Clinical manifestations

Presentation and course

The clinical manifestations of abusive head trauma are nonspecific. They often closely resemble the signs and symptoms observed in other common pediatric disorders. Furthermore, a history of trauma to the infant is usually not forthcoming, and obvious physical signs of trauma, such as bruising or swelling of the skin, are often minimal or absent. For these reasons, clinicians caring for young children must have abusive head trauma in mind for all infants under their care (47).

History. In cases of inflicted trauma to infants and young children, an accurate history is rarely provided. Instead, caregivers often provide a history that excludes any mention of trauma and, instead, focuses on symptoms. The symptoms of abusive head trauma often match those commonly observed among ill infants, such as lethargy, irritability, and vomiting (Table 1). In other cases, more severe symptoms may be described, such as seizures and apnea. In these cases, a history of blunt impact or minor trauma may be obtained. But the degree of injury is often out of proportion to the provided description of the traumatic event. When physical injuries are noted, the caregivers often blame those injuries on resuscitative efforts or on one of the victim’s older siblings. Sometimes, a mechanism of injury incompatible with the child's stage of development is alleged, such as a 4-month-old allegedly crawling and falling down a flight of stairs. Furthermore, in cases of child abuse, the description of the incident from the same individual caretaker often changes over time. A history of shaking or other forms of maltreatment is obtained in only a minority of cases (20; 05; 43).

Table 1. Common Symptoms of Abusive Head Trauma

Examination. The general physical examination of a child with abusive head trauma can vary from normal to profoundly abnormal. Many infants who are later discovered to have abusive head trauma have no external signs of injury on initial examination. Others have nonspecific signs related to brain injury alone, such as vomiting or diminished mental status. Still others have profoundly abnormal examinations reflecting substantial brain dysfunction and possible involvement of other organ systems, such as apnea, bradycardia, and hypotension.

Neurologic examination. Many children with abusive head trauma have an abnormal neurologic examination at presentation (31; 37; 35). However, the type and severity of the neurologic abnormalities vary considerably (Table 2). Some children have only irritability, whereas others are comatose or in status epilepticus.

Table 2. Common Neurologic Findings in Abusive Head Trauma

Cutaneous bruising. Children with abusive head trauma may have cutaneous bruising as an external sign that initially raises the suspicion of child abuse or confirms the diagnosis. In cases of child abuse, the bruising may be present in suspicious patterns, such as linear marks, or in suspicious locations where bruising would not commonly result from normal activities of infants, such as the torso, ears, or neck (53). Bruising in infants before they learn to walk is highly unusual and should raise a strong suspicion of either child abuse or a bleeding disorder (63; 38).

Ophthalmologic lesions. Ophthalmologic injuries are very commonly observed in children suffering from abusive head trauma. A study found that the prevalence of ophthalmologic injuries was greater than 70% among cases of pediatric nonaccidental head trauma (46). Although retinal hemorrhages are the most common type of ophthalmologic lesion, intravitreous hemorrhages and papilledema are also frequently observed.

Retrospective series have suggested that retinal hemorrhages are present in 60% to 85% of abusive head trauma cases (37; 47). Furthermore, the extent of the retinal hemorrhages is often correlated with the severity of intracranial injury and neurologic disability (45; 11). Retinal hemorrhages are uncommon among children who suffer from physical abuse but do not have intracranial injuries (57; 40).

Retinal hemorrhages can occur in other conditions, including accidental injuries. However, the retinal hemorrhages derived from abusive head trauma are typically more numerous and more extensive and involve more layers of retina than those associated with accidents (01; 70; 30).

Retinal hemorrhages are best seen with mydriatic agents, may be unilateral or bilateral, and are described by their location in the retinal layers: subhyaloid when superficial in the retina or flame-shaped when in the superficial nerve fiber layer. Vitreous bleeding may occur as an extension of the retinal bleed (44).

Skeletal injuries. Many children with abusive head trauma have injuries to other organ systems, including the skeletal system and intra-abdominal organs. Injuries to these other organ systems may not be evident on initial physical examination. However, rib fractures and long-bone fractures, each of which is present in about one quarter of all cases of abusive head trauma, may become clinically evident through irritability when the involved body part is moved or manipulated (17).

