undergo a greater degree of deformation, the skull of an infant absorbs the energy of the physical impact and protects the brain better than the skull of an older person.
TABLE 9.1 Causes of Death in Children in 1986
Cause of Death Age <5 years 5–9 years 10–14 years 15–19 years
All causes 46,371 4,082 4,706 16,224
   Accidents 5.5% 20.6% 18.3% 10.8%
   Motor vehicle accidents 2.6% 26.2% 27.2% 43.1%
   Influenza and pneumonia 0.9% 1.2% 0.9% 0.5%
   Congenital anomalies 19.7% 6.3% 0.1% 1.6%
   Malignant neoplasms 2.3% 21.3% 15.8% 7.1%
From U.S. Department of Health, Education and Welfare, Public Health Service, Vital statistics of the United States. Washington, DC: Author, 1986.
The relationship between stress waves and deformity depends mainly on the momentum of the head and of the object at the instant of impact. In a child with intact reflexes, a free fall probably causes more injury to the brain than an aimed object striking the head with even greater speed. Approximately 70% of shear-strain deformation skull fractures are linear and single; the rest are depressed fractures. The faster moving the object delivering the blow, the more likely it is that a depressed fracture will result; a lower-velocity impact with the same momentum tends to produce deformity and linear fracture, sometimes distant from the point of impact but at a weak part of the cranial vault where distortion can occur. In addition, the size of the object that strikes the skull is important. Objects smaller than 5 cm2 tend to produce a depressed fracture; larger objects tend to produce a linear fracture (16). A linear fracture is usually not important in management of the child because outcome of the injury and its complications are determined by damage to the intracranial contents at the moment of impact; an injury can be fatal in the absence of a fracture.
Compression-Rarefaction Strains
In an acceleration injury, in which the momentary compression is greatest, the effects of compression-rarefaction strains are usually maximal at the point of impact; at the same time, an area of low pressure or rarefaction occurs contralaterally, the site of the familiar contrecoup injury (13,14), Thus, a blow to the occiput can result in the major damage in the frontal and temporal regions. Contrecoup injuries generally occur on the undersurface and poles of the frontal and temporal lobes (16). They are relatively rare in infants and young children, presumably because skull distortion rather than pressure waves predominate (19). Contrecoup injuries are most likely to occur when the impact is to the occiput or to the side of the head (16). Distortion of the brainstem is particularly likely in deceleration injuries such as those that occur in falls; any resulting loss of function in this area is liable to have serious or fatal consequences.
Magnetic resonance imaging (MRI) has allowed a better understanding of the lesions that can result from head injury. These were grouped into primary and secondary lesions by Gentry and coworkers (20) (Table 9.2). The injuries expected from various accident mechanisms are depicted in Table 9.3.
TABLE 9.2 Classification of Traumatic Intracranial Lesions
Primary Lesions
   Intraaxial
      Diffuse axonal injury
      Cortical contusion
      Subcortical matter injury
      Primary brainstem injury
   Extraaxial hematomas
      Extradural
      Subdural
   Diffuse hemorrhage
      Subarachnoid
      Intraventricular
   Primary vascular injuries
   Secondary Lesions
   Pressure necrosis (secondary to brain displacement and herniations)
   Tentorial arterial infarction
   Diffuse hypoxic injury
   Diffuse brain swelling
   Boundary and terminal zone infarction
   Others
      Fatty embolism
      Secondary hemorrhage
      Infection
From Gentry LR, Godersky JC, Thompson B. MR imaging of head trauma: review of the distribution and radiopathologic features of traumatic lesions. AJR Am J Roentgenol 1988;15:663. With permission.
TABLE 9.3 Expected Injury Types Associated with Accident Mechanisms in Young Children
Mechanism Injury Types
Fall <4 ft Concussion/soft tissue injury
Linear fracture
Epidural hematoma
Ping-pong fracture
? Depressed fracturea
Fall >4 ft Injuries listed for fall <4 ft plus the following:
   Depressed fracture
   Basilar fracture
   Multiple fractures
   Subarachnoid hemorrhage
   Contusion
   ? Subdural hematomaa
   ? Stellate fracturea
Motor vehicle accident Injuries listed for falls plus the following:
Subdural hematoma
Diffuse axonal injury
aInjury types are uncommonly associated with the given mechanism.
From Dumaine AC. Head injury in very young children. Pediatrics 1992;90:184. With permission.
P.661

Diffuse Axonal Injury
The term diffuse axonal injury (DAI) refers to a clinical-pathologic-radiologic entity that clinically manifests itself by loss or impairment of consciousness.
The lesion usually is not the result of a fall, except when the fall occurs from a considerable height. Instead, it results from severe angular acceleration-deceleration forces (21) and is believed to induce coma through disconnection of the cortex from the lower centers. It is responsible for severe, irreversible, and potentially fatal brain damage occurring at the moment of injury. DAI is frequently unaccompanied by skull fracture, increased intracranial pressure, or cerebral contusion (21).
As viewed microscopically, the hallmark of DAI is the presence of axonal retraction balls identified by silver stains or, as in more recent studies, immunohistochemical staining for beta-amyloid precursor protein (16,22). Postmortem examinations have demonstrated axonal lesions in the inferior portion of the corpus callosum and the dorsolateral quadrant of the rostral brainstem, notably in the region of the superior cerebellar peduncles, and throughout the cerebral hemisphere and cerebellum (23,24,25). Damage to the superior cerebellar peduncles, which are particularly vulnerable to rotational injuries, is responsible for the ataxia commonly observed after major head injuries (26).
The microscopic picture of axonal lesions was first described by Strich (27), who proposed that the shearing forces sustained during injury cause stretching of axons, which might be sufficient to prevent them from functioning. The more subtle changes that precede this final state have been studied in animals by electron microscopy and through the use of immunocytochemical labeling techniques. Although the exact time course for humans has not been determined, in animals an alteration in the structure of the nodes of Ranvier occurs within 15 minutes of the injury (28). This is followed some 12 to 24 hours later by an interruption of antegrade and retrograde axonal transport, the loss of microtubules, and the gradual development of axonal swelling. When axonal swelling exceeds a certain critical level, effective transection of the axon occurs, although without evident tearing or damage to adjacent blood vessels. DAI is frequently accompanied by evidence of neuronal injury (29).
It is unclear how mechanical forces induce the instantaneous perturbation of the cell membrane that in turn initiates these axonal changes. Wolf and coworkers proposed that traumatic deformation of axons induces abnormal sodium influx through mechanically sensitive Na+ channels. This influx subsequently triggers an increase in intraaxonal calcium through the opening of the voltage-gated calcium channel (30). This can result in activation of calmodulin and an increase of extracellular potassium at the damaged node (31). Intracellular calcium, in turn, increases the activity of proteolytic enzymes that disrupt axonal cytoarchitecture. Over time, the swollen axons either degenerate or undergo regenerative changes, as shown by sprouting and growth cones (32).
On MRI, axonal damage is best visualized by diffusion tensor imaging (DTI) or high-spatial-resolution susceptibility-weighted imaging techniques (33,34). By these techniques diffuse axonal injury appears as small, oval, focal abnormalities in white matter tracts, usually adjacent to cortical gray matter, but sometimes in the splenium of the corpus callosum (20). Proton magnetic resonance spectroscopy can provide further information on the extent of axonal injury. A lowered ratio of N-acetylaspartate (NAA) to creatine is believed to be indicative of axonal injury, but there does not appear to be any elevation of tissue lactate (35).
Cortical Contusion
The second-most-common type of brain injury visualized by MRI of patients with severe head trauma is cortical contusions, which tend to be multiple and represent bruises on the surface of the brain. As verified by both imaging and pathologic studies, points of predilection for contusions are the crests of gyri on the orbital surfaces of the frontal lobes and the inferolateral aspect of the temporal lobes. Contusions consist mainly of petechial hemorrhages in the superficial cortical layers occurring at the site of impact (coup injuries) or at contrecoup areas (Fig. 9.1). Impact on the forehead or vertex can send the initial pressure wave caudally, leading to downward displacement of the brain toward and into the tentorial opening. This
P.662

displacement can result in contusions of the hippocampal gyrus, particularly the uncus, the basal ganglia, and the upper part of the brainstem (36). The severity of brain damage caused by contusion depends on the extent of vascular injury. Damage tends to be less common in infants and small children than in older children or adults subjected to comparable trauma. Instead, contusions in infants consist of slitlike tears in the cerebral white matter of the frontal and temporal lobes (16,37). One of the major clinical issues in the management of contusions is their tendency to increase in size, coalesce, or cause a mass effect, especially after the first day following injury. Therefore, a follow-up imaging study is often useful.
FIGURE 9.1. Contrecoup injury. A left-sided epidural hematoma not seen in the picture caused severe compression of the left cerebral hemisphere (white arrow). Hemorrhage in the right inferior temporal lobe (black arrow) probably represents a contrecoup injury. (Courtesy of Dr. Harry V. Vinters, Division of Neuropathology, UCLA Center for the Health Sciences, Los Angeles, CA.)
Cerebral Laceration
Cerebral lacerations are usually the result of damage from penetrating wounds or depressed skull fractures, but they can occur without fracture in small children, whose skulls tend to become more grossly distorted at the moment of injury. Lacerations frequently involve the frontal and temporal poles and are associated primarily with tears of the dura and tears or other injuries of the major vessels and secondarily with thromboses, hemorrhages, or focal cerebral ischemia. Tears in the white matter are seen commonly in infants after blunt trauma even without fracture (38) (Fig. 9.2). They result from the soft consistency of the poorly myelinated cerebrum and from the pliancy of the immature skull.
MRI has demonstrated that macroscopic traumatic injuries to subcortical structures such as the thalamus and to the brainstem are rare. Instead, the pathologic change that results in the clinical picture generally attributed to primary brainstem damage is diffuse axonal injury.
Concussion
The physical processes within the skull caused by trauma induce numerous changes within the brain. The most common is concussion. Although there is some disagreement with respect to the definition of concussion, the American Academy of Neurology defines it as a trauma-induced alteration in mental status with or without loss of consciousness (39). In addition, there also can be post-traumatic amnesia for the moment of injury and for a variable period before it (retrograde amnesia).
The pathogenesis of concussion is under debate. From a structural viewpoint, concussion is believed to result from minor degrees of diffuse axonal injury. With more severe concussive injuries, there is a massive release of excitatory neurotransmitters, notably glutamate (40). As a result, there is a synchronized depolarization of neuronal membranes with subsequent increase of extracellular potassium (41), resulting in a functional deafferentation of the cortex (42). Compounding the disruption of ionic homeostasis is an imbalance between energy demands and supply, a loss of cerebral autoregulation in a significant proportion of patients with minor head injuries (43),
P.663

the presence of vasospasm, and a global increase in intracranial pressure, which reduce the supply of substrates to the tissue.
FIGURE 9.2. Closed head injury. Note the petechial hemorrhages in the corpus callosum and a tear in the white matter of the right centrum semiovale (arrow). (Courtesy of Dr. Harry V. Vinters, Division of Neuropathology, UCLA Center for the Health Sciences, Los Angeles, CA.)
From a clinical point of view, concussion is marked by the immediate loss of consciousness, a suppression of reflexes, a transient arrest in respiration, accompanied by bradycardia and a fall in blood pressure. Vital signs quickly stabilize, and there is a gradual return of consciousness. However, neuropsychologic testing will show deficits for as long as 15 minutes after the injury. With increasing injury, the loss of consciousness is prolonged and there is an increasing duration of post-traumatic amnesia (44). The clinical features of postconcussion syndrome are covered in another part of this chapter.
FIGURE 9.3. Proposed mechanisms for secondary damage after traumatic brain injury (TBI). (From Bayir H, Kochanek PM, Clark RS. Traumatic brain injury in infants and children: mechanisms of secondary damage and treatment in the intensive care unit. Crit Care Clin 2003;19:529–549. With permission.)
Secondary Effects of Brain Trauma
Secondary effects of trauma develop as a consequence of at least five factors: cerebrovascular dysregulation, excitotoxicity, free radical formation leading to oxidative stress, energy failure, and inflammation (Fig. 9.3; Table 9.4) (45). Cerebrovascular dysregulation, excitotoxicity, and energy failure are of primary importance in the evolution of cerebral swelling. This condition has been defined as an
P.664

increase in volume caused by an increase in brain water and sodium content. Brain swelling results from increased cerebral blood volume, cerebral edema, or a combination of the two (45).
TABLE 9.4 Secondary Effects of Brain Trauma
Cerebral Blood Flow Dysregulation
   Hypoperfusion in the first 24 hours following trauma
   Reduced response to endogenous vasodilators (adenosine, nitric oxide, cyclic AMP)
   Reduced NO production
   Increased production of endogenous vasoconstrictors (endothelin-1)
Excitotoxicity
   Increased glutamate after head injury
   Causes sodium-dependent neuronal swelling
   Followed by calcium-dependent activation of intracellular proteases, lipases
   Production of reactive oxygen nitrogen species (free radicals)
   Results in mitochondrial and DNA damage
Oxidative Stress
   Free radical production important in early neuronal and vascular damage
Inflammation
   Increase in cytokines, soluble adhesion molecules
Apoptosis
   Activated by mitochondrial or DNA damage
Cerebral Swelling
   Due to cellular swelling, injury to blood–brain barrier, osmolar swelling
   Results in intracranial hypertension, secondary ischemia
Fishman distinguished three major categories of cerebral edema: vasogenic edema, cytotoxic or cellular edema, and interstitial edema (46). Vasogenic edema is characterized by increased permeability of brain capillary endothelial cells. This results from defects in the tight endothelial cell junctions and an increased number of pinocytic vesicles, which are responsible for the transport of macromolecules across the blood–brain barrier. Vasogenic edema fluid is extracellular and accumulates primarily in white matter because resistance to fluid flow is less in white matter than in gray matter. It occurs around brain tumors, notably glioblastomas and metastatic tumors, after cerebral infarction, and in lead encephalopathy. Cytotoxic edema is marked by swelling of all cellular elements of the brain with reduction in the volume of extracellular fluid. This last form of edema is characteristic of energy depletion such as occurs during the initial stages of cerebral hypoxia. Fishman postulated a third form of brain edema in which there is an increase in water and sodium content of periventricular white matter (46). This condition is seen in obstructive hydrocephalus.
In brain trauma, the prominence of astrocytic swelling, the integrity of the interepithelial tight junctions, and the relative paucity of protein-rich fluid suggest that edema is mainly cytotoxic (47), although massive swelling of perivascular astrocytic foot processes can compress the microvasculature and reduce tissue perfusion, inducing secondary vasogenic edema.
The importance of cytotoxic edema has been confirmed by diffusion-weighted MRI performed on experimental animals. These studies demonstrate that immediately after injury, there is predominantly vasogenic edema, which, over the next few hours to days, becomes superseded and overshadowed by cytotoxic edema (48).
A number of biochemical changes follow severe traumatic head injury (Fig. 9.3, Table 9.4). In both clinical and experimental settings, good correlation exists between the severity of traumatic brain injury and the amount of neuroexcitatory amino acids, such as glutamate and aspartate, that are released (40,49). As a result, neuronal membranes depolarize with subsequent increase of extracellular potassium (42). This is accompanied by the entry of calcium into nerve terminals and is followed by further glutamate release and potassium flux. At the same time, arachidonic acid is released, which in turn induces a cascade of reactions with the ultimate formation of free radicals (50). These changes are accompanied by increased energy demands as brain cells attempt to reinstate normal ionic membrane balance (51). With an increase in energy consumption there is a rapid decrease in adenosine triphosphate (ATP) and an increase in lactic acid (52). These biochemical alterations, which suggest impaired mitochondrial function, are confirmed by the electron microscopic observations of swollen neuronal mitochondria and an increased permeability of the organelles’ outer membranes (53). These factors all contribute to astrocytic swelling, as does an energy shortage–induced failure of the sodium pump (54). In addition, there is a local inflammatory response with complement and microglial activation and an increased release into cerebrospinal fluid (CSF) of a variety of cytokines (interleukin-1, interleukin-6, and interleukin-10) (56). These substances also contribute to the evolution of diffuse cerebral edema. The role of a trauma-induced upregulation of aquaporins, a family of transmembrane proteins that selectively allow the passage of water through the plasma membrane, in the formation of brain edema has not been fully clarified (56a).
The effects of traumatic brain injury on cerebral blood flow are critical to the development of cerebral edema. Early after severe traumatic brain injury, marked hypoperfusion develops with reduction in oxygen delivery and cerebral ischemia (57). Some 24 hours later cerebral blood flow increases, and there is uncoupling of cerebral blood flow and oxidative cerebral metabolism (1). Post-traumatic hyperemia is more common in infants and children than in adults and is probably the consequence of a loss of cerebral autoregulation (45). Hyperemia in turn contributes to massive brain swelling, which is not uncommon in infants and young children and has been termed malignant brain edema (58).
P.665

Anatomically, cerebral edema involves principally the subcortical white matter and the centrum semiovale. Less often, cerebral edema surrounds an area of contusion or an intracerebral hematoma. Brain swelling also can occur in response to the evacuation of a large extracerebral clot or after other major intracranial surgery (59).
When the additive effects of injury and brain swelling are severe, a self-perpetuating sequence develops, which can lead to further increases of intracranial pressure, with collapse of cerebral venules. This collapse in turn reduces cerebral perfusion and causes tissue hypoxia. This leads to further cerebral edema (60). A loss of selective permeability of cell membranes results, with increased loss of fluid from the vascular compartment into the parenchyma, thereby increasing cerebral swelling. Recovery has not been seen when intracranial pressure equals or exceeds mean systemic arterial pressure, at which point cerebral perfusion ceases.
Numerous changes occur secondary to cerebral edema. Herniation of the uncus over the tentorial edge compresses the midbrain and often occludes the posterior cerebral arteries. Edema and infarction of the occipital poles can occur, contributing further to supratentorial pressure and thus to herniation. Petechial hemorrhages develop in the midbrain and pons, with infarctions in the areas of the basal ganglia supplied by the anterior choroidal artery. A decrease in blood pressure, commonly seen with a severe injury, potentiates the vicious circle.
Clinical Conditions
Closed Head Injury
More than 90% of major pediatric head injuries are nonpenetrating and closed (i.e., no scalp wound exists). The clinical picture is highlighted by alterations in consciousness (14). As is the case for head injuries at all ages, boys are involved three times more often than girls.
Clinical Manifestations
When the injury is mild, initial unconsciousness is brief and is followed by confusion, somnolence, and listlessness. Vomiting, pallor, and irritability are common, and particularly in infants can occur in the apparent absence of an initial loss of consciousness. By definition, except for transient nystagmus or extensor plantar responses, neurologic signs are not observed in concussion. As a rule, a computed tomography (CT) scan clarifies the differential diagnosis of contusion, cerebral laceration, or other complications of closed head injury, and a lumbar puncture is not warranted without specific indications.
From 7% to 40% of children with mild head injuries have associated linear fractures of the skull (61). These fractures are most common in the parietal region. According to most authorities, such fractures do not, by themselves, affect the clinical course or prolong the period of morbidity; children with and without simple fractures have the same incidence of serious sequelae (14,61,62).
Electroencephalography (EEG), when performed soon after injury, can reveal striking abnormalities, such as generalized and focal slowing, prolonged reaction to hyperventilation, and even hypsarrhythmia. In milder head injuries, these changes tend to be transient. Taking into account the time elapsed between injury and recording, Mizrahi and Kellaway found that the degree of EEG abnormality correlated with the severity of the injury (63). Nevertheless, the EEG does not predict the development of post-traumatic epilepsy, and there is no correlation between the appearance of the EEG immediately after the head injury and development of early post-traumatic epilepsy (64). Proton resonance spectroscopy appears to be more valuable in predicting outcome. In the experience of Ashwal and coworkers, the presence of lactate had a poor prognosis in that 91% of head-injured infants and 80% of head-injured children with poor outcomes had a lactate peak; conversely, none of the infants and children with good outcome showed the presence of lactate (65).
In major closed head injuries, consciousness is interrupted more profoundly and for longer periods than in minor head injuries, and focal neurologic signs point to localized brain contusion. The clinical picture in such cases is outlined in Table 9.5 (66). Generally, the greatest neurologic deficit is found at the time of injury. New neurologic
P.666

signs appearing subsequently indicate progressive brain swelling or, if localized, indicate secondary intracranial hemorrhage, vasospasm, or thrombosis. The duration of coma depends on the site and severity of injury.
TABLE 9.5 Clinical Findings in 4,465 Children with Major Closed Head Injuries
Finding Percentage
Initial level of consciousness
   Normal 56.0
   Drowsy, confused 30.2
   Major impairment 13.8
Vomiting 30.3
Skull fractures 26.6
   Linear 72.8
   Depressed 27.2
   Compound 19.7
Seizures 7.4
Paralyses 3.8
Retinal hemorrhages 2.3
Pupillary abnormalities 3.6
Papilledema 1.5
Extradural hematoma 0.9
Subdural hematoma 5.2
Mortality 5.4
Major neurologic residua 5.9
This series includes 243 infants with major birth injuries. This group had 50% mortality and a higher incidence of paralyses, retinal hemorrhages, and major residua.
From Hendrick EB, Harwood-Nash DC, Hudson AR. Head injuries in children: a survey of 4,465 consecutive cases at the Hospital for Sick Children, Toronto, Canada. Clin Neurosurg 1963;11:46. With permission.
The clinical picture follows one of several courses (67). Many children die without recovering consciousness. In a smaller group of patients, coma persists. Prognosis for survival is relatively good in children alive 48 hours after injury. In one-half of the surviving children, consciousness is regained in less than 24 hours. Recovery is often complete or nearly complete, although transient sequelae are not unusual. These include CSF leakage often complicated by secondary meningitis, post-traumatic epilepsy, and the development of a carotid artery–cavernous sinus fistula. Communicating hydrocephalus resulting from subarachnoid bleeding more commonly follows perinatal trauma and is discussed in Chapters 5 and 6.
Nonaccidental trauma
Child abuse is the most common cause of head injury in infants younger than 2 years of age. In the experience of Duhaime and coworkers, in 24% of infants admitted to hospital with head injuries the injuries were presumed to have been inflicted, and in another 32% the injuries were suspicious for child abuse (68). Although the term “shaken baby” has been used in the past, pathologic data collected on a larger Scottish series has made it clear that a whiplash shaking injury could be documented in only 6% of cases (69). The hyperextension and hyperflexion whiplashing forces injure the brainstem or upper cervical spinal cord with consequent acute respiratory failure, hypoxia, and brain swelling. Subdural bleeding was generally trivial. These pathologic findings are commensurate with the clinical presentation of collapse or respiratory arrest (70). Shannon and Becker also stressed the frequency of brainstem and spinal cord injuries in infantile child abuse (71).
The more commonly encountered clinical picture is characterized by depressed consciousness, increased intracranial pressure, seizures, and subdural and retinal hemorrhages. These symptoms are frequently accompanied by other nonaccidental injuries and an inconsistent or unreliable medical history. This constellation of symptoms has been termed the shaken impact syndrome. The encephalopathy developed acutely in 53% of the Scottish cases, subacutely in 19%, and chronically in 22% (69). The less acute the presentation, the better is the outcome with respect to long-term morbidity. In the series of Bechtel and coworkers, the clinical picture of nonaccidental head injury could be distinguished from that seen in accidental head injury by more frequent bilateral retinal hemorrhages, a greater likelihood of presenting with an altered mental state and seizures, and a lesser incidence of scalp hematomas (71a).
Pathologic examination of the brain of infants subjected to nonaccidental head injury disclosed skull fractures in 36%, acute subdural bleeding in 72%, and retinal hemorrhages in 71% (70). Severe hypoxic brain damage was present in 77%, and the usual cause of death, seen in 82%, was increased intracranial pressure. Eighty-five percent of the children had signs of impact to the head at autopsy, and 51% had significant extracranial injuries (70). The clinical presentation of infants who showed no signs of impact was of collapse or respiratory arrest; the pathologic picture did not differ from the majority of infants who showed evidence of impact (72).
There is considerable uncertainty as to the mechanism of brain damage in shaken impact syndrome, and according to Donohoe there is inadequate scientific evidence to come to a firm conclusion on most aspects of causation (73). Geddes and Plunkett (74) shared this skepticism and stressed that although most infants do indeed show signs of inflicted violence, serious injury or death from a low-level fall is possible, and contrary to Harding and coworkers, who believed that retinal hemorrhages result from rotational acceleration and deceleration forces (75), they doubted that shaking can induce the characteristic retinal hemorrhages.
The combination of a subdural hematoma and retinal hemorrhages can also result from a variety of bleeding disorders. It is also encountered in Menkes disease and in type 1 glutaric aciduria (see Chapter 1) (76).
Nonsurgical Treatment
Major Head Injury
A recommended sequence consists of the following four steps: a rapid initial evaluation, resuscitation, a more extensive secondary evaluation including radiologic assessment, and definitive treatment (1,45,76a). In many instances of major closed head injury, diagnosis of the injury is performed in parallel with emergency treatment.
Because in children the clinical condition can change rapidly and frequently, a high degree of alertness and preparedness is required of both medical and nursing attendants.
Neurologic Examination
Careful, but not elaborate, neurologic examination must be made and recorded, with particular emphasis on the state of consciousness; pupillary size, equality, and response to light; the extent and symmetry of spontaneous movements; and the reflex responses. Recording of blood pressure, pulse, and respiration is likewise essential. Caloric and optokinetic responses are useful for evaluating brainstem function. Some or all of these observations, especially the level of consciousness and motor activity, are principally of value when made serially at intervals that
P.667

