Pediatric Annals

Imaging in Suspected Child Abuse: What to Expect and What to Order

Marguerite Caré, MD

Abstract

Physical child abuse (nonaccidental injury) remains a serious health risk for young children and infants in the United States. A recent report indicates that there has been an increase in both suspected and confirmed cases of child abuse and neglect during the past 5 years.1 Estimates suggest that more than 1,400 children, almost 4 per day, died of child abuse or neglect in 1999. The vast majority of these children were younger than 5 years.1

Although child abuse encompasses physical abuse, emotional and physical neglect, and sexual abuse, radiologists most often encounter cases of physical child abuse in their practices. During the last century, radiologists have played a primary role in the diagnosis and evaluation of children suffering from physical abuse. The early writings of Dr. John Caffey2 in the 1940s described a previously unknown association of long-term subdural hematomas and long bone fractures in infants. In the following decades, both he3,4 and Dr. W. Silverman5 became instrumental in describing the association of these injuries and their suspected etiology-inflicted injury.

In the past, use of the names "whiplash shaken infant syndrome,"3 "battered-child syndrome,"6 and "shaken baby syndrome" have been used to describe cases of physical abuse in infants and young children. These terms have tried to suggest the etiology or to simplify the mechanism by which these children sustained their injuries. Because of the complex ways these children are injured, including shaking, slamming, throwing, and suffocating, it is prudent today to refer to these cases as inflicted, abusive, or nonaccidental injury, terms that put aside some of the medical and legal controversy regarding whether children can sustain the observed subdural hematomas from shaking alone3 or require shaking and impact.7

CLINICAL PRESENTATION

Prior to undergoing diagnostic imaging, many of these young children present to emergency departments or to pediatricians with vague clinical symptoms. Infants who have sustained intracranial injuries as a result of abusive injury may present with seizures, lethargy or irritability, or both, misleading clinicians to consider suspected infection or sepsis. Extremity injuries may be relatively asymptomatic, often with little or no overlying bruising or swelling. Rib fractures often are found incidentally on chest radiographs obtained for wheezing or fever.

In recent years, there has been some debate over the quality of digital radiographs for the detection of these often subtle fractures. In 1998, Youmans et al.29 suggested that digital imaging may be inadequate for the diagnosis and evaluation of suspected child abuse. However, many fractures are found incidentally on routine imaging, including digital imaging. The ability to manipulate digital images by magnifying and optimizing window and leveling techniques allows enhanced evaluation for subtle fractures. Better archival systems for legal images, more availability of images for involved clinicians, and fewer number of lost films are other distinct advantages of digital imaging.

With the higher quality skeletal surveys and the use of follow-up skeletal surveys in cases of suspected abuse, routine imaging with bone scans has decreased. Bone scans may be helpful with confusing skeletal cases or circumstances where placement of the child is in question. If there is a case of clinically suspected abuse but no definite fractures are seen on skeletal survey, bone scan may provide additional information and documentation to support removal of the child from a potentially dangerous situation. Bone scans often demonstrate subtle rib fractures more easily, as well as subtle injury to the spine and scapula. Metaphyseal injuries may be well demonstrated, but occasionally are obscured by the normally increased radiotracer amount in the metaphyseal locations of growing children. Bone scans should be supplemented with skull radiographs because scans are not as sensitive as radiographs…

Physical child abuse (nonaccidental injury) remains a serious health risk for young children and infants in the United States. A recent report indicates that there has been an increase in both suspected and confirmed cases of child abuse and neglect during the past 5 years.1 Estimates suggest that more than 1,400 children, almost 4 per day, died of child abuse or neglect in 1999. The vast majority of these children were younger than 5 years.1

Although child abuse encompasses physical abuse, emotional and physical neglect, and sexual abuse, radiologists most often encounter cases of physical child abuse in their practices. During the last century, radiologists have played a primary role in the diagnosis and evaluation of children suffering from physical abuse. The early writings of Dr. John Caffey2 in the 1940s described a previously unknown association of long-term subdural hematomas and long bone fractures in infants. In the following decades, both he3,4 and Dr. W. Silverman5 became instrumental in describing the association of these injuries and their suspected etiology-inflicted injury.

In the past, use of the names "whiplash shaken infant syndrome,"3 "battered-child syndrome,"6 and "shaken baby syndrome" have been used to describe cases of physical abuse in infants and young children. These terms have tried to suggest the etiology or to simplify the mechanism by which these children sustained their injuries. Because of the complex ways these children are injured, including shaking, slamming, throwing, and suffocating, it is prudent today to refer to these cases as inflicted, abusive, or nonaccidental injury, terms that put aside some of the medical and legal controversy regarding whether children can sustain the observed subdural hematomas from shaking alone3 or require shaking and impact.7

CLINICAL PRESENTATION

Prior to undergoing diagnostic imaging, many of these young children present to emergency departments or to pediatricians with vague clinical symptoms. Infants who have sustained intracranial injuries as a result of abusive injury may present with seizures, lethargy or irritability, or both, misleading clinicians to consider suspected infection or sepsis. Extremity injuries may be relatively asymptomatic, often with little or no overlying bruising or swelling. Rib fractures often are found incidentally on chest radiographs obtained for wheezing or fever.