Neuroimaging findings. Neuroimaging is obtained early in the evaluation of children suspected of suffering from abusive head trauma. There is not a pattern of cranial injury unique to abusive head trauma. However, subdural hemorrhage (multiple in number, along the convexities, or interhemispheric), cerebral ischemia, cerebral edema, and skull fractures are more common in abusive head trauma than in accidental injury.

Autopsy findings. Despite variations in the mechanism of injury and clinical presentation, autopsy findings in infants who have died from abusive head trauma are remarkably similar. External injury is found in 85% of cases. The extensive nature of scalp and subcutaneous trauma is often discovered on a shaved scalp or skin incision. Fractures are detected at autopsy in 25% of infants (56). Subdural hemorrhage in the parietooccipital convexity or the posterior interhemispheric fissure is the most consistent finding at autopsy (20). Cerebral edema is common if survival occurs for a few hours or days, and necrosis is often observed. Superficial contusions in the olfactory bulbs and gyrus rectus are frequent. Gliding contusions, tears of the corpus callosum, or diffuse axonal injury are common and suggest severe rotational forces (02). Ventricular tears, vein of Galen tears, or even cervical artery tears may be seen occasionally. The most common site of spinal cord injury is in the cervical region, C1-C4. Acute hemorrhage along the optic nerve sheath and retinal hemorrhages in any or all retinal layers are frequent findings.

Prognosis and complications

The mortality rate associated with abusive head trauma is 20% to 30% (34; 35). Factors associated with mortality include an initial Glasgow coma score of 5 or less, the presence of retinal hemorrhages or intraparenchymal hemorrhages, and the development of cerebral edema (58).

Outcomes among those who do survive are usually poor. One retrospective study found that 55% of survivors have persistent neurologic deficits and 65% have persistent vision impairment (37). Predictors of poor neurologic outcomes include occurrence of cardiorespiratory arrest and a low Glasgow Coma Score at admission (54). Neurologic deficits vary among patients, but the most common sequelae include delayed psychomotor development, spastic hemiplegia or quadriplegia, epilepsy, microcephaly, hydrocephalus, vision impairment, language disorders, and learning and behavioral disturbances, including hyperactivity and attention deficits (07; 67; 54). Although neurologic sequelae of abusive head trauma can occur in children of any age, they are most prominently observed in children ages 2 to 4 years (25). These severe neurologic deficits are often reflected in long-term neuroimaging studies by the presence of cystic encephalomalacia.

Retinal hemorrhages can likewise have serious long-term consequences. Large proportions of children with retinal hemorrhages suffer vision deficits that range in severity from mild impairment to complete blindness (36). In a retrospective study evaluating long-term vision loss in children with retinal hemorrhages sustained as a result of abusive head trauma, 46% had vision impairment in at least one eye 30 months following injury (67).

Clinical vignette

A 12-month-old boy was brought to the pediatric urgent care by his parents in the evening with complaints of fussiness and vomiting. According to the parents, the child had been in his normal state of good health that morning, aside from a runny nose. The parents took the child to daycare, which was located at a home in the family’s neighborhood. When the parents picked their child up after work, the babysitter stated that the child had been acting “sick” since mid-morning and had vomited several times. As the evening progressed, the parents became increasingly concerned because the child vomited recurrently and began to appear dehydrated. Worried about their baby, the parents sought medical attention.

On physical examination in urgent care, the patient was mildly tachycardic and had a low-grade fever. The symptoms were felt to be consistent with dehydration secondary to viral gastroenteritis. The parents were told to offer Pedialyte every few hours and seek emergency care if their child became more lethargic or was not improving. The patient was discharged home.

In the early morning hours, the parents checked on their child to offer a bottle. They found their son unresponsive and limp. The parents called 911, and the child was transported to the emergency department. Physical exam in the emergency department revealed lethargy, labored breathing, a tense fontanel, and bruising on the abdomen and scalp. A head CT scan revealed a large subdural hemorrhage over the right frontal region and a smaller subdural hemorrhage over the left. Pediatric neurosurgery was consulted, and the patient was urgently taken to the operating room for evacuation of his subdural bleeds. Following surgery, he was admitted to the pediatric intensive care unit for monitoring.

