Rockwood & Wilkins’ Fractures in Children
6th Edition

Chapter 7
Child Abuse
Robert M. Campbell Jr
Timothy Schrader
Child abuse is maltreatment of a child by either parents or caretakers, and includes physical, sexual, and emotional abuse, as well as emotional and physical neglect (1). The incidence of physical abuse nationally is estimated to be 4.9 children per 1,000, and 1 of every 1,000 abused children die (2). The epidemic of child abuse continues to worsen in the United States, with approximately 2.9 million reports now filed annually compared with 1.2 million in 1982 (3). In the United States 3 children

a day die of abuse or neglect (4). In Texas alone, 50,746 children were confirmed victims of child abuse or neglect in fiscal year 1995, with 101 deaths recorded (5). The types of abuse these children were subjected to included physical abuse in 31.5%, sexual abuse in 15.6%, emotional abuse in 6.8%, abandonment in 2%, medical neglect in 4.7%, physical neglect in 15.4%, and neglectful supervision in 40.3%. The minimal annual cost of child abuse in the United States is estimated to be 9 billion dollars (3), but additional costs, both direct and indirect, exist. The estimated national cost of child abuse for the child welfare system is 14 billion dollars, law enforcement 24 million, and the court system 341 million (4). The median cost of hospital admission for child maltreatment in Pennsylvania in 1995 was $15,587 (6). The long-term social costs of child abuse are impossible to estimate: one third of the victims of child abuse grow up to be seriously inept, neglectful, or abusive parents; one third are at high risk for eventually becoming abusive parents; and only one third do not repeat the destructive patterns they were exposed to as children (7). Indirect estimated national costs of child abuse, including special education for learning disorders of abused children, maternal mental and health care, legal costs of juvenile delinquency, lost productivity to society of abused children as unemployed adults, and later adult criminality of abused children are 94 billion dollars (4). The orthopaedist becomes involved in the care of 30% to 50% of abused children (8). Early recognition of these abused patients by the orthopaedist is critical, because the children who are returned to their homes after an unrecognized episode of child abuse have a 25% risk of serious reinjury and a 5% risk of death (9).
In 1946, Caffey (10) described six infants with long-bone fractures, chronic subdural hematomas, and intraocular bleeding without a history of trauma to explain the injuries. Although his work is often cited as the first report in the English literature of child abuse, Caffey did not speculate about the etiology of the children’s injuries. In 1953, Silverman (11) characterized the unique metaphyseal fractures found in abused children in the pediatric literature and clearly emphasized that these were due to nonaccidental trauma. Altman and Smith (12) in 1960 published the first series in the orthopaedic literature of injuries caused by child abuse. General public awareness of child abuse increased with the 1962 publication of a report by Kempe et al (13) characterizing the problems as the battered child syndrome. In 1974, Caffey introduced the term whiplash-shaken infant syndrome to the literature to emphasize the etiology of subdural hematomas in infants caused by shaking episodes (14). In 1974, Congress acknowledged the national importance of the prevention of child abuse by the passage of the Child Abuse Prevention and Treatment Act.
The Home at Risk
In assessing where abuse of children may occur, households in turmoil from marital separation, job loss, divorce, family death, housing difficulties, or money problems are more likely to have abusive episodes (15). Families with two unplanned births are 2.8 times more likely to have an episode of child abuse than families with no unplanned births (16). In these homes at risk, stepparents, babysitters, boyfriends, relatives, and even larger siblings frequently are abusers (17,18,19). The parents of battered children may themselves have been abused when they were children (20). High levels of parental stress and high belief in the worth of corporal punishment are predictors of child abuse (21). Parental substance abuse, whether alcohol or other drugs, makes child abuse more likely (22). The risk of physical abuse is fivefold more likely with maternal cocaine use (23). Young, unmarried mothers are more likely to have their infants die from intentional injury, with a peak incidence of 10.5 intentional deaths per 10,000 live births in one study (24). Violence in the home is not directed solely toward the child; in one study (25) of families with substantiated child abuse, 30% of the mothers had been battered. Although the youngest, poorest, most socially isolated and economically frustrated caretakers are the most likely to act violently toward their children (26), any adult from any social or economic level may be guilty of abusing a child (17). Primary parental predictors of child abuse are listed in Table 7-1 (27).
The Child at Risk
Certain children of all ages are more likely to experience abuse, and younger children are particularly at risk. Most reported

cases of child abuse involve children younger than 3 years of age (28). In one report of child abuse (29), 78% of all fractures reported were in children younger than 3 years of age and 50% of all fractures occurred in children younger than 1 year of age. Infants younger than 1 year are especially at risk for the most severe form of child abuse: infant homicide (30,31). With severe shaking injuries by caretakers, 30% of infants are disabled and another 30% die (32). The problem probably is more widespread than suspected. There is evidence that suggests that in the United Kingdom, 10% to 20% of children diagnosed as having sudden infant death syndrome may be intentional infant homicide (33). In one report (34), covert video recordings of adults attending their children who were hospitalized for suspicious illness documented 14 separate instances of caretaker attempts at upper airway obstruction. First-born children (29), premature infants, stepchildren, and handicapped children are at a greater risk for child abuse (8). Benedict et al (35), in a longitudinal study of 500 handicapped children followed from birth to age 10 years, documented a 4.6% incidence of physical abuse. Surprisingly, they found that the most severely involved children were less likely to be abused, whereas the marginally functioning children were at greater risk, with parental frustration possibly being a factor.
TABLE 7-1 Parental Predictors of Child Abuse
  • Age <20 years
  • Lower educational achievement
  • History of sexual abuse
  • Child guidance issues
  • Absent father during childhood
  • History of psychiatric illness
  • Age <20 years
  • Lower educational achievement
  • Child guidance issues
  • History of psychiatric illness
  • Parent history of child abuse
  • Divorce or separation of mother’s parents
  • Maternal history of being separated from mother, parental alcohol or drug abuse
  • Maternal history of depression
From Sidebotham P, Golding J. The ALSPAC Study Team: child maltreatment in the “Children of the Nineties”–a longitudinal study of parental risk factors. Child Abuse Negl 2001;25:1177–1200; with permission.
The Risk for Munchausen Syndrome by Proxy
Children who are persistently presented by parents for medical assessment of vague illness, and have a history of multiple diagnostic or therapeutic procedures with unclear outcome, are at risk for having a form of child abuse known as Munchausen syndrome by proxy (36,37). Munchausen syndrome is named for Baron von Munchausen, an 18th-century mercenary whose exaggerated tales of adventure were viewed with great suspicion. Asher (36), in 1951, described Munchausen syndrome in adults who presented with apparent acute illness accompanied by dramatic, untruthful medical histories to gain attention. In Munchausen syndrome by proxy, children become the victims of this adult behavior when parents with a misguided sense of purpose fabricate a wide range of childhood illnesses for their children, often subjecting them to needless diagnostic workups and treatment (37). Symptoms of the child’s “illness” are based on an imaginary medical history given by the parent, with signs of the illness either simulated or induced by the parent. For example, a child may be brought into the emergency room by a parent with a complaint of vomiting. In Munchausen syndrome by proxy, this complaint may either be a total fabrication by the parent or the parent may simulate the complaint by producing “vomitus” from some source as proof of illness. In one report, bloodstained material has been presented by a caretaker as proof of a child’s “gastrointestinal bleeding,” but DNA testing revealed that the source was actually the caretaker (38). Conjunctivitis from a caustic agent placed on an infant by a caretaker has been reported (39). Children have been given clozapine and clonidine by caretakers to simulate illness (40). A parent has caused vomiting in a child by the administration of salt (41) or ipecac. In the most extreme situations, a rodenticide-induced coagulopathy was induced in a 2-year-old child (42), a deliberate self-induced preterm labor was caused by a parent (43), and another gave chronic illicit insulin to a 1-year-old child (37). Over half of reported cases of Munchausen syndrome by proxy involve induced symptoms, whereas another 25% involve a combination of both simulation and induction of symptoms (44).
The biologic mother is almost always the perpetrator of this pattern of abuse (45), but men, especially those with a history of Munchausen syndrome themselves, also can be responsible for this form of child abuse (46). Caretakers in Munchausen syndrome by proxy almost all have a medical background: 35% to 45% are nurses, 5% are medical office workers, 3% are social workers, and 1% are orderlies (45). The perpetrator of the child’s illness denies the knowledge of its etiology, but the acute signs and symptoms of the child’s illness in Munchausen syndrome by proxy will resolve if the syndrome is recognized and the child is separated from the parent (45). Follow-up of families with this disorder is crucial. Davis et al (47) reported a reoccurrence of abuse rate of Munchausen by proxy of 17%, new sibling abuse in 50% of families with a suffocated child, and 40% of families involved with nonaccidental poisoning. Failure to diagnose this condition places a child at risk for either serious long-term sequelae or death in approximately 9% of cases.
The diagnosis of Munchausen syndrome by proxy remains difficult. Covert in-hospital video surveillance (CVS) of caretakers with their children may be a valuable means to substantiate or disprove this diagnosis. Hall et al (48) reported that CVS with audio surveillance added allowed diagnosis of Munchausen syndrome by proxy in 56% of patients monitored and was supportive of the diagnosis in another 22% of children. Both suffocation and injection of bodily fluids were documented by this approach. Privacy concerns were addressed by routine consent of the caretakers on the admission form that “closed-circuit monitoring of patient care may be used for educational or clinical purposes,” and a multispecialty team approved the 24-hour use of security officers to monitor selected patients and make log entries on at least an hourly basis. The approach, though, is expensive and cannot be covered by third party payers. The diagnosis of Munchausen syndrome by proxy depends on health care workers maintaining a high degree of suspicion when children present with repetitive illness with no physiologic explanation.
The Risk for Sexual Abuse
Although the orthopaedist usually considers child abuse in the context of fractures and other obvious injuries, an increasingly important aspect of child abuse to recognize is sexual abuse. It is estimated that 25% of abused or neglected children have been sexually abused (49). Physically abused children have a 1 in 6 chance of being sexually abused, whereas sexually abused children have a 1 in 7 risk of being physically abused (50). Children

living with nonbiologic parents or with caretakers who are substance abusers are most at risk. The child usually discloses sexual abuse under three types of circumstances. The child may have just experienced an argument with the abuser and may “accidentally” reveal the existence of the abusive relationship, the child is permanently separated from the abuser, or the abusive adult is shifting attention to a younger sibling (51). The incidence of sexual abuse of handicapped patients has been estimated as ranging from 25% to 83%, with handicapped males being more likely to be sexually abused than nonhandicapped males in the general population (52). Both sexual and physical abuse may be associated with cults, which may be identified by their symbols. Satanic cults may use a pentagram (an encircled five-pointed star with an inverted apex), the eye of Horus (an eye within a triangle), or various symbols that modify a cross, and the middle or little finger of the left hand may be painted black (2). Sadistic ritual abuse has been reported in preschool and elementary school children (53).
The Risk for Atypical Child Abuse
Bizarre forms of child abuse may cause a confusing illness or injury. There are reports of young children forced to aspirate pepper, of intentional burns of infants by placing them in microwave ovens (54), of children denied water who developed hypernatremic dehydration, and of infants poisoned with cocaine given to them by caretakers (55). In one case of abuse an infant presented with bloody stools and persistent acidosis due to intentional iron poisoning (56). A high index of suspicion is needed to avoid missing the diagnosis of nonaccidental trauma in these situations.
The history is critical in the diagnosis of child abuse. It usually is taken in the chaotic environment of a busy emergency room, so it is important to find a quiet area for the interview where tempers can cool and distractions can be minimized. In addition to taking a thorough history with regard to the mechanism of injury, the orthopaedist must question the patient and the family in an investigative fashion to establish accurately the circumstances of the injury and the environment in which the child lives. These skills rarely are taught in residency training. In a survey of pediatric residents (57), 42% of them had 1 hour or less in training for detection of child abuse, and most orthopaedic residents likely have even less training in this area. Unfortunately, little progress has been made. In a study comparing the documentation of physical abuse between 1980 and 1995 in a teaching hospital, very little improvement was noted (58). It has been found that use of a structured clinical form increased the information collected to support the diagnosis of child abuse (59). Precise documentation in child abuse is vital for reasons beyond medical care. Although most subpoenas for testimony by physicians in child abuse cases do not result in courtroom appearances (60), it must be clear that all documentation in child abuse cases is clearly at risk for becoming evidence in courtroom proceedings and clear, detailed records are the best defended in courtroom testimony by physicians (61).
The Investigative Interview
The extensive history needed to detect child abuse is termed the investigative interview. It begins with documenting the history (or the lack of history) of the injury and, more importantly, goes on to uncover enough details about the child’s life so that plausible scenarios can be created to explain the injury. A full physical examination and skeletal survey should precede this interview to search for any possible evidence of additional undocumented child abuse. The orthopaedist must become social worker and detective, calmly and methodically establishing how the injured child lives; finding out which family members, friends, or other caretakers have access to the child; and how likely it is that they might have contributed to the child’s injuries. A detailed history of injury is obtained individually from each adult family member in a private setting. If the patient and siblings can communicate, they should be interviewed separately from the parents and other members of the family.
Once full injury physical and x-ray assessments of the child are complete, the investigative interview is begun by asking the primary caregiver in a nonjudgmental way how the child was acutely injured. To avoid provoking emotions, any additional soft tissue or skeletal trauma discovered should be brought up at the end of the interview for explanation once the presentation injury has been thoroughly discussed. Each investigative interview should follow a systematic review of symptoms: what happened, who was there, when was the injury recognized, and how long before medical treatment was sought. Inconsistencies are not challenged. Leading questions are avoided in favor of open-end questions. Medical terms should be explained in plain English, with care taken to avoid medical jargon. More plausible explanations for the injury are not volunteered. The place where the injury occurred and which individuals were actually present are determined. Open prompts can enhance the interview (62). Never ask who caused the injury, but rather ask what they saw or what they think happened. Exactly who discovered the injury and how soon the child received medical care after onset of symptoms should be established. Delays in seeking medical care for an injured child are very suggestive of child abuse (15). The crucial questions to be answered are not only whether the given history of trauma is sufficient to explain the severity of injury but what other possible scenarios could explain the injury if the volunteered explanation is not plausible. This requires acquiring a working knowledge of the child’s environment, which can be obtained by asking specific, detailed questions (Table 7-2).
When interviewing injured children, it is essential to be as gentle as possible, asking how they got hurt rather than who hurt them. Any question asked should be appropriate for the child’s age. The child’s account of what he or she was doing at time of injury should be compared with the accounts of the adult witnesses. If possible, the siblings of the injured child