Fig. 1. Skeletal fractures in a 7-month-old abused infant who presented with failure to thrive. (A) Radiograph demonstrates a healing left clavicle fracture and an anterior left rib fracture. (B) Follow-up skeletal survey demonstrates additional healing posterior rib fractures that were not visible on the original examination.

Fig. 1. Skeletal fractures in a 7-month-old abused infant who presented with failure to thrive. (A) Radiograph demonstrates a healing left clavicle fracture and an anterior left rib fracture. (B) Follow-up skeletal survey demonstrates additional healing posterior rib fractures that were not visible on the original examination.

After imaging, radiologists play a primary role not only in diagnosing physical injuries but also in suggesting the diagnosis of inflicted injury to referring clinicians. Unsuspected or unexplained injuries with the lack of a traumatic history or inconsistent history are often a primary clue to the diagnosis of abusive injury. The incompatibility between the observed injuries and the reported history is often the primary consideration prompting a report of suspected physical abuse.8 Therefore, verbal communication between the radiologist and the referring clinician is essential to discuss the imaging, the reported clinical history, and physical findings.

Fig. 2. Metaphyseal fracture in a 2-month-old abused infant presenting with seizures and apnea. (A) Frontal radiograph of the right tibia demonstrates classic "bucket-handle" metaphyseal fracture. (B) Lateral radiograph of the same extremity demonstrates more prominent periosteal reaction and healing extending along the shaft. The metaphyseal injury has a more classic "comer fracture" appearance on this image.

Fig. 2. Metaphyseal fracture in a 2-month-old abused infant presenting with seizures and apnea. (A) Frontal radiograph of the right tibia demonstrates classic "bucket-handle" metaphyseal fracture. (B) Lateral radiograph of the same extremity demonstrates more prominent periosteal reaction and healing extending along the shaft. The metaphyseal injury has a more classic "comer fracture" appearance on this image.

RADIOLOGIC MANIFESTATIONS

The skeletal manifestations of abusive injury are found much more commonly in infants and young children than in older physically abused children (Fig. 1). Rib fractures, classically posterior in location; long bone injuries, specifically metaphyseal fractures (Fig. 2); and skull fractures are the most common sites of injury in abused children.9 Kleinman10 states that rib fractures and metaphyseal fractures (classic metaphyseal lesions) are highly characteristic injuries for abuse. Other fractures that are highly suspicious include scapular (Fig. 3), spinous process, and sternal injuries; however, these are infrequently encountered fractures. Multiple fractures, especially those of varying ages, are also highly suspicious. Midshaft long bone and clavicle fractures, commonly found in both accidental and abusive injury, are less specific sites of injury.10

Fig. 3. Scapular fracture in a 7-month-old abused infant who presented with failure to thrive. Frontal radiograph of the right shoulder demonstrates healing fracture of the acromion, and healing metaphyseal fracture of the proximal right humerus.

Fig. 3. Scapular fracture in a 7-month-old abused infant who presented with failure to thrive. Frontal radiograph of the right shoulder demonstrates healing fracture of the acromion, and healing metaphyseal fracture of the proximal right humerus.

Skull fractures are found in abused infants and young children but more commonly occur as the result of accidental injury.11 Fractures occurring in accidental injury, often in the household setting from a short distance fall, are usually benign neurologically and rarely result in intracranial injury.11

Linear, parietal skull fractures are the most common fractures in both accidental and abusive injury.12 Therefore, distinguishing the etiologies of the fractures is difficult. However, Hobbs12 and Meservy13 suggest that multiple and bilateral skull fractures and fractures that cross suture lines are encountered much more commonly in abusive situations than in accidental injury.

Fig. 4. Extensive intracranial injury in a 2-month-old abused infant with only mild extracranial soft tissue swelling. (A) Axial bone windows demonstrate a linear high right parietal bone fracture with mild overlying soft tissue swelling. (B) Axial CT image demonstrating extensive parenchymal infarction and edema with loss of gray-white matter differentiation, effaced basilar cisterns, trapped right temporal horn, midline shift and herniation, and acute extra-axial hemorrhage most prominent along the left tentorium. (C) Axial CT more cranialry demonstrates edema, extensive infarction, brain shift, and acute posterior interhemispheric subdural hemorrhage.

