Keywords

Child abuse is an important social and medical problem which represents a major cause of morbidity and mortality among children. Battered Child Abuse (BCA) is a comprehensive term to indicate classical features at first described by Ambroise Tardieu (1818–1879), a French forensic pathologist, in 1860 in a series of 32 cases of cruelty to children, resulting to death in 21 cases [1].

In 1946, John Caffey (1895–1978) published his first paper where he described six infants with multiple fractures in the long bones, who additionally had chronic subdural haematoma and no history of injury [2]. He recommended that unexplained fractures of the long bones warranted investigations for subdural haematoma (SDH). In 1962, Kempe and co-workers published their study on ‘The battered-child syndrome’, first real recognition of child abuse as a disease and of the responsibility physicians held for its diagnosis and prevention [3,4,5]. Finally, Kempe et al. provided radiographic clues to whether trauma was accidental or non-accidental [6]. Worthy of noting, this article was considered one of the best paediatric research articles in the last 150 years, as Kempe et al. established that physicians have a special responsibility to children—a responsibility to help keep them safe, sometimes even from their own parents [1]. During decades scientific interest as reflected by online search of the MEDLINE database yields relevant data about the development of child maltreatment awareness by health care professionals since the article by Kempe et al. was published in 1962. In 1963, the keyword child abuse was added to the MEDLINE system [7]. Following keywords were first assigned to battered-child article by the National Library of Medicine (wounds and injuries, child, child welfare, and infant). Twelve articles were categorized under this keyword, in 2006 almost 600 articles were listed in MEDLINE under the keyword child abuse and 1989 in 2016. Not surprisingly, the marked increase in knowledge about child maltreatment has led to the development of a new paediatric subspecialty, child abuse paediatrics [8,9,10]. In 2009, the American Board of Pediatrics will administer the first examination for board certification in this subspecialty, legacy related. A significant result of Kempe’s battered-child syndrome article has been the raising of paediatricians [11,12,13] who are dedicated to diagnosing, treating, and preventing child abuse and neglect. Over the years, diagnosis of child abuse becomes more sophisticated within methods such as biomechanics [14], proteomics [15], biochemistry [16] and genetics [17] also in perspective of forensic molecular approach as epigenetic modifications [18, 19].

The ‘child abuse syndrome’ (also known as the ‘battered baby’ or ‘non-accidental injury in childhood’) is a clinical condition in young children who have received serious physical abuse and is a frequent cause of permanent injury or death. The syndrome should be considered in any child exhibiting evidence of fracture of any bone, subdural haematoma, failure to thrive, soft tissue swellings or skin bruising, in any child who dies suddenly, or where the degree and type of injury is at variance with the history given regarding the occurrence of the trauma [20, 21] (Fig. 11.1). After hundreds of researches and contribution of both clinical and forensic interest in decades, as emphasized (Knight’s 2016) the BCA happen when an infant or child suffers repetitive physical injuries inflicted by a parent or guardian, in circumstances that exclude accident.

Fig. 11.1
figure 1

Clinical and forensic constellation related to Battered child syndrome (authors’ observation). Male patient, 6 years old at Emergency Department observation, presenting multiple skin bruises (periorbital and on the back), adult bite mark on genital area

Full size image

11.1 Fatal Physical Abuse of Children Epidemiology

Child maltreatment—the physical, sexual, mental abuse and/or neglect of children younger than 18 years—exists in every society. It is common in the WHO European Region [22] and globally, often occurring with other negative experiences (mental illness, drug or alcohol problem, prison, witnessing intimate partner, domestic violence, parental separation). While severe child maltreatment may come to [23] the attention of child protection agencies, more hidden forms that progress over many years also exist [24].

Assessments of child abuse involve the interaction of multiple disciplines, including medicine, social work, law enforcement, and the judicial system [25, 26]. This interdisciplinary approach, which is facilitated by Child Advocacy Centers or similar multidisciplinary models, can be challenging because of differing definitions of child abuse, expectations regarding information that can be determined during the medical evaluation, or interpretations of findings [22, 27]. Severe abuse can lead to homicide [28,29,30,31,32,33,34]. While homicide rates for children aged under 15 in the Region appear low at about 850 deaths per year, many child deaths are not investigated and the numbers may be much higher [35,36,37]. National statistics on child abuse in the USA show that in 2013 approximately 679,000 children were victims of maltreatment, and approximately 1520 of them died. Children in the first year of their life had the highest rate of victimization of 23.1 per 1000 children in the national population of the same age. Of the children who experienced maltreatment or abuse, 18% suffered physical abuse. Although figures vary yearly, approximately 700,000 cases of child abuse and neglect are reported annually in the United States, of which 117,772 are physical abuse, as documented by the Department of Health and Human Services in 2015 [23, 24].

Child maltreatment is considered an important public health issue in the European Region. Within data from European Report on Preventing Child Maltreatment, 2013, child maltreatment leads to the premature death of 852 children under 15 years in the European Region every year. In that document it is also properly observed that ‘not all deaths from maltreatment are properly recorded and this figure is likely to be an underestimate’. Deaths are the tip of the iceberg, as it is estimated that for every death, there are between 150 and 2400 substantiated cases of physical abuse [38]. The number of children suffering from maltreatment whose plight goes unrecognized is likely to be very much higher and may only come to light through population surveys [39]. Global estimates state that prevalence ranges from 4 to 47% for moderate-to-severe physical abuse, 15 to 48% for emotional and 20% for sexual abuse in girls and 5 to 10% in boys [40], suggesting that tens of millions of children in the Region suffer different forms of maltreatment. Differences also exist within countries and child death rates are several times higher in disadvantaged populations than wealthier communities; this is also true for hospital admissions, with children from deprived neighbourhoods more likely to be admitted for assaults. Deprivation exposes children to more risk factors for abuse: these can grow over time, increasing the likelihood of violence and neglect [41].

