Orthopedic aspects of child abuse

INTRODUCTION — An overview of the orthopedic aspects of child abuse will be presented here.

The differential diagnosis of the orthopedic manifestations of child abuse, recognition of physical child abuse, and the diagnosis and management of physical child abuse are presented separately. (See "Differential diagnosis of the orthopedic manifestations of child abuse" and "Physical child abuse: Recognition" and "Physical child abuse: Diagnostic evaluation and management".)

EPIDEMIOLOGY — The epidemiology of the orthopedic aspects of child abuse will be discussed below. The epidemiology of child abuse, including risk factors in the perpetrators, victims, and environment, is discussed in detail separately. (See "Physical child abuse: Recognition", section on 'Epidemiology'.)

Soft-tissue injuries are the most common injuries identified in physical abuse and are present in as many as 92 percent of victims [1]. Fractures are the second most common injury and are present in as many as 55 percent of physically abused children, depending upon the type of abuse and the method of fracture detection [2-4].

Most inflicted fractures occur during infancy and early childhood [3], with as many as 85 percent occurring in children younger than three years, and 69 percent in children younger than one year [2,5,6]. The estimated annual incidence of fractures attributable to abuse is 36 cases per 100,000 children younger than 12 months, and 5 cases per 100,000 children between 12 and 35 months [7].

Data from the Kids' Inpatient Database (discharge data on 80 percent of acute pediatric hospitalizations in the United States) for 1997, 2000, and 2003 indicate that child abuse accounts for approximately 12 percent of fracture-related hospitalizations in children younger than 36 months [7]. The proportion of fracture hospitalizations attributed to abuse decreased with increasing age (25 percent for children <12 months of age, 7 percent for children 12 to 23 months of age, and 3 percent for children 24 to 35 months of age) [8]. In a retrospective review of 258 patient visits with abusive fractures, 20 percent had at least one previous physician encounter for the fracture at which the possibility of abuse was not raised [9].

In the largest study of inflicted fractures (429 fractures in 189 children), the median age was seven months (range <1 month to 13 years); 69 percent of the patients were younger than one year of age, 17 percent were between one and two years of age, and 14 percent were older than two years of age [2]. This study also provided a description of some of the characteristic features of the fractures associated with child abuse, including the following:

●50 percent of the patients had only a single fracture; 21 percent had two fractures, 12 percent had three, and 17 percent had more than three.

●Among the patients with a single fracture, the femur was the most common location (35 percent), followed by the humerus (29 percent), and the skull (16 percent); all of the single long-bone fractures occurred in the middle one-third of the bone.

●Among long-bone fractures, transverse fractures were most common (48 percent), followed by spiral fractures (26 percent), avulsion fractures (16 percent), and oblique fractures (10 percent).

MEDICAL EVALUATION — The medical evaluation of suspected child abuse includes (see "Physical child abuse: Diagnostic evaluation and management"):

●A complete history, and physical examination

●Observation of the interactions between the parents/primary caregivers and child

●Radiographs and laboratory tests as indicated

●Information from the child's primary care provider about past history, injuries, medical problems, immunization status, and frequency of health maintenance visits

The goal of the medical evaluation of children with suspected child abuse is to determine whether the level of suspicion for abuse is sufficiently high to warrant further evaluation. Whenever possible, this determination should be made in consultation with a multidisciplinary child abuse team (eg, social worker, nurse, and child abuse specialist). It is rarely possible (and almost never necessary) to definitively conclude that physical abuse has occurred in the first few hours of an evaluation. (See "Physical child abuse: Diagnostic evaluation and management", section on 'Approach'.)

The potential consequences for the child of unreported child abuse include life-threatening injury and death [9-12]. Thus, mandatory reporting of a suspicion of abuse to Child Protective Services is required for physicians and other medical providers in many regions. Clinicians must know and abide by mandatory reporting statutes in the jurisdictions where they practice. (See "Child abuse: Social and medicolegal issues", section on 'Reporting suspected abuse'.)

History — It is essential to obtain a detailed account of the mechanism of injury from the parents or caregivers. It may be necessary to ask specific questions as the interview progresses (eg, if the mechanism of injury is a fall down the stairs, it is important to know how many steps, what the steps are made of, and whether or not they are carpeted). It is also important to ask about the initial position of the child, the dynamics of the fall, and the final position and location of the child relative to the steps [13].

