Child Abuse (Radiology)

Background

Among the first to identify issues related to child abuse, Dr. John Caffey studied and published results in 1946 detailing multiple unexplained long-bone fractures of apparent traumatic origin in infants with chronic subdural hematomas. His work is cited in the landmark 1962 article, “The Battered-Child Syndrome,” by Dr. C. Henry Kempe et al. Kempe and colleagues (1962) stated that “the physician’s duty and responsibility to the child requires a full evaluation of the problem and a guarantee that the expected repetition of trauma will not be permitted to occur.” Within several years of the publication of Kempe et al.’s article, every state in the United States mandated that medical professionals report all suspected cases of child abuse. In 1974, the Child Abuse Prevention and Treatment Act (CAPTA) was signed into law.

CAPTA has been amended several times, most recently by the CAPTA Reauthorization Act of 2019. CAPTA defines child abuse and neglect as “any recent act or failure to act on the part of a parent or caretaker which results in death, serious physical or emotional harm, sexual abuse or exploitation; or an act or failure to act, which presents an imminent risk of serious harm.” Child maltreatment encompasses neglect, physical abuse, emotional abuse, and sexual abuse.

In 2019, an estimated 1840 children died as the result of abuse and neglect in the United States. Based on this estimate and population data, the national rate of child fatalities due to abuse and neglect was 2.5 deaths per 100,000 children. The youngest children are at greatest risk: 7% of all child maltreatment fatalities involved children younger than 3 years.

Based on US Child Protective Services (CPS) data, there were 674,000 victims of child abuse and neglect in 2017, equating to a rate of 9 victims per 1000 children in the US population. This number includes only those cases reported to and substantiated by CPS. Many additional cases go unreported. Children in their first year of life are at greatest risk, with a rate of victimization of 25.3 per 1000 children. Of reported cases of child maltreatment, 74.5% involved cases of neglect. Child physical abuse accounted for 18.3% of reports.

A majority of child physical abuse involves cutaneous injuries that never come to imaging. These injuries include bruises, bite marks, and burns. There are patterns of cutaneous injuries, which pediatricians and emergency medicine providers are trained to recognize as suggestive of physical abuse. Abusive head trauma, thoracic and abdominal organ injury, and orthopedic injuries with their associated soft tissue injures are in the realm of imaging evaluation of child maltreatment. Imaging plays an important role in the diagnosis and follow-up care of these injuries, and it is our duty and responsibility to assist in diagnosing child physical abuse. In a battered child, a missed fracture or an injury misclassified as accidental exposes a child to further abuse and potentially death. This chapter details thoracic, abdominal, and orthopedic injuries of physical abuse. Abusive head trauma is addressed in Chapter 20 .

Fractures Associated With Child Abuse and Fracture Dating

Any fracture can be the result of abuse, but some fractures are found almost exclusively in abused children. These fractures are known to have a higher specificity for abuse than others and when seen by the radiologist should raise higher suspicion for child abuse in infants and toddlers ( Table 21.1 ). The classic metaphyseal lesion (CML) is the most specific type of fracture one can see with child abuse up to the age of 1 year. This injury is caused by shearing, traction, or twisting ( Fig. 21.1 ). Rib fractures (specifically posterior) have high specificity for child abuse, especially in infants ( Fig. 21.2 ). Other highly specific types of injury that are less commonly seen are scapular fractures, spinous process fractures, and sternal fractures.

Multiple fractures or fractures of different ages are moderately specific for child abuse. Other fractures that have moderate specificity include epiphyseal separations, vertebral body fractures, digital fractures in infants, and complex skull fractures. In these cases the radiologist should always take into account the clinical situation and the patient’s age when considering whether to raise concern for child abuse. One example of inappropriate history may be a clinical history that is not commensurate with the patient’s developmental stage, such as a 1-month-old who rolled off of a changing table and presents with multiple injuries.

Although common, the fractures with low specificity for child abuse include long-bone fractures, linear skull fractures, clavicle fractures, and isolated subperiosteal new bone formation; however, keep in mind that abused children may have any or all of these fractures. The femur, humerus, and tibia are the most commonly injured long bones in child abuse. Again, the child’s age is important in determining whether to raise concern for child abuse. For example, a femoral fracture in a nonambulatory child is more suspicious for abuse than in a child who is ambulatory, where the fracture is more likely to be accidental. Fractures of the humeral shaft in a child younger than 18 months is suspicious for child abuse compared with a supracondylar fracture in an ambulatory child. Linear skull fractures are usually not inflicted as opposed to complex or bilateral skull fractures that are more commonly associated with child abuse.

The important thing to remember is that although some fractures may be more specific for child abuse, we must always think about the possibility of child abuse, especially in children younger than 2 years.