Trauma is the most common cause of death in childhood, and inflicted head injury is the most common cause of traumatic death in infancy [40]. The physician may be asked to render a legal opinion as to whether medical findings indicate abuse [42]; many published reports on medical findings indicative of abuse are based on observational data—primarily from case series—and on clinical judgment [43]. In addition to a medical evaluation to guide treatment, findings that do not require therapy but that support an inflicted cause must also be documented [44]. The legal mandate for physicians to report suspected child abuse requires a reasonable suspicion of abuse, which is sometimes a difficult criterion to meet because of uncertainty regarding the diagnosis [45], particularly when the physician is also a paediatrician caretaker of family, possibly the physician may want to be more certain of the diagnosis [46]. Sometimes the history and/or examination findings facilitate a prompt accurate diagnosis of assault, but this is an uncommon scenario [47,48,49,50]. More commonly, the suspicion of child abuse arises after a doctor is told an uncommon story about how an injury occurred or the doctor discovers an injury frequently attributed to assault [51,52,53]. Suspicion can arise when the pattern of injury seems discordant with the alleged mechanism, especially after consideration of injury biomechanics. Discordance is only one of many factors [30, 54,55,56] that raise concern about child abuse but it is an important consideration when evaluating children’s fractures [57, 58].

Once child abuse is suspected, forensic practitioners must remain open to the possibility that the history provided may be truthful, fabricated, deliberately misleading or incomplete. A careful search for other evidence of injury, such as patterned bruising [59, 60] from fingertip pressure, wounds of different ages, cigarette burns and signs of neglect must form part of the child’s clinical examination [61]. Object of protection of such vulnerable persons as a minor must be an absolute imperative.

11.2 Clinical Points on Physical Abuse of Children

The diagnosis of child abuse is often not just a simple diagnosis but requires knowledge from different medical disciplines (paediatrics, neurology, ophthalmology, dermatology, surgery, forensic medicine, toxicology) to reveal a solid diagnostic basis taking into account all differential diagnoses of accidental trauma or confounding diseases [62,63,64] (Table 11.1). The diagnosis of child abuse may have a number of legal consequences [65]. To avoid legal consequences against the treating physicians—in cases of unreported suspected child abuse as well as in the event of reported but not proven child abuse—the diagnosis has to be confirmed and validated [66, 67].

Table 11.1 Assessment for suspected physical abuse of a child
Full size table

The legal mandate for physicians to report suspected child abuse requires a reasonable suspicion of abuse [68], which is sometimes a difficult instance to meet because of uncertainty regarding the diagnosis, particularly when the physician has a relationship with the family, in which case the physician may want to be more certain of the diagnosis. Following key points and assessment for Suspected Physical Abuse of a Child (Table 11.1) are to be carefully considered in diagnostic approach [7]:

  • Physically abused children, particularly infants, may present with non-specific symptoms and signs, such as vomiting or apnoea; the possibility of abusive head trauma requires consideration in such cases.

  • Physical findings, such as bruising of the face, neck, or torso, or intraoral lesions, such as torn frenula, in infants who are not yet ambulatory should arouse suspicion of inflicted trauma.

  • The evaluation of infants and young children for suspected inflicted trauma should include a complete physical examination of the child, with particular attention to the skin, oral cavity [69], and abdomen; imaging of the brain; an examination for retinal haemorrhages; a skeletal survey; and measurement of hepatic and pancreatic enzymes [70,71,72].

  • Physicians are mandated to report to child protective services cases in which they have a reasonable suspicion of child abuse [73].

Detecting fragile bones is a part of clinical assessment, difficult to determine, particularly in children. Neither X-ray images nor bone densitometry scans provide a clinically useful measure of bone strength [57]. Bones have two main structural components: mineral and protein. Abnormality of bone mineralization in children is called osteomalacia or rickets, most commonly related to vitamin D deficiency but there are multiple congenital and acquired causes of rickets. Abnormality of the protein component of bones can also be due to a large number of congenital or acquired conditions and present either as part of a generalized bone disorder effecting multiple bones or joints or incidentally as a localized deformity [74]. The bones of children with reduced bone strength can fracture as a result of lesser forces than children with normal bones (Table 11.2).

Table 11.2 Conditions that may affect bone strength in children
Full size table

Blood tests that should be performed to evaluate bone metabolism include blood levels of calcium, phosphate, alkaline phosphatase, urea, electrolytes, creatinine and vitamin D [75, 76]. There is an ongoing debate regarding the association between vitamin D levels [77], bone mineral density and bone strength at different ages [78] and the AAP have recently recommended that breast-fed infants receive vitamin D [79].

Second-line blood tests include parathyroid hormone levels and urine metabolic screening. Tests for metabolic disease, osteogenesis imperfecta, copper deficiency, syphilis and scurvy might be considered when there are suggestive clinical or radiological features [80]. Additional tests for rare disorders should be considered when additional information related to family history (such as family members with a history of abnormal bones) and the child’s particular circumstances (such as abnormal diet, clinical findings and concurrent illness) warrant further consideration.

Bone densitometry tests are not recommended as a useful tool in the routine forensic investigation of bone injury in childhood because these tests have controversial reference ranges in children. In addition, the tests have such low sensitivity and specificity for reduced bone strength that they are of limited use in forensic evaluation of injury.

In circumstances when the diagnosis of osteogenesis imperfecta or other rare metabolic conditions are being considered, advice from an expert in metabolic and genetic conditions should be obtained [81, 82]. The diagnosis of osteogenesis imperfecta is usually a clinical one. Specific tests of collagen synthesis and the genes that code for type 1 collagen (COL1A1 and COL1A2) might be recommended when the diagnosis of osteogenesis imperfecta is being seriously considered [83]. In literature report of a case of OI type I misdiagnosed as child abuse in which treatment was successful despite a tardive diagnosis are still present [84, 85].

