Child abuse is one of the leading causes of morbidity and mortality in the United States, resulting in approximately 1,800 deaths annually1. In comparison, the American Cancer Society estimates that in the year 2024, 1,590 children will die from pediatric cancer2. The largest proportion of physical abuse related death in young children is due to abusive head trauma (AHT), with the majority of victims being under 3 years of age1,3. Most of the data regarding AHT is from the Unitied States. However, AHT is an internationally accepted diagnosis with similar rates of incidence documented in multiple countries4.

Infants have the highest rates of AHT, with AHT seen in up to 38 per 100,000 children in the first year of life, with an estimated 25% of cases being fatal5. Given the frequency and potential severity of AHT, it is vital that these cases are quickly identified and properly managed. Through this article we hope to provide the reader with the information and confidence needed to successfully recognize AHT in the acute setting.

The term “shaken baby syndrome” (SBS) originated in the 1980s based on literature regarding shaking or acceleration-deceleration injuries in a child6. SBS remained the prevailing term for inflicted head injury until 2009, when the American Academy of Pediatrics published a policy statement advocating for the use of the term AHT7. The change was recommended not to diminish the role that shaking has in cases of AHT, but to be inclusive of all inflicted mechanisms leading to traumatic brain injury in children (i.e. shaking or inertial events as well as blunt force trauma)5,7,8. The term SBS is no longer a generally accepted medical term among Child Abuse Pediatricians, though other specialties or non-medical professionals might still unknowingly use the outdated nomenclature.

Not all cases of AHT are clearly identified since many victims present with only non-specific symptoms such as fussiness, irritability, emesis, difficulty with feeds, changes in sleep, or other vague complaints.8,9,10. Subtle signs of injury may lead to delayed or missed diagnosis of inflicted intracranial trauma5,8,9,10,11,12.

One study by Jenny et al. analyzed 173 cases of AHT, with 54 (31%) cases misdiagnosed by medical providers13. Of the misdiagnosed cases, 15 were subsequently re-injured and 5 died. The 5 deceased patients initially presented with non-specific complaints such as vomiting and irritability, and were diagnosed with influenza, otitis media, or gastroenteritis13. In a later study, Letson et al. found that 25% with AHT had a missed opportunity to identify abuse in a medical setting3.

Missed diagnoses of AHT can expose children to subsequent or escalating abuse if they remain in the same care environment, with recurring injuries becoming more severe or potentially life-threatening10. One study found that 40% of victims of non-accidental trauma (NAT) were subsequently reinjured in the following two years, while another estimated the time between initial abusive injury and death to be approximately 9 months14,15. In fact, Deans et al. showed children who presented to medical care for recurrent NAT had 15% higher mortality when compared to victims of one episode of abuse16.

Many AHT patients present with overt signs of head injury including altered mental status, apnea, or coma, with some literature reporting initial neurologic abnormality in 50-65% of AHT cases8,9,17,18. AHT cases may have higher rates of loss of consciousness and lower Glasgow Coma Scale scores than accidental injuries19. Patients with AHT are also at high risk for seizures,which are identified in 27-90% of cases19,20,21,22,23,24,25. The presence of seizures has been correlated with worse clincal outcomes, as demonstrated by greater duration of hospital stay and increased necessity of rehabilitiation20. One study by Chou et al. found that EEG demonstration of focal seizures and diffuse cortical dysfunction was a predictor of poor prognosis. Additionally, Dingman et al. found that increased seizure severity was associated with worse hypoxic-ischemic injury in patient with AHT24. Given the prevalence of seizures in cases of AHT and the utility of seizure evaluation for patient prognosis, EEG monitoring should be considered in all cases of AHT.

AHT should also be on the differential in young infants who present with isolated seizures, as this age group is at high risk for physical abuse. One study evaluating 31 infants less than 6 months with first time seizures found that of the 22 with imaging, 12 had abnormal results.26 One patient's seizures were ultimately attributed to trauma.

Presentations labeled as BRUE or ALTE could also represent missed cases of AHT27,28. One study found that 6 of 243 (2.5%) ALTE admissions were ultimately diagnosed as AHT, two of whom subsequently died28. Doswell et al. evaluated 2036 patients presenting with BRUE and found that 7 (0.3%) were victims of physical abuse, including 1 with cutaneous findings concerning for trauma29. The remaining 6 patients were diagnosed with physical abuse days to months later, five of whom had AHT. Additionally, Bonkowsky et al. identified 471 patients <12 months of age who were admitted for ALTE and followed these patients for an average of 5.1 years. The authors found that 11% of those patients experienced abuse30. These percentages may be inaccurate representations of true missed cases, however, as evaluation for child maltreatment, such as head imaging, is uncommon during the evaluation of a BRUE29.