should be interviewed because they also are at risk for child abuse. Nonvisual clues during the interview also may be helpful (Table 7-2).
TABLE 7-2 Child Abuse: Investigative Interview
Environmental Issues
Primary caretakers
   Responsible for feeding, discipline, toilet training
   Easy or difficult child
Home environment
   Place of residence
   Living conditions
   Adults employed or unemployed
   Sleeping arrangements
   Marital status of parents
   Boyfriend or girlfriend of single parent
   Substance abuse
Home stress level
   Recent job loss
   Marital problems (separation or divorce)
   Death in the family
   Housing problems
   Inadequate funds for food
Parental or Caregiver Responses and Attitudes
Evasive, not readily responsive to questions
Irritated by questioning
Contradictory in responses
Hostile and critical toward child
Fearful of losing child or criminal prosecution, or both
Unconcerned about child’s injuries
Disinterested in treatment and prognosis
Intermittently unavailable for interview (without valid reason)
Unwilling to give medical information
Unwilling to give consent for tests
Indifferent to child’s suffering (seldom touches or looks at child)
Selected data from Akbarnia BA. The role of the orthopaedic surgeon in child abuse. In: Morrissy RP, ed. Lowell and Winter’s pediatric orthopaedic. Philadelphia: JB Lippincott; 1990; and Green FC. Child abuse and neglect: a priority problem for the private physician. Pediatr Clin North Am 1975;22:329–339; with permission.
To make the diagnosis of child abuse, the orthopaedist must make one crucial decision: Is the history of trauma adequate to explain the severity of injury? That decision should be based on experience in the care of fractures with knowledge of their mechanisms of injury and special insight into the types of trauma most likely to cause significant injury. In addition, it is also extremely important to have knowledge of the developmental abilities of a child when a caretaker states the child’s injuries are self-inflicted (2). It is patently absurd for the parents to explain that an infant’s femoral fracture occurred in a fall while the infant was standing alone when the infant is too young to even roll over. Details of the accident given as the reason of the injury should be carefully considered. Although it is not unusual for a young child to sustain an accidental fall, it is unusual to sustain a serious injury from that fall alone. Infants fall from a bed or a raised surface during a diaper change fairly frequently. In a study of 536 normal infants (63), nearly 50% of them had fallen accidentally from an elevated surface, usually after the first 5 months of life, when the children were able to roll over and pull themselves up. Significant injury in such falls is, however, extremely rare. Combining two studies (64,65) of 782 children younger than 5 years of age who accidentally fell off an elevated surface, such as bed or sofa, injuries were limited to three clavicular fractures, six skull fractures, one humeral fracture, and one subdural hematoma. In another report a much higher rate of fracture was seen in falls from furniture with 98% having fractures, most in the upper extremity (66). More, injuries occur in falls from greater heights. In a report of 363 stairway injuries (13), 10 were in infants who fell with their caretaker while being carried on stairways and four of those sustained skull fractures. In patients aged 6 months to 1 year, 60% were using walkers at the time of injury. Only 4% of patients had extremity fractures, and 1% had skull fractures. Abrasions and contusions were present in 55% of patients, whereas 26% had lacerations. In another study (67) of 76 children who had fallen from a bed, a crib, or a chair (estimated height 1 to 3 feet), while in the hospital for other illness, 18% had scalp or facial hematomas and 12% had lacerations about the head. There was only one nondisplaced skull fracture and one long-bone fracture in a patient with osteogenesis imperfecta.
Additional important information about the child and the family may be obtained by a review of past medical records or by contacting social workers who may have been involved with the family. Conferring with the family’s primary health care provider also may be extremely helpful. A medical release from the family is necessary before these sources of information can be explored (15). The physician or social worker should be asked if there has been a suspicious pattern of injury, illness, ingestion of objects or medications, or noncompliance with health care recommendations; whether the family is receiving counseling or other support from any community groups; and whether the family has any previous involvement with child protective services or the police (15).
Documentation Requirements
Careful documentation is critical. Chart notes may later be needed in court as evidence for either custodial hearings or criminal trial (58), and defending inaccurate or partial chart notes in court can be extremely embarrassing. Each account should be recorded in as much detail as possible, using quotation marks for exact quotes and specifying who is giving the history. Particularly with crucial answers, the exact question preceding the response should be documented. In addition, the general emotional state of the individual providing the account,

as well as the individual’s reactions to emotionally charged questions should be documented to assist in later evaluation of the credibility of the account. If the family wishes to change their story after the initial account, no changes should be made to the earlier record, but an addendum should be placed detailing the new account. The completed record should include several specific items (Table 7-3). If child protective services recommends emergency transfer of custody of the child to a foster home or a shelter, then the orthopaedist should summarize chart documentation in a separate notarized narrative, which is preferred by most courts.
TABLE 7-3 Documentation of Child Abuse
  • Mechanism of injury
  • Family social history
  • Past medical history
  • Family history of diseases such as osteogenesis imperfecta or other metabolic disease
  • Physical examination findings (positive and negative)
  • X-ray findings
  • Laboratory results
  • Photographs of soft tissue injuries
  • Results of consultations
  • Diagnosis of child abuse
  • Treatment recommendations
  • Acute injuries
  • Investigation by child protective services
After the initial musculoskeletal evaluation for acute fracture assessment, an extremely thorough physical examination of the entire patient should follow, systematically going from the head to the toes, to detect any signs of additional acute or chronic injury. Additional acute and subacute fractures may cause local tenderness and swelling, whereas chronic fractures may produce swelling from the presence of callus and clinical deformity from malunion. Specific x-rays should be used to confirm clinically suspected fractures, with the skeletal surveys being used to show most occult fractures not detectable by clinical examination. The thorough review should focus specifically on the other body areas commonly involved in child abuse including the skin, the central nervous system, the abdomen, and the genitalia. These areas should be carefully evaluated for signs of acute and especially chronic injury, because 50% of verified abuse cases show evidence of prior abuse (20).
Soft Tissue Injuries
In addition to examination of the soft tissue around the acute fracture site for swelling and bruising, the patient’s entire body should be carefully and systematically evaluated to detect acute and chronic soft tissue trauma. Deliberate soft tissue injuries are present in 81% to 92% of abused patients (28,68) making these the most common physical examination findings in abused children. The types of skin lesions commonly encountered include ecchymosis, welts, abrasions, lacerations, scars, and burns.
The typical toddler often has multiple accidental bruises over bony prominences such as the chin, the brow, the knees, and the shins (9,17,69). Babies who are not yet mobile have a much lower prevalence of accidental bruising compared to mobile toddlers, and accidental bruises in babies are also typically noted over bony prominences on the front of the body (70). Bruises on the back of the head, neck (17), arms and legs, on the buttocks, abdomen, cheeks, or genitalia may be suspicious for abuse, although accidental bruises can also occur in all these locations (9). Accidental bruising of the face, though, is much less common and should be carefully evaluated. Although nonaccidental bruises often are concentrated on the trunk and buttocks, they are also commonly present on the head and proximal extremities. In a study of 400 nonabused children, Robertson and Barbor (71) found an overall incidence of just 7% for accidental soft tissue injuries of the face and head, with the peak incidence of 17% seen in toddlers. Soft tissue injuries were present on the lower extremities and buttocks in only 31% of children and on the upper extremities of only 9%. In a study of 1,467 patients seen for reasons other than trauma at a medical center over a 1-year period, 76.6% had at least one recent skin lesion, 17% had at least five injuries, 4% had at least 10 injuries, and fewer than 1% had more than 15 injuries. In children over 9 months of age, skin lesions were rare (11.4%) and they were concentrated on the head and face. In children over 9 months of age, the skin lesions were mostly on the lower extremities, with fewer than 1% on the chin, ears, or neck (72). It is clear that there is no threshold number of bruises that are consistent with abuse; some children will only have one bruise from abuse, whereas others may have 20 accidental bruises. What is crucial is the location and configuration of the bruises and the mobility of the child, taken together with the rest of the medical and social history.
The configuration of the bruise resulting from abuse may resemble the implement used to inflict the injury (Fig. 7-1). One study, however, determined that the soft tissue injuries of abuse are weapon specific in fewer than 10% of patients (68). The weapons used to abuse children can be almost any common

object, but they often include belt buckles, staplers, ropes, switches, coat hangers, ironing cords, and the open or closed human hand (73,74). Bruises inflicted by an open hand may appear on the face or a flat area of skin, and grasp marks may appear as ovoid lesions when the fingertips are deeply embedded in the extremities or the shoulders of the child during extreme shaking (75). The injury pattern and the severity of the ecchymosis depend on the amount of force used, how directly the instrument made contact, and the specific type of implement used to strike the child (75). Welts are more complex skin lesions in which swelling accompanies bruising from injury through lashing or whipping.
FIGURE 7-1 Healing soft tissue injuries may resemble the instrument used to inflict the lesion.
The type of soft tissue injuries seen in child abuse may depend on the age of the child (Table 7-4). In the pediatric dentistry literature, in a series of 266 children suspected of being abused, Jessee and Rieger (76) also found that bruises were the most common soft tissue injury, with the most common facial injury being ecchymosis of the cheek, forehead, periorbital area, or the lip.
The age of a cutaneous contusion reportedly can be estimated by a change in its color over the 2 to 4 weeks following injury, with fading of the lesions beginning at the periphery. An acute contusion is blue or reddish purple; it gradually changes to green, then to yellow, with final resolution as a brownish stain

as the hemoglobin is finally broken down (77). Langlois and Gresham (78) noted that a yellowish bruise must be older than 18 hours; a red, purple, blue, or black coloration of the bruise may be present from 1 hour after injury to resolution; red is always present in bruises regardless of the age; and bruises of identical age and etiology on the same person may be of different appearances and may change at different rates. Although the age of a superficial contusion may be approximately dated by its appearance, a deep contusion may take some time to rise to the skin surface because of fascial planes and thus delay visible ecchymosis. Even though the color of a bruise can aid in determining the length of time it has been present, the clinical precision of this practice has been questioned (79).
TABLE 7-4 Distribution of Soft Tissue Injuries in 371 Abused Children
Age Bruises/Abrasions Lacerations Scars Burns Head/Neck Injury
≤9 mo 71% 8% 16% 50%
10–35 mo 64% 13% 6% 6% 12% 38%
≥36 mo 61% 21% 4% 11% 3% 24%
Data from McMahon P, Grossman W, Gaffney M, et al. Soft tissue injury as an indication of child abuse. J Bone Joint Surgery Am 1995;77:1179–1183; with permission.
Natural skin lesions should not be mistaken for bruises. Mongolian spots are naturally occurring deep-blue pigmented areas that are present on the lower back at birth, usually just proximal to the buttocks. They occur more commonly in black and Asian infants than in white infants (80). Unlike traumatic bruises, they do not change in color but gradually resolve as the child matures (75). Cultural differences should always be considered when unusual skin lesions are noted. Vietnamese children may be subjected to a folklore medical practice known as cao-gio, which causes suspicious scratches and bruises on the back and may be mistaken for child abuse (81).
Acute lacerations in various stages of healing and chronic scarring may be present in an abused child. Like bruises, the configuration of the injury can resemble the weapon used to inflict the injury on the child. Although minor lacerations around the eye are fairly common, multiple scars resulting from either lacerations or burns (82,83) are suspicious for abuse.
Burns are found in approximately 20% of abused patients (28) and are most likely to be found in patients younger than 3 years of age (68). Burn evaluation should include configuration, approximate percentage of body surface area, location, distribution, uniformity, length of time the child was in contact with the burning agent, temperature of the burning agent, and presence or absence of splash marks when hot liquids are involved (75). Most self-inflicted accidental pour or spill burns occur anteriorly on the child, but accidental burns can also occur on the back. In accidental burns, the injury usually has an arrowhead configuration in which the burn becomes shallower and more narrow as it moves downward, and there may be splash marks surrounding the lesion (75). The burn pattern in accidental burns may also be indicative of flowing water (84). Abuse should be suspected when deep second- or third-degree burns are seen well demarcated with circumferential definition. In accidental hot water immersion, an indistinct stocking or glove configuration may be seen with varying burn depths and indistinct margins. In deliberate immersion burns, the burn demarcation shows uniform depth and a well-demarcated water line (75). Particularly severe burns occur when a child’s buttocks are immersed in hot tap water as abusive punishment. The central aspect of the buttocks may be spared, giving a doughnut-like appearance to the burn. Galleno and Oppenheim (28) noted that in accidental hot water immersion, the child is uniformly scalded about the lower extremities as the legs are quickly extended by the child to climb out of the water, but in deliberate, abusive immersion the children are lowered into the water, so they instinctively pull up their legs to avoid the scalding hot water, and knee flexion may protect the popliteal areas from burn (Fig. 7-2). The average child abused by scalding burns is an undernourished 2-year-old child with burns involving 15% to 20% of the body, usually the buttocks; these children have a 10% to 15% mortality rate from burns owing to sepsis (84).
Burns also can be inflicted by many objects commonly found in the household. Sometimes the object can be identified by the configuration of the burn. Intentional burns by cigarettes are circular, deeply excavated, and sometimes repetitive, usually about 8 mm in diameter (75). Isolated cigarette burns may be accidental, and in such cases they are usually superficial, pointing downward, and often appearing triangular. Lesions of impetigo resemble those of cigarette burns, but they are much more superficial. Severe eczema may mimic burns suspicious for child abuse (85).
Contact with heated objects may cause burns of unique shape that allow identification of their etiology (Fig. 7-3). Children accidentally grasping curling irons sustain burns of the palms, whereas burns on the dorsum of the hands are more suspicious for abuse (2). Hair dryers can be used to inflict burns on children, and full-thickness skin burns can result from contact with the grill of a dryer up to 2 minutes after it has been turned off (86). Such burns can resemble those caused by hot water or could have a grill pattern caused by direct contact with the dryer. Abuse burns have also been inflicted by stun guns (87). These devices deliver a high-voltage impulse of up to 100,000 volts at 3 to 4 mA, incapacitating the individual and leaving hypopigmented burn scars on the skin 0.5 cm apart. Rope burns, sustained when children are restrained about the wrists for beatings, may be seen as circular scars about the wrists (2). Full-thickness skin burns have been reported in small children who were placed in microwave ovens (54).
Certain folklore practices may cause lesions simulating abusive burns. Round burns on the abdomen, buttock, or between the thumb and forefinger of Southeast Asian children may be due to a variant on the Chinese medical practice of moxibustion. Folk medical practitioners burn balls of the moxa herb on the surface of the skin for therapeutic purposes, and both cigarettes and yarn have been similarly used in refugee camps. The knowledge of these practices may help to avoid inappropriate accusations of child abuse (88). Infants may sustain second-degree burns when they accidentally come in contact with the hot vinyl upholstery of a car (89).
It is critical that the orthopaedist examine and carefully document all soft tissue injuries that are present before treating the acute fractures. The cast applied in the treatment of the fracture, especially a spica cast, may cover up potentially incriminating

skin lesions and will preclude other members of the child advocacy team from being able to identify or document them. Photographs can be used to document lesions before cast placement.
FIGURE 7-2 A. In accidental hot water immersion, the child is uniformly scalded about the lower extremities as the legs are extended quickly by the child to climb out of the water, with burns occurring behind the knee (curved arrow). B. In nonaccidental immersion, the child instinctively pulls up the legs to avoid the hot water, and knee flexion may protect the popliteal areas from burn (curved arrow).
Head Trauma and Spinal Fracture
In physical abuse the most common cause of death is head trauma (90). An excellent review on fatal abusive head injury was published by the National Association of Medical Examiners Ad Hoc Committee on shaken baby syndrome (91). A complete neurologic examination is required in any child suspected of being abused. This should include assessment of the child’s mental status, motor function and sensation, reflexes, and gait, if possible. Any abnormal findings warrant further investigation. Skull x-rays, a head computed tomography (CT) scan and spinal x-rays should be considered in the work up of such patients. Also included should be an ophthalmologic evaluation looking for retinal hemorrhages.
According to the American Academy of Pediatrics Section on Radiology (92), anteroposterior (AP) and lateral skull x-rays are mandatory in head trauma of children. CT scans alone may miss skull fractures. All children with suspected intracranial injury should also undergo either cranial CT scan or magnetic resonance imaging (MRI), or both, to document soft tissue brain injury. Unenhanced CT scans are readily available and can diagnose acute parenchymal, subarachnoid, subdural, and epidural hemorrhage better than MRI. Even in abused children without neurologic findings or retinal hemorrhages, occult head injury should be suspected. One study using both CT scans and skeletal surveys to evaluate 51 neurologically intact abused children identified skull fractures in 74% and intracranial injury in 53% in high-risk groups of those with rib fractures, multiple fractures, facial injury, or those younger than 6 months (93). These high-risk patients without neurologic findings should be screened by CT scan or, if inconclusive, also with MRI. The MRI can be a sensitive means of diagnosing small extra-axial cranial hemorrhage (94). If routine MRI is inconclusive, diffusion-weighted MRI can show subtle hypoxic-ischemic brain injury (95,96). Jaspan et al (96a) noted that cerebral contusion tears could be noted by sonography, and in another study Jaspan et al suggested a detailed protocol for imaging work-up of head injury in child abuse (97).
When an infant with musculoskeletal injury presents with altered mental status, whiplash-shaken infant syndrome should be suspected. Violent shaking of a small child whiplashes the relatively large head back and forth over the thorax, with possible