11.2.1 Child Fatalities Related to Physical Abuse and Head Trauma

Deaths due to child abuse can occur as a consequence of intracranial or extra cranial injuries [86]. Overall, head injuries are the leading cause of death in abused young children [87], and extensive research has been published describing the epidemiology, patterns and mechanisms of injury associated with paediatric abusive head trauma [88,89,90,91,92,93,94,95]. Uncertainties continue to surround determinations of abusive head trauma [96, 97]; the dispute about the aetiology of these injuries relies both with the advances of forensic knowledge about this matter (way and manner) [98,99,100] and duty against the law.

The names applied to the syndromes of Inflicted Head Injury in infancy reflect the evolving and sometimes controversial [101] understanding of the actions necessary to cause the types of injuries seen [11, 102,103,104,105], such as shaking an infant held by the arms or trunk or forcefully striking an infant’s head against a surface [40, 106, 107]. A special focus concerns the triads of Subdural Haematoma, Retinal haemorrhage, brain injury patterns and Shaking syndrome [99, 108,109,110,111,112,113,114], with a more recent caveat of the possibility that uncommon [115], or silent ‘pathological’ [112, 116,117,118,119,120,121,122,123,124,125,126] causes determining retinal haemorrhages.

During inflicted head injury, a distinct type of trauma occurs causing more global brain injury with hypoxic-ischaemic brain injury (HII) and more severe retinal haemorrhages [127,128,129,130]. HII is not a necessary factor for severe retinal haemorrhage to develop from inflicted trauma [131].

Understanding of paediatric abusive head trauma has evolved over the last five decades. In 1962 issue of Kemp identified intracranial haemorrhage in young children as a hallmark sign in many cases. In 1971, Guthkelch suggest shaking as a form of abusive injury, reporting on 23 children (22 < 18 months of age) presenting with various combinations of subdural haemorrhage, fractures, parenchymal brain injury, and retinal haemorrhages [113]. Shortly thereafter, Caffey coined the term Whiplash-Shaken infant syndrome [1]. Both Authors noted a frequent absence of external signs of trauma and suggested the role of torn bridging vessels in the brain as the cause of the intracranial haemorrhage [3, 103, 104]. Using autopsy evidence and a dummy model, Duhaime and colleagues in 1987 suggested that blunt impact trauma may be a prerequisite to generate sufficient deceleration forces for the characteristic injuries to occur [132]; however, consistency across perpetrator confessions suggests that shaking alone is sufficient to cause such injuries, and actual injury threshold levels for infant brains have yet to be established [133]. There are currently multiple hypothesized factors in the pathogenesis of brain pathology and retinal haemorrhage in abusive head trauma [134,135,136], including deceleration and sheering injury, hypoxic-ischaemic injury (from decreased perfusion or apnoea), blunt impact, neck flexion-extension, and raised intracranial or venous pressures [137]. However, the relative importance of these factors cannot be determined precisely based on the published data [138,139,140].

The availability of diffusion-weighted magnetic resonance imaging (DW-MRI) and non-invasive vascular imaging techniques now make it possible to evaluate the role of hypoxic-ischaemic injury (HII) in traumatic paediatric head injuries [141, 142]. DW-MRI enables identification of acute cellular injury and cytotoxic oedema, which in the context of head trauma may result from hypoxic-ischaemic injury, direct traumatic injury, or both. Brain tissue damage causes shifting of water molecules from extracellular to intracellular compartments, which can be identified as reduced diffusion of water on DW-MRI, in comparison to undamaged areas. Such changes can be identified early, hours or days before changes in the appearance of tissue on T2-weighted sequences. DW-MRI (versus T1 or T2) is particularly helpful in infants, whose brains have a high water content and immature myelination. Biousse and colleagues reported a high incidence of possible HII in a cohort of infants with presumed abusive head trauma. In 2007, Ichord and colleagues demonstrated a relationship between HII and inflicted trauma using DW-MRI in a cohort of children with both accidental and inflicted head injuries [143].

figure a

(a) Axial T1, (b) axial T2 and (c) coronal flair MR images of a 12-month boy affected by bilateral subdural chronic haematoma (arrows) with involvement of the interhemispheric fissure (curved arrows). (d) SWI MR image demonstrating retinal haemorrhage (arrow). Courtesy of Dr. Andrea Rossi Giannina Gaslini Children’s Hospital, Genoa

It has been clearly established that both the presence and increasing severity of retinal haemorrhages are highly associated with abusive versus accidental injury in children presenting with traumatic intracranial haemorrhage [144,145,146]; however, the mechanisms underlying retinal haemorrhages are still not clearly established, and there is limited information in the literature addressing HII as it relates to ocular findings in the setting of paediatric head trauma [147, 148].

Although the available evidence suggests that it is the sudden deceleration associated with the forceful striking of the head against a surface that is responsible for most, if not all, severe, inflicted brain injuries [149, 150]. Because the histories given when infants with such injuries present for medical attention are often vague or unreliable, the events must be inferred from knowledge of the causative forces in witnessed cases of accidental trauma and experimental models of injury [151]. Studies of the biomechanics of brain injury have established that forces applied to the head that result in a rotation of the brain about its centre of gravity cause diffuse brain injuries. A differential diagnosis that includes non-accidental as well as accidental causes of skull fractures in child deaths is always required in forensic practice [94, 149, 152, 153]. Accidental deaths due to severe trauma (i.e. motor vehicle collisions, vehicles striking pedestrian, skull fractures from heavy falling objects) usually result in extensive multiple injuries together with skull fractures and intracranial findings. In accidental cases such as these, the investigative history usually corroborates the manner of death, since it corresponds with the injuries found, and excludes child abuse [154,155,156]. Conversely, a fall from a short height (either unwitnessed or witnessed by a single caretaker) can result in skull fractures in children with a very similar pattern to the ones found in inflicted trauma. Confounding this diagnostic dilemma, the majority of head injuries in child abuse and accidental head trauma are often both explained by parents as accidental [157,158,159]. According to the literature, accidental skull fractures will rarely lead to serious or life-threatening intracranial injury. Further, skull fractures due to accidental falls are rarely seen in concert with simultaneous fractures in other skeleton segments (i.e. ribs or extremities) [160, 161]. The investigative history is a fundamental part of the diagnostic process, as an accidental skull fracture can nearly always be routinely absent [157, 159, 162].