A thorough medical history is crucial in identifying potential AHT. When obtaining a history, ask broad and open-ended questions of the child's caregivers. Providers should focus on ascertaining any potential mechanisms of injury, subtle symptoms that could indicate head injury, or a possible underlying medical explanation for the child's presentation.

When a potential trauma history is identified, additional information is helpful including how the child was acting prior to the trauma, how the trauma was sustained, who (if anyone) witnessed the injury, the position the child was found in following the trauma, the height of a proposed fall, what a child fell on during a proposed fall, how the child displayed subsequent symptoms such as loss of consciousness, breathing, cyanosis, seizures, changes in behavior, and how soon after the onset of symptoms did the family seek medical care8,31,32.

Children with AHT may have no reported history of preceding trauma. In those cases, ask when the child was last at their normal state of health. Then clarify the sequence of events and symptoms that led up to the child's presentation—did the symptoms develop over time, worsen, or change in any way? Are caregiver's histories consistent with each other8,11,31?

Providers should ask caregivers about additional symptoms that might point to previous neurologic injury or other injuries preceding the presentation to care, particularly in infants or nonmobile children. History of prior vomiting, changes in breathing, color changes, hyper- or hypotonia, unresponsiveness, limp or painful extremities, eye movement changes, bleeding from the nose or mouth, easy bleeding, previous bruising, unexplained fussiness, or listlessness are all helpful8,31.

The assessing physician should inquire about the developmental abilities of a child. For example, if the patient has yet to demonstrate the ability to stand, sit, or roll, he or she would be less likely to accidentally fall.

In very young infants, questions regarding birth may help to explain injuries or symptoms, though would not account for findings weeks to months later in a previously asymptomatic baby. Questions include: presence of traumatic birth, medical complications during the pregnancy, prematurity, instrumentation including if vacuum assistance or forceps were needed during delivery, vitamin K administration at birth, or if the child required neonatal intensive care8,31

A family history should include questions about bleeding disorders, childhood deaths in the family, frequent miscarriages, frequent fractures, any genetic or metabolic conditions, or other congenital childhood conditions8,31.

In cases of AHT, it is most common to obtain either no history of an injury, an inconsistent history of injury, or a relatively benign history of trauma, such as a short fall8,9. One study by Hymel et al. found that in cases of head injury, a lack of a trauma history, an inconsistent accidental history, or an accidental history that was unlikely based on the child's developmental abilities showed high specificity for AHT33. Social factors such as substance use, domestic violence, and previous child protective service (CPS) involvement may indicate risk factors for AHT, but the absence of these issues does not exclude possible abuse31.

Many victims of AHT have additional injuries identified on physical examination. It is critical that the assessing medical provider complete a full skin examination, particularly in infants and non-mobile children. Bruising can co-occur with AHT; among children less than 6 months old with bruising, 27% were found to have injury on neuroimaging8,11,34. Any bruise in an infant less than 4.99 months or bruising on the torso, ears, neck, frenulum, angle of the jaw, fleshy part of the cheek, eyelids, or subconjunctiva of a child under age 4.0 years carries a high specificity for abuse35,36.

The physical examination should also include vital signs, growth parameters such as weight, height, and head circumference with comparison to growth curve if available, skin perfusion, level of consciousness, mouth, frenula, conjunctiva, pupils, eye movements, fontanelle, and assess for any bony tenderness31,37. A thorough neurologic exam is essential, including mental status, fussiness, movements, tone, strength, and gait if the child is ambulatory38. It is important to note that symptoms such as bony tenderness and mental status changes may be difficult to assess in infants, and injuries may still be present without these findings.

Following a thorough history and physical examination, the acute evaluation of AHT requires a combination of imaging and labs. CT or fast MRI may be used as the initial neuroimaging of choice, depending on the patient's clinical presentation, stability, and the resources available. Brain MRI is helpful in clarifying abnormal findings on CT or fast MRI8,39,40. Additional evaluation for occult injuries includes a skeletal survey—not simply a “babygram”—following American College of Radiology guidelines for children under 2-3 years of age41,42. Lab work including complete blood count (CBC), coagulation studies, lipase, chemistry, and liver function panel for young children can help identify additional concerns for intra-abdominal injuries or signs of bleeding disorders37,43.