Admission labs included a complete blood count, platelet count, electrolytes, liver and pancreatic enzyme levels, and urinalysis. All were normal except for a moderate anemia with a hemoglobin concentration of 10g/dl and hematocrit of 31%.

After stabilization, the child abuse specialist within the hospital was consulted, and further workup was obtained, including evaluation by the ophthalmology, hematology, and neurology services. The child was found to have bilateral retinal hemorrhages that were too numerous to count and in multiple retinal layers. A skeletal survey revealed no fractures. His postoperative course was complicated by frequent seizures, which required several anticonvulsants, including phenobarbital, levetiracetam, and fosphenytoin, to control. Hematologic workup was negative. He was discharged home from the hospital after 3 weeks. MRI scan prior to discharge revealed regions of encephalomalacia at the sites of the previous subdural hematomas.

Legal proceedings followed, and several pediatric specialists were called to testify in court. The child’s babysitter was found guilty of inflicting abusive head trauma on the child.

Biological basis

Etiology and pathogenesis

Abusive head trauma is due to the infliction of blunt force, vigorous shaking, slamming, or other inappropriate and assaultive behaviors by an adult on a young child. However, the exact etiological mechanism of abusive head trauma is controversial (69). In particular, it remains unclear whether the injuries are due principally to rotational forces (shaking) or to translational forces (blunt trauma or slamming of the body and head). The term "shaken baby syndrome," as coined by Caffey was based on evidence that angular deceleration can lead to cerebral concussion and subdural hematoma. Because angular deceleration occurs with shaking, it was assumed that the injuries were due principally to shaking of the baby. However, external injuries observed on many babies suggested that infliction of blunt trauma was also present in many cases and might play an important role.

Part of the reason why the pathogenesis of abusive head trauma remains unclear is that animal models are unable to reliably replicate the full spectrum of neuropathology found in human children with abusive head trauma. Much of this is due to simple differences in neuroanatomy. In particular, although laboratory rodent models are valuable for the study of many neurologic disease processes, their value for the study of abusive head trauma is diminished by the fact that rodents are lissencephalic and have relatively small brains with little white matter. Their lack of gyri and sulci diminishes movement of brain tissue within the skull, and their paucity of white matter makes it difficult to study the diffuse axonal injury that occurs in humans. Domestic animal species, such as sheep and pigs, have larger brains with gyral convolutions, which makes them somewhat more valuable as model systems. But the orientation of the neuraxis in these species, which is almost linear, is very different from humans and makes them less vulnerable to rotational shearing injuries and to concussion (23). Nevertheless, animal models have provided some useful insights.

To examine the relative roles of shaking (rotational deceleration) and blunt trauma (translational deceleration), investigators have developed experimental model systems of these injuries. Utilizing a lamb model of nonaccidental head injury, Finnie and coworkers demonstrated that angular deceleration forces, which cause the head to rotate on its axis, differentially moves the skull, dura, and intracranial contents (24). This leads to parenchymal brain injury, including diffuse axonal injury, thus, showing that rotation alone can lead to serious intracranial injury. However, other investigators have shown the importance of blunt trauma. Duhaime and colleagues constructed mechanical models approximating the size, weight, and density of a 1-month-old infant (20). They subjected the models to vigorous shaking, alone or in combination with impaction against a surface. They found that shaking alone did not generate accelerations sufficient to produce concussion, subdural hematoma, or diffuse axonal injury. However, the combination of shaking and impaction generated much greater tangential and angular accelerations that were capable of producing those injuries.

The mechanism of injury has also been investigated for the cervical spine. In biomechanical studies using animals and neonatal cadaveric models, Bandak observed that the head velocities and accelerations reported in cases of shaken baby syndrome far exceed the structural limits of the cervical spine (06). In addition, in their clinical series, Alexander and colleagues compiled data from 24 infants diagnosed with “shaken baby syndrome” and concluded that shaking alone could account for all of the injuries observed. However, they emphasized that impact may also have been involved, even in those cases where no external evidence of injury was observed (05).

Primary injury. It is likely that both shaking and blunt trauma underlie the injuries of abusive head trauma in young children. Regardless of which forces are involved, structural injuries to the brain constitute the first step in the pathogenesis of abusive head trauma. These structural injuries, which frequently include subdural hemorrhages, subarachnoid hemorrhages, multilayered retinal hemorrhages, intraparenchymal contusions, skull fractures, and spine injuries, constitute the primary injury. Infants and young children possess developmental features that make them more prone than adults to these primary injuries (Table 3).