Fig. 4. Extensive intracranial injury in a 2-month-old abused infant with only mild extracranial soft tissue swelling. (A) Axial bone windows demonstrate a linear high right parietal bone fracture with mild overlying soft tissue swelling. (B) Axial CT image demonstrating extensive parenchymal infarction and edema with loss of gray-white matter differentiation, effaced basilar cisterns, trapped right temporal horn, midline shift and herniation, and acute extra-axial hemorrhage most prominent along the left tentorium. (C) Axial CT more cranialry demonstrates edema, extensive infarction, brain shift, and acute posterior interhemispheric subdural hemorrhage.

Fig. 5. Epidural hemorrhage in an 8month-old child who fell from a physician's examination room table. Axial computed tomography demonstrates a large, mixed attenuation, acute/hyperacute epidural hematoma causing mass effect and midline shift.

Fig. 5. Epidural hemorrhage in an 8month-old child who fell from a physician's examination room table. Axial computed tomography demonstrates a large, mixed attenuation, acute/hyperacute epidural hematoma causing mass effect and midline shift.

As opposed to long bone and rib injuries, skull fractures do not demonstrate the typical radiographic signs of healing, such as periosteal new bone formation or callus. Therefore, dating of skull fractures is difficult. Use of overlying soft tissue swelling is sometimes helpful (Fig. 4). However, in this young population, soft tissue swelling may be remarkably absent both clinically and on imaging, even in cases of significant injury. In some cases, scalp injury or swelling may be noted only at autopsy.7

Although skull fractures are seen more commonly in accidental injury, intracranial manifestations of trauma - including both parenchymal injury and subdural and subarachnoid hemorrhages - are encountered much more frequently in abusive head injury or significant accidental injury such as a highspeed motor vehicle collisions.11 These intracranial injuries, particularly the parenchymal injuries, lead to the greatest morbidity in infants and young children who sustain abusive head injury.

Subdural hematomas are the classic intracranial imaging finding associated with abusive head injury. Subdural hemorrhage occurs when rotational forces applied to the infant's head result in tearing of bridging arteries or veins with bleeding into the "potential" subdural space. These forces are different from the translational forces usually encountered in everyday accidental injuries (eg, short-distance falls) that result in fractures or epidural bleeding (Fig. 5).11,14

On imaging, subdural hematomas (Fig. 6) have a crescentic appearance, with a concave margin against the surface of the brain. Because the sutures do not confine them, subdural hematomas can extend along the entire surface of a cerebral hemisphere. Location of a subdural hematoma along the posterior interhemispheric fissure (Fig. 7) is a classic location for inflicted head trauma.15 However, the subdural collections may be small and located over the convexities or in the posterior fossa. Small collections may cause only minimal adjacent mass effect or may be large, leading to effaced sulci and ventricles, midline shift, and herniation of the brain. Although bilateral collections are classically associated with abuse, in the author's experience many acute subdural hematomas are unilateral and not large enough to warrant acute surgical drainage.

Fig. 6. Subdural hemorrhage in a 6-month-old with suspected abuse. (A) Axial CT demonstrates a mixed attenuation subdural collection consistent with hyperacute bleeding. The hemorrhage has a cresentic shape and is not confined by the sutures. Extra-axial hemorrhage is also seen along the interhemispheric fissure. (B) More cranially, extra-axial hemorrhage is seen adjacent to the right cerebral convexity. This blood is likely both subarachnoid and subdural in location. More prominent acute, posterior interhemispheric subdural hemorrhage is present.

Fig. 6. Subdural hemorrhage in a 6-month-old with suspected abuse. (A) Axial CT demonstrates a mixed attenuation subdural collection consistent with hyperacute bleeding. The hemorrhage has a cresentic shape and is not confined by the sutures. Extra-axial hemorrhage is also seen along the interhemispheric fissure. (B) More cranially, extra-axial hemorrhage is seen adjacent to the right cerebral convexity. This blood is likely both subarachnoid and subdural in location. More prominent acute, posterior interhemispheric subdural hemorrhage is present.

Fig. 7. Acute interhemispheric subdural hemorrhage in a 2-monthold abused infant. Axial CT image demonstrates thin band of high density extending along the right occipital lobe and extending into the posterior interhemispheric fissure. Minimal extra-axial, likely subarachnoid hemorrhage is also seen adjacent to the right frontal lobe.

Fig. 7. Acute interhemispheric subdural hemorrhage in a 2-monthold abused infant. Axial CT image demonstrates thin band of high density extending along the right occipital lobe and extending into the posterior interhemispheric fissure. Minimal extra-axial, likely subarachnoid hemorrhage is also seen adjacent to the right frontal lobe.

Acute subdural hematomas are bright or hyperdense on computed tomography (CT) scanning. Hyperacute hematomas may be mixed in attenuation because of active bleeding, unclotted blood, or blood mixing with cerebrospinal fluid. As the hematoma ages, usually over 7 to 14 days, it gradually becomes less dense. In the past, aging of hematomas has been attempted by noting the appearance on both magnetic resonance imaging (MRI) and CT. However, many variables exist in the aging of these collections, leading to variable signal on MRI1617 and differences in density on CT. Unless a subdural collection or other hematoma is observed from its hyperdense stage on CT, the hematoma cannot be reliably dated. This has important legal implications in the setting of child abuse, where estimates of the timing of injuries is important. An MRI often provides no additional information as to the age of a hematoma but may demonstrate subtle subdural or subarachnoid clot that is not visible on CT (Fig. 8).