A dilated eye exam with an ophthalmology evaluation for retinal hemorrhages should be completed as quickly as is medically feasible, for patient's with intracranial injury44,45. However, it is important to note that retinal hemorrages are rarely seen in cases with normal neuroimaging or with an isolated simple skull fracture. Therefore, an examination is likely unnecessary for those patients46,47.

If non-accidental trauma or AHT is felt to be a possible diagnosis, completing the medical evaluation should not delay mandated reporting to CPS and law enforcement. Timely involvement of CPS and police can identify crucial information regarding the patient's home or the scene of an injury, and can expedite care for any other involved children who might be at risk of abuse8,37. If a child dies and AHT has been either diagnosed or considered, a coroner or medical examiner should perform an autopsy.

Clinical prediction rules exist to guide medical providers as they evaluate patients with potential intracranial findings and possible AHT48. One such tool, the Pittsburgh Infant Brain Injury Score (PIBIS), estimates the likelihood of abnormal head imaging in infants and can aid in the decision to obtain a head CT or MRI49. The tool examines five variables: abnormality on skin examination (2 points), age greater than 3 months (1 point), head circumference that is greater than the eighty-fifth percentile (1 point), and a serum hemoglobin level of less than 11.2g/dL (1 point). With a score of 2, the sensitivity and specificity for abnormal neuroimaging were 93% and 53%, and with a score of 3, sensitivity was 81% and specificity 75%49.

Another clinical prediction rule proposed by Hymel and the Pediatric Brain Injury Research Network (PediBIRN) found that bilateral or interhemispheric fissure SDHs, respiratory compromise noted prior to admission, bruising to the torso, ears, or neck, and a skull fracture that is not unilateral, simple, and parietal had 96% sensitivity for AHT. Specificity, however, was only 43%50.

Notably, pediatric acute care clinical tools including Pediatric Emergency Applied Care Research Network (PECARN), Canadian Assessment of Tomography for Childhood Head Injury (CATCH), and Children's Head Injury Algorithm for the Prediction of Important Clinical Events (CHALICE) were not developed specifically for patients with possible AHT. These rules rely on accurate injury histories, which is often not available in AHT cases10.

A full discussion of radiologic findings is outside the scope of this article. However, several types of injury are worth discussing here as they may be identified on evaluation and may inform your level of concern for an abusive injury.

Skull fractures are frequently seen in cases of AHT8. In acute presentations, they may be found beneath an overlying scalp hematoma. Scalp swelling may be delayed from the initial impact, or may be hidden under hair51. The most common fracture is a simple parietal fracture, which can be seen with accidental trauma or short falls and may not need additional evaluation if there is a plausible history52. Skull fractures that are complex, non-parietal, bi-parietal crossing midline, depressed, diastatic, or multiple are not always definitive for inflicted injury, though they should be assessed in the context of the provided history and other injuries present8,51,53,54,55.

The most common intracranial finding in AHT is a subdural hemorrhage (SDH)8. The mechanism is thought to be from the tearing of bridging veins due to shearing forces from acceleration-deceleration with or without an impact. Children and infants are particularly susceptible to SDH from acceleration-deceleration forces due to their relative large head size and weak cervical musculature8.

Non-acute SDH may have insidious symptoms like fontanelle bulging or increasing head circumference, and can be labeled as subdural effusions, hygromas, or hematohygromas on imaging18,56. Consultation with an experienced pediatric neuroradiologist is highly recommended for these cases.

Epidural hemorrhages (EDH) are typically related to an impact injury damaging meningeal arteries or dural venous sinuses, and they are often seen with skull fractures. EDH are more frequently seen in accidental injuries8,57. EDH may have a short “lucid interval” before symptoms appear. In children with SAH, trauma is the most common cause, though other etiologies exist including aneurysms, vascular malformations, and arterial dissections8,18. Shearing forces from acceleration-deceleration can cause brain tissue injury including cortical contusion or occasionally intraparenchymal laceration, along with SAH and SDH58.

Imaging in AHT patients may show signs of brain swelling, mass effect, and a loss of gray-white matter differentiation. In abuse, hypoxia causing brain injury can result from neurologic insult after trauma, or there could be intentional asphyxia from suffocation or strangulation, though this can be difficult to diagnose without other signs or injuries59. DAI and HIE have been shown to be correlated with poor prognosis60.

