Abusive Head Trauma
AntoniaChiesa, MDa,*, Ann-Christine Duhaime, MDb
Child physical abuse that results in injury to the head or brain has been described
using many terms, which haveevolved over the past half-century or more. These items
have included the battered child syndrome, whiplash injuries, shaken infant or shaking
impact syndrome, and nonmechanistic terms such as abusive head trauma or nonac-
cidental trauma.1–7This evolution has occurred as the spectrum of injuries—and the
mechanisms that are potentially responsible for them—have been studied in
increasing detail in multiple clinical series from around the world, as well as with path-
ophysiologic and biomechanical modeling. Because children may present with
varying histories, physical findings, and radiologic findings, the terms ‘‘inflicted head
injury,’’ ‘‘nonaccidental trauma,’’ and ‘‘abusive head trauma’’ are used in this article
to reflect those constellations of injuries that are caused by the directed application
of force to an infant or young child resulting in physical injury to the head and/or its
contents. These injuries most often include subdural and/or subarachnoid hemor-
rhage, with varying degrees of neurologic signs and symptoms. A high proportion of
children also present with retinal hemorrhages, physical or radiologic evidence of
contact to the head, upper cervical spine injuries, and skeletal injuries. These features
are described in more detail in the following section.
Use of the more general terms reflects an attempt to avoid the pitfalls of assuming
the exact mechanism of injury; the general terms also encompass a wide range of
traumatic forces, which are potentially harmful and can result in different patterns of
neurotrauma. These forces include: blunt force trauma, acceleration/deceleration
(inertial) forces, penetrating trauma, and asphyxiation.
Establishing incidence data for abusive head trauma has been challenging, in part
because of the definitional issues noted; however, several studies have attempted
to examine issues of epidemiology. Early studies revealed that inflicted injuries
aDepartment of Pediatrics, Kempe Child Protection Team, The Children’s Hospital, 13123 E.
16th Avenue, Box 138, Denver, CO 80045, USA
bDepartment of Pediatric Neurosurgery, Children’s Hospital at Dartmouth, Dartmouth
Hitchcock Medical Center, One Medical Center Drive, Lebanon, NH 03756, USA
* Corresponding author.
E-mail address: Chiesa.Antonia@tchden.org (A. Chiesa).
? Child abuse ? Abusive head trauma
?Pediatric traumatic brain injury ?Shaken baby syndrome
? Nonaccidental trauma
Pediatr Clin N Am 56 (2009) 317–331
0031-3955/09/$ – see front matter ª 2009 Elsevier Inc. All rights reserved.
make up a significant portion of traumatic brain injury in children younger than 2 years
of age, and such injuries account for serious morbidity and mortality in that group.8,9
When compared with accidental head injury, the hospital length-of-stay and medical
costs incurred from abusive head trauma are higher.10
A recent study out of North Carolina found an incidence of inflicted brain injury in the
first two years of life of 17.0 per 100,000 person-years.11Another prospective study
from Scotland during 1998–1999 found an annual incidence of 24.6 per 100,000 chil-
dren younger than 1 year (a higher rate than a previous 15-year retrospective study
done in the same county).12The authors of that study suggested that the discrep-
ancies between the prospective and retrospective study outcomes reflect the chal-
lenge of tracking the problem caused by lack of a single international classification
of diseases (ICD) code to describe the medical findings.
This challenge and others were addressed at a 2008 symposium and later summa-
rized in articles to a supplement to the Journal of Preventative Medicine. The sympo-
sium was convened, in part, to discuss definitional issues regarding inflicted brain
injury, as well as methods for measuring its incidence. In his commentary, Alexander
Butchart, PhD, argues that determining the epidemiology of child maltreatment will
help elucidate the issue as a public health concern and lead to the formation of larger
scale prevention efforts.13
In 1946, Dr. John Caffey first recognized a possible traumatic association between
head injuries and fractures in infants.14In the following three decades, important
work by Silverman,15Ommaya,16and Guthkelch,3contributed to the acknowledg-
ment of child abuse as a medical condition. Noting that many of his patients presented
without a clear mechanism of trauma to explain their injuries, in 1974, Caffey coined
the term ‘‘the whiplash shaken infant syndrome.’’4He used the term to describe the
constellation of injuries that includes subdural hematoma, long bone fractures, and
retinal hemorrhages; these are symptoms that, in the absence of a reasonable history
of trauma or other medical condition, are still considered hallmarks for abusive head
injury. The idea that shaking might be causative was first proposed by Norman Guth-
kelch, a neurosurgeon who, working with pediatricians and social workers, obtained
some histories of violent shaking as a part of the injury scenario.3In contrast, Caffey’s
initial concept was that shaking might be injurious even when performed by well-
meaning caretakers as a generally accepted form of discipline, because of the
presumed inherent fragility of young infants.17These authors were aware of experi-
ments in primates, whose heads were subjected to large magnitude angular deceler-
ations involving crashes in high-velocity sleds, leading to unconsciousness and
subdural hemorrhage.18Thus, the idea of angular deceleration as the causative mech-
anism for subdural hematoma was hypothesized as the necessary mechanism in
infant shaking injuries.