can be as often as every 5 minutes. Such examinations allow trends away from or toward normal functioning to be detected, thus providing warning of any need to intervene surgically.
TABLE 9.6 Pediatric Coma Scale
Eyes Open
   Spontaneously 4
   To speech 3
   To pain 2
   Not at all 1
Best Verbal Response
   Oriented 5
   Words 4
   Vocal sounds 3
   Cries 2
   None 1
Best Motor Response
   Obeys commands 5
   Localizes pain 4
   Flexion to pain 3
   Extension to pain 2
   None 1
Normal Aggregate Score
   Birth to 6 months 9
   6–12 months 11
   1–2 years 12
   2–5 years 13
   Older than 5 years 14
Modified from Simpson D, Reilly P. Pediatric coma scale. Lancet 1982;2:450.
Several pediatric modifications of the Glasgow Coma Scale (77), which is widely used for the assessment of head-injured adults, have been proposed. One example is presented in Table 9.6 (78).
The Glasgow Coma Scale measures three neurologic responses: eye opening, verbal response, and limb movement. Each response is given a score; the higher the score, the better is the condition of the patient.
Additionally, the extent of retrograde and post-traumatic amnesia should be recorded when possible. Spontaneous or provoked episodes of decerebrate activity or hypotonia are associated with a poor outcome (79); the length of pre- and post-traumatic amnesia is a useful indicator of the severity of the head injury (80).
Maintenance of Airway and Circulation
After the immediate evaluation of the child’s general condition, an adequate airway should be established and maintained. Airway obstruction is the most frequent cause of respiratory failure. Maintenance of patency requires suction of mouth or pharyngeal contents or endotracheal intubation followed, when necessary, by artificial respiration. Tracheostomy may be indicated to bypass mechanical obstruction of the airway caused by facial or mandibular injuries. Because an injury to the cervical spine must be presumed until proven absent, the neck must be stabilized while an airway is established. Arterial pressure is monitored by cannulation of a peripheral artery and maintained by administration of crystalloid, colloid, or blood products. Gas exchange also must be monitored.
Shock in closed head injuries is usually caused by blood loss elsewhere in the body. Rarely, it indicates damage to the medullary cardiovascular centers. Infants and young children have a higher incidence of shock than those in older age groups (1). In infants, in particular, subgaleal, subperiosteal, subdural, or extradural hemorrhages can be sufficiently extensive to induce shock. The injured brain is highly susceptible to episodes of systemic hypotension, in part because cerebral edema in conjunction with systemic hypotension lowers cerebral blood flow, and perhaps because an impairment of autoregulation of cerebral blood flow also occurs. Whatever the physiologic mechanism, when head injury is accompanied by systemic hypotension, the outcome is significantly worse (81).
After the initial emergency measures have been completed, the child’s neurologic status should be reevaluated. The importance of repeated observations of vital signs and neurologic status cannot be overemphasized. Slowing of the pulse rate and widening of pulse pressure often accompany an increase of intracranial pressure (Cushing effect) (see Chapter 11). An irregular respiratory pattern is also common after severe head injuries (82).
Neuroimaging Studies
CT of the head has revolutionized the clinical assessment of head injuries, and it is the most appropriate study for the assessment of intracranial complications, taking precedence over skull radiography and MRI. The need for and timing of such imaging studies depend mainly on the clinical findings. A suggested strategy is outlined in Table 9.7 (83). Some centers suggest that in children who show signs of brainstem compression and a Glasgow coma score of 3, precious time can be saved by foregoing CT and proceeding directly with exploratory surgery to remove a hematoma. Although emergency trephination may benefit the occasional patient, scanning is still the most important diagnostic study for the majority of children (84).
CT can delineate the presence and the extent of fractures and can demonstrate the presence of diffuse cerebral swelling and hyperemia of the brain (85), hemorrhages in the epidural and subdural space, and intracerebral hemorrhages (Fig. 9.4) (86).
In some injuries, the condition of the sinuses should be noted for evidence of a fracture, which makes the injury compound, or the presence of blood. Ethmoid and maxillary sinuses are pneumatized at birth, the sphenoid sinuses become pneumatized at approximately 5 years of age, and the frontal sinuses, the last to develop, are not visualized until 4 to 6 years of age (87).
TABLE 9.7 Management Strategy for Radiographic Imaging in Pediatric Patients with Head Trauma
Low-Risk Group Moderate-Risk Group High-Risk Group
Possible findings
   Asymptomatic
   Headache
   Dizziness
   Scalp laceration
   Scalp hematoma
Possible findings
   Change of consciousness at time of injury or subsequently
   Unreliable history
   Younger than 2 years of age
   Post-traumatic seizure
   Vomiting
Signs of basilar skull fracture
Fracture into an air sinus
Possible depressed fracture
Suspected child abuse
Possible findings
   Depressed or decreasing level of consciousness
   Focal neurologic signs
   Penetrating skull injury
   Depressed fracture
Recommendations
   Observation
   Discharge with head injury information sheet, and have family observe child
Recommendations
   Hospitalization
   Close observation
   CT scan and neurosurgical consultation
Recommendations
   Emergency CT scan
   Neurosurgical consultation
CT, computed tomography.
Adapted from Masters SJ, McLean PM, Arcarese JS, et al. Skull x-ray examinations after head trauma. Recommendations by a multi-disciplinary panel and validation study. N Engl J Med 1987;316:84–91.
P.668

MRI usually contributes little to the initial management of major closed head injury. It is, however, more sensitive than CT in detecting an extraaxial hematoma, particularly a chronic subdural hematoma (88). Additionally, MRI provides better visualization of subacute and chronic contusions and of shearing lesions of the white matter such as result from child abuse (89). MR angiography (MRA), performed at a subsequent time, can be of use in delineating such vascular abnormalities as arteriovenous fistulae, venous sinus occlusion, and arterial occlusions.
FIGURE 9.4. Computed tomographic scan after head trauma. Head injury shown by computed tomographic scan without contrast. The soft tissue swelling and the left parietal skull fracture are readily evident. An intracerebral hematoma is seen with surrounding cerebral edema resulting in ventricular compression and shift. A small subdural hematoma (s) is also evident.
One must guard against excessive complacency on the basis of a single normal CT obtained in the early hours after an injury because hemorrhagic complications can evolve subacutely after an interval of several hours or days. A repeat CT is indicated under the following circumstances: (a) when doubt exists about the presence of a mass lesion, (b) when intracranial pressure monitoring demonstrates an increase in pressure, (c) when a patient is unconscious despite an initial benign-appearing CT, and (d) when a contusion is accompanied by a neurologic deficit. A normal CT result also does not exclude the presence of cerebral edema. In the series of O’Sullivan and coworkers, 88% of head-injured patients with a Glasgow coma score of 8 or less and no evidence of cerebral edema by CT had increased intracranial pressure when recorded directly (90).
Fluid and Electrolytes
The appropriate fluid management of the hypovolemic head-injured patient remains controversial, although both clinical and experimental studies indicate that the most important goals are to correct any hypovolemia and prevent reduction of osmolality. A period of hyponatremia with natriuresis is a common response to brain injury (91). Hyponatremia can result from either a cerebral salt-wasting syndrome characterized by hypovolemia or from an inappropriate excess of antidiuretic hormone that causes water retention (syndrome of inappropriate
P.669

antidiuretic hormone secretion [SIADH]). Whereas the cerebral salt-wasting syndrome is treated by fluid and salt supplementation, SIADH requires fluid restriction. In the majority of patients, hyponatremia during the first week after a head injury is caused by excessive antidiuretic hormone, whereas hyponatremia that develops later probably results from chronic salt-wasting (92). Berkenbosch and colleagues advocated the measurement of urine output to distinguish between the two conditions. In SIADH, urine output is reduced, and at times there is a frank oliguria, whereas cerebral salt-wasting syndrome is marked by polyuria (93).
The relationship between fluid load, sodium balance, and intracranial pressure is unclear. The traditional view is that during the early post-traumatic period, there is a danger of overloading the patient with fluids, which can increase intracranial pressure and thus diminish the level of consciousness. Therefore, fluid intake for the first 2 to 4 days should be between 35% and 50% of the average normal daily fluid requirements, as long as urine volume remains adequate (94). This view has been challenged, and more recent work indicates that rapid, continuous infusion of hypertonic (3%) saline with a view of maintaining intracranial pressure below 20 mm Hg is safely tolerated and possibly improves the outcome of severely brain injured children without incurring a significant risk for renal failure or extrapontine myelinolysis (95). Most neurosurgeons currently recommend that hypertonic crystalloid solutions be given intravenously to maintain a normal intravascular volume and adequate cerebral perfusion pressure (96). Both clinical and experimental studies suggest that glucose-containing solutions should be avoided because these tend to increase cerebral lactic acidosis (97).
Despite the evidence of sodium retention in the early post-traumatic period, salt should be administered to avoid hypotonicity of extracellular fluids. A total of 30 mEq of sodium chloride per liter of calculated fluid requirements meets the usual electrolyte requirements. As a result of a catecholamine-induced intracellular potassium shift, hypokalemia can accompany severe head injuries (98). Although hypokalemia was documented in 7% of traumatic brain injuries admitted to the Boston Children’s Hospital, it corrected spontaneously, and in almost all instances did not require potassium supplementation (99).
Less commonly, hypernatremia and dehydration follow closed head injury. These are occasionally seen in a subfrontal injury that has caused hypothalamic damage and diabetes insipidus, or they are the result of a failure in the thirst response or inadequate hydration of an unconscious patient.
Coagulation Defects
A significant proportion of patients with head injuries severe enough to result in destruction of brain tissue show clinical and laboratory evidence of impaired coagulation (100). This results from structural damage, hypoxia, or elevated catecholamine and steroid levels. These patients have low fibrinogen levels, diminished amounts of factors V and VIII, and thrombocytopenia. Treatment can require emergency replacement of hemostatic factors with fresh-frozen plasma or recombinant activated factor VII (101). Disseminated intravascular coagulation was seen in approximately one-third of children evaluated within 2 hours of major head injury (102).
Seizures
Seizures can occur shortly after the injury or can appear after an interval of days to years. Seizures appearing during the acute stage can increase intracranial pressure by Valsalva effects. They can increase cerebral blood flow, cause the release of neuroexcitatory transmitters, and aggravate any preexisting hypoxia. They generally are managed with intravenous phenytoin (5 mg/kg loading dose). Intramuscular phenytoin, which can crystallize within muscles, is unreliable. Diazepam (0.1 to 0.3 mg/kg) and lorazepam (0.05 to 0.25 mg/kg) are suitable alternatives. Some authorities are loath to administer phenytoin to infants and toddlers, preferring carbamazepine. In adults, phenytoin appears to reduce the likelihood of early post-traumatic seizures but does not decrease the risk of late post-traumatic epilepsy (103,104). EEG monitoring is desirable to detect electrical seizures even when paralysis of the patient prevents overt seizures. The role of anticonvulsants in the treatment of post-traumatic seizures is discussed in Chapter 14. In view of the relative rarity of post-traumatic seizures, we do not advocate preventive anticonvulsant therapy, except in the presence of a major brain laceration. Chadwick is of the same opinion (103). For this purpose we prefer carbamazepine or valproate to phenytoin.
Increased Intracranial Pressure
Children tend to have a lower incidence of surgically treatable mass lesions than adults (less than 10% in a CT-verified series) but a higher incidence of increased intracranial pressure (105). Therefore, in the infant or child who has sustained a severe head injury, control of increased intracranial pressure becomes the most important problem of medical management. For this purpose, continuous pressure monitoring is necessary for the more seriously injured patients and for infants who are more likely to experience increased intracranial pressure (1,45,94). A variety of monitoring techniques are in use, each with its advantages and disadvantages. These include an indwelling ventriculostomy with direct monitoring from the ventricular system; monitoring from the subarachnoid, subdural, and epidural spaces (106); and fiber-optic monitoring by placement of a device into the frontal parenchyma (107). Of these methods, subdural monitoring using the Camino or Codman fiber-optic pressure
P.670

transducer is preferred at the University of California, Los Angeles, and the University of Washington. This allows an indirect, continuous measurement of intracranial pressure, which, when combined with the value of mean arterial blood pressure, permits calculation of cerebral perfusion pressure (cerebral perfusion pressure equals mean systemic arterial blood pressure minus intracranial pressure). When possible, estimation or measurement of cerebral blood flow allows detection of global cerebral ischemia, which generally worsens the clinical course and outcome (108). Focal cerebral ischemia is more difficult to assess. After major head injuries, some regions of the brain may require higher levels of cerebral perfusion pressure to maintain adequate cerebral blood flow.
In some head-injured children CSF drainage via a ventriculostomy is a useful adjunct for controlling elevated intracranial pressure in those patients whose ventricles can safely be cannulated. Ventriculostomy remains the most effective way to monitor intracranial pressure and treat elevated pressure via continuous or intermittent CSF drainage. In selected patients who do not have mass lesions but have elevated intracranial pressures that are refractory to all other therapies, treatment with controlled lumbar CSF drainage has been encouraging (109).
Continuous monitoring of jugular venous oxygen saturation using an indwelling fiber-optic catheter also has been suggested, but the usefulness of this technique in the pediatric population has yet to be assessed (1). Venous oxygen saturation in the adult should lie between 55% and 85%, and values below 45% indicate global cerebral ischemia (110).
Intracranial pressure is usually maintained below 15 mm Hg, although temporary increases to 20 mm Hg are often unavoidable and can occur in the course of nursing procedures. The initial step in lowering intracranial pressure is to induce hypocarbia by hyperventilation and reduce the carbon dioxide partial pressure (pCO2) to between 25 and 30 mm Hg. With profound hyperventilation and pCO2 values below 25 mm Hg, cerebral ischemia can develop (111).
Historically, hyperventilation has been part of the treatment algorithm for trauma patients with elevated intracranial pressures. This maneuver causes cerebral vasoconstriction, thereby decreasing cerebral blood flow and volume. The response to hyperventilation is rapid but short lived, and hyperventilation must be continued or increased to remain effective. Studies show that the prolonged and routine use of hyperventilation may be deleterious in some patients by exacerbating ischemic brain injuries. Therefore, in the absence of elevated intracranial pressure, we no longer recommend the routine use of hyperventilation. Substantial beneficial effects can be realized when refractory elevated intracranial pressure is treated by moderate hyperventilation, accomplished by maintaining the arterial carbon dioxide pressure between 28 and 35 mm Hg. This therapy can be administered safely, with careful monitoring to avoid the complications associated with prolonged, profound hyperventilation, notably cerebral ischemia (112,113). Should hyperventilation be ineffective in reducing increased intracranial pressure, other measures must be undertaken.
Raising the head between 0 and 30 degrees in the neutral, midline position is commonly performed after severe head injury. This maneuver facilitates cerebral venous drainage and thus theoretically decreased intracranial pressure. Feldman and coworkers showed in their randomized study of head position of the trauma patient that by elevating the head from 0 to 30 degrees, the intracranial pressure was decreased without a decrease in cerebral perfusion pressure of cerebral blood flow (114). When the head is raised approximately 60 degrees, there appears to be a deleterious effect on cerebral blood flow. However, these studies have not been confirmed for children. Thus, the most advantageous head position in a child with an intracranial pressure monitor (i.e., the best way to maximize cerebral perfusion pressure) can be determined by raising the head and observing the concomitant change in intracranial pressure and cerebral perfusion pressure.
Diuretics are used in a number of centers. Mannitol and urea are the most commonly used osmotic diuretics. Mannitol, given in amounts of 1.5 to 2.0 g/kg body weight in the form of a 20% solution, has a slower effect than urea, but maintains the lowered intracranial pressure at its minimum level for 2 to 4 hours. It has the disadvantage of leaking into the brain through areas where injury has damaged the blood–brain barrier and thus inducing local edema. In addition, by slowly diffusing into the intercellular space, mannitol carries water with it, thereby inducing a rebound effect. For these reasons diuretics are used mainly as a temporary measure to gain time while the patient and the operating room are being readied for surgery, while diagnostic studies are being performed, or for acute control of elevated intracranial pressure. Mannitol, furosemide, or both are favored in most institutions; however, there are few well-controlled studies to show their effectiveness in the pediatric population (45). One appropriate dosing regimen for furosemide is 0.5 to 1.0 mg/kg administered every 4 to 6 hours.
Several well-controlled clinical studies have failed, however, to demonstrate any significant effectiveness of corticosteroids in counteracting brain swelling in head-injured children. The agents did not improve outcome or lower intracranial pressure and in some cases may have been deleterious. For these reasons, corticosteroids are not routinely employed in the treatment of pediatric head injuries. A multicenter, randomized, placebo-controlled trial of corticosteroids after significant head injury in adults with a Glasgow Coma Score of 14 or less within 8 hours of injury actually demonstrated an increase in mortality with methylprednisolone (114a,114b).
P.671

The clinical effectiveness of various oxygen-derived free radical scavengers, when given to patients with severe head injuries within 8 to 12 hours of the injury, has not been established in terms of improving neurologic outcome (115). Agents that inhibit lipid peroxidation, block excitotoxicity, and block apoptosis or promote brain cell regeneration have been employed in various experimental situations, but as yet have no clinical application.
High doses of barbiturates, generally pentobarbital, have been advocated whenever intracranial pressure does not respond to other forms of therapy. However, there is no convincing evidence that barbiturate coma improves outcome, and the initial enthusiasm for this form of treatment has waned because barbiturates can cause myocardial depression (1,116). For this reason barbiturates require cautious use in children, especially those with hemodynamic instability. Currently, barbiturates are used as a last resort to help control intractable elevations of intracranial pressure and to help place patients who are in status epilepticus into EEG burst suppression. Under these circumstances, bedside EEG monitoring is mandatory.
Sedation with pharmacologically induced paralysis is often used in the treatment of the severely head-injured child. This reduces the increase in intracranial pressure associated with agitation, suctioning, and ventilation. In our institutions, sedation and paralysis is titrated with intracranial pressure. Intravenous narcotics in small doses by continuous or bolus infusion, short-acting benzodiazepines, and nondepolarizing muscle relaxants are used effectively for this purpose. Titration of short-acting drugs permits intermittent examinations and rapid reversal, when necessary. In nonventilated, awake, but intermittently confused patients, medication is used sparingly because the neurologic examination is the most important parameter for the management of the head-injured patient. Small boluses of narcotics and short-acting benzodiazepines are used in some institutions to sedate the agitated pediatric patient without a mass lesion, but this issue remains controversial among neurosurgeons and critical care physicians.
Induced mild hypothermia (cooling to 32°C to 33°C) in conjunction with hyperventilation (pCO2 between 25 and 30 mm Hg) and barbiturates (4 to 6 mg/kg intravenous thiopental, followed by 6 to 8 mg/kg per hour of continuous thiopental) has reemerged as another measure to lower increased intracranial pressure and increase cerebral perfusion by lowering metabolic demands (45). Hypothermia attenuates the increases of excitatory amino acids, decreases endogenous antioxidant consumption, and has antiinflammatory effects. The effectiveness of this measure in the pediatric population has not been clearly demonstrated (45). Preliminary results suggest that this method can improve neurologic outcome in younger patients, particularly those who were already hypthermic on admission. From these studies it is not clear whether this represents a beneficial effect of early hypothermia or a detrimental effect of passively warming a previously hypothermic patient (117).
Neurogenic pulmonary edema is a rare and potentially fatal complication of head injury. Its pathogenesis is unknown, but focal brainstem lesions, particularly in the region of the nucleus solitarius, can increase pulmonary arterial pressure and capillary permeability. Symptoms appear within the first day after injury and are managed with diuretics and positive end-expiratory pressure (118).
More detailed treatment schedules for the child with a major closed head injury are given by Jennett and Teasdale (119) and Bayir and colleagues (45). All infants with head injuries and all older children with severe head injuries, as defined by a Glasgow coma score of 8 or less, or who have sustained a major skull fracture should be seen by a neurosurgeon.
Minor Head Injury
The American Academy of Pediatrics issued guides for the management of minor head treauma in children older than age 2 years (120) and younger than age 2 years (121). These are summarized in Table 9.8. Considerable judgment is
P.672

required to avoid unnecessary hospitalization and expensive diagnostic studies, at the same time keeping in mind the possibility of post-traumatic complications that require emergency surgery. In general, children who have experienced only a momentary loss of consciousness are better managed at home. Parents are instructed to note at regular intervals the child’s state of alertness and ability to move his or her extremities. The parents should be told to contact the physician if diminished consciousness or limb weakness occurs.
TABLE 9.8 Guidelines for Management of Children with Minor Head Injuries
Children Older Than 2 Years of Age
   Minor closed head injury—no loss of consciousness
      History
      Physical and neurologic examinations
      Observation
   Minor closed head injury—loss of consciousness less than
      1 minute
      History
      Physical and neurologic examinations
      Computed tomography scan
Children Younger Than 2 Years of Age
   Low-risk group: falls less than 3 feet, no signs or symptoms
      2+ hours after injury
      Observation
   Intermediate-risk group: loss of consciousness less than
      1 minute, episodes of vomiting, lethargy resolved by time of exam, behavior not at baseline
      Prolonged observation (4 to 6 hours) OR computed tomography scan
   High-risk group: depressed mental status, focal neurologic signs, irritability, full fontanel
      Computed tomography scan
      Neurosurgical consultation
Adapted from Committee on Quality Improvement, American Academy of Pediatrics. The management of minor closed head injury in children. Pediatrics 1999;104:1407–1415; and Shutzman SA, Barnes P, Duhaime AC, et al. Evaluation and management of children younger than two years old with apparently minor head trauma: Proposed guidelines. Pediatrics 2001;107:983–993.
The role of routine CT in the management of the child with minor head injuries continues to be controversial. As indicated in Tables 9.7 and 9.8, the infant younger than 2 years of age or the child who sustains loss of consciousness should undergo CT. The younger the child, the lower the threshold should be for imaging studies. In the small infant the incidence of significant intracranial injuries is greater, and there is a higher likelihood of asymptomatic intracranial injuries (121). The cost-effectiveness of CT as a screening for safe discharge as compared with hospitalization for observation has been affirmed in a number of centers both in the United States and Europe (122,123,124). These studies have not addressed the costs of follow-up visits for false-positive imaging studies or the costs of return visits and subsequent hospitalization for a youngster who is handled with observation only (120). The use of skull radiography as a screening device for children with minor head injuries has serious limitations and should no longer be used inasmuch as a CT with bone windows to show the presence of fractures is superior diagnostically (122). According to some authorities, the presence of a linear skull fracture should not influence the decision about whether to send the child home; others suggest that a child found to have a skull fracture is at increased risk for an extradural hematoma and therefore should be observed in the hospital (125). We believe that whenever the fracture line crosses the groove of the middle meningeal artery or crosses the path of the sagittal or other major venous sinus, the possibility of an extradural hemorrhage is increased, so that the child should be hospitalized for the initial 24 hours. If the fracture involves an air sinus, checking for the next 10 days for signs of intracranial infection is recommended even if the child is not hospitalized.
The rehabilitation of a child with a major head injury is summarized in Table 9.9. A more detailed discussion is presented in a book edited by Ylvisaker (126).
Prognosis
There has been continuing improvement in terms of mortality and ultimate neurologic status in the outcome of severely brain-injured infants and children as defined by a postresuscitation Glasgow coma score of 8 or less (127). The outlook for full intellectual function in children experiencing minor head injuries (i.e., head injuries without associated neurologic manifestations) is generally excellent (128). The prognosis for major head injuries, although less certain, is far better in children than in infants or adults. As a rule, prediction of outcome with respect to the severity of ultimate disability is accurate in only 70% by the end of the first week after the injury (129). In the experience of the Traumatic Coma Data Bank, children younger than age 4 years who have sustained a severe head injury have a cumulative mortality of 62% during the first year after their injury. This compares with a mortality during the same period of 22% in children aged 5 to 10 years and 48% in adults (130). In this series, only 1 in 5 infants had a favorable outcome. These figures, published in 1992, should be contrasted with a mortality of 10% in a cohort gathered in the years 1994 to 1996 (127). A low postresuscitation Glasgow coma score and the presence of cerebral swelling, particularly when it is accompanied by a shift of midline structures, are indices for a poor outcome. These indices have not changed in the last few decades. The poor outcome in infants in all series probably reflects the higher incidence of child abuse and multiple injuries during the first year of life. Keenan and colleagues, working in North Carolina, found that 53% of serious or fatal head injuries incurred by infants younger than 2 years of age were caused by child abuse (131).
In a series of children treated for major head injuries at the Johns Hopkins Hospital and reported in 1983, 29% were normal on follow-up and 53% had returned to school with mild behavioral or cognitive problems. In the latter group, however, 61% had shown evidence of significant language delay, minimal cerebral dysfunction, and learning problems before the accident (132). The outcome tended to be worse in children who sustained prolonged coma. When outcome is stratified according to age, all studies show that morbidity as well as mortality are greater when the injury is experienced in infancy or in early childhood.
Generally, children with only behavioral abnormalities during the early post-traumatic period later achieve normal functioning. No significant improvement in cognitive function can be expected after the first 6 to 12 months after injury. Improvement in speech and motor function, however, can continue for several years (133,134). Decerebrate rigidity during the early postinjury phase, although associated with an early mortality in 25% to 50% of patients, by itself does not preclude functional survival; approximately one-half of survivors have a fairly good quality of life (134). Excessive daytime sleepiness is fairly common after head trauma, especially in adolescents. In some of these, notably when there has been an associated whiplash injury, sleep apnea is responsible; in others, the condition resembles narcolepsy (see Chapter 15) (135). In the experience of Guilleminault and coworkers, daytime sleepiness
P.673