Fig. 8. CT versus MRI for the detection of extraaxial hemorrhage. (A) Initial CT image demonstrates no appreciable extraaxial hemorrhage. (B) Axial MPCR image obtained 2 days later demonstrates scattered foci of marked hypointense signal consistent with hemorrhage or clot in the high interhemispheric fissure and adjacent to the high posterior frontal and parietal lobes. Blood is likely both in the subdural and subarachnoid space. (C) Axial proton density-weighted image shows bilateral subdural collections that are hyperintense to cerebrospinal fluid suggesting proteinaceous fluid.

Fig. 8. CT versus MRI for the detection of extraaxial hemorrhage. (A) Initial CT image demonstrates no appreciable extraaxial hemorrhage. (B) Axial MPCR image obtained 2 days later demonstrates scattered foci of marked hypointense signal consistent with hemorrhage or clot in the high interhemispheric fissure and adjacent to the high posterior frontal and parietal lobes. Blood is likely both in the subdural and subarachnoid space. (C) Axial proton density-weighted image shows bilateral subdural collections that are hyperintense to cerebrospinal fluid suggesting proteinaceous fluid.

Benign macrocrania is an entity that potentially may be confused with low attenuation subdural collections. This clinical entity usually presents at about 6 months, resolves at 24 to 36 months, and has imaging findings of enlarged subarachnoid, but not subdural spaces (Fig 9, page 657). The ventricles may be mildly enlarged. The children are developmentally normal, and usually present with an enlarged head circumference, a finding that may have been present in their parents as infants.

Like subdural bleeding, subarachnoid blood is rarely found in cases of accidental injury, but is a common finding in inflicted head trauma (Fig. 10, page 657).11 The subarachnoid space lies immediately adjacent to the brain on imaging, and extends into the cortical sulci. Often, small foci of subarachnoid clot may be observed, particularly over the cerebral convexities. Here, MRI may be helpful, because of its multiplanar capabilities. Acute subarachnoid clot is hyperdense on CT and markedly hypointense on gradient echo sequences on MRI. One advantage of demonstrating acute subarachnoid blood is that it suggests acute injury, as rebleeding into chronic subdural collections does not involve the subarachnoid space.

Uncommonly, epidural hematomas may be present in cases of abusive head trauma, but they more often occur as a result of accidental injury.11

Fig. 9. MRI in a child with probable benign macrocrania and a subdural collection. Child presented because of a reported stairway injury. Abuse was suspected, but not confirmed. Axial T2weighted image demonstrates enlarged right subarachnoid space likely related to benign macrocrania. Vessels course through this space, and the fluid extends into the cortical sulci. On the left, an isointense to cerebrospinal fluid subdural collection displaces the vessels against the surface of the brain.

Fig. 9. MRI in a child with probable benign macrocrania and a subdural collection. Child presented because of a reported stairway injury. Abuse was suspected, but not confirmed. Axial T2weighted image demonstrates enlarged right subarachnoid space likely related to benign macrocrania. Vessels course through this space, and the fluid extends into the cortical sulci. On the left, an isointense to cerebrospinal fluid subdural collection displaces the vessels against the surface of the brain.

Although subdural hematomas are the classic intracranial imaging finding associated with child abuse, it is the cerebral parenchymal injuries that lead to the greatest morbidity. Unlike the extra-axial hematomas and collections that may be relatively asymptomatic or present with delayed mental status changes, children who sustain parenchymal brain injury often develop acute neurologic changes. This is especially true in children sustaining hypoxic /ischemic or hypoperfusion injuries, diffuse edema, and axonal injury. The primary parenchymal injuries include cortical contusions, axonal injury, cerebral edema, and hypoxic-ischemic injury.

Fig. 10. Subarachnoid hemorrhage in a 3-month-old abused infant who initially presented with questionable seizures. (A) and (B) Axial CT images demonstrate high attenuation acute subdural hemorrhage adjacent to the high interhemispheric fissure, as well as acute subarachnoid blood that conforms to the surface of the brain, extending into the cortical sulci of the high left parietal lobe. Low attenuation subdural collections are also present.

Fig. 10. Subarachnoid hemorrhage in a 3-month-old abused infant who initially presented with questionable seizures. (A) and (B) Axial CT images demonstrate high attenuation acute subdural hemorrhage adjacent to the high interhemispheric fissure, as well as acute subarachnoid blood that conforms to the surface of the brain, extending into the cortical sulci of the high left parietal lobe. Low attenuation subdural collections are also present.