development of subdural hematomas (Fig. 7-4). Some authors (94) believe these head injuries actually occur when the child’s head is slammed onto a soft surface such as a mattress. On impact, deceleration forces approaching 400 G’s may be seen, tearing the bridging vessels between the skull and the brain and producing intracranial hemorrhage and cerebral edema. Skull fractures, though, are rare unless the child is thrown onto a hard object. Simple linear skull fractures usually are due to a fall of less that 4 feet and may be accidental, but complex skull fractures without a history of significant trauma, including comminuted, diastatic (separated sutures), displaced fractures, and fractures crossing suture lines, are suspicious for abuse (44). Children with head injuries caused by direct impact trauma are described as having shaken impact syndrome (94). Children with head injuries have fevers, bulging fontanelles, and sometimes enlargement of the head. A pattern of paresis may be present, and reflexes may be increased (14). Skull fractures generally are uncommon, and either ultrasound or CT scans are commonly used to diagnose subdural hematomas. Cerebral edema may be lethal (98), and emergency neurosurgical consultation is needed. Funduscopic examination by an ophthalmologist is indicated because of concurrent retinal hemorrhage (14). The examiner should be aware, however, that retinal hemorrhages resulting from normal vaginal birth are present in 40% of newborns, but these injuries resolve by 3 to 4 weeks of age (99). Shaken baby syndrome has a high mortality rate: 30% of infants die of their injuries (32).
FIGURE 7-3 Contact burns by heated objects may cause burns of unique shape that allow identification of their cause.
Fatal cerebral injury also occurs in the abusive tin ear syndrome: a clinical triad of unilateral ear bruising, CT scan evidence

of ipsilateral cerebral edema, and hemorrhagic retinopathy, with a 100% mortality rate in one series (100).
FIGURE 7-4 Violent shaking of a small child held by the chest whiplashes the head back and forth over the thorax, with the possible development of subdural hematomas as well as rib fractures. Infants held by the extremities are also at risk for long-bone fracture.
Pediatric spinal fractures and spinal cord injuries are rare with nonaccidental trauma.
An excellent review of pediatric spinal fractures in general was published by Akbarnia (101). These injuries often can be subtle, and a high index of suspicion is necessary. The actual incidence of these injuries may be underreported (102). Spinal fractures may occur in child abuse when a child is forcibly slammed onto a countertop with hyperflexion of the spine (17). Although neurologic deficit with this type of injury is uncommon (103), children should be carefully evaluated for signs of spinal cord injury secondary to spinal fracture when abuse is suspected.
Abdominal Injuries
Trauma to the abdomen is the second most common reason for death from child abuse (44). Careful evaluation is needed to rule out occult abdominal injury in an abused child, and often abdominal trauma is the presenting complaint in abused children. In a review of the National Pediatric Trauma Registry, 16% of all blunt abdominal trauma in a group of children 0 to 4 years of age was attributable to child abuse (104). These are injuries due to beatings with the hand or fist and also can occur when the child is thrown into a fixed object. External abdominal bruising is present in only 12% to 15% of cases of major abdominal injuries (75). Patients with abdominal injury resulting from child abuse may have fever, vomiting, and anemia, with abdominal distention, localized involuntary spasm, and absent bowel sounds (19). One of the most common abdominal injuries is a ruptured liver or spleen, and the hypovolemic shock from blood loss can be fatal (105). Blunt trauma to the abdomen also may cause intestinal perforation, usually involving the large intestine, and the physical examination suggests peritonitis with free air seen on abdominal x-ray. Intramural duodenal hematoma may cause obstruction and projectile vomiting (75). More severe trauma may cause duodenal avulsion or transection with nausea, vomiting, and clinical acute abdomen (106). Severe blunt trauma to the abdomen or a blow to the back may cause renal or bladder contusion with hematuria. Blunt trauma to the abdomen also may result in shearing of the mesenteric vessels, leading to hypovolemic shock due to blood loss. Pancreatic transection can occur where the pancreas overlies the vertebral column, and a pseudocyst may form (75).
Liver function tests are useful in detecting abdominal injury resulting from occult liver injury in child abuse. In one study (107), elevated aspartate aminotransferase, alanine aminotransferase, and lactic dehydrogenase enzyme levels were useful markers for occult liver lacerations in abused children who had false-negative abdominal examinations.
The American Academy of Pediatrics Section on Radiology (92) recommends CT scans with nonionic intravenous contrast to define injury to abdominal organs. Contrast should not be used if there is a history of iodine allergy, severe shock, or renal failure. The use of oral contrast is debatable with CT scans and may place the patient at risk of aspiration. If CT scan is not possible, peritoneal lavage may be considered in rare cases. Acute abdominal sonography is controversial and may not add useful information after CT scan, but may have worth in follow-up evaluations. An upper GI series is sometimes used to evaluate duodenal hematoma. If a bone scan is obtained to diagnose occult fracture, there may be abnormal patterns of isotope uptake in the abdomen as a result of intestinal injury, renal contusion, or muscle trauma of the abdominal wall (108).
When abdominal injury is suspected in an abused child, the hematocrit and hemoglobin levels are checked, the child is typed and cross-matched for blood, and an intravenous line is placed to provide replacement fluids. Emergency general surgery consultation should be obtained. The overall mortality rate associated with visceral injury in child abuse is 40% to 50% (107). In fatal cases with liver injury, hepatic glycogen staining may be helpful in establishing time of death for legal reasons (109).
Genital Injuries
Sexual abuse should always be considered when evaluating a physically abused child. Children who have been sexually abused can have symptoms of bed wetting, fecal incontinence, painful defecation, pelvic pain, abdominal pain, vaginal itching and bleeding, sexually transmitted diseases, and pregnancy in postmenarchal women. Types of sexually transmitted diseases found in abused children include gonorrhea, syphilis, chlamydiosis, trichomoniasis, and lymphogranuloma venereum. Although the percentage of sexually assaulted children with obvious physical trauma to the genitalia is low, failure to document such findings is a serious matter because sexual abuse is always a criminal offense and must be reported to legal authorities.
The genitalia should always be examined in a chaperoned setting. Infant and toddler girls are placed in a supine frog-leg position, and boys are placed in either a prone knee-chest position or in a lateral decubitus position (110). Patterns of injury that suggest sexually motivated assault include bruises, scratches, and burns around the lower trunk, genitalia, thighs, buttocks, and upper legs, including the knees. Pinch or grip marks may be found where the child was held. Attempted or achieved penetration may involve the mouth, vagina, or anus (50). Sexually abused boys may have poor rectal sphincter tone, perianal scarring, or urethral discharge. Female genital examination findings that are consistent with, but not diagnostic of, sexual abuse include chafing, abrasion, or bruising of the inner thighs or genitalia; distortion of the hymen; decreased or absent hymen; scarring of the external genitalia; and enlargement of the hymenal opening (111). The size of the transverse hymenal orifice does not correlate as a marker of child abuse (112). The examination of the female genitalia can be normal even when there has been penetration, because hymenal tissue is elastic and there can be rapid healing. In a study of 36 adolescent pregnant girls evaluated for sexual abuse evaluation only two

of 36 had genital changes diagnostic of penetrating trauma, suggesting injuries either may not occur or may heal completely (113). There also is a wide variability of appearance of normal female genitalia (98,114), but posterior hymen disruption is rare and should raise suspicion for abuse (115). If the sexual assault occurred within 72 hours of evaluation, then a rape kit must be used by the evaluating physician or nurse examiner to provide medical evidence of the attack (49).
The physical signs of sexual abuse, including genital trauma, sexually transmitted diseases, or presence of sperm are present in only 3% to 16% of verified sexual assaults (110,116), but even this minority of patients will be undiagnosed if sexual abuse is not considered when a child presents with musculoskeletal injury resulting from abuse. The orthopaedist or a consultant such as a pediatrician or a gynecologist must perform and document the genital examination in children with physical abuse.
Fractures documented on plain x-ray studies are present in 11% to 55% of abused children and are most common in children younger than 3 years of age (17,20,117). Fractures resulting from abuse should be suspected in a young child if the caretaker brings in the child for evaluation, reporting no accident, but does report a change in the child such as extremity swelling or decreased movement of the limb (118). Unexplained systemic hypertension in infants may be due to the pain of occult long-bone fractures (119). Femoral fractures are especially suspicious for child abuse in younger children. One study (120) found that 79% of patients younger than 2 years of age with femoral fractures were battered, and, of those, two thirds had femoral fractures as their only sign of abuse. Accidental femoral fractures can occur in children old enough to stand or run who fall with a twisting injury to the lower extremities, but femoral fractures in children younger than 1 year of age are most likely due to abuse (121). Fractures of both the lateral clavicle and the scapula are suggestive of abuse in young children (122). Infants may normally have a separate ossification center adjacent to the tip of the acromion, simulating a fracture (123), but a true fracture has sharp, demarcated edges, may be positive on bone scan, and will show callus on healing (Fig. 7-5). Although fractures of the sternum are believed to be specific for child abuse by Kleinmann (124), accidental midsternal fractures in children have been reported (125).
Fractures of the hands and feet are most commonly due to accidental trauma in older children (126) but are suspicious for abuse in infants. Nimkin et al (127) reviewed 11 hand and foot fractures in abused children younger than 10 months of age and found mostly torus fractures either of the metacarpals or the proximal phalanges of the hand and similar fractures of the first metatarsals of the feet. Clinical signs of fracture were present in only one patient, and bones scans were insensitive to the presence of the fractures in all patients.
All types of fractures have been reported in the child abuse literature. In one of the largest series, King et al (128) reported 429 fractures in 189 abused children. Fifty percent of these patients had a single fracture, and 17% had more than three fractures. Approximately 60% of fractures were found in roughly equal numbers in the humerus, femur, and tibia. Fractures also occurred in the radius, skull, spine, ribs, ulna, and fibula, in order of decreasing frequency. Another study (19) found a similar incidence of fractures of the humerus, femur, and tibia in abused children, with skull fractures seen in 14% of patients (Fig. 7-6). In contrast, Akbarnia et al (129) found that rib fractures in abused patients were twice as prevalent as fractures of any one long bone; the next most frequently fractured bone was the humerus, followed by the femur and the tibia. Nearly a third of their patients had skull fractures. Loder and Bookout (130) found the tibia to be the bone most commonly fractured in their series of abused children, followed by the humerus, the femur, the radius, and the ulna.
New Developments: Biomechanical Evaluation of Abuse Fractures and Computer Models of Childhood Injury
The past literature of child abuse orthopedic injury is based solely on clinical series and pathology specimens from autopsy specimens. For example, injury from falls from furniture by children are almost always accidental (64), except in infants (66), but little is known about the biomechanics of these injuries. Recently there have been attempts to study the biomechanics of the injuries of children in an abuse scenario using a test dummy and computer models of injury (131). Bertocci et al (132) conducted feet-first free-fall experiments using an instrumented Hybrid II simulated 3-year-old test dummy using varying heights and impact surfaces. They found that in falls on playground foam, increasing fall heights increased femoral compression loads and compression moments. Falls from short distances of 27 inches were associated with a low risk of contact-type head injury, but they also found that head acceleration was less for playground foam falls compared with falls on wood, linoleum, or padded carpet. In another study (133), the same authors examined the biomechanics of the same test dummy falling 0.68 meters, an estimation of typical bed height, through a roll onto different surfaces. They found that either the pelvis or the legs made first contact, and head deceleration, femoral loading, and pelvis deceleration were below known head injury criteria and lower extremity thresholds for injury. In 2001 they reported a computer simulation of pediatric stair falls (134) and in 2005 reported an “injury plausible” biomechanical model to assess femoral fractures in 29 children who fell down stairs. They found that with scoring for detail of history, biomechanical compatibility of fracture pattern, time to seeking care, and presence of other injuries, their model allowed them to distinguish between two groups of either plausible injury or suspicious

injury. In the plausible group a transverse femoral fracture had a 10-fold greater linear momentum compared to spiral or buckle fractures (135). Much more research is needed in this area.
FIGURE 7-5 A. A 7-week-old girl was seen in the emergency department with swelling of the left shoulder. History revealed her being roughly grabbed by the shoulder by one of her parents. The anteroposterior x-ray study shows a transverse lucency of the acromion, which was thought to represent either a congenital lesion or a traumatic fracture (arrow). B. A technetium bone scan was obtained. Although the delayed anterior image was normal, the posterior view showed increased uptake in the left shoulder in the area of the acromion (arrow) and a fracture was diagnosed. The incident was investigated. C. A follow-up x-ray 9 days later showed callus bridging the fracture site of the acromion (arrow).
Skull Fractures
Infants in the first year of life with fractures of the skull or the extremities have an equal risk of the etiology being either accident or abuse (136). Skull fractures were the most commonly reported fracture in one series (68), but only 10% of their patients underwent x-ray evaluation. Skeletal surveys missed 26% of skull fractures confirmed on CT scan in one study (93).
Extremity Fractures
There is no predominant pattern of diaphyseal fracture in child abuse. Traditionally, a midshaft spiral fracture is believed to be caused by a violent twisting injury to the extremity of the child. In a study of 23 long-bone fractures in battered children, Herndon (18) found spiral fractures in 78%. He suggested that in

children younger than 3 years of age, a spiral fracture of a long bone is highly suggestive of child abuse. Other authors (28), however, found that 71% of diaphyseal fractures were transverse in abused children. In another study of 34 patients in the first year of life, there was no difference in fracture patterns between those injured by accidental trauma and those injured by abuse. Loder and Bookout (130) reviewed 69 long-bone fractures in abused children and found that 56% were transverse, 36% oblique, and only 8% spiral. In another study of 429 fractures (88), 48% of fractures were transverse and 26% were spiral. Most of these long-bone fractures were in either the middle or distal third of the shaft. Transverse fractures are most commonly associated with either violent bending force or a direct blow to the extremities, whereas spiral or oblique fractures of the long bones are due to twisting injuries or a fall. Significant rotational force is required, however, to produce a spiral fracture with a fall (17). Humeral shaft fractures in children under age 3 years have a 18% risk of being due to probable abuse (137). In delayed follow-up, long-bone fractures may show exuberant callus because of a lack of immobilization, and multiple fractures may be present in different stages of healing (8). Juxtacortical calcification may be seen without fracture when there is diaphyseal periosteal separation resulting from tractional or torsional force when the limb is grasped or pulled along the shaft of the bone (138).
FIGURE 7-6 A. A 4-month-old male without history of trauma was brought into the emergency department by his parents with a history of decreased use of the arm. This distal humeral shaft fracture was seen on x-ray. B. Skeletal survey disclosed a posterior skull fracture (arrow), and the injuries were investigated.
Femoral fractures are especially suspicious for abuse when seen in infants; whereas children old enough to run can fall and accidentally fracture their femurs if there is a significant twisting motion at the time of injury (121). Comparing accidental femoral fractures to those of child abuse, there was no difference in site or pattern of fractures between groups in one study, but 93% of femoral fractures in those younger than 1 year were due to abuse (139). In one report (140) there was equal risk of having a spiral or transverse femoral fracture as a result of abuse. Gross and Stanger (140a) found that 65% of femoral shaft fractures were due to abuse in children younger than 1 year. Schwend et al (141) reported that 42% of femoral fractures in infants not walking were due to abuse while only 2.6% of femoral fractures in ambulatory toddlers were due to abuse. Blakemore et al (142) noted that only 2% of femoral fractures from age 1 to 5 years were due to abuse. It appears the most important predictor of abuse in childhood femoral fractures is an age under 1 year.
Metaphyseal and epiphyseal fractures of the long bones are