Diastases of cranial sutures are more common than skull fractures and could be either a direct consequence of trauma or due to raised intracranial pressure from any cause, because suture diastases occurred both in cases with signs of impact to the head as well as no signs of impact [163,164,165]. Their frequency is significantly higher in infants than in toddlers, because of the progressive fusion of the sutures during growth. Extradural haemorrhages (EDH) is rarely seen in paediatric cases due to the tight adherence of the dura mater to the skull and because of the elasticity of the young skull and not always associated with skull fractures. EDH does not occur in subjects with no signs of an impact. EDHs are typically associated with accidental trauma but have been also described in abused children. EDH may occur with relatively minor trauma to the parietal or temporal skull if the vulnerable middle meningeal artery is torn, so it is frequently observed in accidental trauma (Fig. 11.2).

Fig. 11.2
figure 2

Accidental head trauma in a 3-year-old boy. Fall from a short height (witnessed by caretakers and children) (Authors’ observation). (a) Axial non-contrast CT scan shows a huge epidural right-side epidural haemorrhage, with inhomogeneous density of haematoma, due to different timing of haemorrhage; (b) axial CT scan slab-VR 3D reconstruction shows deformity of parietal and frontal bone along the squamosal (^) suture, compared to left side (∗); (c) 3D VR image clearly shows the fracture along the squamosal suture

Full size image

Intra-dural haemorrhages are caused by physical or physiologic damage to the dural capillary plexus that, according to some researchers, can lead to SDH [166]. SDH are observed both in the subjects with and in those without visible evidence of impact injury. Infants are more affected by subdural haematoma than toddlers and older children, because the brain of infants has more space than the brain of older children to move around in the skull upon impact. It has been reported in the literature that subdural haemorrhages in cases of abusive paediatric head trauma are rarely massive [137, 167]. This seems to confirm that SDH is not typically a lesion producing increased intracranial pressure, but rather a marker of brain movement within the cranial cavity, which may be associated with some shearing brain injury (e.g. diffuse axonal injury). In fact, because the dura is firmly attached to the skull and the arachnoid to the cerebral cortex, most brain motion occurs across the potential subdural space. The thin-walled bridging veins are thus easily vulnerable to tearing. Therefore, in every paediatric autopsy it is extremely important for the pathologist to remove the brain personally or directly observe its removal when performed by a technician. Otherwise, a thin layer of blood from subdural bleeding could easily be missed as it will tend to slide quickly off the surface of the brain as the calvarium is removed. Subdural haemorrhages caused by accidental trauma are typically produced by severe force such as a motor vehicle accident, ejection from a motor vehicle, or a fall from a significant height. Accidental SDHs usually occur at the site of impact, are limited to the cerebral convexities, and are often isolated and associated with an overlying fracture.

11.2.2 Pathology Findings at Autopsy

Closed head injury in early infancy (5 months of age and younger) produces focal lesions, parenchymal laceration—also known as contusional tears—and diffuse astrocytic reaction [168,169,170].

The presence of skull fractures, diffuse axonal injury, and subdural haemorrhage suggest that abused children are subject to several forms of injury, direct trauma to the skull causing fractures; brain acceleration over a short arc producing subdural haematoma; and slower acceleration in shaking where axons are damaged [171]. This inertial effect may also produce acute subdural haemorrhage as the subdural veins are sensitive to shearing forces. Diffuse axonal injury was not usually seen in the brain stem. In the diffuse axonal injury described in adults axonal discontinuities are typically seen in the dorsolateral quadrant of the brain stem [172]. In all cases in which contusional tears had occurred, diffuse axonal injury was evident [141, 173]. Thus, the presence of a contusional tear was focal evidence of more diffuse damage. The experimental studies of axonal damage to subhuman primates, using angular acceleration, suggest that there is a slower component of the acceleration which produces the diffuse axonal injury, but it is impossible to identify the varieties of trauma used in the cases [174, 175] (Fig. 11.3).

Fig. 11.3
figure 3

A 3-month-old infant with accidental head trauma, with clinical and circumstantial suggesting infant shaking syndrome; the child showed coma and apnoea at presentation to first aid of health professional and died after few minutes; at autopsy gross examination showed a thin layer of subarachnoidal haemorrage (∗) on cerebral convexities and left side (a) and dramatic brain swelling (b), with histology evidence of hypoxic ischaemic brain damage (c, H&E, 25×)

Full size image

11.2.3 Child Abuse Imaging Protocol and Forensic

When bone trauma is suspected, the critical investigations are radiological and a classic aphorisms in the study of child abuse was stated by the forensic pathologists, Cameron, Johnson and Camps (‘The skin and bones tell a story which the child is either too young or too frightened to tell’). The essence of forensic evaluation of bone trauma in children is being able to determine whether bone trauma has occurred, when bone trauma might have occurred and being able to determine the likely mechanism of the injury [56, 176, 177].

Forensic evaluation of bone injury in children requires a preliminary understanding of bone metabolism and growth, an understanding of injury biomechanics causing fractures and an appreciation of the power and limitations of radiological investigation to accurately detect the presence and estimate the time of bone injury.

The forensic opinion about the likely cause(s) of injury can then be compared to the offered explanation and their mutual compatibility analysed. To establish whether bone trauma has occurred, it is also essential to have an understanding of the normal structure of different bones and how they deform with application of different forces [178]. Infants and very young children have significantly different bone structure, metabolism, bone strength and reaction to mechanical trauma than older children, adolescents and adults [179]. In addition, there are a number of metabolic processes and congenital abnormalities that can affect the strength of children’s bones and therefore their bones’ susceptibility to injury when mechanical forces are encountered [180, 181]. Children’s bones are softer, contain more cartilage than adults’ bones, the growth plates are relatively fragile and the structure of the bone matrix alters with maturity.