Contusions are commonly observed in adults sustaining head injury18 but are less commonly seen in the pediatric population.1119 On imaging, contusions may be hemorrhagic or bland (Fig. 11), involve the cortex, and are often located in the frontal and temporal lobes. Contusions often bloom or increase in size during the first several days, leading to delayed neurologic compromise. Therefore, routine follow-up imaging is suggested after the first 24 to 48 hours, or if acute neurologic changes occur. Contusions can be as large as several centimeters and may lead to effacement of the adjacent cortical sulci, ventricles, or cisterns.

Fig. 11. Cortical contusions in a 2-1/2-year-old-victim of physical abuse. Axial CT image with heterogeneous contusions involving the right temporal lobe and cerebellar vermis. The image also shows obstructive hydrocephalus with dilation of the temporal horns related to progressing mass effect, as well as diastatic right occipital bone fracture and extracranial swelling.

Fig. 11. Cortical contusions in a 2-1/2-year-old-victim of physical abuse. Axial CT image with heterogeneous contusions involving the right temporal lobe and cerebellar vermis. The image also shows obstructive hydrocephalus with dilation of the temporal horns related to progressing mass effect, as well as diastatic right occipital bone fracture and extracranial swelling.

Traumatic axonal injury, frequent in patients with severe head trauma, is found only infrequently in cases of abusive head injury.20 The axonal injury found in the latter population at autopsy is probably the result of hypoxic /ischemic injury.20 However, when present, this type of injury leads to severe and permanent brain injury. On imaging, axonal injury often appears as small foci of attenuation or signal abnormality at the gray-white matter interface, brainstem, and corpus callosum. The lesions may be hemorrhagic or bland. Magnetic resonance imaging has been shown to demonstrate these lesions more readily than CT.18,21

Edema or swelling is a frequent complication of pediatric head injury.22,23 On imaging, edema results in loss of gray and white matter differentiation and effacement of adjacent cortical sulci and ventricles. It may lead to brain shift and herniation with resulting infarction of the affected brain (Fig. 12). On CT, abnormal low attenuation can be observed within hours of the insult. On MRI, because of the normally hyperintense signal in the infant's white matter, edema may be more difficult to observe. However, T2-weighted images normally demonstrate striking contrast between the hypointense cortex and hyperintense, unmyelinated white matter. Therefore, when this distinction is obscured, edema is often demonstrated readily.

Hypoxic/ischemic injury was recently reported to be present in 27% of infants admitted for suspected inflicted head injury (Fig. 13).24 With the exception of direct vascular compromise from herniation, the etiology of the diffuse pattern of hypoxic/ischemic injury observed in this population is not known. It is suspected to result from axonal injury in the brainstem and cervical cord that results from stretching of these tissues during inflicted head trauma.20 Although not unique to abusive injury, these children often demonstrate a pattern of diffuse hypoperfusion with loss of gray and white matter differentiation, but preservation of the size of the ventricular system. As with traumatic axonal injury, the hypoxic /ischemic injury results in devastating and often permanent brain injury.24

Fig. 12. Hyperacute subdural hematoma and massive cerebral edema in a 1-year-old abused infant. Axial CT image demonstrates heterogeneous, hyperacute subdural hematoma adjacent to the left cerebral hemisphere. There is mass effect with midline shift and effaced basal cisterns. Poor gray-white matter differentiation from edema, predominantly in the frontal and left temporal lobes, contributes to the mass effect and markedly elevated intracranial pressure.

Fig. 12. Hyperacute subdural hematoma and massive cerebral edema in a 1-year-old abused infant. Axial CT image demonstrates heterogeneous, hyperacute subdural hematoma adjacent to the left cerebral hemisphere. There is mass effect with midline shift and effaced basal cisterns. Poor gray-white matter differentiation from edema, predominantly in the frontal and left temporal lobes, contributes to the mass effect and markedly elevated intracranial pressure.

Fig. 13. Global hypoxic/ischemic injury in a 4-month-old abused infant who initially presented with irritability and tense fontanel. (A) Axial CT image from the initial examination demonstrates diffuse loss of gray-white matter differentiation throughout both cerebral hemispheres. The ventricular and extra-axial spaces are preserved. The child also has bilateral low attenuation subdural collections, as well as scattered acute subdural blood. (B) Follow-up examination obtained 2 weeks later demonstrates prominent parenchymal volume loss with more localized encephalomalacia in the left frontal and both parietaloccipital lobes. The subdural collections have increased in size, in part related to extensive volume loss of the brain.