Table 3. Features of a Young Child that Increase Susceptibility to Injury from Abusive Head Trauma

Secondary injury. Although the primary injury to the young child’s brain can be life-threatening and cause serious and long-term disability, it is the secondary injuries that are often even more devastating. In particular, hypoxia and ischemia commonly occur in children with abusive head injury (33). The hypoxic-ischemic injuries can be regional and global. These secondary brain injuries often worsen cerebral edema, exacerbate the seizures, and may be the major cause of morbidity and mortality in children with abusive head trauma. Several mechanisms underlie these secondary hypoxic-ischemic injuries, as shown in Table 4.

Table 4. Mechanisms of Secondary Hypoxic-Ischemic Injury in the Child with Abusive Head Trauma

Epidemiology

Abusive head trauma has an incidence of 32 to 38 cases per 100,000 children per year in the first year of life in the United States, and nearly one quarter of the cases are fatal (47). Abusive head trauma is largely restricted to children under 4 years of age, with the majority being younger than 1 year of age (48). Although children younger than 1 year of age are most likely to suffer abusive head trauma, it is children aged 2 to 4 years that are most likely to die from it. One study found that, among children suffering abusive head trauma, the mortality was 10% for children younger than 2 years of age but was 22% for children 2 to 4 years of age (48). This difference in mortality rates may be due to differences in the types of injuries sustained. Children younger than 2 years suffered more fractures and subdural hemorrhages, whereas children 2 to 4 years suffered more internal injuries and cerebral edema.

The majority of abusive head trauma occurs in private homes. One study found that 98.5% of assaults occurred in private homes, whereas far less than 1% occurred in day care centers (55).

Victim risk factors. Infancy is believed to be the age of highest risk for abuse because that is the stage of childhood when prolonged crying is most common. A prolonged episode of crying, especially if it is recurrent, is a commonly reported trigger for physical child abuse (10; 03). For this reason, babies with colic or other conditions that result in incessant crying, are at high risk of abuse. Other risk factors for infants include prematurity, major birth defects, and male sex (65; 28).

Perpetrator risk factors. The perpetrators of abusive head trauma are usually male. Most commonly, they are the infant’s father. In one study of abusive head trauma, the perpetrator, in decreasing order of frequency, was the father (50%), stepfather or male partner of the mother (20%), or mother (12%) (37). Babysitters (most of whom are female) are also common perpetrators and were identified as the perpetrator in 17% of cases in one study (61).

Other risk factors. Children diagnosed with abusive head trauma have often been maltreated before (59). One large study found that 60% of children with abusive head trauma had a history or clinical evidence of previous physical abuse, and child welfare agencies had been previously involved with the family 22% of the time (37). Other risk factors include alcohol or drug abuse in the family (65), young maternal age (28), and history of abuse in other family members (27). Cases are more common in urban regions than in rural or suburban regions, and more cases are seen in autumn and winter than in spring and summer (08). Financial and other stressors in the family and in society at large can also increase the incidence of abusive head trauma. For example, reports suggest that the incidence of abusive head trauma rose dramatically during the COVID-19 pandemic, probably as a result of social isolation and distancing, along with the stresses of unemployment and threat of illness (60).

Prevention

Abusive head trauma is an entirely preventable disorder. In light of the high incidence of abusive head trauma and its substantial negative impact on the lives of its victims and perpetrators, there is perhaps no other disorder for which preventive education could be more consequential. However, despite prevention efforts, the incidence of abusive head trauma remains high.

The American Academy of Pediatrics Committee on Child Abuse and Neglect has published recommended strategies preventing abusive head trauma (47). There are two general strategies for prevention. The first is a nontargeted strategy, in which large-scale public education campaigns are addressed to the broad population. The advantages of this strategy include the fact that the message reaches many people at a relatively low cost per recipient. Unfortunately, the efficacy of this strategy is questionable. The broad-based applications of these interventions have not been shown to decrease hospitalizations for abusive head trauma (71; 18).