classically associated with child abuse (10,11). In toddlers, these fractures can occur when the child is violently shaken by the extremities (Fig. 7-7) with direct violent traction on or rotation of the extremity (126). Metaphyseal fractures may be characterized by impaction into the epiphysis, with profound production of new periosteal bone. Buckle fractures may be present in multiple sites, and these injuries seldom produce exuberant callus. Repeated injury causes irregular metaphyseal deformities. Periosteal avulsion produces new bone formation within 2 to 3 weeks of injury and can be confused with osteomyelitis (8). New bone formation may be delayed, however, in children with malnutrition. Metaphyseal fractures constituted 4% of fractures in one series (28) but fewer than 15% in another (65).
FIGURE 7-7 When a toddler is violently shaken by the extremities, long-bone fractures can occur through either direct violent traction or rotation of the extremity. These children are also at risk for closed-head injury because of violent motion of the head during the shaking episode (arrow).
Kleinman (124) ranked the specificity of skeletal trauma for abuse (Table 7-5). Distinguishing between an accident and child abuse is based on both location and type of fracture. He cautioned that both moderate- and low-specificity x-ray findings of child abuse become more highly specific when there is an inadequate explanation for the injury.
Corner Fracture of Child Abuse
The so-called pathognomonic fracture of child abuse is the corner or bucket-handle fracture of the metaphysis (8,17). On x-ray study, there is no gross displacement of the epiphysis, but a corner or chip fracture is seen at the edge of the involved metaphysis (Fig. 7-8). If a significant portion of the metaphyseal rim is involved, a bucket-handle fracture pattern is produced. Another fracture pattern that may be seen is a faint metaphyseal lucency just adjacent to the physis. Classically, these fractures have been believed to result from avulsion of metaphyseal fragments through the periosteal attachment of the epiphysis (10,143). Kleinman et al (144,145) challenged this view based on their histopathologic study at autopsy of metaphyseal fractures in abused infants. They found that in bucket-handle or corner fractures, there was actually a full-thickness metaphyseal fracture extending through the primary spongiosa of bone just above the zone of provisional calcification. This fracture corresponded to the zone of metaphyseal lucency occasionally seen on x-ray. Centrally, the amount of metaphysis remaining attached to the physis was thin, but peripherally the fracture line curved away from the physis so that a substantial metaphyseal rim remained attached to the physis. On x-ray study, this metaphyseal rim formed the basis for both corner and bucket-handle fractures (Fig. 7-9). In healing fractures, biopsy specimens showed metaphyseal extension of hypertrophied chondrocytes (146,147). These fractures are most likely caused by either violent shaking or traction injuries to the extremity and are specifically suggestive of child abuse (17) (see Fig. 7-8); however, they are not the most common fractures in abused children, with their incidence in large series ranging from 15% to 32% (28,65,128,130). Subepiphyseal-metaphyseal lucency


also can be caused by systematic disease such as rickets and leukemia. “Corner fractures” of the distal radius, ulna, tibia, and proximal humerus also have been reported with developmental coxa vara associated with spondylometaphyseal dysplasia (148).
TABLE 7-5 Specificity of Skeletal Trauma for Abuse
High specificity
  • Any metaphyseal lesion
  • Posterior rib fracture
  • Scapular fracture
  • Spinous process fracture
  • Sternal fracture
Moderate specificity
  • Multiple fractures, especially bilateral
  • Fractures of different ages
  • Epiphyseal separation
  • Vertebral body fracture or subluxation
  • Digital fracture
  • Complex skull fracture
Low specificity
  • Clavicular fracture
  • Long-bone shaft fracture
  • Linear skull fracture
Data from Kleinman PK, ed. Diagnostic Imaging of Child Abuse. Baltimore: Williams & Wilkins, 1987; with permission.
FIGURE 7-8 A. An 11-month-old girl with a history of right thigh swelling was seen. X-rays of the right femur show corner or “chip” fractures of the distal femur (bottom arrows). Faint periosteal calcification is also visible along the lateral cortex (top arrow). Such fractures raise suspicion for child abuse. B. A technetium bone scan showed increased uptake of isotope in the distal and proximal femur (arrows). C. A healing physeal fracture of the proximal femur is seen (arrow).
FIGURE 7-9 In the corner or bucket-handle fracture, a full-thickness metaphyseal fracture extends through the primary spongiosa just above the zone of provisional calcification. Centrally, the amount of metaphysis remaining attached to the physis is thin, but peripherally the fracture line curves away from the physis so a substantial metaphyseal rim remains attached to the physis (arrows). On x-ray study, this metaphyseal rim (dark shaded area) forms the basis for both the corner and bucket-handle fractures.
Rib Fractures
Rib fractures are uncommon in childhood accidents, so their presence is very suspicious for child abuse, especially when other long-bone fractures are present (149). Abusive rib fractures may be caused by squeezing of the chest by a caretaker (14), hitting the child from behind, or stepping on the chest (150,151). Kleinman et al (152) postulated that severe shaking of an infant (see Fig. 7-1) can cause front-to-back chest compression, which levers the posterior rib over the transverse process of the vertebral body, causing fractures of the posterior rib shaft at the transverse process and of the rib head adjacent to the vertebra (Fig. 7-10). These posterior rib fractures are difficult to diagnose acutely because of a lack displacement. Even with healing, the callus on x-ray study may be obscured by the overlying transverse process (31), and a bone scan may be needed to make the diagnosis (Fig. 7-11). Posterior rib fractures are believed to be most common in child abuse, but fractures may occur anywhere along the arc of the rib, including disruption of the anterior costochondral junction (Fig. 7-12). Posterior paraspinal rib fractures tend to occur between T4 and T9, rib fractures in the axillary line occur mostly in the lower rib cage, and disruption of the anterior costochondral junction usually occurs from the second to the ninth ribs (153).
Lateral rib fractures, both acute and chronic, may be difficult to see on standard AP x-rays, so oblique views may be necessary. Healing fractures show early callus, but healed fractures may be subtle, with only a fusiform thickening of the rib (see Fig. 7-12). Old fractures of the ribs in child abuse may form lytic, expansile lesions that mimic a tumor (154). Fractures along the lateral arc of the rib are likely due to AP compression of the chest (Fig. 7-13), while lateral compression of the chest likely causes fractures of the rib along its posterior arc lateral to the transverse process as well as disruptions of the costochondral junction (Fig. 7-13B). Acute anterior costochondral separations of the ribs may be difficult to see on x-rays, but ultrasound can show costochondral dislocation (155), and, with healing, the anterior end of the osseous rib becomes widened and clubbed on x-rays (19,65). They are commonly associated with abdominal injury and can be detected on CT scan, resembling bucket-handle metaphyseal fractures (156).
FIGURE 7-10 With anterior compression of the chest, the posterior rib is levered over the transverse process of the vertebral body (open arrow), causing fractures of the rib head adjacent to the vertebra (shaded arrow) as well as avulsion fractures of the rib at the transverse process.
In the rare instances when rib fractures are discovered in abused infants undergoing resuscitation for cardiac arrest, there may be confusion about the etiology of the fractures (157), but the elasticity of the infant chest seems to enable it to tolerate compressions, with only 2% of 94 nonabused infants resuscitated having rib fractures in one series (158). Death from cardiac arhythmia from a blow to the chest has been reported in a 7-week-old abused infant whose rib fractures at autopsy were initially thought to be due to resuscitation efforts (159). In addition to rib fractures, abused infants can sustain severe lung contusion and respiratory distress from chest wall trauma (160), with fatal fat embolus reported (161). Necklace calcifications may be present in strangulation cases (162).
In the case of infant fatalities of suspicious origin, postmortem high-detailed skeletal surveys and specimen x-rays can be helpful in diagnosing child abuse (163). In a postmortem study of 31 infants who died of inflicted skeletal injury, Kleinman et al (164) found a total of 165 fractures (51% rib fractures, 39% metaphyseal long bone fractures, 5% long-bone shaft fractures, 4% fractures of the hands and feet, 1% clavicular fractures, and less than 1% spinal fractures). It is important to obtain skeletal


x-rays before autopsy to avoid artifact (P. Kleinman, Personal communication, 1998).
FIGURE 7-11 A. This 5-week-old infant was presented by her parents with a complaint of irritability when her left leg was handled. Her mother stated the infant’s 18-month-old sibling may have “kicked” her in the leg. X-rays showed an acute fracture of the proximal tibia. B. Skeletal survey showed a healing fracture of the left distal clavicle (arrow), a healed fracture of the anterior left 5th rib (arrow), and probable healing of the posterior right 6th, 7th, 8th, and 9th ribs (arrows). C. The posterior image of the delayed phase of the bone scan confirmed the presence of healing posterior rib fractures (arrows).
FIGURE 7-12 A. An anteroposterior “baby-gram” x-ray of an infant who died without clear-cut reason. Note the “halo” calcifications around the anterior rib costochondral junctions, equivocal at the second rib (black arrows), definitive at the 3rd and 4th ribs of the right hemithorax (white arrows) suggestive of anterior costochondral junction rib fracture. On the patient’s left side, note the fusiform swelling of the 3rd and 4th ribs posteriorly (black arrows), which suggests nondisplaced posterior rib fractures, and a minimally displaced fracture of the 10th rib posteriorly (large black arrow). B. A close-up view of the right hemithorax. Note the subtle calcification around the 2nd rib costochondral junction, suggesting fractures from the 2nd to the 4th rib.
Spinal Fractures
Spinal fractures in child abuse are rare but can occur when a child is forcibly slammed onto a flat surface, with hyperflexion of the spine (17,101). Based on autopsy findings (165), spinal fractures of fatally abused children generally involved 25% or less compression of the vertebrae. Half of the fractures involved the anterosuperior end plate associated with a compression deformity, 30% had pure compression fractures, and 20% had fractures of the superior end plate without significant compression. In another study of fractures of the cervical spine, prevertebral soft tissue edema on x-ray was the only sign of cervical injury, because spontaneous reduction of the cervical vertebrae after dislocation was common (166). In a report of 103 children with cervical spine injury, only three patients had injury due to abuse and all had spinal cord injury without radiographic abnormality (SCIWORA) (167). Thomas et al (168) reported a 9-week-old boy with spinal cord injury resulting from cervical spine fracture who presented as a floppy infant. Although routine cervical x-rays were normal, MRI studies showed retropulsion of a fragment of the primarily cartilaginous C3 vertebrae into the spinal canal. Kleinman and Shelton (169) reported a hangman’s fracture in a 6-month-old infant with multiple rib fractures that went on to nonunion and required fusion. This must be distinguished from C-2 primary spondylolysis which may be associated with pyknodysostosis (170). Ravandrin et al (171) reported a 23-month-old child with a hangman’s fracture resulting from abuse that healed with immobilization. The CT scan was helpful in showing the fracture, but a bone scan was nondiagnostic. Rooks et al (172) reported a compression fracture of C-5 with anterior subluxation of C-4 on C-5 in a 3-month-old premature twin requiring decompression and cervical

fusion. The other twin had a C-5 on C-6 fracture-subluxation treated with casting but later required surgery to reduce and fuse the subluxation. MRI was very helpful in showing cord compression in both cases. Carrion et al (173) reported circumferential physeal fractures of the spine associated with child abuse that required open reduction. Positioning premature infants in extreme flexion for lumbar puncture may cause iatrogenic lumbar spine fracture (173a). Bone scans using single photon emission computed tomography can be helpful for diagnosing occult compression fractures of the thoracic spine (153). Although neurologic injury in spinal fractures resulting from child abuse is uncommon (103), any patient with abusive spinal injury should undergo complete neurologic examination.
FIGURE 7-13 A. Anteroposterior compression of the infant chest causes fractures both at the rib head and adjacent to the transverse processes (open arrow), but fractures along the lateral arc of the rib are also possible because of the acute bending of the ribs at this point by compression (shaded arrow). B. Lateral compression of the chest likely causes fractures of the rib along its posterior arc lateral to the transverse process, as well as possible disruption of the costochondral junction (open arrows).
The Skeletal Survey
In addition to standard x-ray studies of the acute injury, a skeletal survey is used to detect the presence of additional fractures in battered children (143). The standard skeletal survey recommended by the American College of Radiology (174) includes the following:
  • AP of the humeri
  • AP of the forearms
  • Oblique and AP of the hands
  • AP of the femurs
  • AP of the lower legs
  • AP of the feet
  • AP and lateral of the chest
  • AP of the pelvis, including mid and lower lumbar spine
  • Lateral of lumbar spine
  • Lateral of cervical spine
  • AP and lateral of skull
The technique should have high resolution, high contrast, screen/film speed of no more than 200, low kVp (bone technique), and single emulsion or special film-screen combination.
An oblique view of the thorax can be helpful in diagnosis of subtle rib fractures and should be added to the skeletal survey when indicated. Oblique x-ray studies of the hand are recommended to detect subtle torus fractures of the metacarpals and the phalanges (127). The indications for skeletal survey are not completely clear. Merten and Carpenter (138) recommended skeletal survey in infants 1 year of age or younger when there is evidence of neglect and in children age 2 years or younger when clinical abuse is evident. The American Academy of Pediatrics Section on Radiology (92) considered a “baby gram” to have no value in diagnosing the fractures of child abuse because the obliquity of the angle at which the x-ray transverses the skeleton may obscure subtle fractures (153). Imaging systems should have a spatial resolution of at least 10 line pairs per millimeter and should be used without a grid (174).
A skeletal survey is mandatory in all cases of suspected physical abuse in children younger than 2 years of age (92), but the sensitivity of skeletal surveys is unclear in older patients. The cost-effectiveness of such skeletal surveys appears to be low because in one study of 331 children, only eight patients without overt physical signs of child abuse had occult fractures revealed by the survey (175), but the use of the skeletal survey in these eight patients possibly prevented both reinjury and death. Perhaps when the death of a child is the ultimate risk of the misdiagnosis of child abuse, we should not play the odds but use very best medical test at our disposal to identify the child at risk for abuse so we can try to protect him or her from further possible fatal abuse.
Sty and Starshak (176) reported a false-negative rate of 12.3% in skeletal surveys of abused children and suggested that a technetium bone scan is the best screening test for occult

fractures. Technetium bone scans, however, may fail to diagnose either epiphyseal or metaphyseal fractures and often fail to show skull fractures (126,176). Although CT scans may miss skull fractures, they may conversely show skull fractures even when skeletal surveys are negative (93). Jaudes (177) found that when results of either a bone scan or a skeletal survey were normal in an abused child, the use of both tests often revealed additional occult fractures. Technetium bone scans are especially useful in the diagnosis of occult rib fractures (151,153), but consistent interpretation is difficult in children younger than 18 months of age. Technetium scans are not useful for dating fractures, because increased isotope uptake may occur at a fracture as early as 24 hours after injury and scan abnormalities may persist for years (178). Kleinman et al (150) reported that a follow-up skeletal survey 2 weeks after the initial series detected 27% more fractures and provided assistance in dating 20% of previously detected fractures.
Digital radiography through the picture archiving and communication system (PACS) is gradually replacing standard film-screen imaging in many hospitals, and its role in the detection of the subtle fractures of child abuse is becoming controversial. Youmans et al noted that child abuse fractures can be missed by digitalized images (179). Kleinman et al (180) noted that digital imaging of child abuse fractures had a spatial resolution lower than film-screen imaging, but the difference was not appreciable in detecting rib fractures in a postmortem evaluation.
Additional Imaging Studies
Other imaging techniques may be useful in investigating suspicious x-ray or bone scan findings. Markowitz et al (181) found that sagittal and coronal sonograms of injured knees in abused children were helpful in diagnosing epiphyseal fractures when compared with the normal side. They cautioned that epiphyseal fractures resulting from rickets in neonates can be confused with child abuse on ultrasound evaluation. Rogers and Poznanski (182) pointed out that acute interruption of the physis in trauma of any etiology can be seen on MRI. They recommended multiplanar gradient-echo imaging with a repetition time of 50 to 700 ms, an echo time of 200 ms, and a flip angle of 20 degrees for optimal imaging of the physis.
Dating Fractures
X-ray proof of unexplained fractures in various stages of healing is believed to be strong evidence of child abuse (Fig. 7-14) (129). The orthopaedist often is asked to estimate the age of