Fig. 13. Global hypoxic/ischemic injury in a 4-month-old abused infant who initially presented with irritability and tense fontanel. (A) Axial CT image from the initial examination demonstrates diffuse loss of gray-white matter differentiation throughout both cerebral hemispheres. The ventricular and extra-axial spaces are preserved. The child also has bilateral low attenuation subdural collections, as well as scattered acute subdural blood. (B) Follow-up examination obtained 2 weeks later demonstrates prominent parenchymal volume loss with more localized encephalomalacia in the left frontal and both parietaloccipital lobes. The subdural collections have increased in size, in part related to extensive volume loss of the brain.

SEQUELAE OF HEAD INJURY

Children sustaining head trauma as the result of abuse may develop parenchymal volume loss, chronic subdural hematomas or collections, and communicating hydrocephalus. On imaging, these entities can have similar appearances.

Chronic subdural hematomas are often bilateral and low in attenuation. However, they often remain slightly hyperdense to cerebrospinal fluid in the adjacent subarachnoid space and ventricles. When large, the collections may cause some localized mass effect. The collections often contain septations that may subdivide the subdural into compartments of different signal intensity or attenuation. Rebleeding25 into these long-term collections may occur with little or no additional trauma (Fig. 14). However, this rebleeding should not result in the acute deterioration of these children, result in acute parenchymal injury, or result in the presence of acute subarachnoid bleeding.

Fig. 14. Initial injury and subsequent rebleeding into a chronic subdural hematoma in a 5-monthokd abused infant. (A) Axial CT image at the time of initial presentation demonstrates diffuse loss of gray-white matter differentiation consistent with diffuse hypoxic/ischemic injury. There are low attenuation bilateral subdural collections with acute, hyperdense blood within them. (B) CT image obtained 2 months later shows extensive parenchymal volume loss with large, complex bilateral subdural collections. (C) CT obtained for episodes of vomiting 4 months after the initial insult shows an increase in the heterogeneity and complexity of the subdural hemorrhages with some increased attenuation suggesting rebleeding. The large collections do exert some mass effect on the adjacent brain. As in (B), there is extensive parenchymal volume loss related to the initial hypoxic/ischemic insult.

Fig. 14. Initial injury and subsequent rebleeding into a chronic subdural hematoma in a 5-monthokd abused infant. (A) Axial CT image at the time of initial presentation demonstrates diffuse loss of gray-white matter differentiation consistent with diffuse hypoxic/ischemic injury. There are low attenuation bilateral subdural collections with acute, hyperdense blood within them. (B) CT image obtained 2 months later shows extensive parenchymal volume loss with large, complex bilateral subdural collections. (C) CT obtained for episodes of vomiting 4 months after the initial insult shows an increase in the heterogeneity and complexity of the subdural hemorrhages with some increased attenuation suggesting rebleeding. The large collections do exert some mass effect on the adjacent brain. As in (B), there is extensive parenchymal volume loss related to the initial hypoxic/ischemic insult.

Children with long-term subdurals often have an enlarged head circumference. The observed macrocrania may be the reason for the initial imaging referral. In children with hypodense subdural collections, the etiology is unclear, although prior trauma may be suspected. A diagnosis of child abuse in these cases rests on the demonstration of unsuspected skeletal injury or retinal hemorrhages,26 and on a thorough social history and investigation.

Communicating hydrocephalus may also develop following inflicted injury. Children with communicating hydrocephalus have developmental delay and often an enlarged head circumference.

However, confusion may occur if the child has parenchymal volume loss as a result of prior injury. Infants with volume loss usually have a small head circumference, unless coexistent subdural collections or communicating hydrocephalus is present. The volume loss may be localized, with areas of scarring or gliosis related to contusions or localized injury or infarct. However, global ischemic insults and diffuse axonal injury may result in diffuse parenchymal loss.

IMAGING STUDIES

As noted above, many children with suspected abuse actually present with unrelated symptoms. Therefore, the astute radiologist must always examine radiographs for unsuspected skeletal injuries that may be present on chest or abdominal radiographs obtained for routine clinical symptoms. When a child is suspected to have inflicted injuries, a skeletal survey should be obtained.

Currently, the American College of Radiology recommends that a skeletal survey be obtained in all children aged 2 years or younger with suspected abuse.27 The skeletal survey should consist of anteroposterior and lateral images of the axial skeleton and frontal projections of each extremity.27 The examination currently includes 19 separate exposures with dedicated images of each anatomic region (Table, below).27 Definite or possible fractures should be imaged with additional oblique or tangential views. Follow-up skeletal surveys, consisting of a series of similar radiographs without skull images, may be obtained in 10 to 14 days to define possible fractures and to identify additional locations of injury.28

Table

Complete Skeletal Survey

Complete Skeletal Survey

In recent years, there has been some debate over the quality of digital radiographs for the detection of these often subtle fractures. In 1998, Youmans et al.29 suggested that digital imaging may be inadequate for the diagnosis and evaluation of suspected child abuse. However, many fractures are found incidentally on routine imaging, including digital imaging. The ability to manipulate digital images by magnifying and optimizing window and leveling techniques allows enhanced evaluation for subtle fractures. Better archival systems for legal images, more availability of images for involved clinicians, and fewer number of lost films are other distinct advantages of digital imaging.

