Jenny C, Hymel KP, Ritzen A, Reinert SE, Hay TC. Analysis of Missed Cases of Abusive Head Trauma. JAMA. 1999;281(7):621-626. doi:10.1001/jama.281.7.621
Author Affiliations: Department of Pediatrics, Brown University School of Medicine (Dr Jenny), and Lifespan Medical Computing (Mr Reinert), Providence, RI; Department of Pediatrics, National Naval Medical Center, Bethesda, Md (Dr Hymel); Department of Pediatrics, University of Oregon Health Sciences Center, Portland (Dr Ritzen); and the Department of Radiology, University of Colorado School of Medicine, Denver (Dr Hay).
Context Abusive head trauma (AHT) is a dangerous form of child
abuse that can be difficult to diagnose in young children.
Objectives To determine how frequently AHT was previously missed
by physicians in a group of abused children with head injuries and to
determine factors associated with the unrecognized diagnosis.
Design Retrospective chart review of cases of head trauma
presenting between January 1, 1990, and December 31, 1995.
Setting Academic children's hospital.
Patients One hundred seventy-three children younger than 3 years
with head injuries caused by abuse.
Main Outcome Measures Characteristics of head-injured children in
whom diagnosis of AHT was unrecognized and the consequences of the
Results Fifty-four (31.2%) of 173 abused children with head
injuries had been seen by physicians after AHT and the diagnosis was
not recognized. The mean time to correct diagnosis among these children
was 7 days (range, 0-189 days). Abusive head trauma was more likely to
be unrecognized in very young white children from intact families and
in children without respiratory compromise or seizures. In 7 of the
children with unrecognized AHT, misinterpretation of radiological
studies contributed to the delay in diagnosis. Fifteen children
(27.8%) were reinjured after the missed diagnosis. Twenty-two (40.7%)
experienced medical complications related to the missed diagnosis. Four
of 5 deaths in the group with unrecognized AHT might have been
prevented by earlier recognition of abuse.
Conclusion Although diagnosing head trauma can be difficult in the
absence of a history, it is important to consider inflicted head trauma
in infants and young children presenting with nonspecific clinical
head trauma (AHT) is a dangerous form of child abuse. More child abuse
deaths occur from head injuries than any other type of
injury.1 Infants and toddlers who survive AHT often have
serious neurologic sequelae.2,3
Head injury in infants and toddlers can be difficult to diagnose
because symptoms are often nonspecific. Vomiting, fever, irritability,
and lethargy are common symptoms of a variety of conditions seen in
children, including head trauma. When caretakers do not give a history
of injury and the victim is preverbal, an abusive head injury can be
mistakenly diagnosed as a less-serious condition.
In March 1995, we evaluated a 14-month-old child who had sustained an
abusive head injury 4 months previously. Shortly after his initial
injury, he had been examined by his physician and his new-onset
seizures were attributed to his history of prematurity. During the next
4 months, the child had 7 physician visits and 2 cranial imaging
studies. At each visit, the diagnosis of AHT was not recognized. When
we examined him 4 months later, he had multiple old and new fractures
and healing brain injuries, including extensive brain atrophy and
healing brain infarctions. This case encouraged us to review our
experience with AHT cases to determine if the appropriate diagnosis had
been previously missed. We also examined factors that may have
contributed to the unrecognized diagnosis of AHT.
We studied cases of AHT in children younger than 3 years
evaluated at the Children's Hospital, Denver, Colo, from January 1,
1990, through December 31, 1995. The Children's Hospital is an
academic medical center affiliated with the University of Colorado
School of Medicine. It is a referral center for Colorado, Wyoming,
Montana, and western Nebraska.