can persist for months to years, and its duration correlates with the severity of the initial head trauma (135). Methylphenidate or amphetamine appear to be the most effective medications. A significant postural and intention tremor affecting primarily the upper extremities has been observed in a substantial number of children recovering from a serious head injury. In the series of Johnson and Hall, the tremor appeared within 2 months of the injury in 49% of their patients and between 2 and 12 months of the injury in 40%. It subsided spontaneously in 54% (136). Hyperekplexia and myoclonic twitches have also been seen (137). The cause of this complication is not clear, but based on radiouptake studies, it probably reflects striatal dopaminergic denervation (138).
TABLE 9.9 Rehabilitation of Head Injuries at Each Level of Consciousness
Level a Neurologic Characteristic Therapeutic Intervention
V No response to stimuli
Glasgow coma score 3
Keep patient in intensive care unit
IV Responds to pain by flexor withdrawal or increased extensor tone; no visual response to light or threat
Glasgow coma score 4–6
Maintain clear airway and prevent pulmonary complications
Control hypertension and tachycardia
Institute seizure prophylaxis
Institute nasogastric tube feeding
Prevent contractures (range of motion, casting, splinting, medication to reduce spasticity)
Get child out of bed; provide sensory stimulation
Continue crisis intervention counseling with family
III Responds to visual stimuli by blinking to light or threat or tracking
Shows nonpurposeful movements of extremities
Glasgow coma score 8–10
Decrease agitation (do not use drugs)
Stimulate visual and auditory responses
Facilitate purposeful movements
Improve head and trunk control
Begin feeding program (delay gastronomy, try nasogastric tube feeding)
Continue family support
II Coma ends
Demonstrates purposeful movements
Follows commands, imitates gestures, responds verbally
Glasgow coma score 12–14
Reinforce appropriate behavior and decrease agitation
Improve gross and fine motor skills
Begin transfers and ambulation training
Start self-care and self-feeding program
Evaluate for developmental and perceptual deficits
Improve child’s speech, language, and cognitive performance: auditory processing, orientation, attention span
Continue family support
Plan discharge
I Is oriented to person, time, and place; may still have cognitive and perceptual deficits
Glasgow coma score 15
Refine motor skills, community ambulation
Increase independence in self-care and community-living skills
Improve orientation, verbal expression, and cognitive and perceptual skills
Assess vision and hearing
Perform neuropsychologic assessment
Complete discharge planning
Continue family preparation and training, adaptive equipment, outpatient medical and therapy follow-up, school placement, involvement with community agencies
aDepending on the severity of the injury, progress can stop at any of these levels. Adapted from Brink J. Case management of head injuries. With permission of the author.
A substantial proportion of survivors from major head injuries experience subsequent emotional and psychiatric disorders. These are covered in another section of this chapter.
Skull Fractures
Linear Fractures
The immature and more flexible skull of the child can sustain a greater degree of deformation than that of an adult. Most skull fractures are linear and asymptomatic and, in the older child, are readily diagnosed by CT using bone windows. In infants, fractures tend to be irregular, so that on plain skull films they are sometimes confused with a suture or wormian bone (Fig. 9.5). In early childhood the fracture can be diastatic and usually involves the lambdoid suture. The separation of the bones indicates that the intracranial pressure increased at the moment of impact.
FIGURE 9.5. Depressed frontal fracture and linear parietal fracture (arrow) on plain skull films of a newborn infant who had undergone a difficult forceps delivery. Associated left-sided epidural hematoma was demonstrated at angiography.
P.674

Often a subperiosteal or subgaleal hematoma, termed cephalhematoma, in the newborn infant accompanies a linear skull fracture. Palpation of the hematoma may falsely lead the examiner to think a depressed skull fracture exists. Imaging studies disclose the underlying linear fracture. A small number of these hematomas calcify (see Fig. 6.1). There rarely is an indication for aspiration of a traumatic hematoma, and insertion of a needle or drain only increases the risk of introducing an infection into the hematoma cavity (see Chapter 6 for further discussion of this complication).
Closed linear fractures generally heal in 3 to 4 months and, except for breaks crossing the path of major vessels or entering the paranasal sinuses, do not require special therapy or observations.
An infrequent complication of a closed head injury with a linear fracture in an infant is the diastatic or growing fracture (139). It occurs in less than 1% of all fractures and usually represents a complication of a serious head injury (140). It occurs when the fracture tears the dura causing the arachnoid to be trapped in the fracture (16). Rare after age 3 years, it is thought to be more common in victims of child abuse, but its appearance is not confined to victims of nonaccidental trauma (140). Dural tears responsible for a growing fracture occur mostly in the parietal area; their presence, unrecognized because the scalp is intact, leads to the development of a CSF-filled cyst between the cortex and the overlying bone. At the same time, the bone edges along the fracture do not unite, apparently prevented by their direct contact with fluid. The bone is resorbed, so that plain skull radiography or CT scans taken after an interval of several months show an irregular bone defect with scalloped edges that is an elliptical erosive cranial defect (139).
In many instances, an associated cortical injury occurs and beneath the area of the cyst there is usually a porencephalic diverticulum of the lateral ventricle, producing a palpable and sometimes visible bone defect and occasionally leading to seizures and progressive hemiparesis. Leptomeningeal cysts should be differentiated from an encephalocele, which is congenital, usually located in the occipital midline, and associated with a regular bone defect. A growing fracture does not resolve spontaneously and must be treated surgically. Treatment, which should be begun as early as possible, involves the surgical separation of the bone from underlying arachnoid, dural repair or replacement, and closure of the bone defect with autologous bone, bone source, or cranioplasty. In rare cases, normal-pressure hydrocephalus develops, which requires a ventriculoperitoneal shunt.
Basal Skull Fractures
Basal skull fractures involve the floor of the anterior, middle, or posterior fossae; they are uncommon in children. Their presence can be suspected when the child has signs of bleeding from the nasopharynx or the middle ear or has postauricular ecchymoses (Battle’s sign) (141). Epistaxis is frequent in childhood head injury, however, because of the high incidence of nasal fractures.
Fractures of the base of the anterior fossa can lead to hemorrhage into the orbit. Under these circumstances a subconjunctival hemorrhage represents a forward extension of blood behind the optic globe, in contrast to an anterior hemorrhage, which arises from a direct blow to the eye. Exophthalmos and subconjunctival hemorrhage occur in conjunction with “raccoon eyes.” Fractures of the mastoid portion of the temporal bone result in postauricular ecchymoses.
Distinctive unilateral fresh purpuric hemorrhages in the antitragus, triangular fossa, and helix of the ear have been termed the “tin ear syndrome” (142). The condition can be associated with retinal hemorrhages and an ipsilateral subdural hematoma, a syndrome considered to be pathognomonic of child abuse. Rotational acceleration of the head produced by blunt trauma to the ear is believed to produce this syndrome (142).
CSF rhinorrhea can accompany a fracture of the floor of the anterior fossa that has involved the cribriform plate. It represents a rare complication of head trauma in children, but the high risk of intracranial infections (20% to 37% as reported by Wilson and coworkers) makes its recognition imperative (143).
CSF rhinorrhea usually appears within the first 2 days after the injury, but it may not become apparent for up to several years. In 70% of cases, it ceases within 1 week,
P.675

and in a large proportion of the remainder, it ends within 6 months. It is accompanied by anosmia in approximately 75% of cases (144). Because the glucose content of nasal discharge (40 mg/dL) differs little from that of CSF, glucose determinations are of no value in distinguishing CSF rhinorrhea from ordinary nasal discharge (145).
Several imaging modalities have been evaluated for their ability to pinpoint the site of CSF leakage. High-resolution CT and MR cisternography using a water-soluble nonionic contrast medium appear to be equally accurate in localizing the site and the extent of the CSF fistula (146,147) and are superior to CT cisternography (148). In our institutions, high-resolution CT and CT cisternography have proven to be adequate for pinpointing the site of CSF leakage.
When rhinorrhea has not ceased within 1 week to 10 days, surgical repair of the dural and bony defect is usually indicated, inasmuch as any CSF leak, particularly one that results in CSF rhinorrhea, predisposes to meningitis (143). The infecting agents are a variety of gram-positive cocci and gram-negative bacilli (149). Conservative management of CSF fistulas is successful in most instances. Bedrest with the head of the bed elevated is usually sufficient. Continuous lumbar or ventricular drainage often works in persistent CSF fistulas. If conservative treatment fails, surgical repair (intradural, extradural, or both) is required.
Whether chemoprophylaxis should be used on all patients with basal skull fractures has been the subject of considerable debate. Prospective studies have shown that prophylactic ampicillin does not reduce the risk of meningitis but can change the flora so that gram-negative organisms become the infecting agents (150). We therefore prefer to withhold antibiotics until careful observation has revealed evidence of infection. Lumbar puncture to obtain CSF and ascertain the presence of an infection is unwise because it can facilitate the entry of organisms into the anterior fossa by lowering intracranial pressure. The patient should be observed carefully, with therapy undertaken only when indicated by symptoms and subsequently by the CSF examination.
Injury to the cranial nerves, particularly the olfactory, facial, and acoustic nerves, can accompany basal fractures. Complete loss of the sense of smell is usually permanent; 90% of facial nerve injuries recover spontaneously (151). Deafness can be a temporary or permanent sequela to temporal bone fractures.
Labyrinthine disorders, notably vertigo and spontaneous or positional nystagmus, are common. In the experience of Eviatar and coworkers, more than 50% of children who had experienced a head injury complained of dizziness, headache, or both (152). The condition is usually transient. In approximately one-half of the cases, electronystagmography provides objective evidence of the presence of an injury to the labyrinth (152). Less commonly, one may encounter episodic vertigo that resembles Ménierè disease. It can result either from fistulization of the bony labyrinth with disturbed perilymph-endolymph pressure relationship or from direct injury to the membranous labyrinth or to the endolymphatic draining system (153).
CSF otorrhea is seen in approximately 0.5% of childhood head injuries, but in approximately 95% of cases it stops spontaneously within 7 to 10 days. Unlike CSF rhinorrhea, recurrence of leakage is rare (151). Rarely, transient total blindness can follow an apparently mild blunt head injury. The onset can be immediate or can be delayed for days or weeks (154). The child may not complain of vision loss, but can appear restless and disoriented and have an unsteady gait. The cause for blindness is uncertain, and outlook for recovery is poor. In some children the event may recur (155,156).
The diagnosis of basal fractures often depends on the clinical findings because in many instances, the fracture is not readily demonstrated by imaging studies. As a rule, the child with a basal fracture associated with hematotympanum or Battle’s sign is hospitalized and observed. However, when the neurologic examination is normal and imaging studies do not show an intracranial injury, hospitalization might not be required (157). Because the incidence of meningitis in children with basilar skull fractures is only 1%, antibiotics should be withheld.
When meningitis is associated with a CSF leak, antibiotic treatment should be continued well after the leak stops.
Air within the cranial cavity (pneumocele) or air within the brain (pneumocephalus) rarely complicates head injury (158). When it does occur, it most commonly follows a fracture into the frontal sinuses and is usually an incidental finding on radiography. Pneumocephalus can denote either extradural or intradural air. Intradural air is of concern because it signifies a dural tear that can require surgical repair. Extradural air often denotes an air sinus disruption that may well resolve spontaneously. However, follow-up imaging is essential in all cases of pneumocephalus. Because meningitis develops occasionally, some centers suggest prophylactic antibiotic therapy until the air has resolved. Tension pneumocephalus is even more rare; it presents a neurosurgical emergency because of the rapid increase in intracranial pressure (159).
Depressed Fractures
Depressed fracture is a common consequence of perinatal injury, often the result of a difficult forceps delivery (ping-pong fracture; see Chapter 6). It also can occur with any localized skull trauma in later childhood in which there is an impact with a small surface with sufficient force to
P.676

depress fragments of the inner table by at least the thickness of the skull (16). It is often associated with a break in the skin (compound fracture) and localized cerebral injury. The extent of the bony injury is best diagnosed by CT. In the past, elevation and examination of the underlying dura was recommended if the depression was greater than approximately 3 mm and the fracture did not reduce spontaneously. Steinbok and coworkers adopt a more conservative approach, reserving surgery for infants with compound depressed fractures or those with focal neurologic signs. The outcome appears to be the same with or without surgery (160). Ersahin and colleagues concur and suggest that conservative treatment is indicated whenever there is no underlying hematoma and the depression is less than 1 cm (161). We believe that a case can be made for elevating on cosmetic grounds a deep or unsightly depression, particularly when it is located in the frontal region.
Compound Fractures
Compound fractures of the skull are seen in approximately 20% of children with major head trauma (66). In this kind of injury, medical treatment is limited to an initial cleansing of the scalp, institution of antibiotics, and tetanus prophylaxis. Anticonvulsant therapy is used routinely when the bony fragments have penetrated beyond the dura. Traditionally, phenytoin has been the preferred drug. Its use requires an intravenous loading dose of 10 mg/kg and subsequent oral doses to maintain a blood level between 12 and 20 μg/mL. In children, maintaining this level is usually difficult and requires careful, repeated determinations of blood levels and adjustments of dose (162). Because of the unreliability of phenytoin in infants and small children and its relatively high incidence of side reactions, we prefer to use carbamazepine as a maintenance anticonvulsant.
Scalp Lacerations
Scalp lacerations can cause considerable blood loss. If any doubt exists about the presence of a scalp injury, the child’s hair should be clipped and the area around the wound widely shaved. If examination and radiography do not show an underlying fracture, the wound should be closed in anatomic layers after careful débridement with strict adherence to aseptic techniques. Closure of the galeal layer to stem hemorrhage and protect from infection is optimal. Tetanus immunization should be administered.
Complications
Extradural Hematoma
An extradural or epidural hematoma is a localized accumulation of blood between the skull and the dura. It occurs in approximately 1.0% to 3.4% of children hospitalized for head trauma (66,163). According to Matson, nearly one-half of childhood cases occur during the first 2 years of life (164). In this age group, the injury is usually the result of a fall of less than 4 to 5 feet, and no other significant injuries are seen (68,165). In the experience of Shugerman and coworkers, 47% of children who developed an epidural hematoma had falls of 6 feet or less and skull fractures were seen in only 18%. Nonaccidental trauma was diagnosed in only 6% (165).
Pathogenesis
An extradural hematoma develops at or near the point of traumatic impact, usually in the temporoparietal region, less commonly in the frontal region. It is nearly always unilateral. In adults, the hematoma is almost invariably caused by a laceration of the middle meningeal artery or one of its branches. In children, extradural bleeding can be the consequence of even mild injury that has produced a tear in the dural veins, a scenario seen in 15% of patients in Matson’s series (164), in the meningeal artery or its branches, in the accompanying middle meningeal veins (60% of cases), or in the smaller emissary veins of the dural sinuses (15%). The stripping of the dura from the skull successively tears more vessels, contributing to the enlargement of the hematoma. This successive tearing of dural vessels is probably why Mazza and coworkers were unable to identify a source in some 20% of their cases (166). Arterial bleeding results in a rapid deterioration of the clinical condition as the hematoma enlarges and produces acute cerebral compression (Fig. 9.6A and B). The severity of symptoms depends on the size of the hematoma, the speed of its evolution, and the development of transtentorial herniation. Transtentorial herniation is more likely when cerebral edema accompanies the hematoma. When bleeding arises from veins, neurologic symptoms progress gradually, a feature more common to the extradural hematomas of childhood than to those of adult life (167). Rarely, the hematoma can stop growing; it can then resolve or ultimately calcify.
The effects of transtentorial herniation accompanying an epidural hematoma are similar to those seen with other space-occupying lesions and are considered in Chapter 11.
Clinical Manifestations
In adults, an extradural hematoma characteristically is preceded by a temporary loss of consciousness followed by partial or full recovery, with subsequent deterioration of sensorium and appearance of focal neurologic signs. This sequence is rare in children; it did not appear in any of Mealey’s 20 children with extradural hematoma (14). It appeared in only one-third of 125 cases reported by McKissock and colleagues (168) and in 40% of Mazza’s 62 cases (166). More commonly, the child appears to be little affected by the initial injury or, at worst, has a brief period of unconsciousness. After an interval of minutes to
P.677

several days, a progressive impairment of consciousness develops and neurologic signs appear. Of the children reported by McKissock and colleagues, the condition took this course in 67% (167). In general, the younger the child, the longer is the latency. In Mealey’s series, 50% of children younger than 6 years of age remained asymptomatic for 12 hours or longer (14). The clinical picture of extradural hematoma is summarized in Table 9.10.
FIGURE 9.6. Cerebral compression and tentorial pressure cone owing to extradural hematoma. A: Coronal section of the brain. B: Diagram illustrating the swelling and displacement of the involved cerebral hemisphere with distortion of brainstem structures. (From Mealey J. Pediatric head injuries. Springfield, IL: Charles C. Thomas, 1968. With permission.)
Aside from the delayed progressive impairment of consciousness, the most significant neurologic signs are pupillary inequality, hemiparesis, papilledema, and changes in vital signs. As the hematoma enlarges, signs of transtentorial herniation appear. The earliest of these signs is a dilated pupil that soon becomes unreactive to light and, in approximately 90% of instances, is on the side of the lesion (167). This dilation is followed by hemiparesis, usually contralateral to the hematoma, and finally by decerebrate rigidity and cardiovascular signs of decompensated increased intracranial pressure. Ipsilateral hemiparesis can develop caused by compression of the cerebral peduncle against the tentorial edge by a contralateral mass, and, in the days before neuroimaging, resulted in a false localizing sign (Kernohan sign) (168). Blood loss in infants can be sufficient to produce shock. When the latency period in infants is as long as days or weeks, progressive anemia can provide a clue to the diagnosis. Seizures resulting from an
P.678

epidural hematoma are rare and were present in only 7.5% of children seen by Hendrick and colleagues (66).
TABLE 9.10 Common Clinical Features of Extradural Hematoma
Symptoma Percentage of Patients
Vomiting 62.5
Unequal pupils 55.0
Delayed loss of consciousness 48.7
Skull fracture, all types 40.0
Hemiparesis 25.0
Papilledema 22.5
Depressed skull fracture 20.0
Third-nerve paresis, other than pupillary dilation 17.5
Retinal hemorrhages 12.5
aSymptoms listed occur with a significantly greater frequency in children with extradural hematomas than in the general head trauma series.
Modified from Hendrick EB, Harwood-Nash DC, Hudson AR. Head injuries in children: a survey of 4,465 consecutive cases at the Hospital for Sick Children. Toronto, Canada. Clin Neurosurg 1963;11:46.
Occasionally, an extradural hematoma develops in the posterior fossa (169,170). Bleeding usually arises from the lateral dural sinuses, and fracture lines crossing the lateral sinus are seen nearly always. In almost all instances the history includes a severe fall on the occiput, followed by persistent impairment of consciousness, headache, vomiting, and neck stiffness. Only approximately one-half of the children have such posterior fossa signs as nystagmus, cerebellar ataxia, and cranial nerve palsies. Evacuation of a posterior fossa hematoma generally constitutes an emergency procedure.
FIGURE 9.7. Extradural hematoma in a victim of child abuse. The lentiform shape of the hematoma and the fact that the collection of blood crosses to the opposite side are consistent with the presence of an extradural hematoma. Bone films (not shown) disclosed an underlying skull fracture. (Courtesy of Dr. Franklin G. Moser, Division of Neuroradiology, Cedars-Sinai Medical Center, Los Angeles, CA.)
Diagnosis
The diagnosis of an extradural hematoma rests principally on the clinical picture, and in some cases surgical treatment is so urgent that there is no time for imaging studies (84). Extradural hematoma must be differentiated from an acute subdural hematoma, an intracerebral hematoma, and severe brain swelling with or without contusion. Because both an acute extradural hematoma and an acute subdural hematoma require immediate surgical evacuation, the distinction between these two entities is academic. CT almost always detects the extradural hematoma, except when the collection of blood is very thin (Fig. 9.7). The usual appearance on CT scanning is a lenticular hyperdense lesion that crosses the suture lines (Fig. 9.8A). This is in contrast to a subdural hematoma, which tends to be crescent shaped and interdigitates with the cortical gyri (Fig. 9.8B). MRI is equal or superior to CT but is rarely indicated because of the time constraints associated with treating an acute injury. In 20% to 40% of children with extradural hematoma, a skull fracture is not
P.679

detectable by radiographic examination or even at operation (14), so that time should not be spent on searching for the fracture. EEG and MRI, which are also time-consuming procedures, should not be used to further delineate the hematoma.
FIGURE 9.8. A: Left extradural hematoma in a teenager after a skateboard accident. Note the lentiform shape of the lesion. Epidural blood dissects the dura from the skull and causes this lenticular appearance, which spans the dural attachments between suture lines. B: Right subdural hematoma. Note that the bleeding interdigitates with the cortical sulci along the periphery of the hemisphere. The blood has extravasated into the subdural space between the arachnoid and the pia mater.
Posterior fossa extradural hematomas are more difficult to recognize. The biconvex and more focal configuration of the extradural hematoma frequently permits the CT to distinguish between extradural and subdural accumulations.
Treatment
Operative removal of the clot can be performed occasionally through burr holes, but usually requires a craniotomy for complete removal and arrest of the bleeding. If the child’s condition does not allow time for imaging studies, a low temporal burr hole is made for confirmation of the diagnosis and rapid removal of the clot. Further burr holes or craniectomy might then be needed. For details of the neurosurgical procedure, see Marshall (171) and Jennett and Teasdale (119). Repeat CTs are useful in assessing a patient after surgical evacuation of a hematoma.
Prognosis
Although the majority of patients show dramatic improvement, the general prognosis of extradural hematoma is grave if the condition is not treated in a timely fashion. Surgery for a large acute extradural hematoma is a life-saving procedure when performed within the first few hours of injury. In Matson’s series, published in 1969, the mortality was nearly 10%, with another 20% of patients being left with major neurologic residua (164). Generally, infants fare worse than older children. In a 1986 series of children younger than 2 years of age reported by Choux and coworkers, 68% were left without apparent sequelae, some 10% died, and 12% had major residua (172). As a rule, the likelihood for sequelae increases if the child deteriorates acutely or had a depressed level of consciousness or neurologic abnormalities by the time of surgery. Posterior fossa extradural hematomas also have a grave prognosis, partly because they tend to progress rapidly (173). Despite the ready availability of imaging, the mortality in one series was 35%, and 20% of patients were left with a moderate disability (169). In more recently compiled cases the outlook was better, and in the series of Berker and coworkers from Turkey, 87% of children had a favorable outcome (170). It is clear that the prognosis for survival and the extent of the neurologic deficit are related to the early diagnosis of this complication and the presence of any associated brain damage.
Subdural Hematoma
A subdural hematoma is a collection of bloody fluid between the dura and the arachnoid over the cerebral mantle. It is a relatively common complication of recognized and unrecognized head trauma of childhood and represents one of the two major neurosurgical problems of
P.680