The second strategy is a targeted strategy, in which educational campaigns are directed toward a specific subset of the population who are at-risk for child maltreatment. Such targeted populations typically include expectant mothers, young parents-to-be, new parents, or babysitters. The advantage of this approach is that it directs information and warnings to those who need it most--those whose circumstances may put them at the highest of being perpetrators. In terms of time and resources, these targeted approaches are more expensive per recipient than the broad-based approach; however, there is some evidence that they may be more effective.

For example, a study found that prenatal and home visitations for new mothers significantly reduced the rates of reported child abuse and neglect compared to those who did not receive the intervention (49). Another study that targeted new mothers and caregivers provided anticipatory guidance regarding the frustration engendered by inconsolable crying. It emphasized that crying and frustration from it are normal but temporary and will soon come to an end. Materials specifically addressing crying were provided to mothers in the intervention arm, whereas general injury prevention materials were provided to those in the control arm. The study showed that with proper anticipatory guidelines, mothers receiving the counseling on crying did better and walked away from a potentially abusive situation more frequently than those in the control group. The study’s shortcoming is that education was not provided to males, who are the most common perpetrators of abusive head trauma (09).

One promising avenue is the use of educational videos targeted at high-risk adults. A study from Japan evaluated the effectiveness of using an educational video to reduce parental risk of committing child abuse. Parents were shown an 11-minute video at 2 and 4 months postpartum visits that outlined features of infant crying, anger, shaking, and a simulation of the anatomical mechanism of abusive head trauma. Videos also discussed how to respond to a crying infant. Parents who completed the intervention reported that they were 74% less likely to shake their infants, 43% less likely to smother their infants, and 52% less likely to shake and smother their infants (26). Although these results were promising, the study did not evaluate whether or not there was a change in the incidence of child abuse cases.

Another study out of the United States used a similar intervention but instead aimed to prospectively evaluate changes in reported child abuse cases in response to an educational video. After 23 months, there was no change in hospital admissions for abusive head trauma within the observed geographical area. However, only 6% of parents involved in the program completed the entire intervention (18). Thus, although data are not strong that educational videos can reduce the incidence of child abuse, the low-cost and low-risk nature of educational videos may make them a worthwhile intervention. The American Academy of Pediatrics policy statement recommends anticipatory guidance on inconsolable crying. During routine medical visits, clinicians should ask parents how they cope with feelings of stress induced by crying and other aspects of parenthood. They may give parents advice regarding steps to take when they feel frustrated with a crying baby, such as putting the baby safely in a crib on his or her back, walking away, and calling a friend or a helpline while checking on the baby every 5 to 10 minutes. Clinicians can also provide a follow-up and referral for parents who may need counseling or additional support (47).

Differential diagnosis

No other condition mimics all of the features of abusive head trauma. The key diagnostic features include history of trauma, unexplained extracranial or bony injuries, retinal hemorrhages, inflicted soft tissue injuries, and radiological evidence of trauma with subdural hemorrhage. When all of these features are present, the diagnosis of abusive head trauma is usually straightforward. In many cases, the patient with abusive head trauma presents with some combination of these features, but not all of them. The diagnosis of abusive head trauma requires careful consideration of a differential diagnosis in all cases, especially when some of the signs are missing. The diagnoses most closely mimicking abusive head trauma are accidental injury, birth injury, and bleeding disorders.

Accidental injury. Accidental injury is at the forefront of the differential diagnosis for abusive head trauma. Patients with abusive head trauma often present with a history of an accidental injury, such as a fall from a highchair or diaper-changing table. Furthermore, unintentional injuries can result in intracranial hemorrhages, retinal hemorrhages, and skeletal fractures. Thus, at least superficially, both the history and the physical examination findings can be plausibly explained by an accidental injury. However, in accidental trauma, the history is clear, consistent, and easily explains the child's presenting symptoms. In contrast, in abusive head trauma, the history is often vague, obscure, and implausible, and the degree of injury is often out of proportion to the reported degree of trauma. Accidental trauma has often been witnessed by more than one observer, whereas abusive head trauma typically has not. Subdural hematomas are more common following inflicted trauma than following an accident, and mixed density subdural hematomas point more strongly toward abuse than toward accident. Retinal hemorrhages in abusive head trauma are usually numerous, involve multiple retinal layers, and extend beyond the retina’s posterior pole to the periphery (30). In contrast, in accidental injury, retinal hemorrhages are rare (62). When they do occur, they are often few in number, localized to a single retinal layer, and confined to the posterior pole (44). Abusive head trauma is often accompanied by unexplained bruising and skeletal injuries, including posterior rib fractures, whereas these are absent in accidental injuries (22).