fractures with some certainty to corroborate a history of injury given by caretakers (Fig. 7-15). Whereas most experienced orthopaedists can roughly estimate the age of fractures based on x-ray appearance, specific guidelines have been established for estimating the age of fractures in children (183). Growth lines present on x-rays may be helpful in dating fractures (184). In general, fractures on x-rays are acute until callus appears, then their age is based on the presence of either soft or hard callus some weeks later (Fig. 7-16). The most difficult fractures to date are those that are completely healed, with substantial remodeling, and often the only sign of fracturing is a thickened cortex (Fig. 7-17.
FIGURE 7-14 A. A 7-month-old boy presented with a nondisplaced left distal tibial fracture with periosteal calcification extending up the tibial shaft (arrows). B. A skeletal survey revealed dense, sclerotic cortical thickening of the contralateral tibia consistent with healed fracture (arrow). The presence of multiple fractures in different stages of healing is very suggestive of child abuse.
FIGURE 7-15 A 3-year-old female was brought to the clinic with a swollen elbow. The mother stated that the child had fallen only a week ago, but x-rays of the elbow showed a healing supracondylar fracture of the distal humerus that was likely 2 to 3 weeks old based on presence of soft callus (arrows). Child protective services was contacted, and a full examination including skeletal survey was performed. No other signs or risk factors for child abuse were detected. The mother then admitted that the child had indeed fallen 3 weeks ago. She was allowed to take her child home after treatment and was counseled about the need for prompt medical care when her children were injured.
FIGURE 7-16 A. A 6-year-old patient presented with a nondisplaced transverse fracture of the distal radius (arrow). B. At 4-week follow-up, soft callus is seen enveloping the fracture site (arrow). C. At 6-week follow-up, hard callus is seen and early remodeling is occurring at the fracture site.
FIGURE 7-17 A. An 18-month-old child presented with a 1-month-old nondisplaced fracture of the distal tibia (large arrow). The fracture line is beginning to fade, but periosteal calcification is seen medially (small arrows). B. At 5 months after injury, the fracture shows almost complete remodeling (large arrow) with the only sign of past injury a thickened medial cortex of the tibia (small black arrows). Subtle signs of past fracture are best noted by comparing the x-rays of the injured side to the contralateral uninjured side.
An abused child should have a complete blood cell count with sedimentation rate, liver function studies, and urinalysis. Clotting studies should be performed routinely, especially in patients with ecchymosis, to rule out a blood disorder as a cause for the bruises. If there is any suspicion of substance abuse by any family member, a toxicology screen should also be performed on the patient (28).
Any significant nonorthopaedic injury should prompt appropriate consultations by neurosurgery, general surgery, plastic surgery, or ophthalmology specialists (8). Any female patients who require pelvic examination to rule out sexual abuse should have a gynecologic consultation.
Although it is extremely important not to miss the diagnosis of child abuse, it is equally important not to make the diagnosis in error. Kaplan (185) pointed out that overdiagnosing battered child syndrome can be harmful to the family, with the parents being placed at risk of losing custody of their child and also facing criminal charges. Even direct allegations of child abuse may turn out to be false. Bernet (186) pointed out that patients or family friends may make false statements about an abuse situation through misinterpretation, confabulation, fantasy, delusions, and other situations. The American Academy of Child and Adolescent Psychiatry (187) has published guidelines for the evaluation of abuse that state that the possibility of false allegations needs to be considered, particularly if allegations are coming from the parent rather than the child, the parents are engaged in a dispute over custody or visitation, or the child is a preschooler.
Normal metaphyseal x-ray appearance should not be confused with child abuse. The x-ray metaphyseal variants include acute angulation of the ossified peripheral tip of the metaphysis adjacent to the physis (most commonly seen in the proximal tibia, distal femur, proximal fibula, distal radius, and distal ulna), which is bilateral in 41% of individuals. A bony beak may be seen medially in the proximal humerus and tibia in rare cases and is bilateral in 77% of individuals. Cortical irregularity in the medial proximal tibia may be seen in 4% of individuals and is bilateral in 25%. Spurs may extend beyond the metaphyseal margins in both the distal radius and lateral aspect of the distal femur, with bilateral variants in 25% of individuals (188).
The signs of child abuse found on x-ray study also can be mimicked by systemic diseases such as scurvy, osteogenesis imperfecta, Caffey disease, osteomyelitis, septic arthritis, fatigue fracture, osteoid osteoma and other tumors, rickets, leukemia, hypophosphatasia, neuromuscular disease, metastatic neuroblastoma, congenital indifference to pain, osteopetrosis, kinky hair syndrome, and prostaglandin therapy (8,44). Fractures resulting from both vitamin D deficiency rickets and druginduced rickets have been mistaken for child abuse (189). Children with biliary atresia may present with osteopenia and fractures without history of significant injury and should not be

mistaken for child abuse (190). Rib fractures are commonly seen in rickets of prematurity as well as rickets of low birth weight and also have been reported after chest physiotherapy (191). There has been an increase in the incidence of syphilis in females of childbearing age, and congenital syphilis can mimic fractures of child abuse with diaphysitis, metaphysitis, and multiple pathologic fractures in different stages of healing (192). Physiologic periostitis (Fig. 7-18), in contrast to lesions from child abuse, is usually bilateral with no excessive uptake of isotope on bone scan (153). The presence of metabolic disease and pathologic fractures does not exclude the possibility of child abuse. Duncan and Chandry (193) reported a 3-month-old girl with multiple fractures associated with rickets who died suddenly at 5 months of age. Child abuse was suspected but not proven. Three years later, evidence of child abuse was found in a subsequent sibling in the same family. Several diseases are commonly brought up in custodial hearings as alternative possibilities to child abuse to explain trauma findings, and these diseases should be carefully explored in the differential diagnosis.
Line lucencies of the proximal tibia resulting from intraosseous vascular access needles may mimic fracture (194). Metaphyseal corner fractures of the distal tibia and fibula were seen in eight children treated with serial casting for clubfoot, with only one thought to be associated with abuse (195). The senior author (RMC) has seen bilateral fractures of the distal tibia and fibula in an infant undergoing serial casting for clubfoot whose father forcedly pulled off the casts.
McClain et al (196) reported a 2-year-old child who died of undiagnosed acute lymphoblastic leukemia. The child had been reported as a possible victim of child abuse because of ecchymosis on the back and extremities. They emphasized that the clinical signs of leukemia, including fever, pallor, petechia, purpura, adenopathy, hepatosplenomegaly, and bone pain, should be sought in children with bruising of unknown origin. Factor XIII deficiency may cause unexplained bleeding from minor trauma and be mistaken for child abuse because the standard coagulation profile may be negative and factor-specific tests may be negative if performed post-transfusion (197).
FIGURE 7-18 A 10-month-old infant with physiologic periostitis of the femurs (arrows). This type of x-ray appearance may be confused with fractures of child abuse, but this periostitis tends to be bilateral with a uniform appearance in contrast to multiple fractures in different stages of healing seen in battered children.
Osteogenesis Imperfecta
Undiagnosed osteogenesis imperfecta should always be considered when a child presents with multiple fractures of unknown etiology, but it may be a very difficult diagnosis clinically. Osteogenesis imperfecta caused by spontaneous mutation can occur without a family history (198). The so-called hallmark of osteogenesis imperfecta is an intensely blue sclerae, but this feature is consistently present only in type I (199) and may be completely absent in patients with type IV osteogenesis imperfecta (198). Sillence and Butler (200) noted that patients with either type II or III osteogenesis imperfecta may have blue sclerae at birth but the sclerae can become normal by adolescence. The rare type II osteogenesis imperfecta has normal sclerae, but bone abnormalities and osteopenia are severe and early death is likely (198). Blue sclerae may be present in normal young children and can be misleading as a sign of pathologic bone fragility. The presence of abnormal teeth, known as dentinogenesis imperfecta, may be helpful in a diagnosis of osteogenesis imperfecta if the child is old enough for teeth to have erupted. Plain x-ray studies, however, may show long bones of normal density in both types I and IV osteogenesis imperfecta. Another x-ray sign of osteogenesis imperfecta, wormian bones of the skull, is consistently present only in type III and is often absent in types I and IV (198). Some authors believe that the presence of metaphyseal fracture is pathognomonic for child abuse and, therefore, helpful to distinguish abuse from osteogenesis imperfecta (201), but other authors (198,202) believe that there is no particular fracture pattern that renders the diagnosis of osteogenesis imperfecta likely. Patients with osteogenesis imperfecta tend to bruise to excess, which simulates lesions of child abuse (203), and sudden infant death has also been recorded in patients with undiagnosed osteogenesis imperfecta (204).
Sometimes, when the diagnosis of osteogenesis imperfecta cannot be made on clinical grounds, the diagnosis can be made by biochemical assay. Gahagan and Rimsza (205) stated that 87% of patients with osteogenesis imperfecta have abnormal procollagen that can be detected by current techniques. A skin biopsy is performed for fibroblast culture, and fibroblasts are assayed for both abnormally low levels of procollagen and primary abnormal procollagen (206). Steiner et al (207) reported that over a 4-year period, 48 patients were referred to them for collagen analysis to rule out the presence of osteogenesis imperfecta in cases of suspected child abuse. Only six of these children had abnormal collagen test results, and in five of those six patients, the diagnosis of osteogenesis imperfecta could have been made on clinical and x-ray grounds. They concluded that routine collagen biochemical testing for osteogenesis imperfecta

is unwarranted in these children, and collagen analysis should be reserved for the rare instances when diagnostic uncertainty persists in cases of suspected child abuse.
Even when a child has osteogenesis imperfecta, fractures may be due to abuse. Knight and Bennett (208) reported on a young child with osteogenesis imperfecta whose abuse could not be proved until linear bruising of the face suggestive of slapping was documented.
Temporary Brittle Bone Disease
In 1993, Patterson et al (198) described 39 patients with a variant of osteogenesis imperfecta that they described as a temporary brittle bone disease in which fractures were limited to the first year of life and then there was spontaneous improvement. These patients presented with vomiting, followed by diarrhea, anemia, hepatomegaly, episodes of apnea, neutropenia, and edema. The most common x-ray findings were metaphyseal corner fractures, rib fractures, diaphyseal fractures, periosteal reaction of long bones, expanded costochondral junctions, and delayed bone age. Only 31% of patients had osteopenia on x-ray study. They suggested that a self-limiting period of copper deficiency was the cause of this problem, but limited serum copper assays were inconclusive. Other authors (209,210,211) doubt the existence of “temporary” brittle bone disease because of the rarity of fractures associated with proven copper deficiency syndrome, but another author (212,213,214,215) supports the existence with indirect evidence. Judicial authorities (51,216,217) have commented that although one patient in the series of Patterson et al had injuries resulting from child abuse, this fact was not included in the report, and they had concerns that assumptions proposed by medical experts that injuries may be considered solely the result of disease may inhibit full investigation of such injuries by civil authorities and place children at risk for further abuse.
Sudden Infant Death Syndrome
In sudden infant death syndrome (SIDS), there is a distinct possibility of child abuse (33), with one study noting 50% of 81 infants dying from child abuse had an initial diagnosis of SIDS before the correct diagnosis was made (218), but other causes of sudden death must be excluded. Byard et al (219) reported a 5-month-old girl who died suddenly because of spontaneous subarachnoid hemorrhage from undiagnosed Ehlers-Danlos syndrome. They recommended collagen analysis in patients with unexplained multifocal spontaneous hemorrhages to exclude this rare syndrome. Sperry and Pfalzgraf (220) reported a 9-month-old infant whose diagnosis of sudden infant death syndrome became uncertain when postmortem x-rays showed healing symmetric clavicular fractures and a healing left medial humeral epicondylar fracture. Subsequent investigation showed that the child had undergone “chiropractic” manipulation 4 weeks before death by an unlicensed therapist to correct “shoulder dislocations,” and the parents were exonerated of abuse charges.
Accidental Trauma
In considering the differential diagnosis of child abuse, accidental trauma should always be considered. The orthopaedist, however, should be comfortable with the diagnosis of accidental trauma only when the acute injury is brought promptly to medical attention and has a plausible mechanism of injury and there are no risk factors for child abuse.
Once child abuse is recognized, the first step in treatment is hospital admission. This is therapeutic in that it places the child in a safe, protected environment and provides the opportunity for additional diagnostic workup and, more importantly, investigation of the family’s social situation by appropriate personnel. In university settings, multidisciplinary teams often are available to evaluate and treat such children, but in other circumstances, the orthopaedist may be primarily responsible for coordinating both evaluation and treatment. Court custody may be required for children of uncooperative families who refuse admission, and hospitalization should be continued until a full investigation is completed by the appropriate child protective services. In the United States, the physician is required by law to report all suspected child abuse to appropriate child protective services or legal authorities. When the reporting is done in good faith, the physician has immunity against criminal or civil liability for these actions, but in only three states=mOhio, California, and Alabama=mis this protection extended to include absolute immunity (221). The distinction is critical. Absolute immunity means that the physician who reports suspected child abuse cannot ever be held for damages sought by families for allegedly inappropriate reports of child abuse or neglect. The granting of absolute immunity, even for physicians, is not encouraged by the American legal system because in theory it would protect individuals who make false reports of child abuse in order to harass families and would deprive the injured parties their legal right to seek damages for harmful actions. In contrast, physician immunity based on good faith reporting of suspected child abuse is contingent on the physician having a reasonable belief that abuse or neglect has occurred. Although in theory this protection seems to be quite adequate, recently there has been a dramatic rise in the number of lawsuits filed by families seeking damages for alleged, unfounded reports of child abuse and neglect. Although it is true that by the time these lawsuits are eventually resolved, physicians have almost never been held liable for good faith reports of child abuse, in a substantial number of these cases, the physicians first lost at trial level before eventually prevailing at appeal. Considerable expense, frustration, and loss of time can be experienced by the physician in defending against such allegations as the families and their attorneys pursue multiple forms of legal theories in court an attempt to evade the immunity provisions (221). On the other hand, failure to