infancy. In the series of Choux and coworkers, this complication was seen in 4.3% of children with head trauma; 73% of cases occurred in children younger than 2 years of age (172). A subdural hematoma should be distinguished from a postmeningitic subdural effusion containing clear or xanthochromic fluid with a high protein concentration. The etiology and the course of treatment of a postmeningitic subdural effusion differ from those of a subdural hematoma and are discussed in Chapter 7.
FIGURE 9.9. Subacute subdural hematoma. Gross appearance of the brain. Note blood oozing from the subdural compartment. (Courtesy of Dr. Harry V. Vinters, Division of Neuropathology, UCLA Center for the Health Sciences, Los Angeles, CA.)
Pathogenesis
Subdural bleeding usually arises from the veins that pass from the cerebral cortex to the dural sinuses, bridging the potential subdural space. Skull distortion at the moment of injury, particularly in infants, and possibly the relative movement of the brain within the skull can so stretch these veins that they rupture and bleed beneath the dura, separating the dura from the underlying arachnoid membrane. It is also likely that a tear of the arachnoid allows CSF to leak into the subdural space. Venous bleeding also can arise from a laceration of the dura or from a direct injury to a dural sinus, as can happen with depressed fractures. Less often, the bleeding originates from cortical arteries and is associated with cerebral contusion (16). In 80% to 85% of infants, the hematoma is bilateral and located in the frontoparietal region. In a large percentage of infants whose hematomas result from postnatal trauma, parental abuse can be suspected from evidence of soft tissue bruising and radiographic evidence of multiple episodes of skeletal trauma (174,175).
The appearance of the hematoma varies with the age of the lesion (Fig. 9.9). In the acute stage (less than 48 hours), the fluid is composed of blood and clot. Subsequently, in the subacute stage (2 to 14 days), the formed elements break down, producing a fluid that is a mixture of clot and fluid and changes in color from chocolate to straw. It contains few red cells but large amounts of methemoglobin and bilirubin. Within approximately 1 week, a membrane forms from the inner surface of the dura and ultimately envelops the clot or fluid collection. After about 14 days the hematoma is chronic and consists solely of fluid blood (16). Eventually, the lesion can enlarge, probably from the leakage of albumin from the thin-walled and abnormally permeable vessels of the outer subdural membrane (176). The accumulation of albumin in the subdural pocket increases the osmotic pressure and causes an influx of water. The evolution of a subdural hematoma can be documented by CT; it is depicted in Fig. 9.10. Acute lesions tend to be hyperdense, and lesions that have produced symptoms for more than 3 weeks are usually hypodense; when symptoms have been present for 1 to 3 weeks, the fluid collection tends to be isodense. Under these circumstances the
P.681

hematoma is recognized by ventricular compression and distortion.
FIGURE 9.10. Temporal evolution of a subdural hematoma. (Courtesy of Dr. E.C.Alvord, Jr., Department of Pathology, University of Washington, Seattle, WA).
Although the enlarging subdural hematoma can produce symptoms of increased intracranial pressure, no evidence exists that the presence of either a hematoma or the membranes interferes with brain development. Instead, any permanent neurologic deficit results from the original trauma that caused the hematoma. Additionally, often a diffuse damage occurs secondary to increased intracranial pressure and diminished cerebral perfusion. Single photon emission CT (SPECT) scanning has been of assistance in demonstrating both regional and diffuse abnormalities in cerebral perfusion (177). In experimental animals, focal ischemic damage in the hemisphere underlying the subdural hematoma is believed to result from the release of vasogenic substances from the hematoma that enhance vasoconstriction and attenuate vasodilatation (178). The focal ischemia, in turn, induces localized cytotoxic edema and the release of free radicals (179). The biochemical changes of the underlying parietal white matter show an accumulation of lactate and a loss of N-acetylaspartate. The severity of these alterations is believed to be prognostic of the outcome (180).
Unlike subdural hematomas in older children and in adults, the post-traumatic subdural hematoma of infants tends to recollect repeatedly, even after total evacuation. Mealey attributed the reaccumulation of fluid to the disproportion between the enlarged skull and the previously compressed brain (14). Most investigators now believe that rebleeding is the principal cause for recurrence and persistence of the subdural hematoma (181). Although it is unlikely that a subdural hematoma inhibits brain growth, it produces an enlargement of the skull and creates a pocket that tends to refill with blood. A similar disproportion between the volume of the brain and the skull increases the incidence of subdural hematomas in infants with congenital or acquired cerebral atrophy (14).
Clinical Manifestations
To a great extent, clinical manifestations depend on the patient’s age. In older children, as in adults, the disorder can be acute or chronic. In both groups, symptoms of increased intracranial pressure predominate.
When a subdural hematoma after a serious head injury takes an acute course, symptoms develop within the first day or two. Venous bleeding usually does not produce symptoms unless it is accompanied by a major cerebral contusion or laceration. In these conditions, the hematoma is only one of several components of the injury, and its evacuation is not usually followed by rapid recovery (182). In part, the poor outlook reflects the aforementioned reduction of cerebral blood flow and metabolic rate throughout the hemisphere underlying the hematoma. Brain injury, usually in the form of an acute subdural hematoma, is seen in a significant proportion of battered babies, and an acute subdural hematoma is the most common cause of death or physical disability in infants; it must, therefore, be sought in all victims of child abuse.
P.682

Chronic subdural hematomas are the consequence of one of three scenarios: trauma, a complication of ventricular shunt placement, and the consequence of an infectious or parainfectious process (181).
A post-traumatic chronic subdural hematoma is most commonly seen in infants. It is usually encountered between ages 2 and 6 months, the average age at admission being 4 months. In this age group, the lesions are bilateral in approximately 80% of cases. In some 60% of infants, the environmental history or other evidence of recent unreported physical trauma suggests child abuse (68,183). The more serious the head injury, the greater is the likelihood of abuse. Of 45 infants younger than 1 year of age who sustained a skull fracture, only 11% were victims of child abuse, whereas of 19 infants who had a subdural hematoma or other forms of serious intracranial injury, 95% were abused (9). In the experience of Bruce and Zimmerman, 80% of traumatic deaths occurring in children younger than 2 years of age were nonaccidental; in fact, infants rarely sustain an accidental injury that is sufficiently severe to render them unconscious (184). Harcourt and Hopkins observed that intraocular hemorrhages in the absence of subdural effusions and external evidence of ocular trauma are commonly encountered in battered children. They postulated that these hemorrhages result from the gravitational effects of swinging the infant around by its feet (185). Additionally, compression of the thorax inducing an abrupt increase in intracranial pressure can play a role. Whether the abused infant is injured by these means or by violent to and fro shaking has since been disputed: Much greater gravitational forces can be generated by striking the infant’s head against a mattress (183).
A chronic subdural hematoma can also develop in older children, especially during adolescence. In the latter age group, the clinical picture is one of a gradual change in personality and alertness, headaches, and, ultimately, seizures or rapid deterioration of consciousness. Often, there is no history of antecedent head trauma. The hematoma is unilateral in 80% of instances, and its differentiation from a tumor of the cerebral hemispheres is difficult clinically. The diagnosis is usually made by imaging studies. MRI permits better delineation of small subdural hematomas, hematomas that are adjacent to the falx or the tentorium, and isodense accumulations. MRI also delineates any associated cerebral injuries and assists in timing the lesion (89).
Occasionally, an arachnoid cyst in the middle fossa can predispose to a chronic subdural hematoma. The management of this condition is discussed by Swift and McBride (181) and Parsch and coworkers (186).
CT and MRI can be of considerable assistance in confirming that a child has been battered (86,187,188). An interhemispheric subdural hematoma in the subtemporal or the parieto-occipital region accompanied by a skull fracture can be documented in more than 50% of abused children, whereas in trauma unrelated to abuse, bleeding in this region accompanied by a skull fracture is seen in only 13% (187,189). Additionally, the presence of both acute and chronic hematomas supports the diagnosis of child abuse.
TABLE 9.11 Clinical Features of 116 Cases of Infantile Subdural Hematoma
Symptom of Finding Percentage of Infants
Tense anterior fontanelle 73
Vomiting 70
Seizures 60
Retinal or subhyaloid hemorrhages 54
Abnormal skull circumference 40
Impaired consciousness 22
Papilledema 12
Skull fracture 13
Other fractures 17
Modified from Till K. Subdural haematoma and effusion in infancy. BMJ 1968;2:400.
An interhemispheric subdural hematoma should be distinguished from an interhemispheric subarachnoid hemorrhage (falx sign), which can be seen after perinatal trauma. In interhemispheric subarachnoid hemorrhage, the hemorrhage tends to extend along the entire interhemispheric fissure (190).
Although in infancy a chronic subdural hematoma lacks a characteristic clinical picture, certain features should suggest the condition. Lethargy and seizures are the most common presenting features. Seizures occur in approximately one-half of the patients; they are focal or, more commonly, generalized. Vomiting, fever, and hyperirritability or lethargy are other common clinical features (Table 9.11) (191,192). Most often, the infant’s history includes failure to gain weight; refusal of feedings followed by frequent episodes of vomiting, some of which might be projectile; irritability; progressive enlargement of the head; and, ultimately, a seizure. Often, symptoms are present for several months before a diagnosis is made.
On examination, the infant can be febrile as a result of dehydration or blood within the cranial cavity. The head is enlarged, with a prominent parietal or biparietal bulge. The fontanelle is full, and a setting-sun sign of the eyes might be noted. Funduscopy can reveal retinal hemorrhages, subhyaloid hemorrhages, or, less commonly, papilledema. Retinal hemorrhages have been found in more than 50% of infants with a subdural hematoma, and almost invariably indicate a nonaccidental injury (164,174). Focal neurologic signs, including hemiparesis or facial palsy, are present in 15% to 25% of patients.
Laboratory studies are usually of little help in establishing the diagnosis, although approximately 50% of the infants are anemic. It is important to exclude a bleeding
P.683

disorder, and various inborn errors of metabolism, notably kinky hair disease (Menkes disease), a sex-linked disorder of copper transport, and glutaric aciduria type 1, which can present with a subdural hematoma and subperiosteal bleeding (76,193). Lumbar puncture can reveal grossly bloody or xanthochromic fluid, evidence of an associated cerebral injury. The protein level can be elevated, and a pleocytosis can be present if the hematoma has been long-standing. The EEG is of little assistance in either the diagnosis or localization of a subdural hematoma and frequently fails to show any focal abnormality.
Diagnosis
A chronic subdural hematoma should be suspected in an irritable infant who has failed to gain weight and has developed an enlarged head and a tense fontanelle. Characteristically, the head assumes a biparietal bulge that differs from the frontal bulge of early hydrocephalus. Confirmation of the diagnosis and determination of the size of the hematoma and of any associated brain damage rest on CT and, if necessary, MRI.
Treatment
Several regimens have been proposed for the treatment of post-traumatic infantile subdural hematoma. These are based on the classic work of Sherwood (194) and Ingraham and Heyl (195), and are summarized by Swift and McBride (181) and in greater detail in a symposium edited by El-Kadi and Kaufman (196).
We have found that with serial subdural taps the collection of fluid dries up completely in selected patients, thus eliminating the need for a surgical procedure. When fluid re-forms between tapping or when imaging studies reveal the presence of a subdural clot, surgical intervention must be considered. This occurs in approximately 50% of cases. In the remainder of instances, fluid formation gradually decreases and the membranes disappear. However, this does not happen when the brain and skull are disproportionate, as is the case in cerebral atrophy. The operative removal of subdural membranes can never be complete. It is no longer practiced except when craniotomy is being performed for the removal of a large blood clot, which seldom is present in the chronic stage of infantile subdural hematoma. Instead, a subdural-peritoneal shunt (119,197) using simple tubes without valves allows the drainage of fluid and reduces the volume of dead space, allowing it to be obliterated by the growing brain. A unilateral shunt is sufficient in many cases (181,192). Subdural membranes completely disappear and the vascularity becomes reduced with time as long as the subdural space is adequately drained. Improvement can be confirmed by repeated imaging studies. Continuous external subdural drainage is also useful in selected patients before subdural-peritoneal shunting because in approximately one-half of the children the procedure makes it possible to avoid shunt placement (198).
Chronic subdural hematomas occurring in older children and adolescents are drained through burr holes (199). At this age, too, the fluid collection is frequently bilateral.
Prognosis
The prognosis for an infant with a subdural hematoma correlates with the extent of damage sustained by the brain rather than with the volume of subdural fluid itself. If brain injury has been extensive, the brain does not expand and the hematoma can calcify or ossify. Removal of a calcified subdural hematoma is of no advantage (200).
The prognosis for children whose subdural hematoma resulted from nonaccidental trauma has been shown to be relatively poor (201,202). Even when no gross neurologic deficits are present, abused children have a higher incidence of neurologic residua, significantly lower IQ scores, growth failure, and a significantly higher incidence of emotional handicaps compared to children with accidental subdural hematoma (203). These differences may result in part from the repetitive rotational forces experienced by the brain during shaking and in part from the socioeconomic environment of children who have experienced abuse (201,203).
Post-Traumatic Epilepsy
Seizures associated with trauma have been classified according to their time of onset into immediate, early, and late types (64,204). A few patients experience seizures 1 or 2 seconds after their head trauma. Such immediate seizures are most probably the result of direct mechanical stimulation of cerebral tissue having a low seizure threshold.
Seizures can appear during the first week after major cerebral trauma (early post-traumatic seizures). These arise from cerebral edema or from intracranial hemorrhage, contusion, laceration, or necrosis. The convulsions are usually generalized, but unilateral seizures and focal twitching (epilepsia partialis continua) can be seen. Generally, early post-traumatic epilepsy is more common in children than in adults, and in the experience of Jennett it was encountered in approximately 10% of head-injured children aged 5 years or younger (204). A history of prior seizures or developmental abnormalities is seen in approximately one-half of cases (205). Status epilepticus occurs in approximately one-fifth of children and is most likely to occur during the first hour after trauma (64). In the experience of Hendrick and his group, 7.4% of children with head trauma requiring hospitalization had seizures during the early post-traumatic period (66). The incidence was highest in infants younger than 1 year of age; those subjected to perinatal injury were particularly susceptible (see Chapter 6).
P.684

In the experience of Hendrick and colleagues, 24% of patients with early post-traumatic seizures had an associated skull fracture (66). Closed and compound depressed fractures are particularly common, together accounting for approximately one-half the fractures seen in patients with early post-traumatic epilepsy. Seizures, and even status epilepticus, are far more likely to occur in children who sustained relatively minor head trauma than in adults with similar trauma.
Late post-traumatic seizures tend to develop within the first 2 years after the injury. The mechanisms that cause post-traumatic epilepsy are poorly understood. In some instances seizures are believed to originate from a cerebromeningeal scar, with the epileptic focus localized to grossly normal tissue (206). Experimental studies show that head trauma induces long-term alterations in both the excitatory and inhibitory circuits in the hippocampus, resulting in a persistent decrease of the seizure threshold (207). Approximately 75% of children with late post-traumatic seizures had no significant deficits at the time of the injury (208). In approximately 50% of cases, seizures appear during the first 12 months after trauma (64). The overall incidence of late post-traumatic epilepsy is difficult to estimate because the figure is lowered by the inclusion of mild head injuries in any prospective series. In the series of Annegers and colleagues, which included both children and adults, the 5-year and 30-year cumulative incidence after severe head injury was 10% and 16.7%, respectively. It is of note that after severe injury the incidence of new-onset seizures remained elevated throughout the follow-up period, an indication that the interval between serious head injury and the onset of post-traumatic seizures can be many years (209).
According to Jennett, the likelihood of late post-traumatic epilepsy is increased by the presence of any of three factors: an acute hematoma, a depressed skull fracture, and early epilepsy (64). Temkin arrived at similar risk factors (210). These conclusions, derived from his experience with a mixed adult and pediatric population, need some amplification. Annegers and colleagues found that in children the presence of early post-traumatic epilepsy did not predict late post-traumatic epilepsy (205), and Jennett noted that when early focal seizures occured in children, the incidence of late post-traumatic seizures did not increase significantly (64). It is not clear whether the location of brain injury, as judged from the fracture site, affects the likelihood of late epilepsy. Most authorities agree, however, that injuries to the parietal lobe and to the anterior and medial parts of the temporal lobe are most likely to be followed by late post-traumatic epilepsy (210).
In Jennett’s data, only 2% of children who retained consciousness after head trauma developed post-traumatic epilepsy (64). The percentage rose to 5% to 10% when consciousness was lost for 1 hour or longer. By comparison, children who sustained brain laceration had a 30% incidence of post-traumatic epilepsy. In general, if the dura is penetrated, the incidence of post-traumatic epilepsy increases at least twofold. Temkin confirmed these data (210). Another factor that influences the incidence of late epilepsy is the duration of post-traumatic amnesia; late epilepsy is twice as common in children with more than 24 hours of post-traumatic amnesia.
Seizures can take several clinical forms. They can be generalized or focal with secondary generalization. Focal seizures can be preceded by an aura consisting of motor phenomena such as clonic movements of an extremity or by somatosensory phenomena. The seizures also can be focal without secondary generalization, but petit mal (absence) attacks do not occur as a result of trauma (64,211). The EEG has proved to be uniformly unsuccessful in predicting post-traumatic epilepsy in children (64,212).
The diagnosis of post-traumatic epilepsy depends on the antecedent history of head trauma and the absence of any pretraumatic seizure history. The possibility of an intracranial hematoma should always be excluded using imaging studies.
Treatment of post-traumatic epilepsy is similar to that used for focal or generalized seizures of unknown cause (see Chapter 14). Although prophylactic phenytoin and other anticonvulsants reduce seizures during the first week after injury, they do not prevent late post-traumatic seizures (213).
Generally, the prognosis of post-traumatic seizures is good. The clinical impression that seizures that begin within 2 years of trauma have a better likelihood of subsiding than those with a later onset has not been proven by more recent studies. Jennett and Teasdale concluded that once a patient has developed late post-traumatic epilepsy, the patient will always remain prone to a seizure disorder, even though he or she can experience remissions of 2 years or longer (119). In approximately 20% to 50% of all patients, seizures gradually become less frequent after the third year and finally cease completely (64,214). In all instances, medical therapy should be used first, but surgery for excision of the meningocerebral scar and any underlying cysts or gliosis should be considered in patients whose seizures persist for 2 or more years despite adequate anticonvulsant therapy.
Some children are subject to profound but temporary alterations of consciousness after relatively mild head trauma. The attacks, similar to complicated migraine, are marked by a scotoma or other visual symptoms, including cortical blindness, hemianopsia, brainstem signs, confusion or depression of consciousness, headache, nausea, and vomiting (215,216). Convulsions can develop after the onset of the migraine-like symptoms, as can a hemiparesis. Attacks do not develop immediately after head trauma, but rather after a symptom-free interval of several
P.685

minutes to an hour. The great majority of attacks resolve completely within 24 hours. During an attack, EEG tracings show symmetric or asymmetric slowing, whereas cerebral angiography and neuroimaging are generally normal. The mechanism for the attacks is unknown (216).
Major Vascular Injuries
Blunt or penetrating injuries to the major vessels of the head or neck are relatively uncommon in children who have experienced trauma to that area. Blunt carotid injuries can result in a contusion or tear of the wall of the internal carotid artery, with a subsequent dissecting aneurysm, or in a carotid-cavernous fistula (217,218,219). A traumatic carotid-cavernous fistula can also be caused by a sphenoid bone fracture that lacerates the internal carotid artery as it passes through the cavernous sinus. Symptoms include unilateral pulsating exophthalmos, an intracranial bruit, and paralysis of the cranial nerves, most commonly the sixth (220). Traumatic thrombosis of the internal carotid artery has been reported in children as a result of relatively minor injuries to the head or neck, such as following puncture of the soft palate by a lollipop stick (221). Internal carotid artery dissection can appear immediately after what often is relatively minor trauma or can develop after a few hours or days (222,223). Spontaneous internal carotid artery dissection is an important cause of cerebrovascular accidents in children (see Chapter 13). Symptoms include focal ischemia or a headache accompanied by an objective and subjective bruit and by Horner syndrome. Internal carotid artery dissection is treated by anticoagulation and early thrombectomy if the site of obstruction is accessible (223).
Trauma to the cervical spine can produce stretching of the vertebral arteries, with disruption of the endothelium and subsequent arterial dissection. Symptoms of intermittent brainstem and cerebellar dysfunction can develop and last for months or years. When, as often occurs, the patient has no history of antecedent trauma, cervical spinal trauma is difficult to differentiate from complicated migraine (224,225,226). MRI of the upper cervical spine and foramen magnum region and MR angiography can assist in the diagnosis (227). Basilar artery migraine is covered in Chapter 15.
Delayed Deterioration after Mild Head Injury
A relatively common, potentially fatal complication of head injury in the pediatric age group is one of rapid secondary deterioration occurring within minutes or hours after relatively minor trauma and after a lucid interval or a period of improved consciousness (228). A significant proportion of children with this symptom complex have early post-traumatic seizures, sometimes with focal or generalized status epilepticus. The syndrome also has been encountered in youngsters who have experienced repeated concussive injuries in sports (“second-impact syndrome”) (229).
The mechanism underlying this phenomenon is not fully understood. Kors and coworkers noted that in at least some patients with delayed post-traumatic cerebral edema have a mutation in the gene that encodes the α1A subunit of a neuronal calcium channel that is primarily involved in the release of neurotransmitters (CACNA1A) (230). Mutations of this gene are responsible for familial hemiplegic migraine and episodic ataxia.
In the series of Bruce and colleagues, diffuse cerebral swelling after minor head injury was encountered in 15% of children and adolescents whose Glasgow coma scores on admission were 8 or better (231). Most likely, the swelling results from vasodilatation and hyperemia, which is probably the indirect consequence of a channelopathy. This in turn, disrupts neurotransmitter release and causes the loss of cerebral vascular autoregulation, with a subsequent increase in brain bulk. Treatment is directed at constricting brain vascular volume. This is best accomplished by prolonged (24 to 48 hours) hyperventilation of the intubated youngster coupled with the treatment scheme for cerebral edema described elsewhere in this chapter. Kors and coworkers also proposed the use of acetazolamide (230).
Post-Traumatic Mental Disturbances
Whereas it has been argued that because of its greater plasticity the brain of a child recovers more fully after injury than that of an adult, this is true only in terms of gross neurologic function. It is now becoming apparent that early brain damage limits intellectual capacity and, in so doing, constrains the formation of new cognitive products over the remaining years of brain growth and maturation (231a). Furthermore, numerous studies now indicate that the younger the child subjected to head injury, the more likely it is that cognitive and academic development will be compromised, and the more significant will be the long-term consequences of the head injury (232).
In the experience of Koskiniemi and colleagues, who studied the long-term outcome of severe head injury incurred by preschool children, 30% had a below-normal IQ when tested in adulthood (233). Only 21% of those with normal IQ were able to work full time outside the home. None of the children who experienced the head injury before 4 years of age were able to work independently. A secondary attention-deficit hyperactivity disorder is commonly seen in children who suffered head injuries. In the recent series of Max and collaborators, it was encountered in 38% of children with severe head injuries and in 8% of
P.686