Birth injury. As a result of the birthing process, young infants can have findings that overlap abusive head trauma. Most notably, these findings include intracranial hemorrhages and retinal hemorrhages. Both occur more commonly following vacuum- or forceps-assisted deliveries than non-instrumented deliveries (68; 32; 39). Following birth, subdural hematomas can be found in 8% to 17% of term neonates (68; 42). However, unlike that observed in abusive head trauma, these hemorrhages from birthing are usually clinically silent and are not accompanied by skull fractures or other bone injuries (42). Subdural hematomas associated with the birthing process typically resorb by 4 weeks postpartum. Retinal hemorrhages associated with the birthing process can be observed in 20% to 30% of infants during the first 24 hours of life and 10% to 15% of infants during the first 72 hours of life. Unlike those observed with abusive head trauma, most perinatal retinal hemorrhages are small, confined to the posterior pole, clinically silent, and resolve within several weeks (41).

Bleeding disorders. Bleeding disorders can mimic abusive head trauma because these disorders can lead to intracranial hemorrhages, retinal hemorrhages, and bruising. In fact, intracranial hemorrhages and easy bruising are quite common in the setting of coagulopathies. On the other hand, retinal hemorrhages in this setting are quite rare but can occur. Furthermore, in contrast to those associated with abusive head trauma, retinal hemorrhages induced by bleeding disorders are usually low in number and confined to the posterior pole (04). The bleeding disorder induced by leukemia can lead to many and large retinal hemorrhages that extend beyond the posterior pole. However, retinal examination often reveals a leukemic infiltration, and children with leukemia typically have additional clinical findings and laboratory abnormalities that point toward this diagnosis and away from abuse.

Other considerations for differential diagnosis. Depending on the clinical setting, other diagnoses should also be considered in children presenting with histories or physical findings suggestive of abusive head trauma, for example:

Diagnostic workup

Although the history and physical examination are of paramount importance, laboratory and radiographic studies are also indispensable in evaluating children with suspected abusive head trauma. The precise set of laboratory and radiographic studies that are obtained and the order in which they are done depends strongly on the child’s condition. In particular, hemodynamically or neurologically unstable patients will require a different approach than stable and well-appearing patients. For example, an unenhanced head CT scan will likely be the first imaging study obtained on an unstable patient, whereas a brain MRI scan will likely be the preferred initial imaging study obtained on a stable patient. Nevertheless, despite the need for an individualized approach, most patients with suspected or confirmed abusive head trauma will require the following studies.

Neuroimaging. Head CT scans are readily available and can reliably identify injuries that require immediate intervention. Head CT readily identifies large extraaxial hemorrhages (including subdural hemorrhage and epidural hemorrhage), injuries producing mass effect, large areas of cerebral ischemia, cerebral contusion, skull fractures, and scalp hematoma. However, they are relatively insensitive at detecting diffuse axonal injury and early brain edema.

MRI brain scans are almost always acquired during the evaluation of abusive head trauma (17). At some centers, an MRI scan is the preferred initial evaluation when it can be obtained in a timely manner and when a pediatric neuroradiologist is available to interpret it promptly. MRI scan can be the preferred first-line study because it avoids the radiation exposure that inevitably accompanies head CT and because MRI can be more sensitive than CT at identifying subdural hematomas at initial presentation. Eventually, an MRI scan of the brain is obtained on almost all patients with abusive head trauma because it is superior to CT at documenting the pattern and extent of injuries and allowing the timing of head injuries to be determined. MRI is more sensitive than CT at detecting diffuse axonal injury, small areas of infarction, early brain edema, and posterior fossa injuries. However, MRI scans are less sensitive than CT at detecting skull fractures and some subarachnoid hemorrhages (52).

MRI scan of the cervical spine should also be obtained. Many infants who have been physically abused, especially those who have been shaken, will have spinal and paraspinal injuries, including extraaxial hemorrhage, ligamentous injury, trauma to the paraspinal soft tissues, and spinal cord contusion (16).