Bone trauma in children can be difficult to detect with both clinical and radiological examination findings sometimes problematic to analyse. Errors in interpretation of findings can lead to incorrect forensic conclusions and subsequent erroneous action on the part of the state to protect vulnerable children. A misdiagnosis of child abuse can be as harmful as a missed diagnosis of child abuse. Studies attempted to compare magnetic resonance imaging (MRI) using limited sequence protocols to CT (computerized tomography) for evaluating paediatric head trauma. Although studies have some differences in sequences and population demographics, both conclude that the CT and MRI are comparable in terms of detecting acute intracranial haemorrhages. In fact, one institution claims to use MRI as the initial imaging exam in paediatric head trauma except ‘if the wait for MRI is unacceptably long for appropriate patient care’ [182], and moreover doesn’t use ionizing radiation [183].

As it was said, if MRI could fulfil three conditions—be cheap, be available everywhere, with no or minimal wait time, and offer complete scans in very short times—every Emergency Room would have one and would use it almost exclusively for evaluating paediatric head trauma [184]. A full sequence MRI is more sensitive and specific than CT for everything in the head except for some bone lesions. Add to that the capability for diffusion imaging and diffusion tensor imaging for discovering new information and making diagnoses of brain lesions, not detectable on conventional MRI, and the user can appreciate why MRI may eventually become ubiquitous. Using MRI costs more and is more expensive to operate. It has physical limitations to access since the magnet is always on and is waiting to suck in the unwary. Also, the vast and ever growing array of possible MRI sequences causes longer and longer exam times. MRI imaging, furthermore, needs completely immobilization of patients and this condition is difficult to have in children, sometimes making necessary to anaesthetize young patients.

Some authors observe they are aware of and very sensitive to radiation exposure in children, particularly those who will need repeated studies using ionizing radiation. However, with the exception of the lens of the eye, the brain is pretty radiation insensitive, but is always necessary to consider the risk-benefit ratio of using CT imaging instead of MR one. If there is an acutely injured child it makes no sense to not perform a CT if that is what the situation demands. A properly performed head CT is highly unlikely to add much, if any, morbidity to a child’s life, especially one with a critical head injury.

As with other areas of medicine, the diagnostic process builds upon the history provided and the clinical examination findings. The process of forensic evaluation of injury follows a standard pathway; this is no different for evaluation of suspected bone trauma than for any other injury that might have an inflicted cause.

Financial, political, jurisdictional and geographical considerations result in varying recommendations and practices throughout the world.

Recommendations from local health authorities, colleges and special societies guide good practice [185,186,187]. If access to facilities for specialist investigation is limited, alternative means of investigation need to be considered, as the following: it is also questionable the possibility to perform all investigations close to the child’s home, to minimize the child’s exposure to ionizing radiation and to investigate using recommended ‘gold standard’ tools. The long-term risks of ionizing radiation which may have a latent period of decades must be balanced against the short-term risk of further physical harm to the child. Obtaining consent for these imaging procedures should not be overlooked, but will sometimes come from a temporary Court-nominated child carer, instead of child’s parents. Sometimes it is in a child’s best interests for him/her and to assure physical protection to travel to a specialist centre or third level hospital, for investigation using techniques and facilities optimized for paediatric imaging. Decision-making must carefully consider the local national guidelines for health assessment of child protection. For children aged less than 2 years, consensus exists regarding the need for a high-quality radiographic skeletal survey as an essential part of the investigation of suspected non-accidental injury. Even if the injury appears to be localized to one body region, the whole skeleton is surveyed because these children are (usually) unable to give a reliable history, are vulnerable to injury because of their immature and relatively weak skeleton and their small size and the injuries may not be evident from clinical examination of the child. The recommended facilities, equipment, technique and protocols needed to obtain this high-quality skeletal survey vary slightly between countries and professional colleges, however, with each region recommending a standard protocol. The American College of radiology (ACR-SPR, Practice Parameter for the Performance and Interpretation of Skeletal Surveys in Children) recommends high-quality digital X-rays, with lowest possible radiation dose, and concurrent monitoring of the images by a radiologist (in case additional views are required in order to further define perceived abnormality) [12, 13]. A common error in acquisition of skeletal survey images by practitioners unfamiliar with accepted protocols is the failure to obtain coned views of the metaphyses and growth plates of the long bones (especially at the wrist, knees and ankles) and inadequate imaging of the ribs due to poor aeration of the lungs, patient movement, suboptimal exposure factors/image capture or poorly centred X-ray beam. A nuclear medicine scan is recommended as an additional ‘first-line’ investigation because the combination of nuclear medicine scan and radiographic skeletal survey increases the overall detection rate of non-accidental injury in children [180, 188,189,190,191,192], but really Authors think that the radiation exposition of nuclear medicine scan could be too high to be routinary accepted in clinical practice. Other imaging modalities can also be used to provide additional information when clinically indicated. Ultrasound examination of bone, for example, for ribs fractures, can help radiology in detecting fractures not clearly detectable in X-ray imaging [193,194,195].

In children aged more than 2 years the detection rate of occult bone injury is significantly lower and suspected bone injuries are best investigated using an X-ray of the suspected site of injury. For some children aged 2–3 years in whom occult bone injury is strongly suspected, radiological investigation with skeletal survey may be considered, leaving bone scan in doubt cases. The likelihood of detecting occult bone injury in a child aged more than 5 years is small because they are able to provide a better history and their bones require a greater force to cause injury. Consequently, for older children, radiographic skeletal surveys should be reserved for emergency or in other exceptional circumstances when the clinical suspicion is unusually high. Blood tests to evaluate bone metabolism should also be performed as a routine procedure when non-accidental bone trauma is suspected in children of all ages. The authors suggest tests such as serum calcium, phosphate, alkaline phosphatase, urea and electrolytes and serum vitamin D [196, 197]. Sometimes additional tests need to be considered. When an individual clinician lacks adequate knowledge and skill to collate this information and form a forensic opinion, it is imperative that he/she seek advice from a suitably experienced forensically trained colleague. Nuclear medicine bone scan demonstrates bone trauma by identifying the metabolic changes that occur within bone tissues as a consequence of trauma. Sometimes, the bone metabolic changes demonstrated may not be associated with a visible change in appearance on an X-ray examination.