With the higher quality skeletal surveys and the use of follow-up skeletal surveys in cases of suspected abuse, routine imaging with bone scans has decreased. Bone scans may be helpful with confusing skeletal cases or circumstances where placement of the child is in question. If there is a case of clinically suspected abuse but no definite fractures are seen on skeletal survey, bone scan may provide additional information and documentation to support removal of the child from a potentially dangerous situation. Bone scans often demonstrate subtle rib fractures more easily, as well as subtle injury to the spine and scapula. Metaphyseal injuries may be well demonstrated, but occasionally are obscured by the normally increased radiotracer amount in the metaphyseal locations of growing children. Bone scans should be supplemented with skull radiographs because scans are not as sensitive as radiographs in demonstrating skull fractures. One distinct disadvantage of bone scan as the primary imaging modality is the possible need for sedation in older infants and young children.

In children and infants who present with reported minor head trauma, the role of skull radiographs is often confusing. In the pediatric population, skull radiographs are ordered much more frequently for the evaluation of skull fractures than in adults. A recent article by Schutzman et al.30 regarding young children who presented with minor head trauma suggests that skull fractures are a strong indicator of intracranial injury, especially in the younger infants. Unfortunately, skull radiographs do not provide any detail of the intracranial structures.

So when should skull radiographs be obtained? In infants and young children with reported minor head trauma and with some signs of bruising or swelling on physical examination, skull radiographs may be indicated to evaluate for a fracture, even when the child has a normal neurologic examination. However, if the child has an altered mental status or the clinical examination is confusing, CT is more prudent. As suggested in the Schutzman article30 and at the author's institution, a lower threshold for ordering CT exists in young infants in whom the neurologic examination is less reliable. One distinct advantage of skull radiographs over CT is better demonstration of fractures that lie within the axial CT plane.

Computed tomography remains the primary advanced imaging modality for children with suspected physical abuse. Computed tomography readily demonstrates scalp swelling that may not be present clinically, complex or depressed fractures, and, more important, acute and long-term injury to the underlying brain. Computed tomography in the acute setting is excellent for evaluating the intracranial contents for extra-axial hemorrhage and parenchymal injury that will alter or direct acute medical or surgical intervention. Currently, on a conventional scanner, head CT images can be obtained in a matter of seconds, with room time limited to just minutes. Because of rapid imaging times, routine use of sedation is often not necessary. Computed tomography also is the primary imaging modality to evaluate possible coexistent chest and abdominal injury.

The American College of Radiology31 suggests that a head CT should be obtained (in addition to a skeletal survey) in children with suspected abuse and an altered mental status. As noted above, more liberal use of CT should be considered for young infants in whom the neurologic examination is more difficult.

Magnetic resonance imaging, which may provide better delineation of small extra-axial hemorrhagic collections or subtle injury to the brain, is indicated in cases of suspected abusive head injury when the CT is normal or equivocal.31 It should also be considered when the neurologic findings are more severe than or disproportionate to the CT imaging findings. The advantages of MRI include multiplanar imaging, often enhanced demonstration of small extra-axial collections and parenchymal injury, and increased visual clarity and detail.21,32 The disadvantages of MRI compared to CT include longer scanning times, often requiring the need for sedation; lower availability than CT, especially in the acute setting; and difficulty in monitoring acutely injured patients. The author finds that MRI often provides little additional information when the CT clearly indicates head injury and may even lead to some confusion in cases of parenchymal injury where white matter edema may be more difficult to visualize than on CT. If performed, a delay of several days often marks more apparent parenchymal abnormalities.33 Although multiple advanced MR techniques are available including spectroscopy and diffusionweighted imaging, the conventional images often define the parenchymal defects clearly before MRI can be performed.

REFERENCES

1. Peddle N, Wang CT. Current Trends in Child Abuse Prevention, Reporting, and Fatalities: The 1999 Fifty State Survey. Chicago, IL: Prevent Child Abuse America; 2001.

2. Caffey J. Multiple fractures in the long bones of infants suffering from chronic subdural hematoma. Journal. 1946;56:163-173.

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

4. Caffey J. On the theory and practice of shaking infants. Its potential residual effects of permanent brain damage and mental retardation. Am J Dis Child. 1972;124:161-169.

5. Silverman W. The roentgen manifestations of unrecognized skeletal trauma in infants. AJR Am J Roentgenol. 1953,-volume:413-427.

6. Kempe CH, Silverman FN, Steele BF, Droegemueller W, Silver HK. Landmark article July 7, 1962: The battered-child syndrome. By C. Henry Kempe, Frederic N. Silverman, Brandt F. Steele, William Droegemueller, and Henry K. Silver. JAMA. 1984;251:3288-3294.