children with mild to moderate head injuries. In some of the latter group the attention-deficit hyperactivity disorder was preexisting (233a).
As a rule, the capacity for new learning is more affected than retention of previously learned information. Perceptual-motor and spatial skills appear to be particularly susceptible to early insult. These deficits are most marked in children who sustained their injury at a young age and are proportional to the duration of impaired consciousness (233). Attention and verbal and written language abilities are also compromised (234). Deficits in verbal and visual recognition memory are particularly evident in younger children and are proportional to the duration of impaired consciousness. Although severity of injury is an important predictor of outcome, age at insult and psychosocial variables must all be acting both independently and interactively to determine prognosis (235). From these data one must conclude that a child is better able to compensate for focal brain injury than an adult, but tolerates less well a diffuse injury that interferes with learning capacities (231a). Skills that are not yet well developed at the time of the head injury are more susceptible to disruption than those that are already well established (234). The effects of brain trauma on an established skill are illustrated in Fig. 9.11A. Improvements in performance with age occur at a growth rate b. At some point of time t2 recovery reaches a plateau, leaving a residual deficit c. When the brain injury affects a developing skill, as illustrated in Fig. 9.11B, the deficit may not be apparent at time t1. This is because the skill has not yet emerged. Sequelae, however, become more marked over time as the child fails to acquire skills at an expected rate. With further time, the disparity between a normal child and a child who has acquired brain trauma may become even more striking as a skill becomes more complex or as new learning is required (232).
Neuroimaging studies, notably positron emission tomography scans and SPECT scans, can corroborate the various cognitive deficiencies by showing areas of hypoperfusion, but these tests do not assist with prognosis (237).
One of the most obvious cognitive deficits is post-traumatic amnesia, an inability to recall events as a result of injury. In most cases, the length of post-traumatic amnesia is proportional to the severity of brain damage (80,236). One of the common features of concussion injuries is a failure to recall events that occurred just before the injury (retrograde amnesia). Here, too, a relationship exists between the length of time before the accident for which memory is impaired and the severity of the brain injury. Post-traumatic amnesia is usually more extensive than retrograde amnesia (238). When a patient has unusually extensive retrograde amnesia, trauma to the limbic system, particularly the hippocampal formation and the mamillary bodies, should be suspected. We and others have encountered the occasional child in whom a mild head injury triggers a profound global and retrograde amnesia. Whether this represents a form of traumatic migraine is unresolved (239). Generally, the rate of recovery from amnesia is faster in children than in adults; however, in the experience of Harris, major post-traumatic psychological difficulties persisted in one-half of the 13% of children who manifested prolonged retrograde amnesia (240).
FIGURE 9.11. A: Hypothetical developmental changes in established skills in children with brain insults (solid line) and in unaffected children (dotted line). B: Hypothetical developmental changes in to-be-acquired skills in children with brain insults (solid line) and in unaffected children (dotted line). (From Taylor HG, Alden J. Age-related differences in outcomes following childhood brain insults: An introduction and overview. J Internat Neuropsychol Soc 1997;3:555–557. Courtesy Dr. H. Gerry Taylor, Department of Pediatrics, Rainbow Babies & Childrens Hospital, Cleveland. OH.)
The effects of a mild head injury, that is, an injury that is sufficiently severe for the child to be seen in the emergency room but not severe enough to require hospitalization, have been studied extensively. It is the current consensus that if the injury occurs at an important developmental age, children may fail to develop certain skills such as reading abilities as quickly as controls (241,242).
In many instances, however, the appearance of major psychiatric disorders after head injury, including lability of mood, outbursts of anger, increased aggressiveness, sleep disturbances, nightmares, and enuresis, is unrelated to the injury itself, but reflects the child’s family and social
P.687

environment. In this respect, it is undoubtedly significant that a large proportion of children showing these post-traumatic behavior disturbances had a history of previous accidents requiring medical treatment, and that approximately 25% were either mentally retarded or had required psychiatric therapy before their head injury. In the experience of Mahoney and colleagues (132), and Hjern and Nylander (247), the overwhelming majority of children with persistent psychiatric symptoms had similar problems before their accident, which only served to aggravate symptoms. Other studies however, have not found any significantly higher than normal rates of preinjury problems in children who suffered mild head injuries (242a). It does appear however, that head injury increases a child’s vulnerability to subsequent environmental stresses (243). It is also significant that when head injuries are incurred during sports, the clinical picture of the postconcussion syndrome, although similar to that seen after potentially litigious injuries, is strikingly brief (244). This observation lends support to the view that the postconcussion syndrome results from environmental, psychological, and physiologic factors. These are reviewed by Nylander and Rydelius (245), and by McClelland and colleagues (246).
Psychiatric symptoms of head injury can frequently be avoided by giving parents, particularly mothers, reassurance or extensive supportive therapy at an early stage, preferably as soon as the child is admitted to the hospital (247). We have most of our patients return to school as soon as feasible but recommend limiting academic demands, increasing them gradually as warranted by the child’s adjustment and school performance. Based on the considerable evidence that the effects of repeated concussion are additive, children who have had two or more episodes of head injury associated with loss of consciousness or amnesia should not be allowed to participate in contact sports (228). This is consistent with the practice parameters issued by the American Academy of Neurology (248).
In summary, the long-term outcome of severe head injuries is poor when incurred in infancy or during the preschool years in terms of both gross neurologic deficits and developmental delays. In infants, the results of the ocular examination, part of the pediatric Glasgow coma score assessed on admission, appear to be most predictive of the outcome. Additionally, when an initial CT scan demonstrates cerebral swelling and a midline shift, a poor outcome can be expected. The outcome for older children is better; it is a function of the condition of the child on admission to the hospital, the severity of increased intracranial pressure, and the duration of coma. Even prolonged coma is consistent with recovery without major neurologic deficits. For instance, in one study, 26% of children who were in coma longer than 24 hours died and 61% had a good outcome from a gross neurologic standpoint (132).
PERSISTENT VEGETATIVE STATE
Persistent vegetative state (PVS) is a relatively rare sequela to major head trauma in the pediatric population. As defined by Jennett and Plum (249), in this state the patient exhibits periods of apparent wakefulness during which the eyes open and move and responsiveness is limited to primitive postural and reflex movements of the limbs.
In the majority of instances, patients blink in response to painful stimuli, exhibit spontaneous eye movements, and show a sleep–wake periodicity. Less often, yawning, chewing, and eye-following movements occur. Meaningless laughter and weeping are not unusual. Decerebrate rigidity is seen in approximately one-half of cases and reflects damage to the midbrain and pons. This is confirmed by MRI studies that show diffuse axonal injury localized in the majority of instances to the corpus callosum and the dorsolateral aspect of the rostral brainstem (250). PVS should be distinguished from the minimally conscious state in which there is minimal behavioral evidence of self or environmental awareness (251).
Higashi and coworkers proposed three levels of PVS (252). Level I is marked by the presence of a sleep–wake cycle and emotional expression, level II is marked by a sleep–wake cycle without emotional expression, and level III is marked by the absence of the sleep–wake cycle. The EEG does not correspond to the severity of damage or to changes in the clinical status; in fact, 25% of patients demonstrate a significant amount of activity.
The prognosis of PVS in the pediatric patient when it follows head trauma is significantly better than when it follows hypoxic brain injury (253). In the experience of Heindl and Laub, 84% of children who had been in post-traumatic PVS for at least 30 days had left PVS 19 months after the injury. Of all children in post-traumatic PVS, 16% were able to become independent. However, after 9 months in PVS less than 5% were able to leave this state (253). Survival in PVS is conditional on the age of the youngster, with infants younger than 1 year of age having a mean survival of 2.6 years, as contrasted with 5.2 years in children aged 2 to 6 years and 7.0 years in children aged 7 to 18 years (254). The same group found that life expectancy doubled between the 1980s and 1993 to 1996, so that in 1993 life expectancy for a 15-year-old in PVS was 10.5 to 12.2 years. For a 1-year-old it was 7.2 years (255). See Chapter 17 for a discussion of the diagnosis of brain death in infants and children.
SPINAL CORD INJURIES
Because of the spinal cord’s protected location, a considerable amount of direct trauma is required to injure it. In children, therefore, injuries to the spinal cord are relatively
P.688

uncommon. Most frequently, they are the result of indirect trauma. This is seen in accidents marked by sudden hyperflexion or hyperextension of the neck or by vertical compression of the spine resulting from falls on the head or buttocks, as can occur from surfing, diving into shallow water, falling from a horse, or various other athletic accidents (256). In physically abused infants, spinal cord injuries can be induced by violent shaking of the head. Obstetric spinal cord injuries are covered in Chapter 6.
Pathology and Pathophysiology
Based on gross pathologic findings, Norenberg and coworkers divided spinal cord injuries into four groups (257). Most common is contusion of the cord with cavity formation. In this type of injury there is no disruption of the surface anatomy, but there are areas of hemorrhage and necrosis that ultimately evolve into cysts. Laceration of the cord induces a clear disruption of the surface anatomy. This injury is usually caused by sharp fragments of bone. In massive compression of the cord the injury is caused by direct compression from without by bone and intervertebral disc or from within by a hematoma. Solid cord injury refers to a cord that grossly appears normal but on histologic examination shows gross disruption of the normal architecture.
The importance of vascular changes in the induction of spinal cord injury has only recently become evident. In particular, post-traumatic ischemia is of major importance in the evolution of spinal cord lesions (258). The anterior spinal artery and its sulcal branches seem particularly vulnerable; the vascular shed in the upper thoracic cord and the ventral radicular artery (artery of Adamkiewicz), which usually feeds the cord at approximately the tenth thoracic vertebra, are other favorite sites for interruption of the blood supply (259). Relatively minor trauma to this area or to the cervical spine can, on occasion, result in an infarction of the spinal cord, with ensuing paraparesis or quadriparesis, respectively (260). Such injuries constituted 8% of spinal cord injuries at the Toronto Hospital for Sick Children (261). Symptoms do not appear immediately, but after a latent period of some 2 hours to 4 days. Paraplegia or tetraplegia is usually profound and recovery is unlikely. Imaging study results are generally normal, although spinal angiography occasionally visualizes an occlusion of the anterior spinal artery (261).
When the patient survives major spinal cord injury for a long time, the damaged area is found to be softened, gray and white matter are poorly delineated, the myelin sheaths are destroyed, and all cellular elements are lost extensively. Replacement by cavities or fibrous gliosis occurs ultimately. In less than 5% of paraplegics, this leads to post-traumatic syringomyelia with worsening of symptoms starting one to several years after the initial injury. There is dispute as to the mechanism of the syringomyelic cavity (262). It can arise from a hematoma at the site of the original injury or from softening of the cord and liquefaction, or it may result from the cord being pulled open by meningeal fibrosis or arachnoiditis. If the injury is symptomatic and of significant diameter, most authorities advise syringo-subarachnoid or syringo-pleural drainage, although others prefer a wide opening of the subarachnoid spaces to allow free CSF movement past the area of cord damage (263).
In many patients in whom an accident produced early paraplegia, the spinal cord does not show any gross pathologic abnormality. Termed spinal concussion, this condition is characterized by transient loss of spinal cord function (264). The mechanism inducing spinal concussion is not clear, but is believed to involve changes in the microvasculature and in neurotransmitters.
Common sites for childhood spinal cord injuries are the second cervical vertebra (27%), followed by the tenth thoracic vertebra (13%), the seventh thoracic vertebra (6%), and the first lumbar segment (6%) (265). Fractures of the twelfth thoracic and first lumbar vertebrae are relatively common and can produce a conus medullaris and cauda equina syndrome.
In the experience of Hamilton and Myles, fracture of the vertebral body or posterior elements without subluxation was the most common pediatric spinal injury. It was seen in 56% of patients (256). Fracture with subluxation was seen in 29% and subluxation without fracture in only 2%. Major spinal cord injury without any radiologic abnormalities was seen in 13% of children. It was the most common form of spinal injury in children younger than 9 years of age, being encountered in 42% of spinal cord injuries.
With the increased use of lap belts in cars, children can incur a horizontal splitting of the spine, also known as a Chance fracture, in the course of a motor vehicle accident. The typical fracture involves L1 with a transverse fracture through the vertebra, with compression of the anterior portion of the body, and vertical distraction posteriorly. Intraabdominal injuries are common, but the spinal canal can be compromised with resultant spinal cord injury and paraplegia (266). The radiographic findings are often subtle and are best visualized on lateral lumbar spine views (267).
Because of the mobility of the neck, the lower cervical region is particularly prone to fracture and dislocation injuries. Direct violence along the axis of the vertebral column can produce fractures of the vertebral bodies, and the spinal cord can be injured by fragments of bone that enter the vertebral canal.
Discussion of the acute and secondary pathophysiologic responses of the spinal cord to injury, in particular, the contributions of calpain activation, inflammation, oxygen radical formation, and lipid peroxidation to cell damage and death, are beyond the scope of this text. This topic is
P.689

reviewed by Carlson and Gorden (268) and Dumont and coworkers (269).
Clinical Manifestations
The clinical picture depends on the severity of the injury and its location. Concussion can result from apparently minor falls on the back and is characterized by a temporary and completely reversible loss of function below the injured segment. With more extensive injuries, recovery is only partial and permanent residua can be expected.
Evaluation of the patient who has sustained injury to the spinal cord has been facilitated by a grading system devised by Frankel and coworkers (270) and more recently modified by the American Spinal Injury Association (271). The Frankel scheme describes four levels: (a) no sensory or motor function; (b) incomplete sensory function, no motor function; (c) incomplete sensory function, no useful motor function; and (d) normal function with some spasticity.
When the cord is seriously compromised, the clinical picture is highlighted by spinal shock. Spinal shock is described in the classic experimental studies of Sherrington (272) and their clinical application by Riddock (273). The condition is marked by the loss of all reflex function distal to the injury, with the segments closest to the injury being the most severely affected. Spinal shock represents a transient decrease of synaptic excitability of neurons distal to the injury. It is caused by loss of supraspinal impulses, which normally produce a background of partial depolarization of the spinal neurons. Clinically, spinal shock can persist for days or weeks and can be prolonged by sepsis, particularly urinary tract infection. The evolution of spinal shock to the return of reflexes and progression to spasticity is reviewed by Ditunno and coworkers (274). This process probably reflects a reorganization of receptors (275,276).
Immediately after the injury, the patient experiences complete loss of motor and sensory function in the segments caudal to the injury (277). There is complete areflexia of variable duration, usually for at least 2 to 6 weeks. A delayed plantar response, a slow flexion, and relaxation of the toes in response to a very strong stroking from the heel toward the toes along the lateral side of the foot can often be elicited (274). Should reflex activity not return, probably the distal spinal cord has been destroyed as the result of vascular insufficiency.
During the first stage of spinal shock, the stage of flaccidity, complete bladder paralysis, and urinary retention occur (273, 278). Gradually, a muscular response of the lower extremities can be elicited in response to stimulation of the skin or the deeper structures. The earliest movements occur in the legs and are flexor. The deep tendon reflexes reappear and soon become hyperactive. Abdominal reflexes also can return. A typical extensor plantar response can be induced and is often accompanied by flexor withdrawal movements of the foot, ankle, and, subsequently, the knee and hips. Contraction of the extensor muscles of the crossed limb frequently accompanies the mass flexion reflex (273,279). During this stage, the bladder empties automatically, although never completely.
In the majority of patients, extensor reflexes involving the quadriceps and other extensor muscles ultimately appear, becoming the dominant reflex activity. Stimuli eliciting the extensor reflex are more complicated than those inducing the flexion response. They include extension of the thigh, as is seen when the patient shifts from a sitting to a supine position, and squeezing of the thigh.
Depending on the severity of the spinal cord injury, the ultimate result can be purely reflex activity of the isolated cord. With less extensive injuries, muscular function or subjective sensation can return over the course of the next few months up to 1 year.
The neurologic picture of the most common spinal cord injuries is summarized in Table 9.12.
An unusual clinical picture that occurs exclusively in children is a transient apparent subluxation of the atlantoaxial joint, which often follows an upper respiratory infection or trauma, especially sports-related trauma (279,280). It is called rotary atlantoaxial luxation. Children present with a head tilt to the affected side, the “cock robin sign,” after its similarity to the appearance of the bird. The neck is tender laterally and posteriorly over C1 and C2. The bulge of the anterior dislocation can be felt by the examiner through the posterior pharyngeal wall, but is best diagnosed by cine-CT scan. This study demonstrates that the subluxed axis and atlas move as a unit during neck rotation (280). The condition is associated only rarely with root or cord signs and usually resolves with traction and immobilization. When the subluxation is persistent or recurrent, as happens in the occasional patient, surgical immobilization is required.
Dislocation of the atlantoaxial joint and an increased atlantoaxial interval are seen with particular frequency in Down syndrome (see Chapter 4). It is also encountered in neurofibromatosis (NF-1) (281), Marfan syndrome, and Arnold-Chiari malformation type I.
Diagnosis
The history of trauma is usually readily elicitable, and the most common diagnostic problem is to establish the site and extent of the injury. In small children, the physician can best perform a sensory examination by demonstrating impairment of autonomic response. Shortly after the injury, the dermatomes below the lesion are dry and often have a defective vasomotor response. Evaluation of reflexes and motor function should not be particularly difficult because reflex withdrawal is not seen during the acute phase of spinal shock.
TABLE 9.12 Clinical Features of Spinal Cord Injuries
Injury Neurologic Features
Transverse Injuries
   T12–1.1 Flaccid paralysis of lower extremities
Loss of sphincter control
Loss of sensation below inguinal ligament
   C5–6 Flaccid quadriparesis
Sparing of diaphragmatic movements
Sensory level at second rib, with preservation of sensation over upper lateral aspect of arm
Bilateral Horner syndrome
Loss of sphincter control
   C1–4 Respiratory paralysis, complete quadriplegia
Rapid death
Conus Medullaris
   Cauda equina syndrome Urinary retention
Disturbance of renal sphincter
Loss of sensation over lumbosacral dermatomes
Flaccid paralysis of lower extremities
Brown-Séquard Syndrome Unilateral muscular paresis
Contralateral disturbances of superficial sensitivity, especially pain and temperature
Incomplete forms far more common than classic syndrome
Central Cord Lesion Disproportionately more motor impairment of upper extremities (caused by involvement of the more medial segments of the lateral corticospinal tracts)
Lower motor neuron lesion of upper extremities; upper motor neuron lesion of lower extremities
Bladder dysfunction (usually urinary retention)
Varying degrees of sensory loss, usually pain and temperature below level of lesion
Relatively good prognosis
Motor power returns first to lower extremities
P.690

The patient with a suspected cervical spinal cord injury must be moved to the radiography unit with utmost care. After plain films of the spine, including lateral films of the cervical spine, are evaluated, it is usually necessary to establish the presence or absence of a subarachnoid block, which can be caused by bone fragments, disc material, hematoma, or swelling of neural tissues (Fig. 9.12). CT or MRI of the involved and adjacent spinal levels provides the most complete information on the status of the spinal column and on the extent to which the spinal cord and canal have been compromised by the injury (282).
Generally, MRI centers are not equipped to image patients with multisystem injuries and needing complex life support systems, so that this procedure is usually deferred until the patient is stabilized. Each imaging study has its advantages. CT provides a better picture of fractures and of trauma to the osseous elements, whereas MRI is better suited to view disc protrusions and the spinal cord itself and any associated bleeding or edema (Fig. 9.13). MRA can be used to screen the status of the vasculature, in particular, the vertebral arteries, although angiography is the definitive diagnostic procedure.
FIGURE 9.12. Compression fracture of thoracic spine (T6, T7, and T8) with anterior displacement of T6 on T7. This patient was a 9-year-old girl who fell from a tree and had an immediate total motor and sensory paralysis below the level of the injury.
Analogous to the changes seen in craniocerebral trauma, three types of abnormalities of the cord can be distinguished. The most common is hyperintensity on T2-weighted images, represented by edema of the cord. Less often, one sees a central hypointensity on T1-weighted images, which evolves to a hypointensity on T2-weighted images surrounded by a ring of hyperintensity. This finding is consistent with an intramedullary hemorrhage and is a poor prognostic sign. In some 16% to 21% of children, all imaging study results are completely normal (256,259,283).
The radiologic diagnosis of a dislocated cervical spine has many pitfalls (284). A marked anterior displacement of C2 on C3 can be seen in 20% of all children younger than 7 years of age. This variant is particularly common during the first 3 years of life. Displacement of C3 on C4 is also common, as is an apparent hypermobility of the atlas on the axis.
FIGURE 9.13. Spinal cord trauma. A sagittal T1-weighted magnetic resonance image (600/15/2) demonstrates vertebral subluxation at the C4 to C5 level (arrow). The cord at and below the bony injury is expanded and hypointense. This 2-year-old boy was wrenched from his mother’s lap and thrown around the vehicle in a motor vehicle accident. He was rendered quadriparetic. (Courtesy of Dr. John G. Curran, Department of Neuroradiology, Childrens Memorial Hospital, Chicago, IL.)
P.691

The severity of the injury often cannot be determined immediately. An early return of reflex activity, particularly of extensor movements, is encouraging. In general, sensory changes give a clearer indication of the level of the lesions than do motor changes. In cervical cord injuries, bilateral meiosis is a bad prognostic sign because it indicates extensive cord damage (285). Prognosis is better when the cord lesion is incomplete. For instance, in the experience of Hamilton and Myles, 74% of children with a physiologically incomplete spinal cord deficit improved by one or two Frankel levels, with 59% experiencing complete recovery (256). When the spinal cord injury was physiologically complete, only 10% improved by one or two Frankel levels. In this study, the absence of radiologic abnormalities did not influence the outcome.
Treatment
Treatment of the child with a spinal cord injury is essentially surgical, but not always operative, and involves a multidisciplinary approach. Intravenous fluids, colloids, and vasopressors are administered to maintain arterial blood pressure. Guttman advanced the conservative, postural treatment of the patient with spinal cord injury and believed that operative procedures for decompression and stabilization should be used only in selected cases (286). Because excessive movement is likely to aggravate spinal cord injury, special care is required in the handling of the patient, and only the absolutely essential diagnostic studies should be done. In injuries of the cervical spine, the head should be maintained in neutral position (287,288). Skeletal traction, usually by means of tongs inserted into the skull, is required for hyperflexion injuries of the cervical spine, whereas mild traction using a canvas sling is used in hyperextension injuries. The management of cervical spine injuries is reviewed by Sypert (289). Injuries of the lumbar spine and thoracolumbar junction are best stabilized in slight hyperextension.
All patients with open wounds of the spine, with injuries in which imaging studies reveal bony fragments within the spinal canal, and with an apparent total block in the presence of an incomplete transection of the cord should undergo surgery, including débridement, removal of bone fragments, laminectomy, and dural repair, if necessary. Any patient whose neurologic deficit increases after initial assessment, either by extending cephalad or by becoming more complete, also should have the benefit of an exploratory laminectomy. Surgical intervention is needed for dislocations of the spine that cannot be reduced adequately by traction and immobilization and for injuries of the spine known by past experience to be unstable; the surgical intervention often need not be immediate. Reduction of dislocations and internal stabilization are then carried out as indicated.
A number of treatments have been proposed to reverse secondary pathophysiologic processes such as ischemia, excitotoxicity, and lipid peroxidation (290). A high dose of methylprednisolone, a synthetic glucocorticoid drug, given within 8 hours of the injury as a 30-mg/kg bolus, followed by 5.4 mg/kg per hour for 23 hours, induces greater improvement in motor and sensory functions than does placebo in patients with complete and incomplete spinal cord deficits (291). At these doses, methylprednisolone may act as an antioxidant or as a free radical scavenger. The monosialoganglioside GM1, which experimentally has been observed to increase neurite outgrowth and to prevent cell death by inhibiting glutamate-induced neuronal
P.692

excitotoxicity, has been found to improve lower limb function after spinal cord injury. The drug is started within 19 to 72 hours after the injury and is administered at 100 mg/day for 18 to 32 days (292). Optimal doses of these drugs, their optimal initiation time, and duration of therapy are not known, and neither treatment has been approved by the U.S. Food and Drug Administration. Many other drugs, notably estrogens, calpain, other free radical inhibitors, and tirilazad mesylate, a 21-amino corticosteroid that acts as a potent antioxidant with no glucocorticoid receptor activity, await preclinical and clinical investigations (293).
There has been considerable interest in the prospects of functional spinal cord regeneration. Although axons of injured spinal neurons cannot regrow within the spinal cord, they can grow long distances in peripheral nerves outside the cord. The reason for this discrepancy is the presence within spinal cord of myelin-associated neurite growth inhibitors (294,295). The various interventional strategies and the problems that prevent their current clinical applications are reviewed by Bregman and coworkers (296).
The long-term care of the paraplegic child is beyond the scope of this book. It is reviewed by Guttman (286), Short and colleagues (297), and Piepmeier (298). The management of the quadriplegic patient is considered by Whiteneck (299). Generally, the patient requires care of the skin overlying the paralyzed part and prevention of decubitus lesions. Also needed is care of the urinary system (300). During the acute phase, urinary retention is treated by intermittent catheterization or insertion of an indwelling catheter, preferably one with a separate irrigating arm (three-way Foley). Additionally, the child needs regular enemas to treat fecal retention; automatic sphincter function can develop. Ileus, when present, can be relieved with neostigmine or with an indwelling rectal tube.
The ultimate outlook for spinal cord function after injury depends on the extent of the injury. The immediate loss of function is caused by both anatomic alteration and impaired physiologic function of the cord. In general, much of the improvement occurs during the first 6 months after the injury and at a much slower pace from 6 months to 2 years after the injury. As a rule, muscles with no initial motor power have a longer period of recovery than those in which there was some initial motor power. The mechanisms underlying the improvement are not clear; certainly, resolution of edema cannot account for the slow recovery. In rare instances, a progressive myelopathy can be caused by a nonunited or malunited dens fracture (301).
Herniated Intervertebral Disc
Although common in the adult, this entity is rare in children, with most cases occurring during adolescence, usually after trauma (302).
Injuries at L4 to L5 and L5 to S1 occur with approximately equal frequency. Many affected children have an underlying malformation of the vertebral column, most often spondylolisthesis or spina bifida. Herniation of intervertebral discs also has been noted in achondroplasia (303,304). Because the presenting symptoms of extradural neoplasms can be similar, MRI is indicated for all suspected disc lesions.
INJURIES TO THE CRANIAL NERVES
Injury to the cranial nerves is a relatively common complication of head injuries. The mechanisms of cranial nerve paralysis are multiple. As listed by Friedman and Merritt (305), they include injury to the nerve by bony fragments, laceration of the nerve when a fracture involves the canal by which the cranial nerve leaves the cranial cavity, tears or stretching, compression by hemorrhage, edema or arachnoiditis, contusion, and injury to the motor cells in the brainstem. Injuries to the olfactory nerves are the most common and are generally bilateral. They are believed to be caused by trauma to the frontal poles or by tearing of the olfactory filaments. Facial and abducens nerve injuries are also encountered, as are palsies of the trochlear nerve. They are generally accompanied by evidence of fracture of middle fossa. Seventh- and eighth-nerve palsies result from a fracture of the petrous ridge. In children, trauma is the most common cause of isolated, acquired, unilateral, or bilateral trochlear nerve palsy (306). In the experience of Friedman and Merritt, the onset of cranial nerve palsy was immediate in 68% of case and was delayed for several days in the remainder (305). In the majority of instances, dysfunction is permanent when the palsy is immediate. The likelihood for recovery is significantly greater when the palsy is delayed (307).
INJURIES TO THE PERIPHERAL NERVES
Peripheral nerve injuries are relatively uncommon in childhood. The most common postnatal injuries are of the brachial plexus and are caused by severe trauma to the shoulder or sudden traction to the arm. Other injuries include division of the ulnar and median nerves at the wrist, the result of pushing the hand through a pane of glass; division of the radial nerve in the upper arm, associated with fracture of the humerus; division of the ulnar nerve with fracture or dislocation of the medial epicondyle; injury of one or both branches of the sciatic nerve as a consequence of injections into the buttocks; and division of the common peroneal nerve in fractures at the neck of the fibula. Peripheral neuropathies are relatively common in Ehlers-Danlos syndrome, with both brachial plexus palsy and lumbosacral plexus palsies developing because of increased ligament laxity (308).
P.693