Over the ensuing decades, other authors noted a high incidence of contact injuries,
including scalp hematomas, skull fractures, and brain contusions, in abused infants;
the injuries were visible either clinically, radiologically, or at autopsy.19–21Biomechan-
ical models of young infants were developed that suggested that even violent manual
shaking caused angular decelerations that were very low compared with those
required to cause concussive or hemorrhagic injury in primates, but that inflicted
impacts were associated with angular decelerations that were approximately 50 times
greater and within the range thought more likely to be associated with brain injury.19,22
It was suggested that physical evidence of impact might not be seen if the deformable
Chiesa & Duhaime
infant head stopped suddenly against a relatively soft surface, but large magnitude
angular deceleration would still occur from this type of impact. Other authors created
alternative physical models that suggested that shaking alone might generate suffi-
cient force to be injurious.23Additionally, the frequent finding of upper cervical or cer-
vicomedullary injury became increasingly recognized, although the mechanisms
required to cause this type of injury and the exact contribution to the clinical constel-
lation remains incompletely understood. Almost all infants with inflicted injuries who
are found at autopsy to have cervical spine injuries also have subdural hemorrhage
and other typical findings, which seem unlikely to be related only to spinal injury.24–26
Finally, the contributions of the initial history to the understanding of mechanism, as
well as confessions by admitted perpetrators, have suggested that various scenarios
mayoccur in thesetting ofviolent inflictedinjury ininfants andyoungchildren. Although
some caretakers mention shaking, this history is given spontaneously in the minority of
the child presents because of symptoms) are the most common initial histories given.28
Although some confessions mention shaking, others involve throwing, striking, or other
violent mechanisms.Forallthesereasons,terminology that doesnotsuggesta specific
papers published by such entities as the American Academy of Pediatrics29and the
National Association of Medical Examiners30reflect the evolution of the medical litera-
ture over a relatively short period of time.
Determining whether a child’s injuries are the result of child physical abuse can be
a difficult process. If a traumatic cause is not readily apparent, careful assessment
is warranted. Glick and Staley propose that a detailed evaluation by a multidisciplinary
child protection team has become standard of care in many facilities.31Teams are
typically directed by a child abuse trained pediatrician (subspecialty board certifica-
tion available in 2009) and may include members from other medical and surgical
disciplines, as well as medical social work. This approach ensures a thorough consid-
diagnosis is reached objectively.
A core component of the diagnostic process is a comprehensive history of present-
ing illness. The initial history should include details about timeline of symptoms and the
exact events leading up to the present, including a detailed description of the events
before and after the child became symptomatic. When there is a history of trauma,
a detailed description of exactly what happened, what position the child was in,
how the child landed, what the fall height was, how the child acted immediately after-
wards and thereafter, and what the caretakers did is invaluable in reconstructing the
injury events.8Included in this history should be: who has cared for the child; the rela-
tionship between caretakers and the child; birth/past medical history; prior trauma;
and family history, particularly any history of bleeding disorders. Care should be taken
not to lead the history by suggesting whether specific mechanisms or actions might
have occurred; it is preferable to simply ask open-ended questions and to seek
specific answers, such as: ‘‘What happened next?’’ or ‘‘What did you do then?’’ or
‘‘What did he/she look like/do?’’
Focusing the history on the identification of a trigger for abuse by the caretaker may
also be helpful. A commonly described trigger is crying. Abusive head trauma inci-
dence curves correlate with periods of normal crying,32colic, and immunizations.