An understanding of the physical principles of these imaging techniques is required to facilitate the identification of injuries that are not apparent on clinical examination. The changes that are demonstrated must be differentiated from normal variants and pathologies that mimic bone trauma. Some mimics of bone pathology require investigation with other laboratory tests.

The differences in the electron density of different body tissues and the properties of the different detector systems allow five different densities to be identified on conventional X-ray images or radiographs of humans. Bony trauma is most commonly identified when a gap, defect, break or alignment-abnormality is demonstrated in a bone or there are radiological signs of a healing bone injury present. Factors that influence whether bone injury will be demonstrated include the separation of the bony margins of the fracture, orientation of the fracture to the X-ray beam, position of the fracture in the body relative to other structures (is it obscured by overlying bones) and both the stage and extent of bony healing. Technical factors that can influence how well the bones are demonstrated include the resolution of the imaging system used, how well the child is immobilized, whether the images are coned to a region, the number of projections making up the examination and whether exposure factors appropriate for the size of the child are used. Coned views and different projections can be very important to identify subtle injuries of growth plates and ribs.

figure b

Femoral radiogram (AP view) demonstrating metaphyseal distal femoral fracture in a 6-month boy. Diffuse periosteal reaction of the distal femur after 15 days (AP and lateral views). Courtesy of Dr. Claudo Granata Giannina Gaslini Children’s Hospital, Genoa

Plain radiographs are two-dimensional (2D) representations of 3D structures and this causes superimposition of body tissues and organs. This, together with the limited ability of plain radiographs to distinguish between different body tissues, can make identification of some structures difficult. However, when looking at bones in the extremities, this is less of a problem as there are fewer superimposed structures, and radiological orthogonal projections of bones help in fractures detection. Plain X-rays have the ability to produce images with very high spatial resolution (down to 100 μm or less). Despite this very high resolution, some bone fractures are not visible on plain X-rays due to the orientation of the X-ray beam to the fracture and/or the small distance of separation of the fracture fragments. Computed tomography (CT) imaging produces images by using a collimated X-ray beam that is directed through the patient in multiple projections. The data collected from the multiple different projections allows mathematical computation of the attenuation of X-rays in individual points of the patient. A single slice used to take less than 1 min to acquire and a longer time to process. Current multidetector CT scans have a resolution of about 0.3 × 0.3 × 0.3 mm, can simultaneously acquire up to 320 slices in less than 0.5 s and are able to process all the data required to generate/construct the images in a few seconds. The improved temporal resolution reduces movement artefact and the improved spatial resolution in all imaging planes has resulted in an increased use of CT scans to identify complex fractures in multiple planes or oblique to the incident X-ray beam, with very good contrast resolution. CT scans can therefore demonstrate fractures that are not visible on conventional radiographs. Computed tomography utilizes ionizing radiation and involves a higher, more significant dose than plain radiographs. Where plain radiographs are able to differentiate five densities, CT has a much greater contrast resolution and can differentiate between tissues such as blood, white matter, grey matter and cerebrospinal fluid (CSF) in a CT scan image of the brain. The contrast between tissues of different structures can be manipulated by adjusting the image window width and level; this gives rise to terms such as ‘bone window’ settings. CT imaging allows superimposed structures to be individually identified and provides detail of bone anatomy which is very useful in the assessment of bones with a complex shape and around joints. CT very clearly shows focal areas of bone lysis/destruction, areas of bone sclerosis and periosteal new bone formation. CT is able to provide some information about soft tissues, particularly when there is localized soft tissue calcification or in case of subcutaneous emphysema, not easy to depict at X-ray when is of little amount. Ultrasound examinations and MRI studies usually provide superior images of soft tissues that are more likely to identify soft tissue pathology, especially in case of detection and in dating of haematomas. Forces causing bones fracture in infant is a part of diagnostic process. Planar micro-fractures at the metaphyseal-epiphyseal region in the immature primary spongiosum layer of immature long bones are believed to be caused by planar shearing forces. These injuries are referred to as classic metaphyseal lesions [198], but they are frequently referred to as ‘bucket-handle’ or ‘corner’ fractures [199]. Shearing forces occur during rapid acceleration and deceleration when an infant is shaken and when traction, compression or rotation forces are applied to cartilaginous epiphyses. While the X-ray appearance may differ depending on the projection of the X-ray beam, these radiological findings actually represent a single pathological process. Diaphyseal (shaft of long bone) fractures are more commonly seen in abused children than classic metaphyseal lesions. Diaphyseal fractures are commonly caused by accidental childhood trauma such as might be experienced during an accidental fall or sporting injury. In contrast to classic metaphyseal lesions, the site and type of diaphyseal fracture rarely contributes to differentiation of accidental from inflicted trauma. In some circumstances, for example, when the presence of a spiral fracture indicates that a torsional force has been applied, a discrepancy between the caregiver’s offered explanation and the observed pattern of injury might suggest that the caregiver is not being truthful. Additional questions would need to be asked about the alleged mechanism of injury. Transverse fractures of the diaphysis can be caused by the application of force perpendicular to the shaft of the long bone (such as a direct blow) but can also be caused by bending forces transmitted along the shaft of the bone (Fig. 11.4). The periosteum is relatively poorly attached to the diaphysis of children’s bones. Torsional and twisting forces can strip the periosteum from the cortex of diaphyseal bone, resulting in subperiosteal haemorrhage and resultant subperiosteal callus. When child abuse is diagnosed on the basis of skeletal injury, skull fracture is a common cause of presentation [164] (Fig. 11.5). Simple linear skull fractures occur as a result of both inflicted and accidental trauma [200,201,202,203,204,205,206] (Fig. 11.6). Non-parietal fractures, multiple and complex fractures, widely separated and depressed fractures and fractures associated with significant intracranial injury should all raise concern about possible inflicted injury. A single impact can result in a skull fracture that crosses a suture line. Rarely, a single impact may transmit forces such that two skull bones fracture some distance from the impact site. When bones on either side of the skull are fractured, multiple impacts and crush injury should be considered. Ping-pong fractures where there is deformity without a gap in the bone due to the pliability of the immature skull and depressed skull fractures can occur as a result of blunt impact with a contoured object in circumstances of accidental and inflicted trauma.