Pathology
The pathologic changes in an injured peripheral nerve depend on whether the axon remains intact. When the axon is destroyed at the site of injury, wallerian degeneration is induced in the peripheral segment of the nerve. The pathologic, neurophysiologic, and biochemical alterations accompanying wallerian degeneration are reviewed by Stoll and coworkers (309).
Stretch injuries of the peripheral nerves result from damage to the perineurium and the blood vessels (vasa nervorum). Ischemic peripheral nerve injuries also are produced by impairment of the blood supply to the nerves via these vessels. Regeneration of the injured nerve starts from the proximal end of the axon and begins shortly after injury, proceeding in children at approximately 2.5 to 3.0 mm/day. The speed and extent of recovery depend on whether the sheaths of the nerve remain intact; these provide a continuous channel for the young neurites sprouting from the proximal axonal end. When the gap between nerve ends is wide, and particularly when the ends are separated by fibrous tissue, a neuroma can form proximally and spontaneous anastomosis can be delayed or prevented.
TABLE 9.13 Clinical Features of Common Peripheral Nerve Injuries
Nerve Injury Predisposing Factors Clinical Features
Brachial plexus upper root (Erb-Duchenne) Sudden traction to arm Arm internally rotated at shoulder, pronated at forearm
Paralysis of spinati, deltoid, biceps, brachialis, brachio-radialis, extensor carpi radialis
Sensory disturbance minimal or absent
Biceps and supinator jerks lost, triceps preserved
Lower root (Klumpke paralysis) Violent upward pull of shoulder Arm flexed at elbow, forearm supinated, fingers extended, edema and cyanosis of hand
Paralysis of small muscles of hand, finger flexors
Sensory loss of ulnar aspects of fingers, hand, and forearm
Horner syndrome if root avulsed
Long thoracic nerve Carrying heavy weights on shoulder
Postimmunization
Paralysis of serratus magnus with or without trapezius
No sensory symptoms
Circumflex nerve Fracture of humerus (crutch palsy) Paralysis of deltoid
Sensory loss of upper and outer part of arm
Radial nerve Fracture of humerus (Saturday night palsy) Paralysis of triceps uncommon, only if nerve damaged in axilla
Paralysis of brachioradialis, extensors of wrist and fingers
Sensory loss inconstant
Median nerve (310) Cuts at wrist, mucolipidosis III Atrophy of thenar eminence
Paralysis of pronation beyond midposition
Paralysis of flexor of index finger, impaired flexion, and opposition of thumb
Sensory loss of radial aspect of palm
Ulnar nerve Fractures at lower end of humerus
Pressure palsy at elbow
Flattened hypothenar eminence, claw hand
Paralysis of ulnar flexors at wrist and fingers, interossei, adductor of thumb
Sensory loss of ulnar side of arm, hand
Sciatic nerve [common peroneal and posterior tibial nerve (311)]
Femoral nerve (312)
Intramuscular injuries (common peroneal component injured more frequently)
Hemorrhagic disease (hemophilia)
Foot drop, paralysis of peronei, anterior tibialis, extensors of toe
Sensory loss of anterior aspect of lower leg and foot
Absent ankle jerk
Paralysis and atrophy of quadriceps
Defective sensation, anterior and anterior medial aspect of thigh
Common peroneal nerve Fracture at neck of fibula Stepping gait
Loss of dorsiflexion (anterior tibialis) eversion at ankle (peronei), extensors of toe
Incorrectly fitted leg cast Sensory deficit, dorsum of foot and outer side of leg
Achilles tendon reflex lost or reduced
Clinical Manifestations
The salient clinical features of the most frequent peripheral nerve lesions of childhood are presented in Table 9.13. In general, symptoms consist of weakness and sensory disturbances in the area supplied by the individual nerve. Muscular weakness and wasting are characteristic for peripheral nerve injuries. Contractures develop through overaction of unopposed muscle groups.
P.694

When a nerve regenerates, manual pressure on the nerve at the level to which axons have regrown can induce a tingling pain referred distally to an area that is still anesthetic (Tinel’s sign). Pain and paresthesia can be generalized or can be referred to one site along the course of the nerve. These symptoms are aggravated by touch or muscular contractions.
Deep tendon reflexes are diminished or abolished in the affected area. A number of vasomotor symptoms, including mottling and thinning of the skin, edema, cyanosis, and impaired sweating, also are observed.
Diagnosis
Examination of the child with a peripheral nerve injury is directed toward determining the cause and the anatomic site of the injury. Several manuals detail the neurologic examination of such a patient (313). Generally, evaluation of motor function and sensory deficits is more important in the diagnosis of peripheral nerve injuries than is the status of reflexes because the presence or absence of reflexes does not depend on the integrity of a single nerve.
Tinel’s sign, which originally was believed to be evidence of regeneration, does not have much prognostic significance during the first month of injury, unless the most distal point at which it can be elicited moves further down the nerve trunk on progressive examinations.
Electrical studies can delineate the extent of the nerve injury. Somatosensory-evoked potentials can delineate the extent of the injury and distinguish between peripheral nerve and spinal cord injuries (314). Electromyography indicates fibrillation potentials of denervated muscles. Nerve conduction times are either impossible to elicit or reduced. The ability of the nerve below the lesion to respond to direct electrical stimulation despite motor paralysis indicates absence of wallerian degeneration and is a favorable prognostic sign. In the course of regeneration, the conduction times, which initially are impossible to determine, are slow at first and subsequently regain as much as 60% of the original velocity (315). CT myelography and MRI have been used for the diagnosis of cervical root evulsion in brachial plexus injury (316). MR neurography has shown great promise in the diagnosis and management of patients with peripheral nerve pathology (317).
Treatment
Surgical treatment may not be required during the early phase of peripheral nerve injuries. Clean lacerations of a nerve are suitable for early repair. In other types of nerve injury, nothing is gained by immediate exploration of the site. If the nerve presents readily in the wound, a single stainless-steel marking suture can facilitate subsequent repair. Secondary disability, such as contractures or injury of the paralyzed muscles by excess stretching, should be prevented by splinting the affected limb. Electrostimulation of the paralyzed muscles is of no advantage. A severed nerve should be explored and reapproximated 3 weeks after injury, provided the wound is healed. At this time, the extent of nerve injury can be defined more clearly and suturing is technically more satisfactory. A nerve that is known to have been traumatized but not severed should be explored if recovery of motor and sensory functions does not take place or is less complete than anticipated. External and internal neurolysis of such a nerve trunk can allow further recovery of function. A neuroma in continuity can require resection and reanastomosis if nerve function is absent or poor.
Prognosis
The prognosis depends on the extent and nature of the nerve injury. Pressure palsies almost invariably recover. If electrical studies indicate wallerian degeneration of the distal segment, recovery is delayed until the regenerating fibers reach the muscles they innervate. If no recovery can be documented after 3 months by either clinical or electrical examinations, the ultimate prognosis is poor, and few patients benefit from surgical exploration. Different nerves have different capacities for regeneration. Generally, radial nerve injuries fare best and sciatic nerve injuries worst. Spontaneous recovery from sciatic nerve injuries can be extremely slow, but can continue for 1 to 2 years.
COMPLEX REGIONAL PAIN SYNDROME (REFLEX SYMPATHETIC DYSTROPHY)
Complex regional pain syndrome (CRPS) type I is the new designation for what was formerly termed reflex sympathetic dystrophy. It is an an occasional sequela to trauma affecting the limbs without obvious peripheral nerve injury. Whereas in CRPS I, the more commonly encountered syndrome, the pain is not limited to the distribution of a single peripheral nerve, CRPS II, formerly termed causalgia, is a condition that develops following nerve injury. Causalgia was first defined in adults by Mitchell (318) and has been encountered in children, predominantly in girls (319,320). The condition is characterized by constant burning pain and hyperesthesia in an extremity. Most often, the lower extremities, notably the ankle and foot, are affected. Pain is accompanied by swelling, sweating, vasomotor instability, and sometimes trophic changes. There may not be a history of antecedent trauma, or the injury may have been considered minor. Psychological disturbances are common and can become the most important part of the clinical picture. An important early sign of reflex sympathetic dystrophy is piloerection over the hairy aspect of the affected limb (321). Often, associated muscle spasms, myoclonus, or focal dystonia occurs (322,323). When dystonia is part of the condition, it can appear at the same time as the causalgia or up to many months later.
P.695

Many theories exist as to etiology, none of them well supported by clinical evidence. A disorder of the sympathetic system, release of a pain substance, or supersensitivity to neurotransmitters are the most likely (320); however, in many instances surgical or chemical sympathectomy does not relieve the pain, an indication that pain is independent of sympathetic function (324). Pathologic examination of nerve and muscle taken from the affected limb suggest a microangiopathy (325). Symptoms of causalgia can last a few days to as long as a year. A number of therapies have been suggested; none is infallible. The response to intensive physical therapy is often excellent in children (321). Analgesics, nonsteroidal inflammatory drugs, antidepressants, narcotics, and corticosteroids all have their advocates. Blockage of the sympathetic chain with paravertebral or epidural anesthetics has been tried, as has surgical sympathectomy (326). No treatment for the movement disorder has been effective; an occasional patient recovers spontaneously after a few years (322).
INJURIES BY PHYSICAL AGENTS
Injuries of the Nervous System by X-Ray Irradiation
Radiation damage to the central nervous system can occur before birth and can result in a variety of gross and microscopic malformations of the brain. These are discussed in Chapter 5.
The neurologic complications of therapeutic irradiation for intracranial tumors are discussed in Chapter 11.
Heatstroke (Sunstroke)
Heatstroke results from prolonged exposure to direct sunlight and heat, with subsequent failure of the body’s heat-regulating mechanism. Cells respond to excessive heat by producing heat-shock proteins. These function as molecular chaperones and bind to partially folded or misfolded proteins and prevent their irreversible denaturation (327). Presence of hyperpyrexia distinguishes heat stroke from symptoms resulting from sodium loss caused by excessive sweating.
The clinical and pathologic picture of heatstroke results from a combination of hyperpyrexia and shock. Sunlight contributes to the heat load, but probably does not affect the brain directly. In fatal cases, examination of the brain reveals generalized edema and degeneration of cerebellar neurons, particularly the Purkinje cells and the cells of the dentate nucleus, with lesser degrees of neuronal loss in other areas of the neuraxis (328,329).
Clinically, brain dysfunction can range from subtle to severe. The patient has sudden onset of coma, cyanosis, impaired sweating, and hyperpyrexia (330). Generalized or focal seizures may occur, especially during cooling (327). Rhabdomyolysis, hyperkalemia, and metabolic acidosis can complicate the picture. Clinical heatstroke can be associated with an underlying abnormality of skeletal muscle similar to that of malignant hyperthermia (330). Cerebellar symptoms can be evident with recovery. They usually improve considerably in time.
Treatment of the acute condition involves removing the child from the sun, reducing body temperature, and providing intravenous fluids. Seizure control is best achieved with benzodiazepines (327).
Electrical Injuries of the Nervous System
Electrical injuries to children can be caused by household current or lightning. Lightning is responsible for approximately one-fourth of all fatal electrical accidents. Cherington described four clinical pictures. Most often, the patient either dies at once as a result of cardiac arrest or recovers fully. Neurologic symptoms may be immediate and permanent, or they may be delayed or progressive (332). The progressive neurologic signs include hemiparesis, cerebellar deficits, paraplegia, cranial nerve palsies, and other focal neurologic signs. Increased intracranial pressure also has been observed (333,334). Long-term behavioral changes, including disturbances in affect, memory, and mood, are not unusual and can last for months, out of proportion to the duration of any cardiorespiratory arrest (332,334a). Kotagal and colleagues suggested they are the consequence of a direct injury to the limbic system (335).
Neurologic Complications of Burns
Approximately 5% of children develop encephalitic symptoms within the first few weeks after they sustain severe burns. As a rule, these represent complications from the burn. These symptoms include changes in level of consciousness, seizures, aphasia, extrapyramidal disorders, and impaired intellectual function, which result from infections, metabolic encephalopathies, and cerebrovascular accidents. Neurologic symptoms during the first 3 days after the burn are generally caused by shock and derangements of electrolytes. Infections are responsible for neurologic symptoms after the first week after the burn in what is essentially an immunocompromised host. They are most common during the second to third weeks after the injury. Three organisms are usually responsible: Candida, Pseudomonas aeruginosa, and Staphylococcus aureus. Almost invariably they enter the nervous system from a systemic source. Microabscesses of the brain are most likely to be caused by candidiasis, with S. aureus a less common culprit. Meningitis or septic infarcts are caused by P. aeruginosa. Intracranial hemorrhages are less common than infarcts and generally result from disseminated intravascular coagulation or serum hyperosmolarity. Central pontine
P.696

myelinolysis and cerebral edema also can develop, the latter generally caused by anoxia (336,337).
Despite the number of potentially irreversible neurologic complications, the majority of children experience full neurologic recovery (338).
Neurologic Injury from Undersea Diving
Underwater diving places the teenager or young adult at risk for two major types of neurologic injuries. The most common is decompression sickness that results from a too rapid decompression on ascent. It mainly affects the spinal cord and results in a unique multilevel spinal cord disease, with deficits consisting of pain, sensory loss, and motor weakness (339). The condition usually resolves with administration of oxygen and recompression, although there may be a residual patchy sensory loss. Arterial gas embolism, a much rarer condition, is the most serious complication of self-contained underwater breathing apparatus (scuba) diving. It results from pulmonary overpressure on ascent with extravasation of air into the arterial system and occlusion of a major cerebral artery. The first sign of the condition is usually a seizure. This can be quickly followed by an acute, strokelike picture marked by hemiparesis, aphasia, and cortical blindness. It is the major cause of death in diving accidents. MRI studies of the head show ischemic changes in a vascular distribution (340). By contrast, imaging study of brain and spinal cord are usually normal in decompression sickness, although on rare occasions studies demonstrate a swollen spinal cord with increased signal posteriorly on T2-weighted images (341).
REFERENCES
1. Adelson PD, Kochanek PM. Head injuries in children. J Child Neurol 1998;13:2–15.
2. Craft AW, Shaw DA, Cartlidge NE. Head injuries in children. Br Med J 1972;4:200–203.
3. North AF. When should a child be in the hospital? Pediatrics 1976;57:540–543.
4. Kraus JF, Rock A, Hemyari P. Brain injuries among infants, children, adolescents and young adults. Am J Dis Child 1990;144:684–691.
5. Frankowski RF, Annegers JF, Whitman S. Epidemiological and descriptive studies. Part 1. The descriptive epidemiology of head trauma in the U.S. In: Becker DP, Povlishok JT, eds. Central nervous system trauma, status report 1985. Bethesda, MD: NIH, NINCDS., 1985:33–43.
6. Klauber MR, Marshall LF, Barrett-Connor E, et al. The epidemiology of head injury: a prospective study of an entire community—San Diego County, CA, 1978. Am J Epidemiol 1981;113:500–509.
7. Leventhal JM, Horwitz SM, Rude C, et al. Maltreatment of children born to teenage mothers: a comparison between the 1960s and 1980s. J Pediatr 1993;122:314–319.
8. McClain PW, Sacks JJ, Froehlke RG, et al. Estimates of fatal child abuse and neglect, United States, 1979 through 1988. Pediatrics 1993;91:338–343.
9. Billmire ME, Myers PA. Serious head injury in infants: accident or abuse. Pediatrics 1985;75:340–342.
10. Rovi S, Chen PH, Johnson MS. The economic burden of hospitalizations associated with child abuse and neglect. Am J Public Health 2004;94:586–590.
11. O’Rourke NA, Costello F, Yelland JD, et al. Head injuries to children riding on bicycles. Med J Aust 1987;146:619–621.
12. Attewell RG, Glase K, McFadden M. Bicycle helmet efficacy: a meta-analysis. Accid Anal Prev 2001;33:345–352.
13. Kopjar B, Wickizer TM. Age gradient in the cost-effectiveness of bicycle helmets. Prev Med 2000;30:401–406.
14. Mealey J. Pediatric head injuries. Springfield, IL: Charles C. Thomas, 1968.
15. Graham DI. Neuropathology of head injury. In Narayan RK, Wilberger JE, Povlishock JT, eds. Neurotrauma. McGraw-Hill, New York, 1996:43–59.
16. Pearl GS. Traumatic neuropathology. Clin Lab Med 1998;18:39–64.
17. Stålhammar D. Experimental models of head injury. Acta Neurochirurg 1986;36(Suppl):33–46.
18. Ommaya AK, Goldsmith W, Thibault L. Biomechanics and neuropathology of adult and paediatric head injury. Br J Neurosurg 2002;16:220–242.
19. Berney J, Froidevaux AC, Favier J. Paediatric head trauma: influence of age and sex. II. Biomechanical and anatomo-clinical correlations. Childs Nerv Syst 1994;10:517–523.
20. Gentry LR, Godersky JC, Thompson B. MR imaging of head trauma: review of the distribution and radiopathologic features of traumatic lesions. Am J Roentgenol 1988;15:663–672.
21. Adams JH, Graham DI, Gennarelli TA, et al. Diffuse axonal injury in nonmissile head injuries. J Neurol Neurosurg Psychiatr 1991;54:481–483.
22. Reichard RR, White CL, Hladik CL, et al. Beta-amyloid precursor protein staining of nonaccidental central nervous system injury in pediatric autopsies. J Neurotrauma 2003;20:347–355.
23. Adams JH, Mitchell DE, Graham DI, et al. Diffuse brain damage of immediate impact type: Its relationship to “primary brain-stem damage” in head injury. Brain 1977;100:489–502.
24. Adams JH, Graham DI, Murray LS, et al. Diffuse axonal injury due to nonmissile head injury in humans. An analysis of 45 cases. Ann Neurol 1982;12:557–563.
25. Yamaki T, Murakami N, Iwamotor Y, et al. Pathological study of diffuse axonal injury patients who died shortly after impact. Acta Neurochir 1992;119:153–158.
26. Chester CS, Reznick BR. Ataxia after severe head injury: the pathological substrate. Ann Neurol 1987;22:77–79.
27. Strich SJ. Diffuse degeneration of the cerebral white matter in severe dementia following head injury. J Neurol Neurosurg Psychiat 1956;19:163–185.
28. Graham DI, McIntosh TK, Maxwell WL, et al. Recent advances in neurotrauma. J Neuropathol Exp Neurol 2000;59:641–651.
29. Singleton RH, Povlishock JT. Identification and characterization of heterogenous neuronal injury and death in regions of diffuse brain injury: evidence for multiple independent injury phenotypes. J Neurosci 2004;24:3543–3553.
30. Wolf JA, Stys PK, Lusardi T, et al. Traumatic axonal injury induces calcium influx modulated by tetrodotoxin-sensitive sodium channels. J Neurosci 2001;21:1923–1930.
31. Maxwell WL, Povlishock JT, Graham DL. A mechanistic analysis of nondisruptive axonal injury: a review. J Neurotrauma 1997;14:419–440.
32. Povlishock JT, Becker DP. Fate of reactive axonal swellings induced by head injury. Lab Invest 1985;52:540–552.
33. Huisman TA, Schwamm, LH, Schaefer PW, et al. Diffusion tensor imaging as potential biomarker of white matter injury in diffuse axonal injury. Am J Neuroradiol 2004;25:370–376.
34. Tong KA, Ashwal S, Holshouser BA., et al. Hemorrhagic shearing lesions in children and adolescents with posttraumatic diffuse axonal injury: improved detection and initial results. Radiology 2003;227:332–339.
35. Cecil KM, Hills EC, Sandel ME, et al. Proton magnetic resonance spectroscopy for detection of axonal injury in the splenium of the corpus callosum of brain-injured patients. J Neurosurg 1998;88:795–801.
36. Lindenberg R. Significance of the tentorium in head injuries from blunt forces. Clin Neurosurg 1964;12:129–142.
P.697