Other triggers in older children include temperament, behavior, and toileting.
Abusive Head Trauma
68. Green MA, Lieberman G, Milroy CM. Ocular and cerebral trauma in non-acciental
injury in infancy: underlying mechanisms and implication for paediatric practice.
Br J Ophthalmol 1996;80:282–7.
69. Odom A, Shrist E, Kerr N, et al. Prevalence of retinal hemorrhages in pediatric
patients after in hospital cardiopulmonary resuscitation: a prospective study.
70. Kanter RK. Retinal hemorrhage after cardiopulmonary resuscitation or child
abuse. J Pediatr 1986;108:430–2.
71. Rubin DM, Christian CW, Bilaniuk LT, et al. Occult head injury in high-risk abused
children. Pediatrics 2003;111:1382–6.
72. Han BK, Towbin RB, de Courten-Myers G, et al. Reversal sign on CT: effect of
anoxic/ischemic cerebral injury in children. Am J Roentgenol 1990;154:361–8.
73. Dias MS, Backstom J, Falk M, et al. Serial radiography in the infant shaken impact
syndrome. Pediatr Neurosurg 1998;29(2):77–85.
74. Zimmerman RA, Bilaniuk LT, Bruce D, et al. Computed tomography of cranioce-
rebral injury in the abused child. Radiology 1979;130(3):687–90.
75. Fernando S, Obaldo RE, Walsh IR, et al. Neuroimaging of nonaccidental head
trauma: pitfalls and controversies. Pediatr Radiol 2007;38(8):827–38.
76. Steinbok P, Singhal A, Poskitt K, et al. Early hypodensity on computed tomo-
graphic scan of the brain in an accidental pediatric head injury. Neurosurgery
77. Gilles EE, Nelson MD Jr. Cerebral complications of nonaccidental head injury in
childhood. Pediatr Neurol 1998;19(2):119–28.
78. Duhaime AC, Durham SR. Traumatic brain injury in infants: the phenomenon of
subdural hemorrhage with hemispheric hypodensity (‘‘Big Black Brain’’). Prog
Brain Res 2007;161:293–302.
79. American Academy of Pediatrics Section on Radiology. Diagnostic imaging of
child abuse. Pediatrics 2000;105(6):1345–8.
80. Kleinman PK. Skeletal imaging strategies. In: Kleinman PK, editor. Diagnostic
imaging of child abuse. 2nd edition. St. Louis (MO): Mosby; 1998. p. 237–41.
81. Kleinman PL, Kleinman PK, Savageau JA. Suspected infant abuse: radiographic
skeletal survey practices in pediatric health care facilities. Radiology 2004;
82. Demaerel P, Caseteels I, Wilms G. Cranial imaging in child abuse. Eur Radiol
83. Jaspan T. Current controversies in the interpretation of non-accidental head
injury. Pediatr Radiol 2008;38(Suppl 3):S378–87.
84. Vinchon M, Noule N, Tchofo P, et al. Imaging of head injuries in infants: temporal
correlates nd forensic implications for the diagnosis of child abuse. J Neurosurg
85. Gerber P, Coffman K. Nonaccidental head trauma in infants. Childs Nerv Syst
86. Jayawant S, Parr J. Outcome following subdural haemorrhages in infancy. Arch
Dis Child 2007;92:343–7.
87. Bonnier C, Nassogne MC, Saint-Martin C, et al. Neuroimaging of intraparenchy-
mal lesions predicts outcome in shaken baby syndrome. Pediatrics 2003;112(4):
88. Hymel KP, Makaroff KL, Laskey A, et al. Mechanisms, clinical presentations,
injuries, and outcomes from inflicted versus noninflicted head trauma during
infancy: results of a prospective, multicentered, comparative study. Pediatrics
Chiesa & Duhaime
89. Barlow KM, Thomson E, Johnson D, et al. Late neurologic and cognitive sequelae
of inflicted traumatic brain injury in infancy. Pediatrics 2005;116(2):e174–85.
90. Durfee MJ, Gellert GA, Tilton-Durfee D. Origins and clinical relevance of child
death review teams. JAMA 1992;267(23):3172–5.
Abusive Head Trauma