7. Duhaime AC, Gennarelli TA, Thibault LE, Bruce DA, Margulies SS, Wiser R. The shaken baby syndrome. A clinical, pathological, and biomechanical study. J Neurosurg. 1987;66:409-415.

8. Solomons G. Trauma and child abuse. The importance of the medical record. Am J Dis Child. 1980;134:503-505.

9. Kleinman PK, Marks SC Jr, Richmond JM Blackbourne BD. Inflicted skeletal injury: a postmortem radiologichistopathologic study in 31 infants. AJR Am J Roentgenol. 1995;165:647-650.

10. Kleinman PK. Postmortem imaging. In: The Diagnostic Imaging of Child Abuse. 2nd ed. St. Louis, MO: Mosby; 1998:8-10.

11. 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(2 Pt 1):179-185.

12. Hobbs CJ. ABC of child abuse. Head injuries. BMJ. 1989;298:1169-1170.

13. Meservy Q, Towbin R, McLaurin RL, Myers PA, Ball W. Radiographic characteristics of skull fractures resulting from child abuse. AJR Am J Roentgenol. 1987;149:173-175.

14. Ommaya AK, Gennarelli TA. Cerebral concussion and traumatic unconsciousness. Correlation of experimental and clinical observations of blunt head injuries. Brain. 1974;97:633-54.

15. Zimmerman RA, Bilaniuk LT, Bruce D, Schut L, Uzzell B, Goldberg HI. Computed tomography of craniocerebral injury in the abused child. Radiology. 1979;130:687-690.

16. Bradley WG Jr. MR appearance of hemorrhage in the brain. Radiology. 1993;189:15-26.

17. Gomori JM, Grossman RI, Yu-Ip C, Asakura T. NMR relaxation times of blood: dependence on field strength, oxidation state, and cell integrity. J Comput Assist Tomogr. 1987;11:684-690.

18. Gentry LR, Godersky JC, Thompson B, Dunn VD. Prospective comparative study of intermediate-field MR and CT in the evaluation of closed head trauma. AJR Am J Roentgenol. 1988;150:673-682.

19. Merten DF, Osborne DR, Radkowski MA, Leónidas JC. Craniocerebral trauma in the child abuse syndrome: radiological observations. Pediatr Radiol. 1984;14:272-277.

20. Geddes JF, Vowles GH, Hackshaw AK Nickols CD, Scott IS, Whitwell HL. Neuropathology of inflicted head injury in children. II. Microscopic brain injury in infants. Brain. 2001;124(Pt 7):1299-1306.

21. Sato Y, Yuh WT, Smith WL, Alexander RC, Kao SC, Ellerbroek Q. Head injury in child abuse: evaluation with MR imaging. Radiology. 1989;173:653-657.

22. Zimmerman RA, Bilaniuk LT, Bruce D, Dolinskas C, Obrist W, Kuhl D. Computed tomography of pediatric head trauma: acute general cerebral swelling. Radiology. 1978;126:403-408.

23. Bruce DA, Alavi A, Bilaniuk L, Dolinskas C, Obrist W, Uzzell B. Diffuse cerebral swelling following head injuries in children: the syndrome of "malignant brain edema." J Neurosurg. 1981;54:170-178.

24. Gilles EE, Nelson MD Jr. Cerebral complications of nonaccidental head injury in childhood. Pediatr Neurol. 1998;19:119-128.

25. McLone DG. Ultrastructure of subdural membranes of children. American Society for Pediatric Neurosurgery. 1981:174-187.

26. Duhaime AC, Christian CW, Rorke LB, Zimmerman RA. Nonaccidental head injury in infants - the "shaken-baby syndrome". N Engl J Med. 1998;338:1822-1829.

27. American College of Radiology. American College of Radiology standard for skeletal surveys in children. ACR Standards. Reston, VA: American College of Radiology; 1997:1-4.

28. Kleinman PK Nimkin K, Spevak MR, et al. Follow-up skeletal surveys in suspected child abuse. AJR Am J Roentgenol. 1996;167: 893-8%.

29. Youmans DC, Don S, Hildebolt C, et al. Skeletal surveys for child abuse: comparison of interpretation using digitized images and screen-film radiographs. AJR Am J Roentgenol. 1998;171:1415-1419.

30. Schutzman 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.

31. Slovis, TL, Smith W, Kushner DC, et al. Imaging the child with suspected physical abuse. American College of Radiology. ACR Appropriateness Criteria. Radiology. 2000;215(SuppI):805-809.

32. Alexander RC, Schor DP, and Smith WL Jr. Magnetic resonance imaging of intracranial injuries from child abuse. J Pediatr. 1986;109:975-979.

33. Nimkin K, Kleinman PK. Imaging of child abuse. Pediatr Clin North Am. 1997;44:615-635.

Complete Skeletal Survey

10.3928/0090-4481-20021001-09

Sign up to receive

Journal E-contents