Aspects of the history that increase the level of suspicion for inflicted injuries include (table 1) [6,14]:

●Inconsistencies and/or discrepancies in caretakers' accounts of the circumstances surrounding the injury

●Unwitnessed injuries

●Injuries attributed to young children

●Injuries inconsistent with the child's developmental stage

●Injuries inconsistent with the mechanism of injury offered

For example, children who have inflicted musculoskeletal injuries are often said to have fallen from a piece of furniture. However, when young children fall from furniture, the risk of fracture is low (<2 percent) [15-17]. (See "Prevention of falls and fall-related injuries in children", section on 'Falls from furniture'.)

The clinician should also carefully assess the child’s past medical history, family history, and social history to identify other factors associated with abuse as well as risk factors for conditions associated with bone fragility. Children who have pathologic bone disease (eg, children with chronic renal disease, osteogenesis imperfecta, or rickets) may sustain fractures from seemingly trivial incidents, and these fractures can mimic patterns seen in nonaccidental trauma. Although evidence is not conclusive, these children also may be more likely to be abused or neglected than children without chronic medical problems [18-20]. (See "Physical child abuse: Diagnostic evaluation and management", section on 'History' and "Differential diagnosis of the orthopedic manifestations of child abuse".)

Physical examination — The physical examination of the child with suspected inflicted injury should include evaluation for characteristic skin lesions, swelling or deformity, bone tenderness, reluctance to use an extremity, retinal hemorrhages, trauma to the genitals, and signs of neglect (eg, malnourishment, poor hygiene) (table 2). (See "Physical child abuse: Recognition" and "Child neglect: Evaluation and management".)

A systematic, thorough examination of the musculoskeletal system should be performed. Areas that are obviously injured should be examined last in order to minimize the child's anxiety and discomfort.

●The entire surface of the skin should be examined. The location, size, and approximate ages of any skin lesions (bruises, welts, burns) should be documented. Sketches and/or high-quality photographs are helpful in documenting extensive injuries. However, in some states, parental permission must be obtained before photographs may be taken. (See "Physical child abuse: Recognition", section on 'Inflicted bruises' and "Physical child abuse: Diagnostic evaluation and management", section on 'Reporting and documenting suspected abuse'.)

●There may or may not be bruising at the site of fractures [21-23]. Associated bruising is more common with skull fractures than with fractures of the extremities or ribs [21]. The absence of bruising does not help to distinguish inflicted trauma from an underlying bone disorder [24-26].

●The axial and appendicular skeleton should be examined for deformity, decreased range of motion, tenderness, crepitus, and swelling. Fracture callus, which indicates healing fracture, is sometimes palpable.

●A detailed neurologic examination of the extremities appropriate to the child's age and development should also be performed. (See "Detailed neurologic assessment of infants and children", section on 'Neurologic examination'.)

Findings that raise the suspicion of child abuse are discussed in detail separately (table 1 and table 2). (See "Physical child abuse: Recognition".)

RADIOGRAPHIC EVALUATION — The radiographic evaluation for suspected child abuse depends upon the age of the child, the presenting complaints, and the physical examination findings. When the injuries are unintentional, the radiologic evaluation may provide an alternative explanation (eg, osteopenia, skeletal dysplasia). On the other hand, when the injuries have been inflicted, the radiologic evaluation may identify the extent of injuries [27-29].

Plain radiographs — Orthogonal radiographs (anteroposterior [AP] and lateral) should be obtained of all areas of bone tenderness, swelling, deformity, or limited range of motion, and all areas that have a history or evidence of previous trauma [28]. All radiographs should be reviewed by a pediatric radiologist who has experience in the imaging of child abuse.

Skeletal survey — According to the American Academy of Pediatrics (AAP) Section on Radiology, the AAP Committee on Child Abuse and Neglect, and the American College of Radiology, the skeletal survey is "the method of choice for global skeletal imaging in cases of suspected child abuse" [28,30]. All skeletal surveys require interpretation by a radiologist with expertise in the imaging of child abuse.

Indications for skeletal survey depend upon the patient’s age as follows (see "Physical child abuse: Diagnostic evaluation and management", section on 'Skeletal survey'):

●<2 years old – All children <2 years of age in whom child abuse is suspected and children ≤2 years old who are living in the same environment (home or daycare) as an abused child.

●2 to 5 years old – Children 2 to 5 years old must be evaluated on an individual basis, and localized radiographs or skeletal survey may be warranted based upon clinical findings.

●>5 years old – Skeletal survey usually is not necessary in verbal, ambulatory children older than five years of age and, if obtained, has a low yield [31,32]; it may be recommended in children with intellectual disability or who are otherwise unable to give a history or indicate areas of trauma or pain.

In addition, a consensus panel has developed indications for skeletal survey for children with bruising (table 3). (See "Physical child abuse: Diagnostic evaluation and management", section on 'Skeletal survey'.)