The children in this study were evaluated by the hospital's
Child Advocacy and Protection Team (CAP Team). The CAP Team is a
multidisciplinary group that consults on cases of suspected child abuse
and neglect. The team is led by pediatricians whose clinical focus is
child abuse. Social workers, nurses, psychologists, child
psychiatrists, and attorneys also participate. The team routinely
to document medical history and the history of
the acute injury, review previous medical and social service records,
review prior radiological studies, perform a careful physical
examination, and order appropriate new diagnostic studies. In all
cases, organic illnesses that mimic AHT are ruled out. Confirmation
that head trauma was inflicted requires multidisciplinary team
Head trauma cases were identified from the log records of the CAP Team
and charts were reviewed in depth. To ensure concurrence, study cases
were reviewed by at least 2 of the authors (including C.J.) and
radiological imaging studies were reviewed by a pediatric radiologist
(T.C.H.). Permission for the anonymous chart review was granted by the
hospital's human subjects committee. Information gathered included
demographics, social and family data, details of the children's
injuries, presenting complaints, clinical course, and details of
previous medical visits related to head trauma, if applicable.
We limited the study to children with head injuries who were younger
than 3 years for 2 reasons. First, children older than 3 years are not
as likely to sustain severe injury when struck in the head or shaken.
Second, children older than 3 years are more likely to be able to
articulate their experiences. Hence, AHT is much less likely to be
missed as the appropriate diagnosis.
Abusive head trauma was defined as inflicted cranial injury.
Researchers debate whether shaking alone or shaking and impact cause
the signs and symptoms commonly referred to as shaken baby
syndrome.4- 6 The mechanism of injury cannot always be
accurately determined in child abuse cases. Because shaking, impact to
the head, or both are all potentially harmful to infants and toddlers,
we grouped all head injuries caused by abuse into the single category
Factors considered by the multidisciplinary team in reaching the
diagnosis of AHT (rather than nonintentional head injury) included (1)
confession of intentional injury by an adult caretaker; (2)
inconsistent or inadequate histories given by caretakers (the history
given did not explain the nature and severity of the injuries); (3)
associated unexplained injuries, such as fractures or intra-abdominal
injuries; and (4) delay in seeking care.
Cases of AHT were defined as missed if review of medical
records and radiological studies confirmed the following predefined
criteria: (1) Prior to the diagnosis of AHT, a physician evaluated the
child (on ≥1 occasions) for nonspecific clinical sign(s) compatible
with head trauma (ie, recurrent vomiting, irritability, facial and/or
scalp injury, altered mental status, abnormal respiratory status,
and/or seizures). (2) The medical evaluation(s) for these nonspecific
clinical sign(s) did not result in a diagnosis of AHT. (3) Thereafter,
1 or more of the following scenarios occurred: (a) The child
improved clinically, later experienced (repeat) acute trauma confirmed
as abusive, and underwent diagnostic imaging that revealed old cranial
injuries and other new injuries. (b) The child remained
symptomatic or experienced worsening clinical signs until head trauma
was recognized, verified by cranial imaging studies, and confirmed as
abusive. (c) The person who injured the child later admitted
to abusing the child shortly before the onset of the child's
nonspecific clinical sign(s). In all cases, the estimated age of the
cranial injuries documented by imaging studies was consistent with the
prior time of onset of the child's nonspecific clinical sign(s).
All remaining cases of AHT evaluated during the study period were
considered recognized. Children who sustained any new
inflicted injuries during the period of diagnostic delay were
classified as reinjured. Study patients whose medical records
after their inflicted head trauma revealed abnormal head growth,
recurrent seizures, psychomotor delays, chronic anemia, vomiting,
weight loss, and/or sensory deficits were classified as having
medical complications of AHT.
We examined data to determine what factors were associated with a
missed vs recognized diagnosis. We used χ2 testing to
assess the independence of 10 variables on the outcome variable of a
correct diagnosis of head trauma. Variables resulting in
χ2P≤.25 or less were entered into an initial
multivariate logistic regression model. We then used Wald and
likelihood ratio testing to iteratively remove noncontributory
variables from the model.7 Analysis
was performed using
Stata software, Version 5.0 (Stata Corp, College Station, Tex).
A total of 232 children with suspected head injuries were evaluated by
the CAP Team from January 1990 through December 1995. Fifty-nine
children did not meet study criteria. Of these, 8 were eliminated
because they were aged 3 years or older. It was determined that 38 were
not abused. The medical records of 13 children could not be located.
The remaining study sample included 173 abused children with head
The mean age of the 173 children was 247 days (range, 10 days to
2.9 years). Ninety-five (55%) of the children were male and 78 (45%)
were female. The boys' ages at the time they were first seen for
symptoms of AHT were not significantly different than the girls' ages.