report suspected child abuse may expose the physician to charges of malpractice (17). All states require physicians to report not only cases of definitive child abuse or neglect but also cases when abuse is just suspected or is considered a possibility. Physicians have been held liable for damages for their negligence in failing to diagnose child abuse when the child subsequently was reinjured by more abuse, and, ironically, the parents also may be able to collect additional compensation for losses resulting from medical expenses. For families to be successful in these lawsuits, they must be able to prove that the failure to make the diagnosis of child abuse was negligent and that, had the diagnosis been made, steps would have been taken to protect the child from additional abuse. Although the probability of a physician being held liable under such circumstances is low, the amount of damages can be high if the family does prevail when the child has suffered permanent sequelae (221). Researchers of child abuse have no legal requirement to report suspected child maltreatment, so they may be exposed to both civil and criminal liability for even “good faith” reports that may turn out to be false and injurious (222).
After admission, the orthopaedist proceeds with care of the child’s musculoskeletal injuries and coordinates various medical consultations. There should be frequent communications with child protective services to stay current with the results of their investigations. Recommendations for disposition of the child after completion of medical treatment may be a group decision through a multidisciplinary team or, more often, the decision of the primary physician, who may be the orthopaedist. Final disposition choices may include return to the family, return to a family member who does not live in the child’s home, or placement in a shelter or a foster home setting. The risk of reinjury and death is significant if the abused child is returned to the unsafe home, so the orthopaedist must strongly support child protective services in custodial actions when it is believed that a child’s injury truly occurred from abuse at home. Not only must the definitive diagnosis of child abuse be documented in the chart but a separate notarized affidavit may be necessary. Commonly, custodial actions by child protective services are reviewed in a court hearing in a matter of weeks, and the physician is likely to be called to testify in the hearing. Criminal charges also may be brought against the perpetrator of the child abuse, and the physician likely also serves as a witness in these proceedings.
The Orthopaedic Legal Complications of Child Abuse
The orthopaedist fills a dual role in the courtroom in child abuse proceedings. First, he or she serves as a material witness whose testimony is confined to the physician’s personal involvement in the legal matter. The testimony may include clarification to the court of information contained in progress notes in the chart or of other past documentation. As a material witness, the physician, like the layman, cannot render opinions about the facts as stated during his or her testimony. In addition, however, the physician may also be sworn in as an expert witness (223). This is an individual considered by the court to have special knowledge and experience that qualifies him or her to render opinions about certain facts presented in the courtroom. The limits of the physician’s expertise usually are defined by the attorneys in court before the testimony of the expert witness.
Physicians usually are reluctant to testify in court for many reasons. The courtroom is an unfamiliar setting for almost all physicians, and the adversarial nature of the American law system makes it a hostile environment. In the courtroom, the perception of truth is just as important as the truth itself, and opposing attorneys will search for inconsistencies in the testimony or unfamiliarity with the record to discredit the physician witness. To avoid being a poor witness, the orthopaedist must meticulously prepare to give testimony.
The orthopaedist preparing to testify in a child abuse case should begin with a thorough review of the child’s medical records and a review of recent medical literature on the subject of child abuse (223). Often, there is a pretestimony discussion with child protective services counsel in family court cases or the district attorney’s office in criminal cases. Such meetings should preferably be in person, and the orthopaedist’s professional training and expertise are examined to determine whether he or she may serve as a material witness, an expert witness, or both. The attorney should be provided the orthopaedist’s curriculum vitae, and another copy should be made available to the court. If the orthopaedist is to serve as a material witness, the factual information of the case as well as the limitations of the physician’s knowledge are discussed, as are questions that may be posed during testimony. Orthopaedists functioning as expert witnesses should indicate relevant information that should be provided through questioning during testimony. In addition, anticipated testimony from any opposing expert witness and cross-examination questions from the opposing attorney should be discussed. The opposing attorney also may request an informal pretestimony meeting. The orthopaedist should request a list of questions that will be asked in this session ahead of time and request that both the prosecution attorney and the opposing attorney be present during the session, which often is recorded. The next step may be a deposition in which both attorneys question the witness under oath to “discover” the testimony that the witness will provide in court. The primary purpose for a deposition in the discovery process is to keep attorneys from later being surprised in court by testimony of witnesses (224). Preparations for depositions should be meticulous. Any testimony the physician gives during the deposition will be recorded, and later in court any inconsistencies between testimony and prior depositions will be vigorously attacked by attorneys in cross-examination. Depositions rarely are used in criminal prosecutions (223).
A subpoena is issued requiring a physician witness to appear at the courtroom at a certain time, but often there may be hours of delay before the testimony actually begins. Through prior arrangements with the attorneys, the orthopaedist may be placed “on call” if he or she works within a reasonable distance

of the courtroom and can be available a short time before the actual testimony is needed. The physician has no legal right to such treatment and must be prepared to honor the exact conditions of the subpoena if alternative arrangements cannot be made. If significant delays are encountered to giving testimony and the attorneys are not responsive to physician hardship, then the orthopaedist should contact the judge directly to remedy the situation (224). In the courtroom, the orthopaedist should be conservatively dressed and appear attentive, competent, poised, and at ease (223,224).
Once called to the stand, the orthopaedist is sworn in and identified. Next follows qualification, direct examination, and then cross-examination. In the qualification process, the attorney asks the physician fairly detailed questions about the orthopaedist’s training and background to establish whether he or she is a credible witness (223). The attorney wishes to impress the judge or jury with the orthopaedist’s qualifications as a witness, whereas the opposing attorney may challenge the witness with questions to cast doubt on his or her expertise (224). During this phase, the attorneys also may establish the limits of the physician’s expertise as an expert medical witness. Next, the attorney will proceed with direct examination. A series of questions are asked that aim at developing a logical and progressive line of thought leading to a conclusion (224). In child abuse cases, in particular, the testimony will lead to the fact that the abuse has occurred and that it has been appropriately diagnosed. In addition, the physician expert witness may be asked to give an opinion of the risk for subsequent abuse if the child returns to the home where the alleged abuse occurred. Almost never will the physician witness be asked about the guilt or innocence of the caretaker accused of abuse, but the orthopaedist in certain circumstances will come close to answering the “ultimate question” (224), by testifying about a child’s statement of history if it identifies the abuser. Some states, however, restrict such testimony. In Maryland, a physician may not testify regarding any disclosures made by a child abuse victim unless the disclosure is admissible under a recognized exception to the rule prohibiting hearsay evidence (69). The orthopaedist should ask about any possible restrictions on his or her testimony with the attorney in pretrial discussion. In testimony, the orthopaedist will want to use the courtroom setting to advocate for the safety and well-being of the child (223). Questions regarding medical findings often will be prefaced in the courtroom by the words “reasonable medical certainty,” a term that is poorly understood by most physicians. Chadwick (224) offered a definition of reasonable medical certainty as “certain as a physician should be in order to recommend and carry out treatment for a given medical condition.” He offered an example that the certainty for the diagnosis and treatment of leukemia must be much higher than that for diagnosis and treatment of a viral upper respiratory tract infection.
During testimony, the orthopaedist’s words should be carefully chosen and should be understandable by a lay jury. Testimony should be objective, honest, and thorough (223). Attorneys may frame questions in ways that are difficult to understand, and the orthopaedist should not hesitate to ask the attorney to clarify a question (224). Answers should be brief, without volunteering extra information, but the perception listeners will have of the answers should be carefully considered by the orthopaedist. In particular, attorneys may phrase yes or no questions that could place misleading words in the mouth of the orthopaedist. In such situations, when neither response is appropriate, the orthopaedist should answer in a sentence that provides an accurate answer (223). Language should be straightforward, and visual aids may be used in providing clear testimony. The expert should use testimony as an educational process for the court, in which the common experience and knowledge of the jury is used to build understanding with common sense explanations of medical findings (224).
Cross-examination by the opposing attorney follows direct examination. The opposing attorney’s role is to challenge the material presented by the physician witness to protect the defendant (223). This may involve an attempt to bring into question the physician’s credibility, the medical record, the physician’s training or expertise, or the physician’s objectivity or composure and clarity of thought before the jury (223). Attorneys may accomplish this by finding inconsistencies with prior statements, asking leading questions as well as questions that allow only certain desired answers, and minimizing physician qualifications (224). The attorney may frame a question that contains certain elements that the physician agrees with and others that are misleading, and often the question will end with “Isn’t that so, doctor?” The physician witness should be firm in answering such questions, clearly stating what in the question he or she agrees with and what he or she does not. It is also common to encounter questions from attorneys based on hypotheses that are extremely unlikely, and the physician needs to point out that unlikelihood (224). Part of the strategy of aggressive cross-examination is to provoke the physician into arguments or unprofessional behavior that could discredit the physician or his or her testimony before the court. In particular, juries will allow aggression on the part of an attorney, but they expect physician witnesses to respond professionally, even under extreme duress (117). Inexperienced potential physician witnesses can prepare themselves by either watching trials or participating in mock trials (224). Brent (226) assembled an excellent series of vignettes of expert medical witness case studies in court and provided detailed instructions with regard to the responsibilities of such experts. Both redirect examination and recross-examination may follow cross-examination at the discretion of attorneys, but usually these procedures are very short (224).
Disposition Following Custody Hearings
After a hearing or trial, the child historically either remained in the protective custody of the state or was returned to the home, but the danger of further abuse exists in both situations. In a study of recurrent maltreatment in 10 states based on the National Child Abuse and Neglect Data System, Fluke et al (227) found that the recurrence rate was 13% by 6 months after the

first episode of reported abuse, increasing to 17% by 12 months. In a report of 206 care and protection petitions brought to the Boston juvenile courts (228), 31 were dismissed with return of the child to the parents. During a 2-year follow-up of these dismissed cases, 29 had reports of further mistreatment and 16 were returned to court under another care and protection petition. One risk factor identified by the study was a previous appearance in court. Half of dismissed cases with this risk factor returned to court again. Of the children ordered permanently removed from parental custody by the court, six returned to court with evidence of further abuse by another caretaker. Another alternative pathway of custody is gaining popularity with the court systems in which the abused child is released to the custody of a relative of the family with consideration given to the wishes of the parents or other prior custodians of the child. Although in theory this approach may help preserve the integrity of the family unit, the child may still be in danger in this sort of arrangement. Handy et al (229) of the Pediatric Forensic Medicine Program of the Kentucky State Medical Examiner’s Office noted evidence of recurrent abuse 2 to 9 months after the original injury in six patients out of 316 referrals to the program. They emphasized that in two of these cases, the child was in protective custody of a family relative when the original perpetrator was allowed unsupervised access to the child in violation of court order. One 6-month-old infant placed in the custody of a grandmother was bent backward by the original perpetrator until his head touched his feet, sustaining a thoracic spinal fracture with fatal outcome. Another 4-year-old was in the custody of her grandmother when the child’s mother gained unauthorized access to her and the child was reinjured with a burn on the face with an iron. It is possible that such reinjuries occur because either the close relatives of the child abuser may not believe that the caretaker committed the original abuse or the relatives are under emotional pressure to allow the caretaker to have access to the child in spite of court order. It is hoped that the court systems can strike a balance between the need to preserve the family unit and the need to protect the child from further abuse.
Prevention of child abuse lies in early intervention. Home visitor programs can contact a mother immediately after the birth of her child and arrange for a visit in which the mother’s parenting strengths are assessed. Parents requiring additional support are linked to community agencies and family resources (122,230). Such support seems to enhance parent and child interactions, and mothers report a diminished need to punish or restrict their children. Anti-victimization programs teach children certain concepts believed to facilitate self-protection, such as identification of strangers, types of touching, saying “no” to inappropriate advances, and telling someone about inappropriate behavior.
Parenting education offers instruction in specific parenting skills such as discipline methods, basic child care, infant stimulation, child development, education, and familiarity with local support services and introduction to other new parents in the community (105). Continuing abuse can be prevented by the orthopaedist’s prompt recognition of child abuse in the emergency department and appropriate intervention. The most important issue in dealing with child abuse is to help both the child and the family through early recognition of the problem and appropriate therapeutic measures by all health personnel. The failure to diagnose child abuse may result in serious injury to or death of the child.
Special thanks is given to Hope Trevino and Randy Llamas for technical assistance.
1. Kocher MS, Kasser JR. Orthopaedic aspects of child abuse. J Am Acad Orthop Surg 2000;8:10–20.
2. Johnson CF. Inflicted injury versus accidental injury. Pediatr Clin North Am 1990;37:791–814.
3. Texas Department of Health. Children’s Health and Safety Initiative: Building Blocks for Healthy Children. Texas Department of Health 1996:24.
4. Prevent child abuse America: Accessed August, 2005.
5. Texas Medical Association Foundation. A Report by the Blue Ribbon Panel on Family Violence. Texas Medical Association Foundation.
6. Forjuoh SN. Child maltreatment related injuries: incidence, hospital charges, and correlates of hospitalization. Child Abuse Negl 2000;24:1019–1025.
7. Oliver JE. Intergenerational transmission of child abuse: rates, research, and clinical implications. Am J Psychiatry 1993;150:1315–1324.
8. Akbarnia BA, Akbarnia NO. The role of the orthopedist in child abuse and neglect. Orthop Clin North Am 1976;7:733–742.
9. Schmitt BD. Child abuse. In: Green M, Haggerty RJ, eds. Pediatrics. Philadelphia: WB Saunders; 1984:111.
10. Caffey J. Multiple fractures in long bones of infants suffering from chronic subdural hematoma. Am J Roentgenol 1946;56:163–173.
11. Silverman FN. The roentgen manifestations of unrecognized skeletal trauma in infants. Am J Roentgenol 1953;69:413–427.
12. Altman DH, Smith RL. Unrecognized trauma in infants and children. J Bone Joint Surg Am 1960;42:407–413.
13. Kempe CH, Silberman FN, Steele BF, et al. The battered-child syndrome. JAMA 1962;181:105–112.
14. 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.
15. Dubowitz H, Bross DC. The pediatrician’s documentation of child maltreatment. Am J Dis Child 1992;146:596–599.
16. Zuravin SJ. Unplanned childbearing and family size: their relationship to child neglect and abuse. Fam Planning Perspect 1991;23:155–161.
17. Akbarnia BA. The role of the orthopaedic surgeon in child abuse. In: Morrissy RM, ed. Lovell and Winter’s pediatric Orthopaedics. Philadelphia: JB Lippincott; 1996:1315.
18. Herndon WA. Child abuse in a military population. J Pediatr Orthop 1983;3:73–76.
19. O’Neill JA Jr, Meacham WF, Griffin PP, et al. Patterns in injury in battered child syndrome. J Trauma 1973;13:332–339.
20. Green FC. Child abuse and neglect: a priority problem for the private physician. Pediatr Clin North Am 1975;22:329–339.
21. Crouch JL, Behl LE. Relationships among parental beliefs in corporal punishment, reported stress, and physical child abuse potential. Child Abuse Negl 2001;25:413–419.
22. Helfer RE. The epidemiology of child abuse and neglect. Pediatr Ann 1984;13:745–751.
23. Wasserman DR, Leventhal JM. Maltreatment of children born to cocaine-dependent mothers. Am J Dis Child 1993;147:1324–1328.
24. Siegel CD, Graves P, Maloney K, et al. Mortality from intentional and unintentional injury among infants of young mothers in Colorado, 1986 to 1992. Arch Pediatr Adolesc Med 1996;150:1077–1083.
25. Campbell JC. Child abuse and wife abuse. Md Med J 1994;43:349–350.
26. Wolfner GD, Gelles RJ. A profile of violence toward children: a national study. Child Abuse Negl 1993;17:197–212.
27. Sidebotham P, Golding J. The ALSPAC Study Team: child maltreatment in the “Children of the Nineties”=ma longitudinal study of parental risk factors. Child Abuse Negl 2001;25:1177–1200.