The standard skeletal survey includes all of the following (table 4) (see "Physical child abuse: Diagnostic evaluation and management", section on 'Skeletal survey'):

●AP and lateral views of the skull and chest

●Oblique views of the ribs

●Lateral views of the spine

●AP views of the pelvis, long bones of the extremities, and feet

●Posteroanterior oblique views of the hands

All of these views are necessary if occult fractures of abuse are to be detected. Proper positioning of the child and radiographic penetration are essential. Skull radiographs may be omitted if computed tomography (CT) of the head with three-dimensional reconstruction of the skull is planned. (See "Child abuse: Evaluation and diagnosis of abusive head trauma in infants and children", section on 'Skeletal evaluation'.)

Especially in children who do not have head CT with 3D reconstruction of bone windows, bilateral lateral skull films and the Towne's view may increase sensitivity for Wormian bones, which are more commonly seen among children with certain types of osteogenesis imperfecta and other metabolic bone disease. These views are not part of a standard skeletal survey. (See "Osteogenesis imperfecta: An overview", section on 'Bone deformities'.)

A repeat skeletal survey taken two weeks after the initial evaluation may increase the diagnostic yield and is recommended in cases in which the suspicion of child abuse is high [28,33,34]. The second study may demonstrate fractures that were not apparent initially and permit more precise determination of the age of individual injuries (table 5) [28,35,36]. On the other hand, a repeat study that is unchanged from the initial evaluation may suggest an explanation other than fracture for the original abnormalities (eg, normal anatomic variant, bone dysplasia), and may lower the suspicion of abuse [37].

Expanding the skeletal survey to include lateral views of the extremities is performed on all skeletal surveys in some institutions, although evidence supporting these additional views is limited. In a retrospective study of 138 skeletal surveys (most performed in deceased children) that had blinded review by two pediatric radiologists, 2 of 16 fractures (one distal radius and one distal tibia) were not seen on AP views [38]. The addition of lateral extremity views improved the pediatric radiologist's confidence for negative studies. These findings suggest that increased diagnostic accuracy may be obtained by adding lateral views to the standard skeletal survey, but more evidence is needed before these views become routine in all children undergoing evaluation for child abuse.

Abbreviated skeletal surveys, or studies that image the entire infant or young child with one or two radiographic exposures (eg, "babygrams"), have no role in the evaluation of suspected child abuse [28,39]. The importance of the skeletal survey in children undergoing evaluation for physical abuse is demonstrated by an observational study of 703 children; 11 percent had abnormalities, primarily fractures, and the highest frequency of positive surveys was seen in infants under six months of age and in infants and children with an apparent life-threatening event, seizure, or suspected abusive head trauma [40]. Of the 76 patients with positive skeletal surveys, the results of the survey directly led to the diagnosis of abuse in 50 percent of cases. In a separate retrospective study of over 2000 skeletal surveys, 23 percent showed a previously unknown fracture, of which 10 percent were to the hands, feet, spine, or pelvis [41].

Fracture age — Clinicians may be asked to determine the age of fractures diagnosed by skeletal survey. There is a general timetable of the stages of fracture healing by which practitioners can estimate the approximate age of individual injuries (table 5).

●Soft-tissue changes (obliteration of the normal fat planes and muscle boundaries), secondary to hemorrhage or inflammation, are the first radiographically apparent signs of musculoskeletal injury [42]. These changes usually persist for several days [43,44].

●Periosteal new bone formation, which is detectable on plain radiographs only after it becomes calcified, may occur as early as four days after injury, but usually occurs between 7 and 14 days [42].

●As necrotic bone is resorbed, the margins of the fracture lines become blurred, and the fracture gap widens [42]. This process peaks between two and three weeks after injury and may be the only means of dating metaphyseal fractures [43,44].

●Soft callus, from the production and calcification of osteoid, is visible as an increase in density soon after the appearance of periosteal new bone.

●Hard callus (the lamellar bone that bridges the fracture site, filling in the fracture line) appears approximately one week after soft callus. This phase of healing is usually complete between three and six weeks after injury [44,45].

Computed tomography skeletal survey — Limited experience from one case series suggests that CT skeletal survey may be helpful for evaluating children who are too unstable to undergo the recommended plain radiograph skeletal survey [46]. In the series, CT detected a posterior rib fracture not seen on skeletal survey in two infants but did not detect a corner fracture present on skeletal survey in one infant.