Dating the age of fractures is a common request to the radiologist, and care should be used when attempting to determine the approximate age of fractures. There are phases of fracture healing that can be identified on plain radiographs. These include soft tissue swelling, periosteal reaction, soft callus, hard callus, bridging, and remodeling ( Fig. 21.3 ). Even with these stages of fracture healing, it is still difficult to date a fracture exactly, although the fractures can be lumped into relatively broad time frames. Soft tissue swelling is most associated with acute fractures that are less than 1 week old. Periosteal reaction and soft callus with increasing hard callus and bridging are associated with recent fractures that are 8 to 35 days old. Old fractures (>36 days old) show a combination of periosteal reaction, hard callus, bridging, and remodeling. The more important task for the radiologist may be to determine whether fractures are of different ages rather than to specifically date the fractures. Multiple fractures of different ages indicate a pattern of repetitive trauma that places the child at greater risk for additional injuries and possibly even death.

Soft Tissue Injuries in Child Abuse

There are no specific imaging characteristics of soft tissue injury caused by child abuse. Unlike musculoskeletal and neurological findings pathognomonic for child abuse, traumatic radiographic findings of the soft tissues caused by motor vehicle accident or from a child falling on handlebars while bicycling can be identical to intentional injury by a parent, caregiver, or sibling. In addition, children can present with a combination of soft tissue injuries secondary to accidental and nonaccidental mechanisms, confusing the picture. Due to these facts, combined with the knowledge that the radiologist may be the first physician exposed to soft tissue injury caused by child abuse and that we are mandated reporters for intentional trauma, extensive research has been done to help educate radiologists in findings of accidental versus nonaccidental trauma.

Infants less than 1 year of age have the greatest incidence of soft tissue injury caused by child abuse. If a child is younger than 5 years with traumatic soft tissue injury and if the mechanism of injury given by the parent/caregiver does not match the diagnostic findings, abuse must be considered. Research has found accidental trauma causing soft tissue injuries is most often found in older children 7.6 to 10.3 years old. Child abuse that causes soft tissue injuries occurs in younger children, with the average age being 2.5 to 3.7 years. The most common soft tissue injuries associated with child abuse are liver injury, hollow visceral injury (predominately involving the distal duodenum and proximal jejunum), and pancreatic injury. In those scenarios, if there is no reported history of motor vehicle accident, fall onto handle bars while biking, or appropriate high-force blunt trauma history, child abuse should be considered. Suspicious histories that would not be concordant with those injuries include “falling out of bed” or rolling onto a sleeping infant.

As stated, most common injuries are to the liver, duodenum/proximal jejunum, and pancreas ( Figs. 21.4–21.7 ). These injuries are due to lack of coverage by the ribs and fixation of these structures in the epigastric region of the abdomen (e.g., the ligament of Treitz fixates the proximal jejunum). Liver injury associated with child abuse more commonly involves the left lobe than accidental injury; however, abusive or accidental injury can occur in either left or right hepatic lobes. Children who are younger than 5 years and who present with a duodenal injury are most likely to have sustained that injury from abuse with complete transection/perforation being rare and duodenal hematomas causing proximal partial obstruction being the most common presentation (see Fig. 21.7 ). Bowel transections, unless caused by seat belt or bike handlebar injuries, are considered abusive in etiology unless proven otherwise. One-third of pancreatic injuries are nonaccidental, and some researchers claim that if the injury was not caused by motor vehicle accident or bicycle handle bar, then again the etiology is considered to be abuse unless proven otherwise.

Hypoperfusion complex, an entity associated with intracranial injury and severe neurological impairment and/or severe bleeding, can be a consequence of nonaccidental trauma ( Fig. 21.8 ). In hypoperfusion complex, also known as shock abdomen, imaging findings are secondary to severe hypovolemia. Computed tomography (CT) findings include abnormal intense enhancement of the bowel wall, adrenal glands, kidneys, liver, and/or pancreas; decreased caliber of the inferior vena cava and aorta; bowel thickening; and/or dilatation. CT findings may precede clinical ones, making familiarity with this entity critical because the radiologist may be the first person to become aware of a child’s tenuous volume status.

Rarer injuries, involving other organs, have been reported in the literature and include pharyngeal injuries and esophageal rupture (i.e., forced feeding or entry of foreign bodies), lung laceration/contusions associated with rib fractures (although these are uncommon findings in child abuse in comparison with accidental trauma because of the different mechanism causing the rib fractures), chylothorax (associated with multiple bone fractures), cardiac lacerations, commotio cordis, chylous ascites, portal venous gas, aortic pseudoaneurysms, adrenal injury (usually right-sided and accompanied by ipsilateral rib fractures and/or visceral injuries), and stomach, colon, and bladder rupture.

Infants, children, and adolescents who present late after sustaining nonaccidental type injuries can present with pseudocysts of the pancreas secondary to prior pancreatic contusion or partial tear, multiple osteolytic lesions in various bones caused by prior pancreatic injury and fat embolism, or focal posttraumatic bowel strictures, again typically seen involving the junction of the duodenum and jejunum, secondary to healed partial mural bowel tears.