Retinal hemorrhages are present in up to 85% of cases of AHT45. Although retinal hemorrhages due to AHT range in presentation and may demonstrate any pattern, severe retinal hemorrhages, including ones that are too numerous to count, hemorrhages that occur in multiple layers of the retina, hemorrhages that extend to the ora serrata, retinal detachment, and retinoschisis are highly concerning for a traumatic injury61. Differential diagnoses for retinal hemorrhages are described in a later portion of this article.

Spinal injury may also be seen in cases of AHT, with injury identified in 23% to 67% of AHT cases that were evaluated with MRI or CT62. Ligamentous injury or a SDH are the most common spinal injuries in nonfatal AHT cases. Ligamentous injuries are likely to occur in the nuchal ligament, given the relative weakness of young childrens’ and infants’ neck musculature. Spinal SDH are often in the thoracolumbar region, which may be from blood tracking down the spine from an intracranial injury or from direct trauma to the spine. The exact mechanism of spinal SDH remains a topic of continued research. Although it is common practice to image only the cervical spine in cases of suspected AHT, newer literature advises whole spine imaging to avoid a missed injury8,63,64.

The differential diagnosis for AHT can be broken down into three categories: non-accidental trauma, accidental trauma, and an underlying medical etiology. However, underlying medical conditions mimicking AHT do not exclude patients from experiencing abuse.

As stated previously, the history obtained from the caregivers in the context of the patient's developmental abilities, as well as the identification of injuries more commonly seen in abusive trauma, will be most helpful in the differentiation of a non-accidental versus accidental injury.

Infection, bleeding diatheses, genetic diseases, or metabolic disorders can cause intracranial findings similar to AHT. Therefore labs such as CBC, cerebrospinal fluid (CSF), coagulation studies, genetic testing, and the newborn metabolic screen can point to a possible alternate etiology, depending on the clinical presentation65. Genetic conditions such as glutaric aciduria type 1, Menkes disease, Alagille syndrome, and some forms of osteogenesis imperfecta may predispose patients to SDH with minor trauma8,66,67. Enlarged extra-axial spaces in infants and young children has been proposed to predispose patients to SDH with minimal trauma, though this diagnosis should be approached with caution and considered with the context of presenting symptoms, other injuries, and subspecialist collaboration18. If suspicion is high for an alternate underlying medical condition, testing should be completed, though waiting for results should not impede other evaluation or CPS/police reports.

The differential diagnosis of retinal hemorrhages is broad and includes etiologies such as accidental trauma, birth trauma, coagulopathies, meningitis, vasculitis, endocarditis, glutaric aciduria type I, leukemia, and retinal disease. Retinal hemorrhages in these conditions seldom present with severe hemorrhages8,61,68. Increased intracranial pressure is highly uncommon and generally would not lead to extensive retinal hemorrhages69,70. An experienced ophthalmologist can help identify patterns of retinal hemorrhages that are indicative of AHT.

Children and infants with severe intracranial injuries are likely to display symptoms shortly after the inciting event, and symptoms typically persist or worsen8,9. Victims of AHT with less severe or non-acute findings, such as subdural effusions, may present with only mild and non-specific symptoms or are neurologically normal. Precise timing of injuries is extremely difficult if not impossible without an accurate history even with head imaging and neuroradiologist interpretation, particularly when injuries are non-acute71.

A discussion of the identification of AHT would be incomplete without attention given to the disparity that exists among children evaluated and reported for AHT. One study by Wood et al. found that white children with traumatic brain injury (TBI) were less likely to receive a skeletal survey than black or Hispanic children72. Additionally, those with public or no insurance were more likely to receive a skeletal survey. Another article from 2018 supports these findings, demonstrating that minority children were more likely to be both evaluated and reported for AHT than non-minority children73. These studies highlight the role of bias in the evaluation of AHT and provide evidence for the importance of standardized screening protocols, which have been shown to be effective at reducing bias74.

The medical and legal implications of a diagnosis of AHT are great and thus every effort should be made to complete a thorough evaluation with a broad differential diagnosis. It is prudent to involve a multidisciplinary team when evaluating these cases, including but not limited to, emergency medicine providers, intensivists, neurologists, trauma surgeons, neurosurgeons, ophthalmologists, neuroradiologists, geneticists, and hematologists. Social workers add invaluable experience in obtaining psychosocial histories, identifying family stressors, and facilitating difficult discussions regarding potential abuse. Additionally, hospitals may have Child Abuse Pediatrics (CAP) departments. These subspecialists provide guidance on the evaluation and diagnosis of AHT and will often serve as the primary expert witness in subsequent court cases.