In our study sample, minorities were overrepresented (33.5% minority)
compared with the racial distribution of the Denver metropolitan area
The types of injuries noted in the children are shown in Table 1. Many of the children sustained more than
1 type of injury. Eighty-nine children (51.4%) were covered by
Medicaid-funded insurance programs. Twenty-seven children (15.6%) were
uninsured. The remainder had private health insurance.
In the 173 children with AHT, 54 cases (31.2%) were classified as
missed. For children with missed AHT, the mean number of physician
visits before the trauma was recognized was 2.8 (range, 2-9 visits).
For children in whom the diagnosis of AHT was missed, the mean length
of time to diagnosis of head trauma from the day of the first visit was
7 days (range, 0-189 days). In 5 cases, the children were seen twice in
the same day and the diagnosis was made on the second visit; hence, the
designation of 0 days until diagnosis in some cases of missed AHT.
When missed cases were compared with recognized cases, several factors
were found to be significantly different.
Children with missed AHT were much younger than those in whom the
diagnosis was recognized on the first physician visit. The mean age of
missed AHT cases at the time of their first medical visit for head
injury symptoms was 180 days (95% confidence interval [CI],
125-236). The mean age of the recognized cases was 278 days (95% CI,
228-328). The mean ages of children with missed and recognized AHT were
significantly different (independent samples t test,
P = .02).
Abusive head trauma was missed significantly more often in white
children than children of minority races. In white children, 43
(37.4%) of 115 cases of AHT were missed and in minority children, 11
(19%) of 58 were missed (Pearson χ2, P = .01).
Abusive head trauma was more likely to be missed in families in which
both parents lived with the child. Thirty-seven (40.2%) of 92 cases
were missed in intact families. In families in which the mother and
father of the child were not living together, 14 (18.7%) of 75 cases
were missed (Pearson χ2, P = .003).
Not surprisingly, the more severely symptomatic children were more
likely to be recognized as having head trauma at first visit to the
physician. Table 2 summarizes the
number and percentage of children who were missed and recognized as
having AHT compared with their symptoms and signs. At the first visit,
children who were comatose, whose breathing was compromised, who were
having seizures, or who had facial bruising were more likely to be
accurately diagnosed. Conversely, children who presented with
irritability or vomiting at the first visit were less likely to be
identified as having AHT.
Several factors were found not to differ between children with
missed vs recognized AHT. These included whether the parents were
employed, whether the parents had private insurance coverage, the sex
of the child, the birth weight of the child, and whether the child had
been born prematurely (<37 weeks' gestation).
Nine variables were found to be significantly associated with
missing the diagnosis of AHT by univariate analysis. These were
transformed to dichotomous variables and entered into a logistic
regression model. They included age younger than 6 months, minority
race, parents not living together, and 6 signs and symptoms noted at
the first visit, including facial injury, seizures, decreased mental
status, abnormal respiratory status, vomiting, and irritability. Of
these 9 variables, 4 were retained in the multivariate logistic model.
These 4 independent variables predicting the correct diagnosis of AHT
at the first visit included (1) abnormal respiratory status (odds ratio
[OR], 7.23; 95% CI, 2.4-21.3; P<.001); (2) seizures
present (OR, 6.67; 95% CI, 2.5-17.3; P<.001); (3) facial
and/or scalp injury present (OR, 4.81; 95% CI, 2.1-11.0;
P<.001); and (4) parents not living together (OR, 2.49; 95%
CI, 1.1-5.7; P = .03).
Applying the logistic regression model constructed from the data, we
found that if none of these 4 factors were present, the probability
that a physician would make the correct diagnosis of AHT was P
= .20. That is, if a child had normal respirations, had no seizures,
had no facial or scalp injury, and came from an intact family, the
probability that AHT would be recognized was less than 1 in 5.