Contrast-enhanced CT of the abdomen and pelvis is the gold standard whenever soft tissue injury caused by child abuse is suspected. No oral contrast is needed in the emergent setting, and a routine single-phase abdomen and pelvis protocol will suffice. The alliance for radiation safety for pediatric imaging dictates a low-dose radiation protocol for all pediatric patients. If a delayed phase must be acquired to rule out injury to the renal collecting system, consideration to decrease radiation is recommended. Focused assessment with sonography is not recommended because of its limited sensitivity of 50%. Although focused assessment with sonography is easily performed and widely available, positive findings require CT for further investigation to exclude more subtle injuries, while negative findings do not rule out injury. A routine complete abdominal ultrasound is suggested in patients younger than 2 years who are suspected of being abused but with low probability of internal soft tissue injury. For equivocal findings of bowel injury on CT or ultrasound, upper gastrointestinal series can help exclude intramural hematomas involving bowel caused by partial thickness tears and subserosal bleeding. Radiographs, often obtained on initial assessment, are often used to assess for free air or bowel obstruction (the latter can be associated with duodenal or other intramural hematoma of bowel). However, it is important to remember that free air from perforation of retroperitoneal structures (e.g., duodenum) may not show up on radiographs. Significant traumatic injury (i.e., multivisceral) without appropriate history should heighten the radiologist’s suspicion.

American College of Radiology Criteria

The American College of Radiology Appropriateness Criteria are helpful in guiding clinicians and radiologists as to what imaging tests are appropriate given the history and physical examination. Sometimes the diagnosis of child abuse is clear based on the history and physical examination. In these cases, imaging may be used for documentation and for legal purposes. However, some cases are less straightforward, and imaging can be used for detection of injuries. Radiologists must always consider the possibility of metabolic and genetic conditions when interpreting these studies, as described later in this chapter, although children with metabolic and genetic conditions are not exempt from abuse.

A radiographic skeletal survey is the first imaging test ordered for detection of fractures in a child younger than 2 years where there is a suspicion of possible child abuse. The skeletal survey should include frontal and lateral views of the skull, lateral views of the cervical and thoracolumbar spine, and single frontal views of all of the long bones separately using high-resolution including single frontal images of the hands and feet as well as single views of the chest and abdomen. Combining body parts on one image is not acceptable, as the large field of view and lack of attention/centering on each bone may reduce sensitivity. Oblique views of the ribs are also helpful in diagnosing rib fractures that are a strong indication of child abuse (see Fig. 21.2B ). Skeletal surveys are less commonly performed in children older than 2 years; however, they are performed when the clinical findings suggest abuse or when there is a need to document the presence or absence of skeletal injuries. Skeletal survey should be performed in any child who has unexplained head or abdominal injuries that are suspicious for child abuse. Siblings of abused children are also at risk and should also be worked up for abusive injury.

Repeat skeletal surveys may be obtained 2 weeks after the initial survey in cases where the first skeletal survey was equivocal, abnormal, or in cases where abuse is still suspected based on the clinical assessment despite a normal initial skeletal survey. These repeat surveys often detect additional new or healing fractures, as well as confirm fractures of differing ages. The follow-up studies contain all of the same radiographic views as the initial study, omitting repeat imaging of the skull, pelvis, and spine (if no prior injury was seen in those areas) to reduce radiation exposure. Bone scintigraphy can be used (albeit rarely) as a complementary examination in cases where the skeletal survey is negative but clinical suspicion remains high. One may also consider the use of whole-body or focused MRI with Short Tau Inversion Recovery (STIR) imaging to identify injuries related to child abuse, for example, in the case of epiphyseal separation ( Fig. 21.9 ). The usefulness of this technique needs to be weighed against the risks of sedation in the infants and toddlers who are the most at risk for child abuse. Ultrasound may also be a helpful technique, especially in young children where epiphyseal separation is suspected.

If the skeletal survey reveals multiple fractures or rib fractures in a child younger than 2 years or if the child is younger than 6 months, a CT of the brain without contrast should be obtained regardless of neurological symptoms because these patients may have abusive head trauma without clinical symptomatology.

In a child with suspected visceral or vascular injuries to the chest, abdomen, or pelvis secondary to child abuse regardless of age, a CT scan with IV contrast is highly recommended. This allows the radiologist to evaluate for vascular and solid organ injuries. It also allows for better visualization of fractures (particularly rib fractures) compared with the plain radiographs.

Differential Diagnosis of Skeletal Injuries

The differential diagnosis for skeletal injuries that result from physical abuse is broad and includes accidental injuries, birth trauma, normal variants, metabolic diseases of bone, skeletal dysplasia, and infection. A brief discussion of the more important differential considerations follows.