Fig. 11.4
figure 4

Non-accidental fractures in a 4-year-old female (Authors’ observation). (a) Bilateral fractures of distal radius and ulna with metaphyseal dorsal angulation and dislocation. (b) Nevertheless, the altered cortical bone profile, thanks to the physiological bone remodelling, may undergo to a complete resolution

Full size image
Fig. 11.5
figure 5

Non-accidental brain injuries in a 2-year-old female (battered by her mother). (a) Brain CT scan shows right parasagittal and occipital epidural haemorrhage (∗) with ipsilateral tentorial and posterior falcine involvement; (b) small ipsilateral cortical lesion of the occipital/peri-mastoid skull (arrow) is present as well; (c) subarachnoid haemorrhage is still detected in brain MR, performed 3 month after with little haematic residuals (∗) along the occipital bone; (d) 3D VR reconstruction CT image shows an inflexible fracture of occipital bone (arrow)

Full size image
Fig. 11.6
figure 6

18-month-old female. D: 3D VR reconstruction CT image shows a huge fracture of occipital bone (arrow)

Full size image

These fractures show transversal appearances, which are uncommon in accidental traumatic fractures in children.

Indeed, they usually show greenstick appearance, with eccentric far cortex comminution. Alternatively, but less commonly, they may show spiral appearance, caused by the involvement, during traumatic events, of diaphyseal long axis, thus resulting in oblique and longitudinal fracture rime. In this patient instead complete transverse fracture rimes are present and they can be caused by direct injury, probably painful, on the distal diaphyseal bones. Moreover, the bilateral fractures localized at the same bone portion may suggest a non-accidental injury, while they usually show different localization in the same bone, especially in cases of accidental distortion-caused fractures. Patient was treated by orthopaedic casts placement, but without surgical fractures reduction, which has lead, in 30 days, to abundant callus formation and bones deformities.

11.3 Conclusion

There are no particular bone injuries that are pathognomonic of child abuse [202,203,204, 207]. Although there are not bone injuries diagnostic of child abuse [208], however, some patterns of bone injury are seen much more frequently in assaulted children than in children who sustain accidental bone trauma. Classic metaphyseal lesions in children aged less than 2 years are seen far more commonly in abused children than in children experiencing trauma as a result of accidents [209].

Posterior rib fractures, fractures of the scapula, ends of clavicles, sternum and vertebral spinous processes suggest, but do not prove, injury caused by assault. Multiple bilateral symmetrical fractures, fractures judged to be of different ages, fractures of the hands and feet, vertebral body fractures, complex skull fractures and associated non-skeletal injury should raise suspicions of possible assault. Authors have referred to the ‘specificity’ of recognized patterns of bone injury to facilitate a diagnosis of child abuse [198, 210], but caution is imperative. Fractures seen predominantly in children known to have been abused have been categorized as ‘specific’ for child abuse whereas fractures occurring in both assaulted and accidentally injured children have been categorized as ‘non-specific’. The most common types of fractures seen in assaulted children are fractures that are categorized as ‘non-specific’, including diaphyseal fractures and linear skull fractures. When forensic practitioners form an opinion about the cause of a child’s bone injury, their existing knowledge base underpins opinion formation. Unfortunately, the existing literature is replete with examples of poor research methodology, dogmatic statements based on only a small number of cases and faulty reasoning by way of circular logic and other fallacies of logic. For example, the willingness of a forensic practitioner to diagnose child abuse in a toddler with an isolated spiral femur fracture is strongly influenced by the forensic practitioner’s belief that in a non-ambulant child this fracture pattern is more likely to be caused by abuse than an accident. At times, circular logic, the child’s poor social circumstances [211] and the presence of coexisting injuries influence the forensic practitioner’s diagnosis in relation to fracture causation [212]. Knowledge of bone healing in children take a relevant part of diagnostic contribution in a suspected abusive fracture and several authors have made important contributions to our understanding of how bones recover from trauma. The times-since-injury implies particularly those working in a forensic context, are cautioned that these times are only estimates based on limited data and the understanding has been derived from the study of bone injury in children who have died as a result of accidental trauma. Histological analysis suggests that the healing process is usually no different when the cause of injury is inflicted trauma [213]. Bone healing is understood to occur in a number of phases. Firstly, the induction phase extends from the time of injury to the appearance of new bone at the fracture site. An inflammatory response initiated with the associated bleeding may last a few days and reveal itself on an X-ray in the form of soft tissue swelling with displacement and obliteration of normal fat and fascial planes. The initial soft tissue swelling is referred to as the pro-callus. A fracture line that initially appears sharp will change as healing progresses. Damaged bone is first resorbed as part of the inflammatory process which may blur the fracture margins. A nuclear medicine bone scan may detect increased blood flow due to inflammation and an MRI or ultrasound scan may detect soft tissue changes that are not yet evident on an X-ray. Soft or primary callus then starts to form with the fibrocartilaginous tissues laid down to stabilize the bones at the fracture site converting to loosely woven bone. In infants, this calcification/ossification is frequently seen as periosteal new bone and can occur within approximately 7–10 days but tends to occur later in older children (10–14 days after injury). Exuberant callus formation can be a sign of fracture instability and/or repetitive injury. Next, hard callus forms when disorganized periosteal and endosteal bone (sometimes called the provisional callus) begins to convert to lamellar bone with the bony trabeculae orientated along the lines of weight bearing/compression/tension. This phase begins in infants at 14–21 days at the earliest and peaks at 21–42 days. Remodelling occurs with gradual correction of deformity and resorption of the excess callus laid down as part of the initial healing process. This phase begins after bone union and may continue for 1–2 years after the injury. Remodelling of bone following bone fracture in children can result in a bone that heals completely and appears indistinguishable on X-ray from a bone that has not been injured.