Skeletal survey. All young children suspected of suffering from abusive head trauma should undergo a skeletal survey. This study can reveal unexplained fractures, which can strongly support the diagnosis of trauma and abuse. In addition, certain types of fractures, such as posterior rib fractures and metaphyseal fractures, point strongly toward physical abuse. Whether or not the initial skeletal survey reveals any fractures, the skeletal survey should be repeated several weeks later as this delayed study may reveal fractures that were not evident on the initial survey and may play a critical role in dating the injury.

Coagulation studies. Because abusive head trauma commonly includes intracranial bleeding and bleeding at other sites, coagulopathies are prominent components of the differential diagnosis. Thus, coagulation studies are critical in the evaluation of a child with suspected abusive trauma. All children with suspected abusive head trauma should have a complete blood count with platelets, prothrombin time (PT), INR, and activated partial thromboplastin time (aPTT). If a child has prominent bruising, additional coagulation studies should be obtained, including von Willebrand factor (VWF) antigen and activity, factor VIII level, and factor IX level.

Chemistry studies. Serum electrolytes, liver enzyme levels, pancreatic enzyme levels, and a urinalysis are recommended to identify metabolic abnormalities and injuries to other organ systems. Urine organic acids should be considered to evaluate for possible glutaric aciduria as a cause for intracranial hemorrhage. Cardiac enzymes, including troponin and creatine kinase with muscle and brain subunits (CK-MB), should be obtained if cardiac injury is suspected.

Lumbar puncture. Because they are often suspected of having sepsis or meningitis, many young children with abusive head trauma will undergo a lumbar puncture with collection of a CSF sample. Xanthochromia or the presence of red blood cells in the CSF may reflect intracranial injury.

Ophthalmologic examination. All children with suspected abusive head trauma should undergo a funduscopic examination to identify retinal hemorrhages and other eye injuries. An ophthalmologist should preferably perform this examination.

Management

The management of abusive head trauma varies considerably from case to case. The first significant determinant of the direction of management lies with the vital signs and level of consciousness. For patients with impaired consciousness or unstable vital signs, maintenance of airway, breathing, and circulation are the first steps. Fluid resuscitation is necessary in hypotensive infants.

Usually, the second major determinant of the direction of management lies with the results of the neuroimaging studies. In particular, a child with a large intracranial hemorrhage will often require neurosurgical evacuation. This is especially true if the hemorrhage is large enough to produce mass effect and midline shift.

Another common neuroimaging result in abusive head trauma that requires prompt intervention is cerebral edema. Cerebral edema with mass effect and midline shift often requires hyperventilation and osmotic interventions with hypertonic saline or mannitol administration.

Another important management step is the treatment of seizures. Seizures are common in children with abusive head trauma. For this reason, some physicians prefer to treat patients prophylactically with an anticonvulsant before any seizures occur. Others prefer to wait for the onset of seizures. The most commonly used anticonvulsant in cases of abusive head trauma is levetiracetam. The advantages of levetiracetam in this setting are that it can be administered intravenously, and it has little effect on level of consciousness and, thus, does not interfere with neurologic assessment.

A lumbar puncture is often performed as an early management step to evaluate for possible meningitis and to detect the presence of xanthochromia or red blood cells, reflective of intracranial bleeding.

Laboratory studies, especially a CBC, general chemistry, clotting profile, urinalysis, liver enzymes, and pancreatic enzymes, should be checked early in the management course to identify metabolic abnormalities, coagulopathies, and injuries to abdominal or pelvic organs that require prompt attention.

A skeletal survey should be obtained early in the course to identify orthopedic injuries that require stabilization and aid in diagnosing abusive trauma.

After the patient is stabilized, several specialists should be consulted. These include an ophthalmologist to assess for retinal hemorrhages and other forms of ocular trauma and a child abuse specialist to ensure that the workup is complete and accurate, that all documentation has been completed, and that protocols have been followed.

If abuse is suspected, state laws require that appropriate child welfare and law enforcement agencies be notified. It should be explained to caretakers that investigative procedures are necessary for the welfare of the child. Physicians involved in cases of suspected abusive head trauma are often required to testify in court. For this reason, extra effort should be exerted to ensure that the medical records are accurate and thorough (19).