Magnetic resonance imaging (MRI) — MRI can be useful as an additional test in cases where radiographs are indeterminate, and where the significance of a single fracture would be high. However, it is not recommended as a stand-alone full-body study because it is not sensitive for either rib fractures or classic metaphyseal lesions (CMLs), and thus has lower sensitivity for fractures than a skeletal survey [47,48]. MRI does not expose the child to radiation but typically requires sedation or general anesthesia in young children. In one study of 170 children undergoing evaluation for child abuse (64 with at least one fracture) who underwent skeletal survey, whole body MRI, and bone scintigraphy, a skeletal survey combined with whole body MRI achieved a sensitivity of 96 percent (95% CI 91-99 percent) and a specificity of 99.2 percent (95% CI 98.9-99.9 percent), which was not significantly different from the sensitivity and specificity of a skeletal survey combined with bone scintigraphy with a lower radiation dose [48]. Thus, when added to a skeletal survey, whole body MRI, and when interpreted by a skilled pediatric radiologist, appears to have equivalent diagnostic accuracy to follow up radionuclide scanning without further radiation exposure.

Radionuclide scan — Skeletal survey remains the first line of screening for abuse-related fractures. Bone scan serves as a follow-up modality in the setting of a strong suspicion of fracture and negative plain films [28,49]. A repeat skeletal survey performed two weeks after the first evaluation serves a similar purpose without the disadvantages of the bone scan [28,50]. In addition, a skeletal survey combined with whole body MRI appears to have similar sensitivity and specificity when compared with a skeletal survey combined with bone scintigraphy with a lower radiation dose. (See 'Magnetic resonance imaging (MRI)' above.)

In the past, radionuclide bone scanning, or bone scintigraphy, had been recommended by some authorities as an alternative to skeletal survey for the initial evaluation of possible fractures and as a complementary study by others [51-55] because it is more sensitive than plain radiography, especially in the detection of rib and costovertebral fractures, recent nondisplaced fractures that are not yet visible on plain films, and early periosteal elevation [28,52]. In one study of child abuse cases in which the child underwent both skeletal survey and scintigraphy, 124 bony injuries were detected in 30 children [53]. Injuries were detected by both modalities in 70 percent of cases; by scintigraphy alone in 20 percent; and by skeletal survey alone in 10 percent.

However, nuclear scintigraphy lacks sensitivity and specificity relative to plain radiographs for classic metaphyseal lesions (corner fractures), requires sedation, is not rapidly available in many facilities, is more expensive, and delivers a substantially greater exposure to radiation than does skeletal survey with modern imaging systems. Thus, it is not recommended for the initial evaluation for fractures in children with suspected child abuse [27,48,56].

Ultrasonography — Ultrasonography is particularly useful in the diagnosis of costochondral injuries and can sometimes be used to detect subperiosteal hemorrhage, indicative of fracture, before the appearance of the fracture on a plain radiograph [45,57,58]. Ultrasonography also is appealing because it is noninvasive, can be performed in a relatively short time period, and does not expose the child to radiation.

Positron emission tomography — Preliminary results suggest that Fluorine-18 NaF positron emission tomography (PET) has greater initial sensitivity than the initial skeletal survey for detecting thoracic fractures, including posterior rib fractures (92 versus 72 percent, respectively) but lower sensitivity for classic metaphyseal lesions (67 versus 80 percent) and may be a useful adjunct in selected cases. However, initial radiographs are still necessary in all cases [59]. One advantage of Fluorine-18 NaF PET against follow-up skeletal survey is that it can be obtained immediately, which is especially useful when results may determine whether a child will be removed from the home, and when follow-up cannot be assured.

However, the availability of a PET scan and the necessary isotope (Fluorine-18 NaF) may be limited and sedation is required for performance of the study.

DEXA scan — Bone mineral density scanning may have a role in differentiating between child abuse and osteogenesis imperfecta (OI) [60-62]. Most OI patients have osteoporosis, whereas most victims of child abuse have normal bone density. DEXA scans require comparison to age-, sex-, and height-matched reference data and should only be interpreted by physicians with expertise in childhood osteoporosis [63]. Furthermore, limited reference data exist for children younger than three years of age, decreasing the utility of DEXA scans in these patients. This problem can be mitigated somewhat by repeating the DEXA scan six months after the interval study to determine if there has been interval increase in bone density.

Osteogenesis imperfecta is discussed in detail separately. (See "Differential diagnosis of the orthopedic manifestations of child abuse", section on 'Osteogenesis imperfecta' and "Osteogenesis imperfecta: An overview".)

FRACTURE PATTERNS

Overview — Although there are fracture patterns that are suggestive of inflicted injury [5,6,11,27,42,45,63-69], there is no particular fracture pattern, location, or morphology that is absolutely pathognomonic of child abuse. Many physically abused children present with a single fracture, and fractures that are common in abuse (skull fractures, diaphyseal fractures of the long bones) are also common in nonaccidental trauma [2,3,7,70].