37. Freytag E. Autopsy findings in head injuries from blunt forces. Arch Pathol 1963;75:402–413.
38. Calder IM, Hill I, Scholtz CL. Primary brain trauma in nonaccidental injury. J Clin Pathol 1984;37:1095–1100.
39. American Academy of Neurology: Practice parameter: the management of concussion in sports (summary statement). Report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 1997;48:581–585.
40. Bullock R, Zauner A, Woodward JJ, et al. Factors affecting excitatory amino acid release following severe human head injury. J Neurosurg 1998;89:507–518.
41. Shaw NA. The neurophysiology of concussion. Prog Neurobiol 2002;67:281–344.
42. Katayama Y, Becker DP, Tamura T, Hovda DA. Massive increases in extracellular potassium and the indiscriminate release of glutamate following concussive brain injury. J Neurosurg 1990;73:889–900.
43. Junger EC, Newell DW, Grant GA, et al. Cerebral autoregulation following minor head injury. J Neurosurg 1997;86:425–432.
44. McCrea M, Kelly JP, Randolph C, et al. Immediate neurocognitive effects of concussion. Neurosurgery 2002;50:1032–1042.
45. Bayir H, Kochanek PM, Clark RS. Traumatic brain injury in infants and children: mechanisms of secondary damage and treatment in the intenisve care unit. Crit Care Clin 1003;19:529–549.
46. Fishman RA. Cerebrospinal fluid in diseases of the nervous system, 2nd ed. Philadelphia: WB Saunders, 1992:116–138.
47. Bullock R, Maxwell WL, Graham DI, et al. Glial swelling following human cerebral contusion: an ultrastructural study. J Neurol Neurosurg Psychiatry 1991;54:427–434.
48. Barzo P, Marmarou A, Fatouros P, et al. Contribution of vasogenic and cellular edema to traumatic brain swelling measured by diffusion-weighted imaging. J Neurosurg 1997;87:900–907.
49. Ruppel RA, Kochanek PM, Adelson PD, et al. Excitatory amino acid concentrations in ventricular cerebrospinal fluid after severe traumatic brain injury in infants and children. The role of child abuse. J Pediatr 2001;138:18–25.
50. Becker DP, Verity MA, Povlishock J, et al. Brain cellular injury and recovery—horizons for improving medical therapies in stroke and trauma. West J Med 1988;148:670–684.
51. Kawamata T, Katayama Y, Hovda DA, et al. Administration of excitatory aminoacid antagonists via microdialysis attenuates the increase in glucose utilization seen following concussive brain injury. J Cereb Blood Flow Metab 1992;12:12–24.
52. Hovda DA, Becker DP, Katayama Y. Secondary injury and acidosis. J Neurotrauma 1992;9(Suppl 1):S47–S60.
53. Bakay L, Lee JC, Lee GR, et al. Experimental cerebral concussion. Part I: An electron microscopic study. J Neurosurg 1977;47:525–531.
54. Kempsky O. Cerebral edema. Semin Nephrol 2001;21:303–307.
55. Bell MJ, Kochanek PM, Doughty LA, et al. Comparison of the interleukin-6 and interleukin-10 response in children after severe traumatic brain injury or septic shock. Acta Neurochir Suppl (Vienna) 1997;70:96–97.
56. Papadopoulos MC, Krishna S, Verkman AS. Aquaporin water channels and brain edema. Mt Sinai J Med 2002;69:242–248.
56a. Griesdale DE, Honey CR. Aquaporins and brain edema. Surg Neurol 2004;61:418–421.
57. Adelson PD, Clyde B, Kochanek PM, et al. Cerebrovascular response in infants and young children following severe traumatic brain injury: a preliminary report. Pediatr Neurosurg 1997;26:200–207.
58. Bruce DA, Alavi A, Bilaniuk L, et al. Diffuse cerebral swelling following head injuries in children. The syndrome of “malignant brain edema.” J Neurosurg 1981;54:170–178.
59. Langfitt TW, Gennarelli TA, Obrist WD, et al. Prospects for the future in the diagnosis and management of head injury: pathophysiology, brain imaging, and population based studies. Clin Neurosurg 1982;29:353–376.
60. Cruz J, Gennarelli TA, Alves WM. Continuous monitoring of cerebral hemodynamic reserve in acute brain injury: relationship to changes in brain swelling. J Trauma 1992;32:629–635.
61. Choux M. Incidence, diagnosis and management of skull fractures. In: Raimondi AS, Choux M, DiRocco C, eds. Head injuries in the newborn and infant. New York: Springer-Verlag, 1986:163–182.
62. Lloyd DA, Carty H, Patterson M, et al. Predictive value of skull radiography for intracranial injury in children with blunt head injury. Lancet 1997;349:821–824.
63. Mizrahi EM, Kellaway P. Cerebral concussion in children: assessment of injury by electroencephalography. Pediatrics 1984;73:419–425.
64. Jennett B. Epilepsy after nonmissile head injuries, 2nd ed. Chicago: Year Book, 1975.
65. Ashwal S, Holshouser BA, Shu SK, et al. Predictive value of proton magnetic resonance spectroscopy in pediatric head injury. Pediatr Neurol 2000;23:114–125.
66. Hendrick EB, Harwood-Nash DC, Hudson AR. Head injuries in children: a survey of 4465 consecutive cases at the Hospital for Sick Children, Toronto, Canada. Clin Neurosurg 1963;11:46–65.
67. Raimondi AJ, Hirschauer J. Head injury in the infant and toddler—coma scoring and outcome scale. Childs Brain 1984;11:12–35.
68. Duhaime AC, Alario AJ, Lewander WJ, et al. Head injury in very young children: mechanisms, injury types, and ophthalmologic findings in 100 hospitalized patients younger than 2 years of age. Pediatrics 1992;90:179–185.
69. Minns RA, Busuttil A. Patterns of presentation of the shaken baby syndrome. BMJ 2004;328:766.
70. Geddes JF, Hackshaw AK, Vowles GH, et al. Neuropathology of inflicted head injury in children. I. Patterns of brain damage. Brain 2001;124:1290–1298.
71. Shannon P, Becker L. Mechanisms of brain injury in infantile child abuse. Lancet 2001;358:686–687.
71a. Bechtel K, Stoessel K, Leventhal JM, et al. Characteristics that distinguish accidental from abusive injury in hospitalized young children with head trauma. Pediatrics 2004;114:165–168.
72. Geddes JF, Vowles GH, Hackshaw AK, et al. Neuropathology of inflicted head injury in children. II. Microscopic brain injury in infants. Brain 2001;124:1299–1306.
73. Donohoe M. Evidence-based medicine and shaken baby syndrome. Part I. Literature review, 1966-1998. Am J Forensic Med Pathol 2003;24:239–242.
74. Geddes JF, Plunkett J. The evidence base for shaken baby syndrome. BMJ 2004;328:719–720.
75. Harding B, Risdon RA, Krous HF. Shaken baby syndrome. BMJ 2004;328:720–721.
76. Menkes JH. Subdural haematoma, non-accidental head injury. Eur J Paediatr Neurol 2001;5:175–176.
76a. Jaffe D, Wesson D. Current concepts: emergency management of blunt trauma in children. N Engl J Med 1991;324:1477–1482.
77. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A practical scale. Lancet 1974;2:81–84.
78. Simpson D, Reilly P. Pediatric coma scale. Lancet 1982;2:450.
79. Overgaard J, Hvid-Hansen O, Land AM, et al. Prognosis after head injury based on early clinical examination. Lancet 1973;2:631–635.
80. Russell WR. The traumatic amnesias. London: Oxford University Press, 1971.
81. Chesnut RM. Management of brain and spine injuries. Crit CareClin 2004;20:25–55.
82. Plum FB, Posner JB. Diagnosis of stupor and coma, 3rd ed. Philadelphia: Davis, 1980.
83. Masters SJ, McClean PM, Arcarese JS, et al. Skull x-ray examinations after head trauma. Recommendations by a multidisciplinary panel and validation study. N Engl J Med 1987;316:84–91.
84. Johnson DL, Duma C, Sivit C. The role of immediate operative intervention in severely head-injured children with a Glasgow Coma Scale score of 3. Neurosurgery 1992;30:320–324.
85. Zimmerman RA, Bilaniuk LT, Bruce D, et al. Computed tomography of pediatric head trauma: acute general cerebral swelling. Radiology 1978;126:403–408.
86. Johnson MH, Lee SH. Computed tomography of acute cerebral trauma. Radiol Clin North Am 1992;30:325–352.
87. Ginsburg CM. Frontal sinus fractures. Pediatr Rev 1997;18:120–121.
88. Sklar EML, Quencer RM, Bowen BC, et al. Magnetic resonance applications in cerebral injury. Radiol Clin North Am 1992;30:353–356.
P.698

89. Kelly AB, Zimmerman RD, Snow RB, et al. Head trauma: comparison of MR and CT—experience in 100 patients. Am J Neuroradiol 1988;9:699–708.
90. O’Sullivan MG, Statham PF, Jones PA, et al. Role of intracranial pressure monitoring in severely head-injured patients without signs of intracranial hypertension on initial computerized tomography. J Neurosurg 1994;80:46–50.
91. Sivakumar V, Rajshekhar V, Chandy MJ. Management of neurosurgical patients with hyponatremia and natriuresis. Neurosurgery 1994;34:269–274.
92. Vingerhoets F, de Tribolet N. Hyponatremia hypo-osmolarity in neurosurgical patients. “Appropriate secretion of ADH” and “cerebral salt-wasting syndrome.” Acta Neurochir 1988;91:50–54.
93. Berkenbosch JW, Lentz CW, Jimenez DF, et al. Cerebral salt wasting syndrome following brain injury in three pediatric patients: suggestions for rapid diagnosis and therapy. Pediatr Neurosurg 2002;36:75–79.
94. Pascucci RC. Head trauma in the child. Intensive Care Med 1988;14:185–195.
95. Khanna S, Davis D, Peterson B, et al. Use of hypertonic saline in the treatment of severe refractory postraumatic intracranial hypertension in pediatric traumatic brain injury. Crit Care Med 2000;28:1144–1151.
96. Adelson PD, Bratton SL, Carney NA, et al. Guidelines for the acute medical management of severe traumatic brain injury in infants, children, and adolescents. Chapter 11. Use of hyperosmolar therapy in the management of severe pediatric traumatic brain injury. Pediatr Crit Care Med 2003;4(3 Suppl):S40–S44.
97. Feldman Z, Zachari S, Reichenthal E, et al. Brain edema and neurological status with rapid infusion of lactated Ringer’s or 5% dextrose solution following head trauma. J Neurosurg 1995;83:1060–1066.
98. Schaefer M, Link J, Hannemann L, et al. Excessive hypokalemia and hyperkalemia following head injury. Intensive Care Med 1995;21:235–237.
99. MacDonald JS, Atkinson CC, Mooney DP. Hypokalemia in acutely ill children: a benign laboratory abnormality. J Trauma 2003;54:197–198.
100. Brohi K, Singh J, Heron M, et al. Acute traumatic coagulopathy. J Trauma 2003;54:1127–1130.
101. Morenski JD, Tobias JD, Jimenez DF. Recombinant activated factor VII for cerebral injury–induced coagulopathy in pediatric patients. Report of three cases and review of the literature. J Neurosurg 2003;98:611–616.
102. Miner ME, Kaufman HH, Graham SH, et al. Disseminated intravascular coagulation fibrinolytic syndrome following head injury in children: frequency and prognostic implications. J Pediatr 1982;100:687–691.
103. Chadwick D. Seizures and epilepsy after traumatic brain injury. Lancet 2000;355:334–335.
104. Chang BS, Lowenstein DH. Practice parameter: antiepileptic drug prophylaxis in severe traumatic brain injury. Neurology 2003;60:10–16.
105. Zimmerman RA, Bilaniuk LT. Computed tomography in pediatric head trauma. J Neuroradiol 1981;8:257–271.
106. Kanter MJ, Narayan RK. Intracranial pressure monitoring. Neurosurg Clin North Am 1991;2:257–265.
107. Jensen RL, Hahn YS, Ciro E. Risk factors of intracranial pressure monitoring in children with fiberoptic devices: a critical review. Surg Neurol 1997;47:16–22.
108. Bouma GJ, Muizelaar JP, Choi SC, et al. Cerebral circulation and metabolism after severe traumatic brain injury: the elusive role of ischemia. J Neurosurg 1991;75:685–693.
109. Levy DI, Recate HL, Cherny WB, et al. Controlled lumbar drainage in pediatric head injury. J Neurosurg 1995;83:453–460.
110. Miller JD. Head injury. J Neurol Neurosurg Psychiatr 1993;56:440–447.
111. 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.
112. Marion DW, Firlik A, McLaughlin MR. Hyperventilation therapy for severe traumatic brain injury. New Horizons 1995;3:439–447.
113. Cruz J, Raps EC, Hoffstad OJ, et al. Cerebral oxygen monitoring. Crit Care Med 1993;21:1242–1246.
114. Feldman Z, Kanter MJ, Robertson CS, et al. Effect of head elevation on intracranial pressure, cerebral perfusion pressure, and cerebral blood flow in head-injured patients. J Neurosurg 1992;76:207–211.
114a. Roberts I, Yates D, Sandercock P, et al. Effect of intravenous corticosteroids on death within 14 days in 10008 adults with clinically significant head injury (MRC CRASH trial): randomised placebo-controlled trial. Lancet 2004;364:1321–1328.
114b. Alderson P, Roberts I. Corticosteroids for acute traumatic brain injury. Cochrane Database Syst Rev 2005;25:CD 000196.
115. Young B, Runge JW, Waxman KS, et al. Effects of pegorgotein on neurologic outcome of patients with severe head injury. A multicenter, randomized controlled trial. J Am Med Assoc 1996;276:538–543.
116. Ward JD, Becker DP, Miller, JD, et al. Failure of prophylactic barbiturate coma in the treatment of severe head injury. J Neurosurg 1985;62:383–388.
117. Clifton GL, Miller ER, Choi SC, et al. Hypothermia on admission in patients with severe brain injury. J Neurotrauma 2002;19:293–301.
118. Rubin DM, McMillan CO, Helfaer MA, et al. Pulmonary edema associated with child abuse: case reports and review of the literature. Pediatrics 2001;108:769–775.
119. Jennett B, Teasdale G. Management of head injuries. Philadelphia: Davis, 1981.
120. Committee on Quality Improvement, American Academy of Pediatrics. The management of minor closed head injury in children. Pediatrics 1999;104:1407–1415.
121. Shutzman SA, Barnes P, Duhaime AC, et al. Evaluation and management of children younger than two years old with apparently minor head trauma: Proposed guidelines. Pediatrics 2001;107:983–993.
122. Stein SC, O’Malley KF, Ross SE. Is routine computed tomography scanning too expensive for mild head injury? Ann Emerg Med 1991;20:1286–1289.
123. Davis RL, Hughes M, Gubler KD, et al. The use of cranial CT scans in the triage of pediatric patients with mild head injury. Pediatrics 1995;95:345–349.
124. Ingebrigtsen T, Romner B. Routine early CT-scan is cost saving after minor head injury. Acta Neurol Scand 1996;93:207–210.
125. Rosenthal BW, Bergman I. Intracranial injury after moderate head trauma in children. J Pediatr 1989;115:346–350.
126. Ylvisaker M, ed. Traumatic brain injury rehabilitation: children and adolescents, 2nd ed. Boston: Butterworth–Heinemann, 1997.
127. Levi L, Guilburd JN, Bar-Josef G, et al. Severe head injury in children—analyzing the better outcome over a decade and the role of major improvements in intensive care. Childs Nerv Syst 1998;14:195–202.
128. Bijur PE, Haslum M, Golding J. Cognitive and behavioral sequelae of mild head injury in children. Pediatrics 1990;86:337–344.
129. Hall DMB, Johnson SL, Middleton J. Rehabilitation of head injured children. Arch Dis Child 1990;65:553–556.
130. Levin HS, Aldrich EF, Saydjari C., et al. Severe head injury in children: experience of the Traumatic Coma Data Bank. Neurosurgery 1992;31:435–443.
131. Keenan HT, Runyan DK, Marshall SW, et al. A population-based study of inflicted traumatic brain injury in young children. JAMA 2003;290:621–626.
132. Mahoney WJ, D’Souza BJ, Haller JA, et al. Long-term outcome of children with severe head trauma and prolonged coma. Pediatrics 1983;71:756–762.
133. Hjern B, Nylander I. Late prognosis of severe head injuries in childhood. Arch Dis Child 1962;37:113–116.
134. Costeff H, Groswasser Z, Goldstein R. Long-term follow-up review of 31 children with severe closed head trauma. J Neurosurg 1990;73:684–687.
135. Guilleminault C, Faull KF, Miles, L, et al. Posttraumatic excessive daytime sleepiness: a review of 20 patients. Neurology 1983;33:1584–1589.
136. Johnson SL, Hall DM. Posttraumatic tremor in head-injured children. Arch Dis Child 1992;67:227–228.
P.699

137. Kruass JK, Trankle R, Kopp KH. Posttraumatic movement disorders after moderate or mild head injury. Mov Disord 1997;12:428–431.
138. Zijlmans J, Booij J, Valk J, et al. Posttraumatic tremor without parkinsonism in a patient with complete contralateral loss of the nigrostriatal pathway. Mov Disord 2002;17:1086–1088.
139. Kingsley D, Till K, Hoare RD. Growing fractures of the skull. J Neurol Neurosurg Psychiatr 1978;41:312–318.
140. Muhonen MG, Piper JG, Menezes AH. Pathogenesis and treatment of growing skull fractures. Surg Neurol 1995;43:367–373.
141. Battle WH. Hunterian lectures. London: Royal College of Surgeons of England, 1890.
142. Hanigan WC, Peerson RA, Njus G. Tin ear syndrome: rotational acceleration in pediatric head injuries. Pediatrics 1987;80:618–622.
143. Wilson NW, Copeland B, Bastian JF. Posttraumatic meningitis in adolescents and children. Pediatr Neurosurg 1990–1991;16:17–20.
144. Ommaya AK. Spinal fluid fistulae. Clin Neurosurg 1976;23:363–392.
145. Hull HF, Morrow G. Glucorrhea revisited: prolonged promulgation of another plastic pearl. JAMA 1975;234:1052–1053.
146. El Gammal T, Sobol W, Wadlington VR, et al. Cerebrospinal fluid fistula: detection with MR cisternography. Am J Neuroradiol 1998;19:627–631.
147. Shetty PG, Shroff MM, Sahani DV, et al. Evaluation of high-resolution CT and MR cisternography in the diagnosis of cerebrospinal fluid fistula. Am J Neuroradiol 1998;19:633–639.
148. Eberhardt KE, Hollenbach HP, Deimling, M, et al. MR cisternography: a new method for the diagnosis of CSF fistulae. Eur Radiol 1997;7:1485–1491.
149. Baltas I, Tsoulfa S, Sakellariiou P, et al. Posttraumatic meningitis: bacteriology, hydrocephalus, and outcome. Neurosurgery 1994;35:422–426.
150. Rathore MH. Do prophylactic antibiotics prevent meningitis after basilar skull fracture? Pediatr Infect Dis J 1991;10:87–88.
151. Davis RE, Telischi FF. Traumatic facial nerve injuries: review of diagnosis and treatment. J Craniomaxillofac Trauma 1995;1:30–41.
152. Eviatar L, Bergtraum M, Randel RM. Posttraumatic vertigo in children: a diagnostic approach. Pediatr Neurol 1986;2:61–66.
153. Shea JJ, Ge X, Orchik DJ. Traumatic endolympathic hydrops. Am J Otol 1995;16:235–240.
154. Erdilitz-Markus T, Shuper A, Schwartz M, et al. Delayed posttraumatic visual loss: a clinical dilemma. Pediatr Neurol 2000;22:133–135.
155. Griffith JF, Dodge PR. Transient blindness following head injury in children. N Engl J Med 1968;278:648–651.
156. Eldridge PR, Punt JAG. Transient traumatic cortical blindness in children. Lancet 1988;1:815–816.
157. Kadish HA, Schunk JE. Pediatric basilar skull fracture: do children with normal neurologic findings and no intracranial injury require hospitalization? Ann Emerg Med 1995;26:37–41.
158. Dandy WE. Pneumocephalus (intracranial pneumatocele or aerocele). Arch Surg 1926;12:949–982.
159. Keskil S, Baykamer K, Ceviker N, et al. Clinical significance of acute traumatic intracranial pneumocephalus. Neurosurg Rev 1998;21:10–13.
160. Steinbok P, Flodmark O, Martens D, et al. Management of simple depressed skull fractures in children. J Neurosurg 1987;66:506–510.
161. Ersahin Y, Mutluer S, Mizai H, et al. Pediatric depressed skull fractures: analysis of 530 cases. Childs Nerv Syst 1996;12:523–531.
162. Foy PM, Chadwick DW, Rajgopalan N, et al. Do prophylactic anticonvulsants alter the pattern of seizures after craniotomy? J Neurol Neurosurg Psychiatry 1992;55:753–757.
163. Dhellemmes P, Lejeune JP, Christiaens JL, et al. Traumatic extradural hematoma in infancy and childhood. J Neurosurg 1985;62:861–864.
164. Matson DD. Extradural hematoma. In: Matson DD. Neurosurgery of infancy and childhood, 2nd ed. Springfield, IL: Charles C. Thomas, 1969:316–327.
165. Shugerman RP, Paez A, Grossman DC, et al. Epidural hemorrhage: is it abuse? Pediatrics 1996;97:664–668.
166. Mazza, C, Pasqualin A, Feriotti G, et al. Traumatic extradural haematoma in children: experience with 62 cases. Acta Neurochir 1982;65:67–80.
167. McKissock W, Taylor JC, Bloom WH, Till K. Extradural haematoma: observations on 125 cases. Lancet 1960;2:167–172.
168. Kernohan JW, Woltman HW. Incisura of the crus due to contralateral brain tumor. Arch Neurol Psychiatr 1929;21:274–278.
169. Holtzschuh M, Schuknecht B. Traumatic epidural haematomas of the posterior fossa: 20 new cases and a review of the literature since 1961. Br J Neurosurg 1989;3:171–180.
170. Berker M, Cataltepe O, Ozcan OE. Traumatic epidural haematoma of the posterior fossa in childhood: 16 new cases and a review of the literature. Br J Neurosurg 2003;17:226–229.
171. Marshall LF. Surgical treatment of extracerebral lesions in head injury. In: Pitts LH, Wagner FC, eds. Craniospinal trauma. New York: Thieme, 1990:37–48.
172. Choux M, Lena G, Genitori L. Intracranial hematomas. In: Raimondi AJ, Choux M, DiRocco C, eds. Head injuries in the newborn and infant. New York: Springer-Verlag, 1986:203–216.
173. Neubauer UJ. Extradural hematoma of the posterior fossa. Twelve years experience with CT-scan. Acta Neurochir 1987;87:105–111.
174. Duhaime AC, Gennarelli TA, Thibault LE, et al. The shaken baby syndrome. A clinical, pathological, and biomechanical study. J Neurosurg 1987;66:409–415.
175. Wilkins B. Head injury—abuse or accident? Arch Dis Child 1997;76:393–396.
176. Rabe EF, Flynn RC, Dodge PR. A study of subdural effusions in an infant with particular reference to the mechanisms of their persistence. Neurology 1962;12:79–92.
177. Provencale J. The current role of SPECT in imaging subdural hematoma. J Nucl Med 1992;33:248–250.
178. Yakubu MA, Leffler CW. 5-Hydroxytryptamine–induced vasoconstriction after cerebral hematoma in piglets. Pediatr Res 1997;41:317–320.
179. Duhaime AC, Gennarelli LM, Yachnis A. Acute subdural hematoma: is the blood itself toxic? J Neurotrauma 1994;11:669–678.
180. Haseler LJ, Arcinue E, Danielsen ER, et al. Evidence from proton magnetic resonance spectroscopy for a metabolic cascade of neuronal damage in shaken baby syndrome. Pediatrics 1997;99:4–14.
181. Swift DM, McBride L. Chronic subdural hematoma in children. Neurosurg Clin North Am 2000;11:439–446.
182. McLaurin RL, Tutor FT. Acute subdural hematoma: review of ninety cases. J Neurosurg 1961;18:61–67.
183. Litovsky NS, Raffel C, McComb JG. Management of symptomatic chronic extra-axial fluid collections in pediatric patients. Neurosurgery 1992;31:445–450.
184. Bruce DA, Zimmerman RA. Shaken impact syndrome. Pediatr Ann 1989;18:482–494.
185. Harcourt B, Hopkins D. Ophthalmic manifestations of the battered-baby syndrome. Br Med J 1971;3:398–401.
186. Parsch CS, Krass J, Hoffmann E., et al. Arachnoid cysts associated with subdural hematomas and hygromas: Analysis of 16 cases, long-term follow-up, and review of literature. Neurosurgery 1997;40:482–490.
187. Barlow KM, Gibson RJ, McPhillips M, et al. Magnetic resonance imaging in acute non-accidental head injury. Acta Paediatr 1999;88:734–740.
188. Harwood-Nash DC. Abuse to the pediatric central nervous system. Am J Neuroradiol 1992;13:569–575.
189. Zimmerman RA, Bilaniuk LT, Bruce D, et al. Computed tomography of craniocerebral injury in the abused child. Radiology 1979;130:687–690.
190. Dolinskas CA, Zimmerman RA, Bilaniuk LT. A sign of subarachnoid bleeding on computed tomograms of pediatric head trauma patients. Radiology 1978;126:409–411.
191. Parent AD. Pediatric chronic subdural hematoma: a retrospective comparative analysis. Pediatr Neurosurg 1992;18:266–271.
192. Till K. Subdural haematoma and effusion in infancy. Br Med J 1968;2:400–402.
193. Hoffmann GF, Naughten ER. Abuse or metabolic disorder? Arch Dis Child 1998;78:199.
194. Sherwood D. Chronic subdural hematoma in infants. Am J Dis Child 1930;39:980–1021.
P.700

