Chapter 113 Child Abuse and Neglect

Paula M. Mazur, James J. Woytash, Lynn J. Hernan

• Accidental trauma differs from nonaccidental trauma in presentation, workup, treatment, and outcome. In child abuse, the history of trauma is withheld or falsified, and if the presenting symptoms are vague and nonspecific, the diagnosis and treatment of injuries may be delayed. This contributes to the increased death and long-term morbidity seen in abused children.

• Abusive head trauma is the most common nonaccidental injury seen in the pediatric intensive care unit. It is the most common cause of death in abused children. Abusive head trauma is more likely to be fatal and more likely to cause long-term morbidity in survivors than accidental head trauma.

• The infant brain is more vulnerable to injury, especially to shaking (shaken baby syndrome [SBS]). The infant’s neck muscles are underdeveloped and the head is proportionately larger than that of the adult. Large cerebrospinal fluid spaces in the infant allow greater movement of the brain within the skull. The infant brain has greater water content and increased deformability. Neurons and axons are less protected because of incomplete myelinization. Increased cerebral vasoreactivity predisposes the infant brain to cerebral edema.

• Accidental head trauma generates translational forces that result in focal damage. Nonaccidental trauma (e.g., shaking) generates rotational forces from rapid acceleration/deceleration of the head. Rotational forces tear cerebral bridging veins (creating subdural hematoma) and axons (creating diffuse axonal injury).

• Signs of SBS include subdural hemorrhage, retinal hemorrhage, and skeletal injury. A small number of cases will have all three signs. The presence of subdural hemorrhage and retinal hemorrhage, alone or in combination, in the appropriate clinical setting, are suggestive of SBS.

• Posterior rib fractures, metaphyseal fractures, and spinous process fractures are highly specific for abuse because their proposed mechanisms of injury are unlikely to occur accidentally.

• Abdominal trauma is the second leading cause of fatal child abuse, with death rates approaching 40% to 50%. Inflicted abdominal injury is often occult, presenting without obvious signs or symptoms; thus recognition is delayed. Treatment may be delayed because of a delay in seeking medical attention or a failure to consider nonaccidental injury in the differential diagnosis.

• Two types of forces are generated in inflicted abdominal trauma. Compression forces crush viscera against the anterior spine, and this crush causes burst injuries of the solid viscera and perforation of air-filled viscera. Deceleration forces cause shear injuries at the site of fixed, ligamentous attachments, with tears and hematoma formation at the ligamentous attachments of the liver and small bowel. Children with inflicted abdominal injury often have concomitant injuries suggestive of child abuse (bruises, head trauma, long-bone and skull fractures).

• Abusive thermal burns are of uniform thickness and closely replicate the shape of the inflicting object. Accidental thermal burns have varying degrees of thickness and irregular shapes. Abusive scald burns have an immersion pattern with a burn that is circumferential, is of uniform depth with a well-defined edge, and spares body creases. Accidental scald burns have more random patterns, vary in depth, have poorly defined edges, and do not spare body creases.

• All cases of suspected child abuse must be reported to child protective services and law enforcement. All cases of fatal child abuse must be referred to the medical examiner. Detailed, legible medical documentation and good communication between the treating physician and the medical examiner are essential to secure evidence needed for successful prosecution of the case.

Only a small percentage of physically abused children require hospitalization in the pediatric intensive care unit (PICU), but these children have been shown to have higher morbidity and death rates than critically injured victims of accidental trauma.^1,² The types of injuries that result from nonaccidental injury, coupled with delays in diagnosis and management, account for the worse outcome in abuse victims. Diagnosis and management of life-threatening injury is often delayed because the history of preceding trauma is not provided, or is so vague or the trauma is depicted as so minor that the physician is led away from a possible traumatic cause for the patient’s condition. The “golden hour” allotted for successful trauma resuscitation is spent considering an infectious, neurologic, or metabolic cause for the child’s illness. In addition, there can be considerable delay in seeking medical care. Several hours to days may elapse between time of injury and time of presentation to a medical facility.

Head trauma is the most common type of inflicted injury seen in the PICU, followed by abdominal trauma, burns, and thoracic trauma. Child abuse victims should be approached like trauma patients. They may have occult multiple-organ injury. The discovery of one injury demands a thorough evaluation for additional trauma. A meticulous investigation for injury has obvious medical utility, but it also becomes an essential part of the forensic investigation that ensues once a report of suspected child abuse is filed with child protective services and law enforcement.