With every fracture, the possibility of inflicted injury must be considered, based upon the age of the child, overall injury pattern, stated mechanism of injury, and pertinent psychosocial factors [3,70]. Any fracture in an infant or young child should raise concern for child abuse. Fractures suspicious for child abuse include any fracture of the long bones in a child who is too young to walk, any fracture that is inconsistent with the age of the child or the history provided by the caretakers, any fracture that occurs in combination with extraskeletal injuries, and any healing fracture for which there was a delay in seeking medical attention (table 6) [3,6,42,70-73].

Fractures that are highly suggestive of intentional injury and warrant evaluation for abuse include [3,5,11,63,70,71,74-79]:

●Long bone fractures in nonambulatory children

●Metaphyseal corner and bucket-handle fractures (figure 1 and image 1A-C and image 2 and image 3)

●Rib fractures (image 4)

●Fractures of the sternum, scapula, or spinous processes

●Multiple fractures in various stages of healing (image 5)

●Bilateral acute long-bone fractures

●Vertebral body fractures and subluxations in the absence of a history of high force trauma

●Digital fractures in children younger than 36 months of age or without a corresponding history

●Displaced physeal fractures (sometimes called epiphyseal separations), especially transphyseal distal humerus fractures, in nonambulatory children (image 6)

●Complex skull fractures in children younger than 18 months of age, particularly without a corresponding history

Fractures with less specificity for abuse include [3,63]:

●Isolated long-bone fractures in ambulatory children

●Linear skull fractures

●Clavicle fractures

●Subperiosteal new bone formation

Long-bone fractures — The long bones most commonly injured in child abuse are the femur, humerus, and tibia [2,67]. Isolated fractures of the tibia and femur in ambulatory toddlers may result from nonabusive, low-energy mechanisms [80]. (See "Tibial and fibular shaft fractures in children", section on 'Spiral (toddler's) fractures'.)

Femur fractures — The peak age of incidence for femur fractures is between two and three years; boys outnumber girls by approximately 3:1 [81]. Among children younger than one year of age, child abuse accounts for 60 to 80 percent of femoral shaft fractures [67,82,83]. Among children older than one year of age who have femoral shaft fractures, there are no clinical features (eg, fracture type) that absolutely differentiate intentional from unintentional injury [84,85].

Whether or not the child is ambulatory is the single most important predictor of inflicted injury in isolated femur fractures [70,86]. In one review of 139 children younger than four years with isolated fractures of the shaft of one or both femurs, a greater proportion of nonambulatory than ambulatory children had fractures that were caused by abuse (42 versus 2.6 percent) [86].

In a meta-analysis of 13 studies (including the one described above [86]), the probability of confirmed or suspected abuse in children with a femur fracture was 43 percent (95% CI, 32-54 percent) [70]. Children with an overt history of significant trauma (eg, motor vehicle crash) or predisposition to pathologic fracture were excluded from the analysis. Midshaft fractures were most frequent in both abused and nonabused children. Likewise, the distribution of transverse, spiral, and oblique fractures was similar between the groups.

The American Academy of Orthopaedic Surgeons recommends that children younger than 36 months with a diaphyseal femur fracture without an explanatory history (eg, high-force injury) or known predisposition to pathologic fracture be evaluated for child abuse [73].

Humeral fractures — In a meta-analysis of four cross-sectional studies, the probability of confirmed or suspected abuse in children younger than three years with a humeral fracture was 54 percent (95% CI 20-88 percent) [70]. In two of the included studies, the prevalence of abuse was increased among children younger than 15 months [87,88]. Midshaft humeral fractures of the humerus were more common in abuse than nonabuse, whereas supracondylar fractures were more common in nonabuse than abuse [70]. Spiral/oblique fractures were the most common type of humeral fracture related to abuse [70].

Transphyseal distal humerus fractures are also suggestive of child abuse [77,79,89]. These fractures typically occur in children under the age of one year who have an unossified distal humerus epiphysis and so appear similar to an elbow dislocation radiographically (image 6). Because of this, the injury is commonly missed or misdiagnosed until the callus and periosteal reaction are apparent on follow-up radiographs.

Location and type of fracture — The components of a long bone are depicted in the figure (figure 2). Transverse and spiral diaphyseal fractures and metaphyseal corner fractures comprise the most common type of long bone fracture seen in abused children.

Diaphyseal — Single, fresh fractures of the diaphysis (shaft) of long bones are the most common fracture pattern in child abuse [2,6,90,91]. Diaphyseal fractures are of two basic types: spiral and transverse.