195. Ingraham FD, Heyl HL. Subdural hematoma in infancy and childhood. JAMA 1939;112:198–204.
196. El-Kadi H, Kaufman HH, eds. Chronic subdural hematoma. Neurosurg Clin North Am 2000;11.
197. Sauter KL. Percutaneous subdural tapping and subdural peritoneal drainage for the treatment of subdural hematoma. Neurosurg Clin North Am 2000;11:519–524.
198. Ersahin Y, Mutluer S, Koraman S. Continuous external subdural drainage in the management of infantile subdural collections: a prospective study. Childs Nerv Syst 1997;13:526–529.
199. Tabaddor K, Shulman K. Definitive treatment of chronic subdural hematoma by twist-drill craniostomy and closed-system drainage. J Neurosurg 1977;46:220–226.
200. McLaurin RL, McLaurin KS. Calcified subdural hematomas in childhood. J Neurosurg 1966;24:648–655.
201. Haviland J, Russell RI. Outcome after severe non-accidental head injury. Arch Dis Child 1997;77:504–507.
202. Martin HP, Beezley P, Conway EF, et al. The development of abused children. Adv Pediatr 1974;21:25–73.
203. Elmer E. A follow-up study of traumatized children. Pediatrics 1977;59:273–279.
204. Jennett WB. Trauma as a cause of epilepsy in childhood. Dev Med Child Neurol 1973;15:56–62.
205. Annegers JF, Grabow JD, Groover RV, et al. Seizures after head trauma: a population study. Neurology 1980;30:683–689.
206. Hendrick E, Harris L. Posttraumatic epilepsy in children. J Trauma 1968;8:547–556.
207. Santhakumar V, Ratzliff ADH, Jeng J, et al. Long-term hyperexcitability in the hippocampus after experimental head trauma. Ann Neurol 2001;50:708–717.
208. Hendrick E, Harris L. Posttraumatic epilepsy in children. J Trauma 1968;8:547–556.
209. Annegers J, Hauser WA, Coan SP, et al. A population-based study of seizures after traumatic brain injuries. N Engl J Med 1998;338:20–24.
210. Temkin NR. Risk factors for posttraumatic seizures in adults. Epilepsia 2003;44(Suppl 10):18–20.
211. Caveness WF, Meirowski AM, Rish BL, et al. The nature of posttraumatic epilepsy. J Neurosurg 1979;50:545–553.
212. Walton JW, Barwick DD, Longley BP. The electroencephalogram in brain injury. In: Rowbotham GF, ed. Acute injuries of the head: their diagnosis, treatment, complications and sequels. Baltimore: Williams & Wilkins, 1964.
213. Temkin NR. Antiepileptogenesis and seizure prevention trials with antiepileptic drugs: meta-analysis of controlled trials. Epilepsia 2001;42:515–524.
214. Walker AE. Posttraumatic epilepsy: an inquiry into the evolution and dissolution of convulsions following head injury. Clin Neurosurg 1958;6:69–103.
215. Jan MS, Camfield PR, Gordon K, et al. Vomiting after mild head injury is related to migraine. J Pediatr 1997;130:134–137.
216. Haas DC, Lourie H. Trauma-triggered migraine: an explanation for common neurological attacks after mild head injury. J Neurosurg 1988;68:181–188.
217. Dandy WE, Follis RH Jr. On the pathology of carotidcavernous aneurysms (pulsating exophthalmos). Am J Ophthalmol 1941;24:365–385.
218. Fattahi TT, Brandt MT, Jenkins WS, et al. Traumatic carotid-cavernous fistula: pathophysiology and treatment. J Craniofac Surg 2003;14:240–246.
219. Desal H, Leaute F, Auffray-Calvier E, et al. Fistule carotido-caverneuse directe, études clinique, radiologique et thérapeutique. Ápropos de 49 cas. J Neuroradiol 1997;24:141–154.
220. Berger G, Bonilha L, Santos SF, et al. Thrombosis of the internal carotid artery secondary to soft palate injury in children and childhood. Report of two cases. Pediatr Neurosurg 2000;32:150–153.
221. Mokri B, Piepgras DG, Houser OW. Traumatic dissections of the extracranial internal carotid artery. J Neurosurg 1988;68:189–197.
222. Payton TF, Siddiqui KM, Sole DP, et al. Traumatic dissection of the internal carotid artery. Pediatr Emerg Care 2004;20:27–29.
223. Chabrier S, Lasjaunias P, Husson B, et al. Ischaemic stroke from dissection of the craniocervical arteries in childhood: report of 12 patients. Eur J Paediatr Neurol 2003;7:39–42.
224. Ganesan V, Chong WK, Cox TC, et al. Posterior circulation stroke in childhood: risk factors and recurrence. Neurology 2002;59:1552–1556.
225. Khurana DS, Bonnemann CG, Dooley EC, et al. Vertebral artery dissection: issues in diagnosis and management. Pediatr Neurol 1996;14:255–258.
226. Hasan I, Wapnick, S, Tenner MS, et al. Vertebral artery dissection in children: a comprehensive review. Pediatr Neurosurg 2002;37:168–177.
227. Auer A, Felber S, Schmidauer C, et al. Magnetic resonance angiographic and clinical features of extracranial vertebral artery dissection. J Neurol Neurosurg Psychiatry 1998;64:474–481.
228. Snoek JW, Minderhoud JM, Wilmink JT. Delayed deterioration following mild head injury in children. Brain 1984;107:15–36.
229. McCrory PR, Berkovic SF. Second impact syndrome. Neurology 1998;50:677–683.
230. Kors EE, Terwind GM, Vermeulen FLMG, et al. Delayed cerebral edema and fatal coma after minor head trauma: Role of the CACNA1A calcium channel subunit gene and relationship with familial hemiplegic migraine. Ann Neurol 2001;49:753–760.
231. Bruce DA, Alavi A, Bilaniuk L, et al. Diffuse cerebral swelling following head injuries in children: the syndrome of malignant brain edema. J Neurosurg 1981;54:170–178.
231a. Levin HS. Neuroplasticity following non-penetating traumatic brain injury. Brain Injury 2003;17:665–674.
232. Taylor HG, Alden J. Age-related differences in outcomes following childhood brain insults: An introduction and overview. J Int Neuropsychol Soc 1997;3:555–567.
233. Koskiniemi M, Kyykka T, Nybo T, Jarho L. Long-term outcome after severe head injury in preschoolers is worse than expected. Arch Pediatr Adolesc Med 1995;149:249–254.
233a. Max JE, Lansing AE, Koele SL, et al. Attention deficit hyperactivity disorder in children and adolescents following traumatic brain injury. Dev Neuropsychol 2004;25:159–177.
234. Hawley CA, Ward AB, Magnay AR, et al. Outcomes following childhood head injury: A population study. J Neurol Neurosurg Psychiatry 2004;75:737–742.
235. Ewing-Cobbs L, Barnes M, Fletcher JM. Modeling of longitudinal academic achievement scores after pediatric traumatic brain injury. Dev Neuropsychol 2004;25:107–133.
236. Anderson VA, Morse SA, Klug, G, et al. Predicting recovery from head injury in young children: A prospective study. J Int Neuropsychol Soc 1997;3:568–580.
237. Mitchener A, Wyper DJ, Patterson J, et al. SPECT, CT, and MRI in head injury: acute abnormalities followed up at six months. J Neurol Neurosurg Psychiatry 1997;62:633–636.
238. Symonds C. Disorders of memory. Brain 1966;89:625–644.
239. Haas DC, Ross GS. Transient global amnesia triggered by mild head trauma. Brain 1986;109:251–257.
240. Harris P. Head injuries in childhood. Arch Dis Child 1957;32:488–491.
241. Gronwall D, Wrightson P, McGinn V. Effect of mild head injury during the preschool years. J Int Neuropsychol Soc 1997;3:592–597.
242. Wrightson P, McGinn V, Gronwall D. Mild head injury in preschool children: Evidence that it can be associated with a persistent cognitive deficit. J Neurol Neurosurg Psychiatr 1995;59:375–380.
242a. Pelco L, Sawyer M, Duffield G, et al. Premorbid emotional and behavioral adjustment in children with mild head injuries. Brain Inj 1992;6:26–37.
243. McKinlay A, Dalrymple-Alford JC, Horwood LJ, et al. Long term psychosocial outcomes after mild head injury in early childhood. J Neurol Neurosurg Psychiatry 2002;73:281–288.
244. Chadwick O. Psychological sequelae of head injury in children. Dev Med Child Neurol 1985;27:72–75.
244a. Cook JB. The effects of minor head injuries sustained in sport and the postconcussion syndrome. In: Walker AE, Caveness WF, Critchley M, eds. The late effects of head injury. Springfield, IL: Charles C. Thomas, 1969;408–413.
245. Nylander I, Rydelius PA. Post-concussion syndrome. Brain damage, consitutional characterics and environmental reactions. Acta Paediatr Scand 1988;77:475–477.
P.701

246. McClelland RJ, Fenton GW, Rutherford W. The postconcussional syndrome revisited. J R Soc Med 1994;87:508–510.
247. Hjern B, Nylander I. Acute head injuries in children: traumatology, therapy and prognosis. Acta Paediatr 1964 (Suppl):152.
248. American Academy of Neurology. Practice parameter: the management of concussion in sports. Neurology 1997;48:581–585.
249. Jennett B, Plum F. Persistent vegetative state after brain damage. A syndrome in search of a name. Lancet 1972;1:734–737.
250. Kampfl A, Franz G, Aichner F, et al. The persistent vegetative state after closed head injury: clinical and magnetic resonance imaging findings in 42 patients. J Neurosurg 1998;88:809–816.
251. Ashwal S. Medical aspects of the minimally conscious state in children. Brain Dev 2003;25:535–545.
252. Higashi K, Sakata Y, Hatano M, et al. Epidemiological studies on patients with a persistent vegetative state. J Neurol Neurosurg Psychiatry 1977;40:876–885.
253. Heindl UT, Laub MC. Outcome of persistent vegetative state following hypoxic or traumatic brain injury in children and adolescents. Neuropediatrics 1996;27:94–100.
254. Ashwal S, Eyman RK, Call TL. Life expectancy in a persistent vegetative state. Pediatr Neurol 1994;10:27–33.
255. Strauss DJ, Shavelle RM, Ashwal S. Life expectancy and median survival time in the permanent vegetative state. Pediatr Neurol 1999;21:626–631.
256. Hamilton MG, Myles ST. Pediatric spinal injury: review of 174 hospital admissions. J Neurosurg 1992;77:700–704.
257. Norenberg MD, Smith J, Marcillo A. The pathology of human spinal cord injury: Defining the problems. J Neurotrauma 2004;21:429–440.
258. Tator CH, Fehlings MG. Review of the secondary injury theory of acute spinal cord trauma with emphasis on vascular mechanisms. J Neurosurg 1991;75:15–26.
259. Bischof W, Nittner K. Zur Klinik und Pathogenese der vaskular bedingten Myelomalazien. Neurochirurgie 1965;8:215–231.
260. Ahmann PA, Smith SA, Schwartz JF, et al. Spinal cord infarction due to minor trauma in children. Neurology 1975;25:301–307.
261. Choi JU, Hoffman HJ, Hendrick EB, et al. Traumatic infarction of the spinal cord in children. J Neurosurg 1986;65:608–610.
262. Williams B. Pathogenesis of posttraumatic syringomyelia. Br J Neurosurg 1992;6:517–520.
263. La Haye PA, Batzdorf U. Posttraumatic syringomyelia. West J Med 1988;148:657–663.
264. Zwimpfer TJ, Bernstein M. Spinal cord concussion. J Neurosurg 1990;72:894–900.
265. Ruge JR, Sinson GP, McLone DG, Cerullo IJ. Pediatric spinal injury: the very young. J Neurosurg 1988;68:25–30.
266. Greenwald TA, Mann DC. Pediatric seatbelt injuries: diagnosis and treatment of lumbar flexion-distraction injuries. Paraplegia 1994;32:743–751.
267. Taylor GA, Eggli KD. Lap-belt injuries of the lumbar spine in children: a pitfall in CT diagnosis. Am J Roentgenol 1988;150:1355–1358.
268. Carlson GD, Gorden C. Current developments in spinal cord injury research. Spine J 2002;2:116–128.
269. Dumont RJ, Okonkwo DO, Verma S, et al. Acute spinal cord injury, part I: pathophysiologic mechanisms. Clin Neuropharmacol 2001;24:254–264.
270. Frankel HL, Hancock DO, Hyslop G, et al. The value of postural reduction in the initial management of closed injuries of the spine with paraplegia and tetraplegia. Paraplegia 1969;7:179–192.
271. Ditunno JF, Young W. Donovan WH, et al. The international standards booklet for neurological and functional classification of spinal cord injury. American Spinal Injury Association. Paraplegia 1994;32:70–80.
272. Sherrington CS. The integrative action of the nervous system. New York: Scribner, 1906.
273. Riddock G. The reflex functions of the completely divided spinal cord in man, compared with those associated with less severe lesions. Brain 1917;40:264–402.
274. Ditunno JF, Little JW, Tessler A, et al. Spinal shock revisited: a four-phase model. Spinal cord 2004;42:383–395.
275. Atkinson PP, Atkinson JLD. Spinal shock. Mayo Clin Proc 1996;71:384–389.
276. Hiersemenzel, LP, Curt A, Dietz V. From spinal shock to spasticity. Neuronal adaptations to a spinal cord injury. Neurology 2000;54:1574–1582.
277. Chiles BW, Cooper PR. Acute spinal injury. N Engl J Med 1996;514–520.
278. Kuhn RA. Functional capacity of the isolated human spinal cord. Brain 1950;73:1–51.
279. Sullivan AW. Subluxation of atlanto-axial joint. J Pediatr 1949;35:451–464.
280. Menezes AH, Mohonen M. Management of occipito-cervical instability. In: Cooper PR, ed. Neurosurgical topics, Vol. 1. Baltimore: Williams & Wilkins, 1990:65–76.
281. Veras LM, Castellanos J, Ramirez G, et al. Atlanto axial instability due to neurofibromatosis: case report. Acta Orthop Belg 2000;66:392–396.
282. Controversies in imaging acute cervical spine trauma. Am J Neuroradiol 1997;18:1866–1868.
283. Hadley MN, Zabramski JM, Browner CM, et al. Pediatric spinal trauma. Review of 122 cases of spinal cord and vertebral column injuries. J Neurosurg 1988;68:18–24.
284. Cattell HS, Filtzer DL. Pseudosubluxation and other normal variations in the cervical spine in children. J Bone Joint Surg 1965;47A:1295–1309.
285. Jefferson G. Concerning injuries of the spinal cord. Br Med J 1936;2:1125–1130.
286. Guttman L. Spinal cord injuries, 2nd ed. Oxford: Blackwell, 1976.
287. Sonntag VKH, Hadley MN. Nonoperative management of cervical spine injuries. Clin Neurosurg 1988;34:630–649.
288. Vale FL, Burns J, Jackson AB, et al. Combined medical and surgical treatment after acute spinal cord injury: results of a prospective pilot study to assess the merits of aggressive medical resuscitation and blood pressure management. J Neurosurg 1997;87:239–246.
289. Sypert GW. Stabilization and management of cervical injuries. In: Pitts LH, Wagner FC, eds. Craniospinal trauma. New York: Thieme, 1990:171–185.
290. Becker D, Sadowsky CL, McDonald JW. Restoring function after spinal cord injury. Neurologist 2003;9:1–15.
291. Bracken MB. Methylprednisolone and acute spinal cord injury: an update of the randomized evidence. Spine 2001;26(24 Suppl):S47–S54.
292. Geisler FH, Dorsey FC, Coleman WP. Recovery of motor function after spinal cord injury—a randomized, placebo-controlled trial with GM1 ganglioside. N Engl J Med 1991;324:1829–1838.
293. Geisler FH. Clinical trials of pharmacotherapy for spinal cord injury. Ann N Y Acad Sci 1998;845:374–381.
294. Kapfhammer JP. Axon sprouting in the spinal cord: growth promoting and growth inhibitory mechanisms. Anat Embryol 1997;196:417–426.
295. Okada S, Nakamura M, Mikami Y, et al. Blockade of interleukin-6 receptor suppresses reactive astrogliosis and ameliorates functional recovery in experimental spinal cord injury. Neurosci Res 2004;76:265–276.
296. Bregman BS, Coumans JV, Dai HN., et al. Transplants and neurotrophic factors increase regeneration and recovery of function after spinal cord injury. Prog Brain Res 2002;137:257–273.
297. Short DJ, Frankel HL, Bergstrwöm EMK. Injuries of the spinal cord in children. In: Vinken GW, Bruyn GW, Klawans HL, et al, eds. Spinal cord trauma. Handbook of clinical neurology, Rev. Ser. 17, Vol. 61. Amsterdam: Elsevier, 1992:233–252.
298. Piepmeier JM. Late sequelae of spinal cord injury. In: Narayan RK, Wilberger JE Jr, Povlikshock JT, eds. Neurotrauma. New York: McGraw-Hill, 1996:1237–1244.
299. Whiteneck G, Lammertse DP, Manley S, et al., eds. The management of high quadriplegia. New York: Demos, 1989.
300. Lightner DJ. Contemporary urologic management of patients with spinal cord injury. Mayo Clin Proc 1998;73:434–438.
301. Crockard HA, Heilman AE, Stevens JM. Progressive myelopathy secondary to odontoid fractures: clinical, radiological and surgical features. J Neurosurg 1993;78:579–586.
302. Martinez-Lage JF, Fernandez Cornejo V, Lopez F, et al. Lumbar disc herniation in early childhood: case report and literature review. Child Nerv Syst 2003;19:258–260.
P.702

303. Epstein JA, Lavine LS. Herniated lumbar intervertebral disks in teenage children. J Neurosurg 1964;21:1070–1075.
304. Kahanovitz N, Rimoin DL, Sillence DO. The clinical spectrum of lumbar spine disease in achondroplasia. Spine 1982;7:137–140.
305. Friedman AP, Merritt HH. Damage to cranial nerves resulting from head injury. Bull Los Angeles Neurol Soc 1944;9:135–139.
306. Brazis PW. Palsies of the trochlear nerve: diagnosis and localization—recent concepts. Mayo Clin Proc 1993;68:501–509.
307. McKennan KX, Chole RA. Facial paralysis in temporal bone trauma. Am J Otol 1992;13:167–172.
308. Galan E, Kousseff BG. Peripheral neuropathy in Ehlers-Danlos syndrome. Pediatr Neurol 1995;12:242–245.
309. Stoll G, Jander S, Myers RR. Degeneration and regeneration of the peripheral nervous system: from Augustus Waller’s observations to neuroinflammation. J Peripher Nerv Syst 2002;7:13–27.
310. Manabe Y, Sakai K, Kashihara K, et al. Presumed venous infarction in spinal decompression sickness. Am J Neuroradiol 1998;19:1578–1580.
311. Van Meir N, De Smet L. Carpal tunnel syndrome in children. Acta Orthop Belg 2003;69:387–395.
312. Villarejo FJ, Pascual AM. Injection injury of the sciatic nerve (370 cases). Childs Nerv Syst 1993;9:229–232.
313. Medical Research Council. Aids to the examination of peripheral nerve injuries. London: Balliere Tindall, 1986.
314. Papazian O, Alfonso I, Yaylali I, et al. Neurophysiological evaluation of children with traumatic radiculopathy, plexopathy, and peripheral neuropathy. Semin Pediatr Neurol 2000;7:26–35.
315. Hodes R, Larrabee MG, German W. The human electromyogram in response to nerve stimulation and the conduction velocity of motor axons: studies on normal and on injured peripheral nerves. Arch Neurol Psychiatry 1948;60:340–365.
316. Carvalho GA, Nikkhah G, Matthies C, et al. Diagnosis of root avulsions in traumatic brachial plexus injuries: value of computerized tomography myelography and magnetic resonance imaging. J Neurosurg 1997;86:69–76.
317. Aagaard BD, Maravilla KR, Kliot M. Magnetic resonance neurography: magnetic resonance imaging of peripheral nerves. Neuroimaging Clin North Am 2001;11:131–146.
318. Mitchell SW. Injuries of nerves and their consequences. Philadelphia: Lippincott, 1972.
319. Matsui M, Ito M, Tomoda A, et al. Complex regional pain syndrome in childhood: report of three cases. Brain Dev 2000;22:445–448.
320. Gordon N. Reflex sympathetic dystrophy. Brain Dev 1996;18:257–262.
321. Hannington-Kiff JG. Reflex sympathetic dystrophy. J R Soc Med 1987;80:605.
322. Bhatia KP, Bhatt MH, Marsden CD. The causalgia-dystonia syndrome. Brain 1993;116:843–851.
323. Birklein F, Riedl B, Sieweke N., et al. Neurological findings in complex regional pain syndromes–analysis of 145 cases. Acta Neurol Scan 2000;101:262–269.
324. Goldstein DS, Tack C, Li, ST. Sympathetic innervation and function in reflex sympathetic dystrophy. Ann Neurol 2000;48:49–59.
325. van der Laan L, ter Laak HJ, Gabreels-Festen A, et al. Complex regional pain syndrome type I (RSD). Pathology of skeletal muscle and peripheral nerve. Neurology 1998;51:20–25.
326. Honjyo K, Hamasaki Y, Kita M, et al. An 11-year-old girl with reflex sympathetic dystrophy successfully treated by thoracoscopic sympathectomy. Acta Paediatr 1997;86:903–905.
327. Bouchama A, Knochel JP. Heat stroke. N Engl J Med 1002;346:1978–1988.
328. Malamud N, Haymaker W, Custer RP. Heat stroke: a clinicopathologic study of 125 fatal cases. Milit Surg 1946;99:397–449.
329. Freedman D, Schenthal J. A parenchymatous cerebellar syndrome following protracted high body temperature. Neurology 1953;3:513–516.
330. Chavez-Carballo E, Bouchama A. Fever, heatstroke, and hemorrhage shock and encephalopathy. J Child Neurol 1998;13:286–287.
331. Hopkins PM, Ellis FR, Halsall PJ. Evidence for related myopathies in exertional heat stroke and malignant hyperthermia. Lancet 1991;338:1491–1492.
332. Cherington M. Neurologic manifestations of lightning strike. Neurology 2003;60:182–185.
333. Critchley M. Neurological effects of lightning and of electricity. Lancet 1934;1:68–72.
334. Silversides J. The neurological sequelae of electrical injury. Can Med Assoc J 1964;91:195–204.
334a. Duff K, MacCaffrey RJ. Electrical injury and lightning injury: a review of their mechanisms and neuropsychological, psychiatric and neurological sequelae. Neuropsychol Rev 2001;11:101–116.
335. Kotagal S, Rawlings CA, Chen SC, et al. Neurologic, psychiatric and cardiovascular complications in children struck by lightning. Pediatrics 1982;70:190–192.
336. Winkelman MD, Galloway PG. Central nervous system complications of thermal burns. A postmortem study of 139 patients. Medicine 1992;71:271–283.
337. Mohnot D, Snead OC, Benton JW. Burn encephalopathy in children. Ann Neurol 1982;12:42–47.
338. Antoon AY, Volpe JJ, Crawford JD. Burn encephalopathy in children. Pediatrics 1972;50:609–616.
339. Greer HD. Neurologic consequences. In: Bove AA, ed. Diving medicine, 3rd ed. Philadephia: Saunders, 1997:258–269.
340. Reuter M, Tetzlaff K, Hutzelmann A, et al. MR imaging of the central nervous system in diving-related decompression illness. Acta Radiol 1997;38:940–944.
341. Deymeer F, Jones HR. Pediatric median mononeuropathies: a clinical and electromyelographic study. Muscle Nerve 1994;17:755–762.