Recognition of Child Abuse

History of Injury

The history of an injury must account for the type and severity of injury that is seen on physical examination. Suspicion of abuse should arise when any of the following occur:

1. The caretaker is unable to explain the injuries or gives a mechanism of injury that does not match the degree of injury seen. For example, a minor fall would not explain the presence of life-threatening cerebral edema.

2. The timing of the injury does not fit with the time of presentation (e.g., a critical head injury cannot be attributed to a fall that occurred 1 week before presentation).

3. The child’s developmental stage does not fit the history (e.g., rolling off a changing table should raise suspicion if the child is younger than 4 months).

4. The history of injury changes over time or from caretaker to caretaker. A careful review of all histories documented in the medical record may reveal discrepancies.

Patterns of Injury

Inflicted injury may be differentiated from accidental injury by its appearance, location, and distribution on the body (Table 113-1).

Table 113–1 Patterns of Injury

Accidental	Nonaccidental
Unilateral	Bilateral/symmetrical
Isolated injury	Multiple injuries
Amorphous shape	Well-defined shape
Prominent bone areas	Soft tissue areas
Posterior aspect of body	Anterior aspect of body
One age of injury	Multiple ages of injury

Bruising

Inflicted bruises are often bilateral, widely distributed, and located on soft tissue areas of the body that are unlikely to make surface contact during a fall. They may take the shape of the inflicting object (e.g., fingers, a hand print, linear whip marks from a belt, loop marks from a folded belt or cord). They are frequently found on the posterior trunk, buttocks, and the posterior side of the extremities because the victim would naturally be trying to run away from the perpetrator. Bruise color is not a reliable indicator of the time an injury occurred. Bruises resolve and therefore change color at different rates depending on their location and the force with which they were inflicted.¹ Nevertheless, documenting bruise color is important, particularly with the presence of bruises of markedly different colors at the same time, suggesting that the child may have been abused on more than one occasion. A simple gingerbread-man drawing of the child’s body, marked with the locations of all the child’s injuries, is a concise descriptive tool that will quickly jar a physician’s memory before any legal proceeding.

Early photographic documentation of the child’s injuries is essential. Thirty-five mm photography has been the standard for documentation, but digital cameras, both still and video, are replacing 35 mm photography. Polaroid photographs have been used in court, but they are inferior to 35 mm and digital photographs in both clarity and durability. Every effort should be made to obtain the best photographs; 35 mm photographs taken by a medical or law enforcement photographer are ideal. If a professional photographer is unavailable or if the attempt to obtain a professional is unsuccessful, any staff member familiar with the use of the camera should take photographs for the medical record.

Burns

Numerous researchers have attempted to describe the profile of an abused burned child and to characterize inflicted burns. In general, the child is often from a single-parent family of lower socioeconomic status on which previous suspicions of child abuse or neglect may have been filed. Compared with accidental burns, abuse burns are more extensive in degree and distribution, and often require management in a PICU.^2,³ Inflicted burns are of two types. Thermal injuries involve forced contact with a hot object, and scald injuries involve contact with a hot liquid, usually water. Abusive thermal burns are of uniform thickness and closely replicate the shape of the inflicting object. For example, an inflicted cigarette burn is approximately 8 mm round and uniform in depth. Accidental thermal burns have varying degrees of thickness and irregular shapes. If a child accidentally brushes against a hot cigarette, the burn will be more linear and of varying depth along its length.¹

Abusive scald burns have an immersion pattern. Part of the child’s body, usually the buttocks or limbs, are forcefully immersed and held in hot water. The resulting burn is circumferential and of uniform depth with a well-defined edge called a tide mark. Body creases are spared (withdrawal sign) during inflicted scald injury because the child pulls and folds his arms and legs inward to avoid immersion in the hot water. Accidental scald burns have more random patterns, vary in depth, have poorly defined edges, and do not spare body creases.²

Ideally, photograph documentation should be done before the initial burn dressing. Additional photographs taken 24 to 48 hours later, when the burn has fully evolved in depth and distribution, can also be valuable during litigation.

Fractures

Many of the characteristics used to recognize inflicted bruises and burns can be applied to fractures. The mechanism of injury described by the caretaker must remain consistent, be compatible with the child’s developmental stage, and must account for enough force to break the child’s bone. Inflicted fractures can be bilateral fractures of the same age or multiple fractures in different stages of healing. In his textbook Diagnostic Imaging of Child Abuse, Kleinman ⁴ divides fractures according to their degree of specificity for inflicted injury. Posterior rib fractures, metaphyseal fractures, and spinous process fractures are highly specific for abuse because their proposed mechanisms of injury are unlikely to occur accidentally. Scapula and sternum fractures are highly specific for abuse if the caretaker’s history does not account for a tremendous amount of force having been applied to these bones. Fractures of low specificity for abuse are fractures commonly seen after bumps and falls (e.g., clavicle fractures, linear skull fractures, long-bone fractures in ambulatory toddlers). Between high and low specificity for abusive trauma are acute, bilateral fractures, multiple fractures of different ages, widened (diastatic) or depressed skull fractures, and long-bone shaft fractures occurring in the young, nonambulatory infant.