28. Galleno H, Oppenheim WL. The battered child syndrome revisited. Clin Orthop 1982;62:11–19.
29. Beals RK, Tufts E. Fractured femur in infancy: the role of child abuse. J Pediatr Orthop 1983;3:583–586.
30. Desilva S, Oates RK. Child homicide: the extreme of child abuse. Med J Aust 1993;158:300–301.
31. Kleinman PK, Marks SC, Adams VI, et al. Factors affecting visualization of posterior rib fractures in abused infants. Am J Roentgenol 1988;150:635–638.
32. Bruce DA, Schut L, Bruno LA, et al. Outcome following severe head injuries in children. J Neurosurg 1978;48:679–688.
33. Emery JL. Child abuse, sudden infant death syndrome, and unexpected infant death. Am J Dis Child 1993;147:1097–1100.
34. Samuels MP, McClaughlin W, Jacobson RR, et al. Fourteen cases of imposed upper airway obstruction. Arch Dis Child 1992;67:162–170.
35. Benedict RB, Wulff LM, Hall BJ. Reported maltreatment in children with multiple disabilities. Child Abuse Negl 1990;14:207–217.
36. Asher R. Munchausen syndrome. Lancet 1951;1:339.
37. Mehl AL, Coble L, Johnson S. Munchausen syndrome by proxy: a family affair. Child Abuse Negl 1990;14:577–585.
38. Wenk, RE. Molecular evidence of Munchausen syndrome by proxy. Arch Pathol Lab Med 2003;127:e36–e37.
39. Baskin DE, Stein F, Coats DK, et al. Recurrent conjunctivitis as a presentation of Munchausen syndrome by proxy. Ophthalmology 2003;110:1582–1584.
40. Bartsh C, Risse M, Schutz H, et al. Munchausen syndrome by proxy (MSBP): an extreme form of child abuse with a special forensic challenge. Forensic Sci Int 2003;137:147–151.
41. Meadow SR. Non-accidental salt poisoning. Arch Dis Child 1993;68:448–452.
42. Babcock J, Hartman K, Pedersen A, et al. Rodenticide-induced coagulopathy in a young child: a case of Munchausen syndrome by proxy. Am J Pediatr Hematol Oncol 1993;15:126–130.
43. Goss PW, McDougall PN. Munchausen syndrome by proxy: a cause of preterm delivery. Med J Aust 1992;157:814–817.
44. Berkowitz CD. Pediatric abuse: new patterns of injury. Emerg Med Clin North Am 1995;13:321–341.
45. Rosenberg DA. Web of deceit: a literature review of Munchausen syndrome by proxy. Child Abuse Negl 1987;11:547–563.
46. Meadow R. Munchausen syndrome by proxy abuse perpetuated by men. Arch Dis Child 1998;78:210–216.
47. Davis P, McClure RJ, Rolfe K, et al. Procedures, placement, and risks of further abuse after Munchausen’s by proxy, non-accidental poisoning, and non-accidental suffocation. Arch Dis Child 1998;78:217–221.
48. Hall DE, Eubanks L, Swarnalatha M, et al. Evaluation of covert surveillance in the diagnosis of Munchausen syndrome by proxy: lessons from 41 cases. Pediatrics 2000;105:1305–1312.
49. Krugman RD. Recognition of sexual abuse in children. Pediatr Rev 1986;8:25–30.
50. Hobbs CJ, Wynne JM. The sexually abused battered child. Arch Dis Child 1990;65:423–427.
51. Wissow LS. Child abuse and neglect. N Engl J Med 1995;332:1425–1431.
52. Sullivan PM, Brookhouser PE, Scanlan JM, et al. Patterns of physical and sexual abuse of communicatively handicapped children. Ann Otol Rhinol Laryngol 1991;100:188–427.
53. Young WC. Sadistic ritual abuse: an overview in detection and management. Primary Care 1993;20:447–458.
54. Alexander RC, Surrell JA, Cohle SD. Microwave oven burns to children: an unusual manifestation of child abuse. Pediatrics 1987;79:255–260.
55. Reece RM. Unusual manifestations of child abuse. Pediatr Clin North Am 1990;37:905–912.
56. Black J, Zenel JA. Child abuse by intentional iron poisoning presenting as shock and persistent acidosis. Pediatrics 2003;111:197–199.
57. Dubowitz H, Black M. Teaching pediatric residents about child maltreatment. Dev Behav Pediatr 1991;12:305–307.
58. Limbos MAP. Documentation of child abuse: how far have we come? Pediatrics 1998;102:53–58.
59. Bar-on ME, Zanga JR. Child abuse: a model for use of structured clinical forms. Pediatrics 1996;81:17–24.
60. Palusci VJ, Hicks RA, Vandervort FE. “You are hereby commanded to appear”: pediatrician subpoena and court appearance in child maltreatment. Pediatrics 2001;107:1427–1431.
61. Hanes M, Mc Aufliff T. Preparation for child abuse litigation: perspectives of the prosecutor and the pediatrician. Pediatr Ann 1977;26:288–295.
62. Orebach Y, Lamb ME. Enhancing children’s narratives in investigative interviews. Child Abuse Negl 2000;24:1631–1648.
63. Kravitz H, Driessen G, Gomberg R, et al. Accidental falls from elevated surfaces in infants from birth to one year of age. Pediatrics 1969;44:869–876.
64. Helfer RE, Slovis TL, Black M. Injuries resulting when small children fall out of bed. Pediatrics 1977;60:533–535.
65. Kogutt MS, Swischuk LE, Fagan CJ. Patterns of injury and significance of uncommon fractures in the battered child syndrome. Radiology 1974;121:143–149.
66. Hennrikus WL, Shaw BA, Gerardi JA. Injuries when children reportedly fall from a bed or a couch. Clin Orthop Rel Res 2003;407:148–151.
67. O’Connor JF, Cohen J. Dating fractures. In: Kleinman PK, ed. Diagnostic Imaging of Child Abuse. Baltimore: Williams & Wilkins; 1987:103.
68. McMahon P, Grossman W, Gaffney M, et al. Soft-tissue injury as an indication of child abuse. J Bone Joint Surg Am 1995;77:1179–1183.
69. Sugar NF, Taylor JA, Feldman KW. Bruises in infants and toddlers: those who don’t cruise rarely bruise. Puget Sound Pediatric Research Network. Arch Pediatr Adolesc Med 1999;153:399–403.
70. Carpenter RF, The prevalence and distribution of bruising in babies. Arch Dis Child 1999;80:363–366.
71. Robertson DM, Barbor P. Unusual injury? Recent injury in normal children and children with suspected non-accidental trauma. Br Med J 1982;285:1399–1401.
72. Labbe J, Caouette G. Recent skin injuries in normal children. Pediatrics 2001;108:271–276.
73. Johnson CF, Kaufman KL. The hand as a target organ in child abuse. Clin Pediatr 1990;29:66–72.
74. Sussman SJ. Skin manifestations of the battered-child syndrome. J Pediatr 1968;72:99–101.
75. Hyden PW, Gallagher TA. Child abuse intervention in the emergency room. Pediatr Clin North Am 1992;39:1053–1081.
76. Jessee SA, Rieger M. Physical abuse: a study of age-related variables among physically abused children. J Dent Child 1996;14:275–280.
77. Wilson FF. Estimation of age of continuous contusions in child abuse. Pediatrics 1977;60:750–752.
78. Langlois NEI, Gresham GA. The age of bruises: a review and study of the color changes with time. Forensic Sci Int 1991;50:227–238.
79. Stephenson T, Bialas Y. Estimation of the age of bruising, Arch Dis Child 1996;74:53–55.
80. Asnes RS. Buttock bruises = mongolian spot [Letter]. Pediatrics 1984;74:321.
81. Anh NT. “Pseudo-battered child” syndrome. JAMA 1976;236:2288.
82. Purdue GF, Hunt JL, Prescott PR. Child abuse by burning: an index of suspicion. J Trauma 1988;28:221–224.
83. Titus MO, Baxter AL, Starling SP. Accidental scald burn in sinks. Pediatrics 2003;111:E191–E194.
84. Renz BM, Sherman R. Abusive scald burns in infants and children: a prospective study. Am Surgeon 1993;59:329–334.
85. Heider TR, Priolo D, Hultman CS, et al. Eczema mimicking child abuse: a case of mistaken identity. J Burn Care Rehabil 2002;23:357–359.
86. Prescott PR. Hair dryer burns in children. Pediatrics 1990;86:692–697.
87. Frechette A, Rimsza ME. Stun gun injury: a new presentation of the battered child syndrome. Pediatrics 1992;89:898–901.
88. Feldman KW. Pseudoabusive burns in Asian refugees. Am J Dis Child 1984;138:768–769.
89. Schmitt BD, Gray JD, Britton HL. Car seat burns in infants: avoiding confusion with inflicted burns. Pediatrics 1978;62:607–609.
90. Reece RM, Sege R. Childhood head injuries: accidental or inflicted? Arch Pediatr Adolesc Med 2000;154:11–15.
91. Case ME, Graham MA, Handy TC, et al. Position paper on fatal abusive head injuries in infants and young children. Am J Forensic Med Pathol 2001;22:112–122.
92. American Academy of Pediatrics Section on Radiology. Diagnostic imaging of child abuse. Pediatrics 2000;105:1345–1348.
93. Rubin DM, Christian CW, Bilaniuk LT, et al. Occult head injury in high-risk abused children. Pediatrics 2003;111:1382–1386.
94. Duhaime AC, Gennarelli TA, Thibault LE, et al. The shaken baby syndrome: a clinical, pathological, and biomechanical study. J Neurosurg 1987;66:409–415.
95. Pariziel PM, Ceulemans B, Laridon A, et al. Cortical hypoxic-ischemic brain damage in shake-baby (shaken impact) syndrome: value of diffusion-weighted MRI. Pediatr Radiol 2003;33:868–871.
96. Suh DY, Davis PC, Hopkins KL. Nonaccidental pediatric head injury: diffusion-weighted imaging findings. Neurosurgery 2001;49:309–320.
96a. Jaspan T, Narborough G, Punt JAG, et al. Cerebral contusion tears as a marker of child abuse: detection by cranial sonography. Pediatr Radiol 1992;22:237–245.
97. Jaspan T, Griffiths PD, McConachie NS, et al. Neuroimaging for non-accidental head injury in childhood: a proposed protocol. Clin Radiol 2003;58:44–53.
98. Cho D-R, Wang Y-C, Chi C-S. Decompression craniotomy for acute shaken/impact baby syndrome. Pediatr Neurosurg 1995;23:192–198.
99. Eller AW, Brown GC. Retinal disorders of childhood. Pediatr Clin North Am 1983;301:87–101.
100. Hanigan WC, Peterson RA, Njus G. Tin ear syndrome: rotational acceleration in pediatric head injuries. Pediatrics 1987;80:618–622.
101. Akbarnia BA. Pediatric spine fractures. Orthop Clin North Am 1999;30:521–536.
102. Ghatan S, Ellenbogen RG. Pediatric spine and spinal cord injury after inflicted trauma. Neurosurg Clin North Am 2002;13:227–233.
103. Cullen JC. Spinal lesions in battered babies. J Bone Joint Surg Br 1975;57:364–366.
104. Trokel M, DiScala C, Terrin NC, et al. Blunt abdominal injury in the young pediatric patient: child abuse and patient outcomes. Child Maltreat 2004;9:111–117.
105. Touloukian RJ. Abdominal visceral injuries in battered children. Pediatrics 1968;42:642–646.
106. Tracy T Jr, O’Connor TP, Weber TR: Battered children with duodenal avulsion and transection. Am Surgeon 1993;59:342–345.
107. Coant PN, Kornberg AE, Brody AS, et al. Markers for occult liver injury in cases of physical abuse in children. Pediatrics 1992;89:274–278.
108. Howard JL, Barron BJ, Smith GG. Bone scintigraphy in the evaluation of extraskeletal injuries from child abuse. Radiographics 1990;10:67–81.
109. Thogmartin JR, England D, Siebert Jr CF. Hepatic glycogen staining: applications in injury survival time and child abuse. Am J Forensic Med Pathol 2001;22:313–318.
110. Sinal SH. Sexual abuse of children and adolescents. South Med J 1994;87:1242–1258.
111. American Academy of Pediatrics. Guidelines for the evaluation of sexual abuse of children. Pediatrics 1991;87:254–260.
112. Ingram DM, Everett VD, Ingram DL. The relationship between the transverse hymenal orifice diameter by the separation technique and other possible markers of sexual abuse. Child Abuse and Negl 2001;25:1109–1120.

113. Kellog ND, Menard SW, Santos A. Genital anatomy in pregnant adolescents: “normal” does not mean “nothing happened.” Pediatrics 2004;113:e67–e69.
114. Cantwell HB. Vaginal inspection as it relates to child sexual abuse in girls under thirteen. Child Abuse Negl 1983;7:171–176.
115. Berenson AB, Heger AH, Hayes JM, et al. Appearance of the hymen in prepubescent girls. Pediatrics 1992;89:387–394.
116. Bays J, Chadwick D. Medical diagnosis of the sexually abused child. Child Abuse Negl 1993;17:91–110.
117. Dalton HJ, Slovis T, Helfer RE, et al. Undiagnosed abuse in children younger than 3 years with femoral fracture. Am J Dis Child 1990;140:875–878.
118. Leventhal JM, Thomas SA, Rosenfield NS, et al. Fractures in young children: distinguishing child abuse from unintentional injuries. Am J Dis Child 1993;147:87–92.