The 54 children with missed AHT received 98 diagnoses other than AHT
during their 98 patient visits. Table
3 lists the diagnoses applied to the
children with missed AHT. The most common diagnoses made were for viral
gastroenteritis and accidental head injury. In some cases, the
diagnoses were correct, even though coexistent head trauma was not
recognized. For example, in 1 case an infant was accurately assessed to
have a retropharyngeal abscess, but the accompanying subdural hematoma,
retinal hemorrhages, and skull fracture were not recognized. In other
cases, the symptoms of head trauma were attributed to conditions other
than AHT. In 10 cases, the wrong diagnosis was applied more than once
to the same child. We did not count these repeated diagnoses on our
Twenty-five (14.5%) of the 173 children died as a result of
their head injuries. Of the recognized AHT cases, 20 (16.8%) of 119
children died. In the missed AHT cases, 5 (9.3%) of 54 children died.
The percentage of children in the missed AHT group who died was not
statistically different than in the recognized AHT group
(χ2 = 1.712; P = .19). In our estimation, 4 of
the 5 deaths in the missed AHT group might have been prevented by
earlier recognition of abuse (Table
Of the missed AHT cases, 15 (27.8%) of the 54 children
were known to have been reinjured because of the delay in diagnosis.
Twenty-two children (40.7%) had medical complications related to the
delay in diagnosis. These conditions included seizure disorders,
chronic vomiting, and increasing head size because of increasing
untreated subdural hematomas.
In 7 of the children whose diagnosis of AHT was missed, radiological
errors contributed to the delay. These 7 children had 8 studies in
which trauma was missed, including 6 computed tomography scans of the
head, 1 skeletal survey, and 1 long-bone radiograph of the arm. The 2
longest delays in diagnosis (141 days and 174 days) and 6 of 25 cases
in which the diagnosis of AHT was missed for longer than 7 days
involved radiological misreadings. Table
5 summarizes the nature of the errors made
and the time in delay of diagnosis attributed to the
It is difficult to study the cases of child abuse that clinicians do
not recognize. In 1972, Jackson9 reviewed traumatic
injuries in children at King's College Hospital in London, England,
and found 18 of 100 cases to have been missed cases of child abuse.
O'Neill et al10 reported a series of
110 battered children
in 1973. Eighty percent of those children had signs of prior injury.
Alexander et al11 found physical
evidence of previous head
trauma in 8 of 24 children evaluated for head injury due to shaking.
Ewing-Cobbs et al12 discovered
signs of preexisting brain
injury in 45% of children with inflicted traumatic brain injury
compared with none in children with accidental traumatic brain injury.
Incidental cases of missed child abuse have been
published.13 In their study of abusive head injuries,
Benzel and Hadden mention that 9 of 23 abused children with head
injuries ". . . were known to have been seen by other physicians
because of similar problems or other injuries consistent with child
abuse."14 Since then, an increased
awareness of child
abuse has occurred, but similar studies have not been reported.
We do not know how many cases of AHT are never detected. Surely,
the injuries occurring from impact or shaking represent a range of
severity, from no injuries to mild concussion or small subdural
hemorrhage, severe brain damage, extensive intracranial bleeding, and
cerebral edema. Caffey15 speculated
in 1972 that many
children who are found to have mild neurologic abnormalities and
learning disabilities may have been victims of AHT.
Parents who confess to shaking or hitting the heads of their
children frequently report doing the same thing previously. In 1 study
case, an infant was hospitalized 3 times before someone witnessed the
child being shaken violently. On 1 occasion, he was evaluated and
treated for possible sepsis. The other 2 hospitalizations were for
apnea and reflux, respectively. The child's father admitted to
multiple episodes of shaking that led to the infant's various
In the current study, we found that 31.2% of children who were
clinically symptomatic after AHT were misdiagnosed as having other
conditions. Infants have few ways to demonstrate illness or injury.
Nonspecific signs, such as vomiting, fever, and irritability, are seen
in a myriad of conditions, including many benign, self-limited
illnesses. The difficulty, then, is to be able to discern when these
signs and symptoms indicate potentially serious or fatal pathology.
The possibility exists that in some of the visits we classified
as missed, the child had not yet been injured. However, in another
study by our group, we found that patients became symptomatic
immediately after their injuries in 37 cases in which perpetrators
admitted to causing head injuries in infants.16 To guard against misclassification, we examined the medical records extremely
carefully to correlate clinical and radiological findings.