Healing generally occurs more rapidly in younger infants but the rate at which bones heal and remodelling occurs varies with a number of factors: the anatomy of the injured bone; the site and nature of fracture; the degree of angulation and displacement of bone segments; the degree of immobilization of fractured bones during healing; stresses applied across the fracture during healing; and metabolic processes that enable healing of bone injury. Subtle fractures such as classic metaphyseal lesions and rib fractures may only be identified on plain X-rays once the healing process is well established, making bone scans or repeat X-ray examination essential in the evaluation of injury. With optimal healing conditions, a fractured bone may appear normal on an X-ray or nuclear bone scan examination in as short a time as a few months. Other bones, such as the skull, may not show radiological evidence of healing; dating skull fractures is notoriously imprecise. Table 11.3 identifies factors that may influence bone healing.

Table 11.3 Bone healing repairing
Full size table

When a fracture involves the growth plates or is allowed to unite with angulation or displacement, the bone may not grow normally resulting in a permanent deformity and/or limb shortening. When children’s bones have been injured by inflicted trauma, other factors that might influence the healing process are a delay in presentation for medical treatment, coexisting nutritional deficiencies (possibly associated with neglect) and subsequent/repeated trauma at the fracture site. Two 1987 case reports by Perkins and Skirvin [214] and Spencer [215] reported faster healing of femoral fractures in children with coexisting severe head injury. These studies have not been replicated and the clinical significance of this association remains uncertain.

When forensic practitioners form an opinion about the cause of a child’s bone injury, their existing knowledge base underpins opinion formation. Unfortunately, the existing literature is replete with examples of poor research methodology, dogmatic statements based on only a small number of cases and faulty reasoning by way of circular logic and other fallacies of logic. For example, the willingness of a forensic practitioner to diagnose child abuse in a toddler with an isolated spiral femur fracture is strongly influenced by the forensic practitioner’s belief that in a non-ambulant child this fracture pattern is more likely to be caused by abuse than an accident. At times, circular logic, the child’s poor social circumstances and the presence of coexisting injuries influence the forensic practitioner’s diagnosis in relation to fracture causation [212]. Knowledge of bone healing in children take a relevant part of diagnostic contribution in a suspected abusive fracture and several authors have made important contributions to our understanding of how bones recover from trauma. The times-since-injury implies particularly those working in a forensic context, are cautioned that these times are only estimates based on limited data and the understanding has been derived from the study of bone injury in children who have died as a result of accidental trauma. Histological analysis suggests that the healing process is usually no different when the cause of injury is inflicted trauma [213]. Bone healing is understood to occur in a number of phases. Firstly, the induction phase extends from the time of injury to the appearance of new bone at the fracture site. An inflammatory response initiated with the associated bleeding may last a few days and reveal itself on an X-ray in the form of soft tissue swelling with displacement and obliteration of normal fat and fascial planes. The initial soft tissue swelling is referred to as the pro-callus. A fracture line that initially appears sharp will change as healing progresses. Damaged bone is first resorbed as part of the inflammatory process which may blur the fracture margins. A nuclear medicine bone scan may detect increased blood flow due to inflammation and an MRI or ultrasound scan may detect soft tissue changes that are not yet evident on an X-ray. Soft or primary callus then starts to form with the fibrocartilaginous tissues laid down to stabilize the bones at the fracture site converting to loosely woven bone. In infants, this calcification/ossification is frequently seen as periosteal new bone and can occur within approximately 7–10 days but tends to occur later in older children (10–14 days after injury). Exuberant callus formation can be a sign of fracture instability and/or repetitive injury. Next, hard callus forms when disorganized periosteal and endosteal bone (sometimes called the provisional callus) begins to convert to lamellar bone with the bony trabeculae orientated along the lines of weight bearing/compression/tension. This phase begins in infants at 14–21 days at the earliest and peaks at 21–42 days. Remodelling occurs with gradual correction of deformity and resorption of the excess callus laid down as part of the initial healing process. This phase begins after bone union and may continue for 1–2 years after the injury. Remodelling of bone following bone fracture in children can result in a bone that heals completely and appears indistinguishable on X-ray from a bone that has not been injured.

A radiological estimate of the time of injury involves evaluation of the appearance of soft tissues, the fracture line, the deformity and the callus. It is therefore only possible to offer an estimate of the possible date of injury within a range. However, when there are multiple sites of trauma it is often useful to assess whether the injuries are likely to have occurred at the same or different times. Repeated trauma raises suspicions of inflicted injury. In the systematic review of Prosser et al. [216], the dating of fractures in children identified only three articles that met their inclusion criteria. Of the 189 children reported (243 fractures), only 56 children were aged less than 5 years, the age group in which most fractures caused by assault occur. Kleinman’s textbook (1998) offers opinion about dating fractures based on the author’s personal experience. Offiah and Hall [217] recently added significantly to the literature about fracture patterns and bone healing in abused children. However, there remain few cases on which to base firm opinions about the features of healing fractures in children of different ages, in different bones and with different fracture patterns.

In addition to the physiological factors, the radiographic technique and imaging system resolution might not be optimized. It is important that high-resolution techniques be used and that imaging is performed by radiographers familiar with techniques used to optimize the imaging of children. Radiologists who are less familiar in interpreting children’s X-rays might not detect signs of bone injury that are recognized more readily by experienced paediatric radiologists. Ultrasound is very operator-dependent and requires specialist skills. Its nature is more suited to localized examination of a clinically suspicious area rather than generalized screening for evidence of injury. However, it can be used to identify soft tissue changes when there is a focus of clinical concern such as bruising or swelling when the X-ray is normal.