119. Adelman, RD, Fink RA, Rohn RD et al. Child abuse in an infant presenting as unexplained acute systemic hypertension. Pediatr Nephr 200;14:811–812.
120. Ander WA. The significance of femoral fractures in children. Ann Emerg Med 1982;11:174–177.
121. Thomas SA, Rosenfield NS, Leventhal JM, et al. Long-bone fractures in young children: distinguishing accidental injuries from child abuse. Pediatrics 1991;88:1–476.
122. American Academy of Pediatrics: Section on Child Abuse and Neglect. A Guide to References and Resources in Child Abuse and Neglect. Elk Grove Village, IL: American Academy of Pediatrics; 1994:107–190.
123. Kleinman PK, Spevak MR. Variations in acromial ossification simulating infant abuse in victims of sudden infant death syndrome. Radiology 1991;180:185–187.
124. Kleinman PK. Diagnostic Imaging of Child Abuse, 2nd ed. St. Louis: Mosby; 1998.
125. Hechter S, Huyer D, Manson D. Sternal fractures as a manifestation of abusive injury in children. Pediatr Radiol 2002;32:902–906.
126. Merter LDF, Radkowski MA, Leonidas JC. The abused child: a radiological reappraisal. Radiology 1983;146:377–381.
127. Nimkin K, Spevak MR, Kleinman PK. Fractures of the hands and feet in child abuse: imaging and pathological features. Radiology 1997;203:233–236.
128. King J, Diefendorf D, Apthorp J, et al. Analysis of 429 fractures in 189 battered children. J Pediatr Orthop 1988;8:585–589.
129. Akbarnia B, Torg JS, Kirkpatrick J et al: Manifestations of battered-child syndrome. J Bone Joint Surg Am 1974;56:1159–1166.
130. Loder RT, Bookout C. Fracture patterns in battered children. J Orthop Trauma 1991;5:428–433.
131. Pierce MC, Bertocci GE, Vogeley E, et al. Evaluating long bone fractures in children: a biomechanical approach with illustrative cases. Child Abuse Negl 2004;28:505–524.
132. Bertocci GE, Pierce MC, Deemer E, et al. Influence of fall height and impact surface on biomechanics of feet-first free falls in children. Injury 2004;35:417–424.
133. Bertocci GE, Pierce MC, Deemer E, et al. Using test dummy experiments to investigate pediatric injury risk in simulated short-distance falls. Arch Pediatr Adolesc Med 2003;157:480–486.
134. Bertocci GE, Pierce MC, Deemer E, et al. Computer simulation of stair falls to investigate scenarios in child abuse. Arch Pediatr Adolesc Med 2001;155:1008–1014.
135. Pierce MC, Bertocci GE, Janosky JE, et al. Femur fractures resulting from stair falls among children: an injury implausibility model. Pediatrics 2005;115:1712–1722.
136. McClelland CQ, Heiple KG. Fractures in the first year of life: a diagnostical dilemma? Am J Dis Child 1982;136:26–29.
137. Shaw BA, Murphy KM, Shaw A, et al. Humerus shaft fractures in young children: accident or abuse? J Pediatr Orthop 1997;17:293–297.
138. Merten DF, Carpenter BLM. Radiologic imaging of inflicted injury in the child abuse syndrome. Pediatr Clin North Am 1990;37:815–837.
139. Rex C, Kay PR. Features of femoral fractures in nonaccidental injury. J Pediatr Orthop 2000;20:411–413.
140. Scherl SA, Miller L, Lively N, et al. Accidental and nonaccidental femur fractures in children. Clin Orthop Relat Res 2000;376:96–105.
140a. Gross RH, Stanger M. Causative factors responsible for femoral shaft fractures in infants and young children. J Pediatr Orthop 1983;3:341–343.
141. Schwend RM, Werth C, Johnston A. Femur fractures in toddlers and young children: rarely from child abuse. J Pediatr Orthop 2000;20:475–481.
142. Blakemore LC, Loder RT, Hensinger RN. Role of intentional abuse in children one to five years old with isolated femoral shaft fractures. J Pediatr Orthop 1996;16:585–588.
143. Leonidas JC. Skeletal trauma in the child abuse syndrome. Pediatr Ann 1983;12:875–881.
144. Kleinman PK, Marks SC. Relationship of the subperiosteal bone collar to metaphyseal lesions in abused infants. J Bone Joint Surg Am 1995;77:1471–1476.
145. Kleinman PK, Marks SC, Blackbourne B. The metaphyseal lesion in abused infants: a radiologic-histopathologic study. Am J Roentgenol 1986;146:895–905.
146. Kleinman PK, Marks SC Jr, Spevak MR, et al. Extension of growth-plate cartilage into the metaphysis: a sign of healing fracture in abused infants. Am J Roentgenol 1991;156:775–779.
147. Osier LK, Marks SC Jr, Kleinman PK. Metaphyseal extensions of hypertrophied chondrocytes in abused infants indicate healing fractures. J Pediatr Orthop 1993;13:249–254.
148. Currarino G, Birch J G, Herring, JA. Developmental coax vara associated with spondylometaphyseal dysplasia (DCV/SMD): “SMD-Corner fracture type” DCV/SMD-CF) demonstrated inmost cases. Pediatr Radiol 2000;30:14–24.
149. Strouse PJ, Owings CL. Fractures of the first rib in child abuse. Radiology 1995;197:763–765.
150. Kleinman PK, Nimkin K, Spevak MR, et al. Follow-up skeletal surveys in suspected child abuse. Am J Roentgenol 1996;167:893–896.
151. Smith FW, Gilday DL, Ash JM, et al. Unsuspected costo-vertebral fractures demonstrated by bone scanning in the child abuse syndrome. Pediatr Radiol 1980;10:103–106.
152. Kleinman PK, Marks SC Jr, Spevak MR, et al. Fractures of the rib head in abused infants. Radiology 1992;185:119–123.
153. Conway JJ, Collins M, Tanz RR, et al. The role of bone scintigraphy in detecting child abuse. Semin Nucl Med 1993;23:321–333.
154. Magid N, Glass T. A “hole in a rib” as a sign of child abuse. Pediatr Radiol 1990;20:334–336.
155. Smeets AJ, Robben SGF, Meradji M. Sonographically detected costo-chondral dislocation in an abused child. Pediatr Radiol 1990;20:566–567.
156. Ng SC, Hall CM. Costochondral junction fractures and intra-abdominal trauma in non-accidental injury (child abuse). Pediatr Radiol 1998;28:671–676.
157. Cumming WA. Neonatal skeletal fractures: birth trauma or child abuse? J Can Assoc Radiol 1979;30:30–33.
158. Betz P, Liebhardt E. Rib fractures in children: resuscitation or child abuse? Int J Legal Med 1994;106:215–218.
159. Baker AM, Craig BR, Lonnergan GJ. Homicidal commotion cordis: the final blow in a battered infant. Child Abuse Neglect 2003;27:125–130.
160. McEniery J, Hanson R, Grigor W, et al. Lung injury resulting from a nonaccidental crush injury to the chest. Pediatr Emerg Care 1991;7:166–168.
161. Nichols GR II, Corey TS, Davis GJ. Case report: non-fracture-associated fatal fat embolism in a case of child abuse. J Forensic Sci 1990;35:493–499.
162. Carty H. Case report: child abuse=mnecklace calcification=mof strangulation. Br J Radiol 1993;66:1186–1188.
163. Kleinman PK, Blackbourne BD, Marks SC, et al. Radiologic contributions to the investigation and prosecution of cases of fatal infant abuse. N Engl J Med 1989;320:507–511.
164. Kleinman PK, Marks SC, Richmond JM, et al. Inflicted skeletal injury: a postmortem radiologic–histopathologic study in 31 infants. Am J Roentgenol 1995;165:647–650.
165. Kleinman PK, Marks SC. Vertebral body fractures in child abuse: radiologic–histopathologic correlates. Invest Radiol 1992;27:715–722.
166. Swischuk LE. Spine and spinal cord trauma in the battered child syndrome. Radiology 1969;932:733–738.
167. Brown R, Brunn MA, Garcia VF. Cervical spine injuries in children: a review of 103 patients treated conservatively at a level 1 pediatric trauma center. J Pediatr Surg 2001;36:1107–1114.
168. Thomas NH, Robinson L, Evans A, et al. The floppy infant: a new manifestation of nonacccidental injury. Pediatr Neurosurg 1995;23:188–191.
169. Kleinman PK, Shelton YA. Hangman’s fracture in an abused infant: imaging features. Pediatr Radiol 1997;27:776–777.
170. Currarino G. Primary spondylolysis of the axis vertebrae (C-2) in three children, including one with pyknodysostosis. Pediatr Radiol 1989;19:535–538.
171. Ravindran RK, Mullett JH, Burke TE. Hangman’s fracture caused by suspected child abuse: a case report. J Pediatr Orthop B 2002;11:329–332.
172. Rooks VJ, Sisler C, Burton B. Cervical spine injury in child abuse: a report of two cases. Pediatr Radiol 1998;28:193–195.
173. Carrion WV, Dormans JP, Drummond DS, et al. Circumferential growth plate fracture of the thoracolumbar spine from child abuse. J Pediatr Orthop 1996;16:210–214.
173a. Habert J, Haller JO. Iatrogenic vertebral body compression fracture in a premature infant caused by extreme flexion during positioning for a lumbar puncture. Pediatr Radiol 2000;30:410–411.
174. American College of Radiology. Imaging of the child with suspected child abuse. In: American College of Radiology Standards for Skeletal Surveys in Children. Resolution 22. Reston, VA: American College of Radiology; 1997;23.
175. Ellerstein NS, Norris KJ. Value of radiologic skeletal survey in assessment of abused children. Pediatrics 1984;74:1075–1078.
176. Sly JR, Starshak RJ. The role of bone scintigraphy in the evaluation of the suspected abused child. Radiology 1983;146:369–375.
177. Jaudes PK. Comparison of radiography and radionuclide bone scanning and the detection of child abuse. Pediatrics 1984;73:166–168.
178. Fordham EW, Ramachandran PC. Radionuclide imaging of osseous trauma. Semin Nucl Med 1974;4:411–429.
179. Youmans DC, Don S, Hidebolt C, et al. Skeletal surveys for child abuse: comparison of interpretation using digitized images and screen-film radiographs Am J Roentgenol 1998;171:1415–1419.
180. Kleinman PK, O’Connor B, Nimkin K, et al. Detection of rib fractures in an abused infant using digital radiography: a laboratory study. Pediatr Radiol 2002;32:896–901.
181. Markowitz RI, Hubbard AM, Harty MP, et al. Sonography of the knee in normal and abused infants. Pediatr Radiol 1993;23:264–267.
182. Rogers LF, Poznanski AK. Imaging of epiphyseal injuries. Radiology 1994;191:297–308.
183. Dreizen S, Spirakis CN, Stone RE. The influence of age and nutrition on “bone scar” formation in the distal end of the growing femur. Am J Phys Anthropol 1964;22:295–306.
184. Edwards DK. Skeletal growth lines seen on radiographs of newborn infants: prevalence and possible association with obstetric abnormalities. Am J Roentgenol 1993;161:141–145.
185. Kaplan JM. Pseudoabuse: the misdiagnosis of child abuse. J Forensic Sci 1986;31:1420–1428.
186. Bernet W. False statements and the differential diagnosis of abuse allegations. J Am Acad Child Adolesc Psychiatry 1993;32:903–910.
187. American Academy of Child and Adolescent Psychiatry. Policy Statement: Guidelines for the Clinical Evaluation of Child and Adolescent Sexual Abuse. Washington, DC: American Academy of Child and Adolescent Psychiatry; 1990.
188. Kleinman PK, Belanger PL, Karellas A, et al. Normal metaphyseal radiologic variants not to be confused with findings of infant abuse. Am J Roentgenol 1991;156:781–783.
189. Zeiss J, Wycliffe ND, Cullen BJ, et al. Radiological case of the month. Am J Dis Child 1988;142:1367–1368.
190. DeRusso PA, Spevak MR, Schwarz KB. Fractures in biliary atresia misinterpreted as child abuse. Pediatrics 2003;112:185–188.
191. Chalumeau M, Fox-l’Helias L, Scheinmann P, et al. Rib fractures after chest physiotherapy for bronchiolitis or pneumonia in infants. Pediatr Radiol 2002;32:644–647.
192. Lim HK, Smith WL, Sato Y, et al. Congenital syphilis mimicking child abuse. Pediatr Radiol 1995;25:560–561.
193. Duncan AA, Chandry J. Case report: multiple neonatal fractures=mdietary or deliberate? Clin Radiol 1993;48:137–139.
194. Harty MP, Kao SC. Intraosseous vascular access defect: fracture mimic in the skeletal survey for child abuse. Pediatr Radiol 2002;32:188–190.
195. Grayev AM, Boal DKB, Wallach DM, et al. Metaphyseal fractures mimicking abuse during treatment for clubfoot. Pediatr Radiol 2001;31:559–563.
196. McClain JL, Clark MA, Sandusky GE. Undiagnosed, untreated acute lymphoblastic leukemia presenting as suspected child abuse. J Forensic Sci 1990;35:735–739.
197. Newman RS, Jalili M, Kolls BJ, et al. Factor XIII deficiency mistaken for battered child syndrome: case of a commonly accepted qualitative test with limited negative predictive value. Am J Hematol 2002;71:328–330.
198. Patterson CR, Burns J, McAltion SJ. Osteogenesis imperfecta: the distinction from child abuse and the recognition of a variant form. Am J Med Genet 1993;45:187–192.
199. Sillence D. Osteogenesis imperfecta: an expanding panorama of variants. Clin Orthop 1981;159:11–25.
200. Sillence D, Butler B, Latham M, et al. Natural history of blue sclerae in osteogenesis imperfecta. Am J Med Genet 1993;45:183–186.
201. Albin DS, Greenspan A, Reinhart M et al. Differentiation of child abuse from osteogenesis imperfecta. Am J Roentgenol 1990;154:1035–1046.
202. Dent JA, Patterson CR. Fractures in early childhood: osteogenesis imperfecta or child abuse? J Pediatr Orthop 1991;11:184–186.
203. Shoenfeld Y, Fried A, Ehrenfeld NE. Osteogenesis imperfecta: review of the literature with presentation of 29 cases. Am J Dis Child 1975;129:679–687.
204. Ojima K, Matsumoto H, Hayase T, et al. An autopsy case of osteogenesis imperfecta initially suspected as child abuse. Forensic Sci Int 1994;65:97–104.
205. Gahagan S, Rimsza ME. Child abuse or osteogenesis imperfecta: how can we tell? Pediatrics 1991;88:987–991.
206. Byers PH. Disorders of collagen biosynthesis and structure. In: Scriver CR, Beaudet AL, Sly WS, et al, eds. The Metabolic Basis of Inherited Disease. New York: McGraw-Hill; 1989:2814–2824.
207. Steiner RD, Pepin M, Beyers PH. Studies of collagen synthesis and structure in the differentiation of child abuse from osteogenesis imperfecta. J Pediatr 1996;128:542–547.
208. Knight DJ, Bennet GC. Nonaccidental injury in osteogenesis imperfecta: a case report. J Pediatr Orthop 1990;10:542–544.
209. Albin D. Osteogenesis imperfecta: a review. Can Assoc Radiol J 1998;49:110–123.
210. Ablin DS, Sane SM. Non-accidental injury: confusion with temporary brittle bone disease and mild osteogenesis imperfecta. Pediatr Radiol 1997;27:111–113.
211. Chapman S, Hall CM. Non-accidental injury, or brittle bones. Pediatr Radiol 1997;27:106–110.
212. Miller M. Fractures during physical therapy [Letter]. Pediatr Radiol 2002;32:536–537.
213. Miller ME, Hangartner TN. Temporary brittle bone disease: association with decreased fetal movement and osteopenia. Calcif Tissue Int 1999;64:137–143.
214. Miller ME. Temporary brittle bone disease: a real entity? Semin Perinatol 1999;23:174–182.
215. Miller ME. Another perspective as to the cause of bone fractures in potential child abuse. Pediatr Radiol 2000;30:495–496.
216. Cazalet J. Note Re R (A Minor). Expert Evidence. [1991] 1 FLR 291.
217. Lynch MA. A judicial comment on temporary brittle bone disease [Letter]. Arch Dis Child 1995;73:379.
218. Meadow R. Unnatural sudden infant death. Arch Dis Child 1999;80:7–14.
219. Byard RW, Keeley FW, Smith CR. Type IV Ehlers-Danlos syndrome presenting as sudden infant death. Am J Clin Pathol 1990;93:579–582.
220. Sperry K, Pfalzgraf R. Inadvertent clavicular fractures caused by “chiropractic” manipulations in an infant: an unusual form of pseudoabuse. J Forensic Sci 1995;35:1211–1216.
221. Clayton EW. Potential liability in cases of child abuse and neglect. Pediatr Ann 1997;26:173–177.
222. Steinberg AM, Pynoos RS, Goenjian AK, et al. Are researchers bound by child abuse reporting laws? Child Abuse Negl 1999;8:771–777.
223. Halverson KC, Elliott BA, Rubin MS, et al. Legal considerations in cases of child abuse. Primary Care 1993;20:407–415.
224. Chadwick DL. Preparation for court testimony in child abuse cases. Pediatr Clin North Am 1990;37:955–970.
225. Drake JA. Physicians and attorneys: a partnership on behalf of the youngest victims of family violence. Md Med J 1994;43:365–367.
226. Brent RL. The irresponsible expert witness: a failure of biomedical graduate education and professional accountability. Pediatrics 1982;7:754–762.
227. Fluke JD, Yuan Y-Y, Edwards M. Recurrence of maltreatment: an application of the National Child Abuse and Neglect Data System (NCANDS). Child Abuse Negl 1999;23:633–650.
228. Murphy JM, Bishop SJ, Jellinek MS, et al. What happens after the care and protection petition? Reabuse in a court sample. Child Abuse Negl 1992;16:485–493.
229. Handy TC, Nichols GR II, Smock WS. Repeat visitors to a pediatric forensic medicine program. J Forensic Sci 1996;41:841–844.
230. Smith PB, Poertner J, Fields JD. Preventing child abuse and neglect in Texas. Texas Med 1990;86:44–45.
Two references invaluable for preparation for testimony in the courtroom about child abuse:
Chadwick DL. Preparation for court testimony in child abuse cases. Pediatr Clin North Am 1990;37:955–970.
Halverson KC, Elliott BA, Rubin MS, et al. Legal considerations in cases of child abuse. Primary Care 1993;20:407–415.
The bible of child abuse radiology:
Kleinman PK. Diagnostic imaging of child abuse, 2nd ed. St. Louis: Mosby; 1998.
A current, concise review of the orthopaedic detection of child abuse:
Kocher MS, Kasser JR. Orthopaedic aspects of child abuse. J Am Acad Orthop Surg 2000;8:10–20.
A handy reference for courtroom testimony involving dating fractures:
O’Connor JF, Cohen J. Dating fractures. In: Kleinman PK, ed. Diagnostic Imaging of Child Abuse. Baltimore: Williams & Wilkins, 1987.