●Spiral fractures of the long bones result from a torsional force on the limb [92]. Like all long-bone fractures, they are suspicious (but not pathognomonic) in a child who is too young to walk or who lacks a consistent history (ie, of having had a twisting or rotational injury). However, in the presence of such a twisting force, these fractures often occur accidentally, as in the noninflicted "toddler's fracture" (childhood accidental spiral tibial [CAST] fracture), a spiral fracture of the tibia that usually is nondisplaced [93].

●Transverse fractures result from a direct blow with the line of force perpendicular to the shaft of the bone. Transverse fractures are the most common type of long-bone fracture associated with abuse, comprising 48 to 71 percent of abuse-related fractures in various studies [2,3,64,87,90,94,95].

Metaphyseal — Certain types of metaphyseal fractures are considered highly suggestive of child abuse [3,4,50,96,97], particularly when present in more than one bone. These metaphyseal fractures account for between 11 and 28 percent of nonaccidental long-bone injuries [2,66,90].

The classic metaphyseal lesion (CML) is also called a "corner" or "bucket-handle" fracture. CMLs are thought to occur when the extremity is pulled or twisted forcibly, or the child is shaken [97-100]. Experimentally, a porcine model has shown CMLs to occur under tensile stress, with varus and valgus loading [101]. CML fractures that occur in infants who are not yet walking are highly associated with child abuse [97,102]. The resultant shearing force undercuts an isolated peripheral fragment of the metaphysis that includes the subperiosteal bone collar, resulting in a disk-shaped bone fragment that is wider at the edges than at the center (figure 1 and image 2) [54,71,103,104]. Rarely, CMLs can occur from unintentional mechanisms, which result in significant twisting motion at the knee [105]. Iatrogenic CML has occurred in one infant who was subject to cephalic version and subsequent delivery by Caesarean section and in eight young infants receiving treatment for club foot [106,107].

Plain radiographs reveal a disruption of the metaphysis with lucency [4]. The radiographic projection influences the appearance of the lesion. When viewed tangentially, the lesion appears as a triangular fragment of the metaphysis (ie, the "corner" fracture) (image 1A); if an angulated view is obtained, the fragment appears as a curvilinear density adjacent to the metaphysis (ie, the "bucket handle") (image 3). Repeated injury results in widening of the radiolucent zone and traumatic cupping of the metaphysis [4].

Rib fractures — Rib fractures are produced by direct blows to the chest or compression of the chest (eg, as the torso is grasped during squeezing or shaking) [3]. Children's ribs are flexible and difficult to fracture, so a rib fracture in the absence of a history of severe trauma such as a motor vehicle crash or fall from a significant height is strongly suggestive of child abuse [3,50,108], particularly if the fractures are of the first rib [109]. The force used for chest compressions in cardiopulmonary resuscitation of infants is not great enough to cause rib fractures [110-112].

The predictive value of rib fractures as an indicator of child abuse was evaluated in a meta-analysis of seven cross-sectional studies [70]. When children with significant trauma (eg, motor vehicle crash) and postoperative children were excluded, the probability of abuse in children with a rib fracture was 71 percent (95% CI 42-91 percent). Six of the seven studies indicated that the number of rib fractures was greater in abused than nonabused children [70]. Inflicted rib fractures usually are nondisplaced and involve multiple sequential ribs (corresponding to the positioning of the abuser's fingers) (image 4) [3,4,111]. Two studies in the meta-analysis confirmed that anterior rib fractures were more common in abuse and that lateral fractures were more common in nonabuse [113,114]. Information regarding the predictive value of posterior rib fractures for abuse was inconsistent [113-115]. The prevalence of abuse increases as age decreases with over 80 percent of rib fractures in infants younger than 12 months of age being inflicted [78]. The location of the rib fracture was not associated with the likelihood of abuse in this study.

Inflicted rib fractures are associated with a high risk of mortality from related injuries because of the degree of force that is required to produce them [3]. Nonetheless, external signs of injury may be absent. In addition, rib fractures may be difficult to detect on radiographs because of overlapping structures [71]. In one study of 31 infants who died as a result of abuse, 51 percent (84 of 165) of fractures involved the ribs [116]. Only 36 percent of the rib fractures were visible on skeletal survey; the remainder required specialized imaging or pathologic analysis.

Rib fractures are most apparent during the stage of callus formation, 10 to 14 days after the injury (table 5) [4]. Asymmetry of the rib necks from side to side or between upper and lower ribs is a radiographic sign of healing rib fractures. The asymmetry is due to widening of the neck of the rib from apposition of subperiosteal bone [71]. The addition of oblique views and/or radionuclide imaging (image 4) increases ability to detect rib fractures in children suspected to be victims of nonaccidental trauma [3,28,52,117]. Follow-up skeletal survey may also be useful in this regard. (See "Physical child abuse: Diagnostic evaluation and management", section on 'Skeletal survey'.)