Not surprisingly, the infants and toddlers in our study whose head
injuries were misdiagnosed were overall less ill than those whose head
injuries were recognized. The fact that they were not as ill made the
diagnosis of AHT difficult. Also, the children whose AHT was missed
were, as a group, younger. The difficulty of diagnosing serious illness
or injury in young infants is complicated by the limited range of their
normal behavior. With less-sophisticated behavioral and neurologic
signs to assess, the changes in young infants with head injuries are
more difficult to detect.
Striking differences were seen in the race and family composition
of infants with missed and recognized injuries. Infants with recognized
AHT were more likely to be minority children or children whose mothers
and fathers were not living together. We speculate that this may
represent a subtle bias in decision making based on the physician's
assessment of risk. A physician examining a white child from an intact
family may be less likely to think about the possibility of child
abuse. Another hypothesis is that perhaps minority and single-parent
families were more likely to obtain care from public clinics or
hospital emergency departments, where physicians may be more attuned to
abuse issues. In the current study, the children of intact, 2-parent
households were much
more likely to have private insurance (Pearson
χ2, 23.953; P<.001). In addition, white
families were much more likely to have private insurance than minority
families (Pearson χ2, 5.148; P = .02). However,
we did not collect data on the practice setting in which missed and
recognized diagnoses were made.
Are missed cases of AHT inevitable? If a child's caretakers cannot or
will not give an accurate history, making the correct diagnosis is
extremely difficult. Physicians cannot obtain cranial computed
tomographic scans for every infant and toddler who presents with
vomiting, irritability, and fever. Based on this study and on our
experience with these cases, we make the following suggestions to
facilitate the diagnosis of AHT.
Be alert for bruises or abrasions on the faces or heads of
children presenting with nonspecific symptoms. In 20 of 54 missed AHT
cases in this study, facial or head bruising was attributed to
accidental injury unrelated to the presenting illness symptoms. One
study of bruising in healthy, nonabused children found no bruises on
children who were not yet strong enough to pull to
standing.17 The presence of bruises
in infants raises the
possibility of inflicted injury.
When evaluating infants and toddlers with nonspecific symptoms,
such as vomiting, fever, or irritability, consider head trauma in the
differential diagnosis. Perform a head-to-toe physical examination,
palpate the fontanelles, measure the head circumference, and be alert
for signs of trauma.
When collecting spinal fluid in cases of suspected infantile
sepsis, examine any bloody cerebrospinal fluid for xanthochromia. A
supernatant of a spinal fluid contaminated by blood secondary to a
traumatic procedure should be clear in color if the specimen is
examined shortly after it is collected. Xanthochromic spinal fluid can
represent old blood in the cerebrospinal fluid from previous trauma,
although it is not specific for an intracranial bleed.18- 20
Pediatrically trained radiologists should be consulted to interpret
x-ray film and computed tomographic images in cases of suspected child
In addition to these suggestions, other as yet unvalidated
strategies to detect occult abuse could be considered. Dilated retinal
examinations in infants and children with nonspecific symptoms of
illness could increase the recognition of retinal hemorrhages. Retinal
hemorrhages have been reported in the majority of children who are
victims of AHT.21 Other possibilities need further
research. Some markers of brain trauma are known to cross the
blood-brain barrier, such as the BB fraction of creatine kinase. If
rapid tests were available for such markers, a simple blood test
possibly could be done to detect occult trauma. In a recent study by
Hymel and colleagues,22 children with traumatic parenchymal
brain injury were frequently noted to have prolonged prothrombin and
partial thromboplastin times. These tests are generally available and
inexpensive to run. Their sensitivity and specificity as screening
tests for head trauma in infants are not known.
There are other ways for AHT to present clinically that we did not see
in this group of patients. The list of signs and symptoms we examined
is not universally inclusive. Another limitation of our method is that
the study was done retrospectively through record review. However, this
seems to be the only option we currently have for examining diagnostic
errors. Finally, information concerning the training, experience, or
practice setting of the physicians evaluating these patients was not
Although it is difficult to detect all serious AHT in the clinical
setting, an awareness of the nonspecific nature of the signs and
symptoms of AHT, particularly in less-serious cases, could increase the
likelihood that more cases will be detected.