It was long held that spiral fractures were highly suggestive of abuse because their spiral configuration implied that a forceful twisting motion had been applied along the length of the bone. Dalton et al.⁵ looked at femur fractures in children younger than 3 years, dividing the fractures into three types (oblique, transverse, spiral), and into three age categories (0 to 1 year, 1 to 2 years, 2 to 3 years). Their results showed that the incidence of spiral fractures increased significantly with increasing age, whereas the incidence of abuse was highest in the youngest age group regardless of fracture type. Spiral fractures can occur accidentally in vigorous, ambulatory toddlers. The age of the child holds more significance than fracture type when the possibility of inflicted injury is considered.^5,⁶

Posterior rib fractures and metaphyseal fractures are frequently seen in shaken or battered infants. Posterior rib fractures occur when the child’s chest is compressed. This compression causes the rib to rock back over its articulation with the transverse vertebral process. The transverse process acts as a fulcrum for the rib, and a fracture occurs on the rib’s pleural surface. Although this fracture has been seen in pediatric patients with major, high-speed trauma, it does not occur accidentally in healthy children during mild to moderate thoracic trauma. Cardiopulmonary resuscitation (CPR) has not been shown to cause posterior rib fractures. A detailed discussion of posterior rib fractures can be found in Kleinman’s textbook, Diagnostic Imaging of Child Abuse.⁴ Acute rib fractures are difficult to see on a plain film. They may only first be visible 2 to 3 weeks after injury, when a callus has formed around the fracture site. Therefore finding a callused fracture on x-ray film is discovering trauma that occurred at least 2 to 3 weeks before the radiograph.⁴

The periosteum of a pediatric long bone is loosely attached to its cortex. Any violent pull, tear, or twist on the shaft of a child’s long bone displaces the periosteum. The results are subperiosteal hemorrhage and periosteal elevation that can be seen on x-ray film. Conversely, the periosteum is tightly attached at its point of origin, the metaphyseal plate. The metaphyseal plate, which is the most newly laid-down bone above the growth plate, has delicate trabeculations. Violent forces applied to the midshaft periosteum are transferred to its point of origin, the metaphyseal plate, and an avulsion fracture occurs through the delicate trabeculae. Depending on the angle at which the radiograph is taken, a metaphyseal avulsion fracture can appear as a thin line through the metaphysis, as “corners” broken off the edges of the long bone, or as a “bucket handle” attached to the end of the long bone. Like posterior rib fractures, metaphyseal fractures are pathognomonic for abuse. These corner fractures or bucket handle fractures are occult. There is no deformity or swelling, and they are not obviously tender to palpation. Like posterior rib fractures, metaphyseal fractures are usually found on a radiograph obtained for other reasons or on a skeletal survey done during the medical investigation of a suspected abuse case. Spinous process fractures are the remaining type of fractures listed under “high specificity” in Kleinman’s text. They are thought to occur during infant shaking, when the spine is in hyperflexion, causing sudden stress on the posterior spinous ligament as it articulates with the posterior spinous processes.⁴

Skull fractures are the second most common fracture seen in abused children, but skull fractures are also commonly seen in accidental head trauma. A skull fracture is suspicious for abuse when there is no significant trauma to account for the fracture. An unexplained skull fracture obligates the physician to look for additional injuries that are pathognomonic for abuse. For example, a skeletal survey may reveal occult posterior rib or metaphyseal fractures. A noncontrast computed tomographic (CT) scan may reveal bilateral subdural hematomas, which are not commonly associated with accidental skull fractures.

The guidelines for dating fractures are broad. In general, periosteal elevation can occur acutely, within hours to days after injury. Callus formation is seen approximately 2 weeks after injury. Loss of the fracture line begins to occur 3 weeks after the injury, and remodeling of the fracture occurs anywhere from 3 months to a year after injury, depending on the child’s age. Infants will heal and remodel faster than older children. Skull fractures and metaphyseal fractures are difficult to date because they do not show the same periosteal reactions that healing ribs and long bones do.⁴ However, a diastatic skull fracture, which is a linear skull fracture greater than 3 mm wide, is not an acute injury. It takes time for a skull fracture to separate more than 3 mm.