AJR 2002; 178:1517-1521
© American Roentgen Ray Society
Postmortem Radiography After Unexpected Death in Neonates, Infants, and Children: Should Imaging Be Routine?
Elizabeth P. McGraw1,
John E. Pless2,
Debra J. Pennington3 and
Susan J. White4
1 Department of Radiology, Duke University Medical Center, Box 3808, Durham, NC
27710.
2 Department of Pathology and Laboratory Medicine, Indiana University Medical
Center, Medical Science 157, Indianapolis, IN 46202.
3 Austin Radiological Association, 6101 W. Courtyard Dr., Bldg. 5, Austin, TX
78730.
4 Clinical Diagnostic Radiology Associates, 1300 N. 12th St., Phoenix, AZ
85006.
Received June 19, 2001;
accepted after revision December 10, 2001.
Presented at the annual meeting of the International Pediatric Radiology
Society, May 2001.
Address correspondence to E.P. McGraw.
Abstract
OBJECTIVE. The purpose of this study was to determine whether
postmortem radiography of neonates, infants, and children provides additional
information that is not detected at autopsy in cases of unexpected death.
MATERIALS AND METHODS. Inclusion criteria for 106 consecutive
postmortem skeletal surveys (1998-2000) were neonates, infants, and children 2
years old or younger with no preexisting medical condition to account for
mortality. Pediatric radiologists interpreted all the radiographic
examinations, which consisted of high-detail, collimated anteroposterior
radiographs of the appendicular and axial skeleton, lateral radiographs of the
axial skeleton, and oblique radiographs of the ribs. Imaging results were
compared with those obtained from standard protocol autopsies on all children.
Four categories of death were designated: homicide (i.e., abuse, n =
14), accidental (e.g., drowning, n = 28), natural (e.g., acute
illness, n = 43), and undetermined (n = 21).
RESULTS. The causes of death in the 14 child abuse victims were
blunt force injuries to the intracranial (n = 11) and chest and
abdominal (n = 1) areas; asphyxia (n = 1); and shaking
injury (n = 1). In six (43%) of these 14 patients, radiography
detected 26 extremity fractures that had not been detected at autopsy; four
(67%) of these six patients had fractures of different ages that involved more
than one extremity. All fractures carried a high index of suspicion of abuse.
No skeletal injuries were found in cases of accidental, undetermined, and
natural deaths.
CONCLUSION. Postmortem radiography provides important additional
information regarding the extent and chronicity of extremity trauma that may
not be documented at autopsy. This finding supports the routine use of
radiography in cases of suspected child abuse. Normal findings on postmortem
skeletal radiography may help to distinguish cases of natural, accidental, and
undetermined causes of death from those of abuse, aiding in the proper
handling of these cases by medical and law enforcement personnel.
Introduction
According to the latest National Child Abuse and Neglect Reporting System
report, an estimated 1100 children in the United States died of abuse and
neglect in 1999 [1]. These data
correspond to a rate of 1.6 deaths per 100,000 children in the general
population. The true incidence of fatal child abuse is likely underestimated
because not all states contributed data to this report and because many cases
of abuse are not detected. Infants accounted for 42.6% of the reported
fatalities, and children younger than 6 years accounted for 86.1%
[1]. One of the difficulties in
accurately reporting child abuse lies in its detection, which requires a high
index of suspicion by both medical and law enforcement personnel. Once
suspicion is established, a thorough investigation by both parties ensues. The
current standard dictates that suspected victims of child abuse undergo an
autopsy, which may or may not include a radiographic evaluation. The objective
of our study was to determine whether important additional information can be
obtained with high-detail postmortem radiography and how this information
influenced the ultimate disposition of cases of unexpected death in 106
neonates, infants, and children 2 years old and younger.
Materials and Methods
This internal review board—approved investigation consisted of the
evaluation of 106 consecutive postmortem skeletal radiographic examinations
performed from 1998-2000. Neonates, infants, and children 2 years old or
younger who had no preexisting medical condition to account for mortality were
included in the study. None of the patients had undergone antemortem
radiography for the evaluation of suspected abuse. The postmortem skeletal
surveys were performed at no cost to the patients' families.
Skeletal surveys were performed on ultradetail, 10-1p/mm, film (Kodak
InSight; Eastman Kodak, Rochester, NY) with collimated anteroposterior
radiographs of the appendicular skeleton, anteroposterior and lateral
radiographs of the axial skeleton, and oblique radiographs of the ribs. Twenty
images were obtained for each study. Radiologists who had obtained
Certificates of Added Qualification in pediatric radiology and were based in a
university children's hospital interpreted all radiographs and documented
abnormalities. In some cases, the circumstances surrounding the death of the
victim were known at the time of interpretation.
Fractures were classified not only by their specific location but also by
their location in either the appendicular or axial skeleton. The age of each
fracture was estimated; on the basis of fracture age, each fracture was
classified as acute (fracture line clearly delineated with no surrounding
subperiosteal new bone formation), subacute (loss of fracture line definition
with subperiosteal new bone formation), or remote (callus formation
surrounding fracture site). These criteria served as guidelines for fracture
dating, although we realize that dating of classic metaphyseal fractures is
more difficult because the signs of healing previously described are sometimes
inconspicuous [2]. Results were
given to the forensic pathologist at the time of the initial
interpretation.