Spinal fractures — Spinal fractures are uncommon in children, comprising only 2 to 3 percent of all pediatric injuries [3,118,119]. They are caused either by direct trauma to the spinal column or by forceful acceleration/deceleration forces, and are most common in the lower thoracic and upper lumbar spine [4,118,119].

The pediatric spine, especially the cervical spine, is more prone to injury than is the adult spine for several reasons (table 7). Nonaccidental spinal injuries constitute only 0 to 3 percent of all inflicted injuries [3,119]. However, these numbers may significantly underestimate the actual frequency of spinal injuries, because most inflicted spinal fractures are asymptomatic [120] and may elude detection on skeletal surveys that do not include high-quality lateral spine images [119].

The most common type of inflicted fracture of the spine is an asymptomatic compression fracture of the vertebral body that is detected on skeletal survey; multiple vertebrae may be involved, most commonly in the lower thoracic and upper lumbar spine [4,118,121]. Compression or burst fractures of the thoracic, lumbar, or sacral spine may result from forceful "sitting" (ie, the child is thrown down into the sitting position from a height). A burst fracture with formation of a hematoma or retropulsion of a bone fragment may lead to immediate paraplegia or delayed symptoms of spinal cord compression [121,122].

High-velocity hyperflexion and/or hyperextension, as in violent shaking of an infant or small child, may produce a variety of fracture-dislocations of the posterior elements of the spinal column. They usually involve the cervical spine, cervico-thoracic, or thoraco-lumbar junctions [118,120,121]. The hangman's fracture (fracture-dislocation of C2) is a classic example of a hyperextension injury [123]. Compression fractures and notching of the anterior surfaces of the vertebral bodies (caused by disruption of the vertebral end plates and resulting bone loss) also can be produced by this mechanism [121].

Fracture/avulsions of the cartilaginous processes of the posterior spinous processes result from traction and shearing forces on the interspinous ligament during shaking; they usually are not seen acutely because of the lack of visualization of cartilage on plain radiographs. These injuries become visible as ossification of the avulsed cartilage progresses during the healing process [4,120].

Presenting signs and symptoms of spinal injury may be entirely absent [121] or may include:

●Isolated irritability [124]

●Slight abnormality of tone or posture [121]

●Tender spinal mass [118,119,121,125]

●Acute or subacute onset of paraplegia or quadriplegia with urinary retention, loss of anal tone, and loss of bulbocavernosus reflex [125]

●Multiple other fractures that may be visualized on skeletal survey [121,123,124]

Skull fractures — Skull fractures are frequently seen in abused children. Linear, parietal skull fractures are the most common type of skull injury following both unintentional and inflicted trauma. Complicated skull fractures (ie, complex [eg, stellate, branching], multiple, bilateral, crossing suture lines, depressed, or diastatic) are suggestive of abusive head trauma, but also occur following unintentional injury. (See "Child abuse: Epidemiology, mechanisms, and types of abusive head trauma in infants and children", section on 'Skull fractures'.)

Given the potentially similar features of accidental and inflicted skull fractures, these injuries should be evaluated in the context of the history that is provided by the child's caregiver(s). Abusive injury should be suspected if the history is inconsistent with the physical examination findings. If the mechanism of injury is a fall, it is important to determine the height of the fall and the type of surface of impact (eg, concrete, wood, carpet). In general, falls from two feet or less are unlikely to produce head injuries more serious than linear skull fractures. (See "Child abuse: Epidemiology, mechanisms, and types of abusive head trauma in infants and children", section on 'Skull fractures' and "Physical child abuse: Diagnostic evaluation and management", section on 'Evaluation' and "Prevention of falls and fall-related injuries in children".)

Other fractures — Fractures of the sternum, scapula, and pelvis are suspicious for child abuse if they occur in the absence of a plausible history of high-energy injury, such as a motor vehicle accident [3,5,11,27,126,127].

In addition, fractures of the hands and feet, particularly in nonambulatory children, should raise the suspicion of child abuse [54,128]. It may be necessary to obtain oblique or high-detail, coned-down radiographs to detect these subtle lesions, which typically include torus fractures of the metacarpals and metatarsals, and buckle fractures of the proximal phalanges (from forced hyperextension) [128].

Multiple fractures — Multiple fractures (image 1A), old or in various stages of healing (table 5), are highly suggestive of child abuse in the absence of a metabolic disorder [4,7,70,129]. The likelihood of abuse increases four- to sixfold for children with ≥3 fractures compared to those with only one [7]. (See "Differential diagnosis of the orthopedic manifestations of child abuse".)

Multiple fractures are found in approximately 80 percent of abused children younger than one year of age, and more than 50 percent of all abused children [3,96,130]. Nevertheless, many abused children present with only one fracture [2,90].

Periosteal reaction — Although periosteal new bone formation is a common finding in cases of child abuse, it is nonspecific (see "Differential diagnosis of the orthopedic manifestations of child abuse"). It usually occurs in association with fracture, but may be seen in the absence of fracture if there is subperiosteal hemorrhage (eg, after the application of a gripping, twisting, or shaking force) [45,66,71,131]. It can also present as a feature of osteomyelitis, Caffey disease, or malignancy. In one study of 429 fractures in 189 abused children, eight patients (4 percent) had traumatic periosteal reaction without fracture [2].

Periosteal reaction is usually not apparent clinically. It is only apparent on radiographs as the new bone is calcified (eg, 10 to 14 days after the injury) (table 5). Nuclear bone scan or repeat radiographs may be indicated if suspicion of fracture is high and no fracture is seen on initial films [131,132]. In addition, a bone scan may show areas of increased radionuclide uptake at sites of subperiosteal hemorrhage even when the radiographs are normal [133]. Ultrasonography is also particularly useful in demonstrating subclinical periosteal reaction [45]. (See 'Radiographic evaluation' above.)

TREATMENT — As a general rule, fractures resulting from child abuse can be managed nonoperatively, usually with splinting or casting. The rare exceptions are some physeal and intraarticular fractures, which may require operative reduction and fixation. Since child abuse victims are usually young, healing is rapid, the potential for remodeling is great, and the prognosis is good.

Soft-tissue injuries (bruises, lacerations, burns) that are to be covered by a cast or splint should be photographed before treatment begins [71]. In some states, parental permission must be obtained before photographs may be taken. The management of child abuse is discussed separately. (See "Physical child abuse: Diagnostic evaluation and management".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Child abuse and neglect".)

SUMMARY AND RECOMMENDATIONS

●Medical evaluation – Features of the medical evaluation that are suspicious for inflicted orthopedic injury include (see 'Medical evaluation' above):

•Inconsistencies and/or discrepancies in caretakers' accounts of the circumstances surrounding the injury

•Unwitnessed injuries

•Injuries attributed to the patient's siblings

•Injuries inconsistent with the child's developmental stage

•Injuries inconsistent with the mechanism of injury offered

Red flag injuries that should raise concern for physical child abuse are provided in the table (table 6). (See "Physical child abuse: Recognition".)

●Imaging – Biplanar radiographs (AP and lateral) of the injured area and the joint above and below it should be obtained in all patients. We recommend a skeletal survey (table 4) in all children younger than two years who have any evidence of physical abuse, children younger than five years with a suspicious fracture, and older children who are unable to give a history or indicate areas of trauma or pain (ie, intellectually disabled children). A repeat skeletal survey, whole body MRI, or radionuclide bone scan can be helpful when the suspicion for inflicted injury is high but the initial skeletal survey is negative. (See 'Radiographic evaluation' above.)

●Fracture patterns – Although there are no absolutely pathognomonic fracture patterns, fracture patterns that are highly suggestive of physical child abuse and warrant further evaluation for child abuse include (see 'Overview' above and "Physical child abuse: Diagnostic evaluation and management"):

•Metaphyseal corner fractures (figure 1 and image 1A and image 2 and image 3) (see 'Metaphyseal' above)

•Rib fractures (image 4) (see 'Rib fractures' above)

•Fractures of the sternum, scapula, or spinous processes (see 'Spinal fractures' above and 'Other fractures' above)

•Multiple fractures in various stages of healing (image 5) (see 'Multiple fractures' above)

•Bilateral acute long-bone fractures and isolated long-bone fractures in nonambulatory children (see 'Long-bone fractures' above)

•Vertebral body fractures and subluxations in the absence of a history of high force trauma (see 'Spinal fractures' above)

•Digital fractures in children younger than 36 months of age or without a corresponding history (see 'Multiple fractures' above)

•Displaced physeal fractures, including transphyseal distal humerus fractures in nonambulatory children (image 6)

•Complex skull fractures in children younger than 18 months of age, particularly without a corresponding history (see 'Skull fractures' above)

●Fractures with less specificity for abuse include (see 'Overview' above):

•Isolated long-bone fractures in ambulatory children

•Linear skull fractures

•Clavicle fractures

•Subperiosteal new bone formation

With every fracture, the possibility of inflicted injury must be considered, based upon the age of the child, overall injury pattern, stated mechanism of injury, and pertinent psychosocial factors.

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Erin Endom, MD, who contributed to earlier versions of this topic review.