Forensic pathologists performed autopsies on all children using a routine
protocol that included examination of the axial skeleton and visual
examination of the external surfaces of the extremities. The pathologist
designated four categories of manner of death: homicide (i.e., child abuse,
n = 14), accidental (e.g., drowning, n = 28), natural (e.g.,
acute illness, n = 43), and undetermined (n = 21). Cases of
sudden infant death syndrome (n = 20) were classified as natural
deaths. Radiographic and autopsy results were then compared.
Results
Of the 106 postmortem skeletal surveys and autopsies performed, 14 cases
were categorized as homicides; all were determined to have been caused by
child abuse. Of the 14 abused children, six were male and eight were female,
ranging in age from 2 weeks to 17 months. The average age was 3.6 months.
Eleven died from blunt force injury of the head. Five of these 11 children
also had coexisting blunt force injuries of the chest, abdomen, and back.
Additional causes of death included asphyxia (n = 1), blunt force
injury of the chest and abdomen (n = 1), and shaking injury
(n = 1).
The postmortem skeletal surveys detected skeletal abnormalities in seven
(50%) of the 14 abuse victims. Six (43%) of these 14 patients had fractures
involving both the appendicular and axial skeleton, and one victim had
fractures involving only the axial skeleton
(Fig. 1).
View larger version (9K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1. —Flowchart of all child abuse deaths (n = 14) from
106 consecutive cases shows seven patients had fractures present on postmortem
radiography. Six (43%) of 14 patients had fractures of both appendicular and
axial skeleton. One (7%) of 14 patients had fractures involving only axial
skeleton. Seven (50%) of 14 abuse victims had no skeletal survey findings.
|
|
Twenty-six extremity fractures were detected radiographically
(Fig. 2) in the 14 victims of
child abuse. Twenty-four (92%) of these fractures were metaphyseal corner
fractures, the type of fracture most associated with abuse (Fig.
3A,3B,3C,3D).
A total of two diaphyseal fractures were detected in two infants. This type of
injury, although not as specific for abuse as the classic metaphyseal corner
fracture, also carries a high index of suspicion of abuse in the nonambulatory
infant [2]. Interestingly,
postmortem radiography revealed classic metaphyseal fractures in addition to
the diaphyseal fractures in both of these patients. Only subacute metaphyseal
fractures were present in two patients, only acute metaphyseal fractures were
present in one patient, and a combination of both healing and acute fractures
was seen in three patients (Table
1).
View larger version (25K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2. —Diagram of infant skeleton reveals specific sites and total
number of all extremity fractures detected on postmortem radiography in abuse
victims. All fractures were classic metaphyseal lesions, which are associated
with high index of suspicion of abuse, except two fractures that involved
proximal right tibial diaphysis in one infant and left distal radial diaphysis
in another infant. In two infants with these fractures, classic metaphyseal
fractures were also present in other extremities. R = right.
|
|
View larger version (135K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A. —11-week-old male infant with blunt force injuries of chest
and abdomen. Extremity fractures and healed rib fractures were detected only
on postmortem radiography. Acute rib fractures were revealed both at autopsy
and on postmortem radiography. Anteroposterior radiograph of right distal
femur shows corner fracture (classic metaphyseal lesion) of distal right
medial femoral metaphysis (arrow).
|
|
View larger version (157K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B. —11-week-old male infant with blunt force injuries of chest
and abdomen. Extremity fractures and healed rib fractures were detected only
on postmortem radiography. Acute rib fractures were revealed both at autopsy
and on postmortem radiography. Anteroposterior radiograph of distal right
tibia shows classic metaphyseal fracture of distal right tibial metaphysis
(arrow).
|
|
View larger version (127K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3C. —11-week-old male infant with blunt force injuries of chest
and abdomen. Extremity fractures and healed rib fractures were detected only
on postmortem radiography. Acute rib fractures were revealed both at autopsy
and on postmortem radiography. Anteroposterior radiograph of distal left
femoral metaphysis shows classic metaphyseal fracture of distal left medial
femoral metaphysis (arrow).
|
|
View larger version (121K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3D. —11-week-old male infant with blunt force injuries of chest
and abdomen. Extremity fractures and healed rib fractures were detected only
on postmortem radiography. Acute rib fractures were revealed both at autopsy
and on postmortem radiography. Anteroposterior radiograph of chest reveals
fractures with callus formation of right lateral third and fourth ribs and
right posterior ninth ribs (large arrows). Note acute fractures
involving left posterior third, fourth, seventh, eighth, and ninth ribs
(arrowheads) and left lateral sixth and seventh ribs (small
arrows). Endotracheal tube terminates in distal trachea. Gastric
dilatation resulted from resuscitative efforts at scene before intubation.
|
|
Fifty-three fractures of the axial skeleton were identified at autopsy.
Twenty (38%) of these 53 fractures were acute fractures of the ribs
(n = 19) and thoracic vertebral bodies (n = 1) that were not
detected on postmortem radiography. These findings support previously
published data [3,
4]. The rib fractures were
nondisplaced, and 12 (63%) of these 19 fractures involved the anterior ribs.
Acute, nondisplaced rib fractures were identified at autopsy by the presence
of acute hemorrhage and edema overlying the fracture site. These findings then
led the pathologist to digitally examine the rib cortex. Any cortical
irregularity felt by the pathologist at digital examination of the site of
overlying acute edema or hemorrhage was determined to represent an acute rib
fracture. Healed fractures involving the posterior ribs at the costovertebral
junction and lateral ribs were identified on radiographs by callus formation
surrounding the prior fracture site. No fractures of the axial skeleton were
detected solely on postmortem radiography.
In cases of accidental, natural, and undetermined causes of death,
postmortem radiography detected no osseous injuries.
Discussion
More than 50 years ago, the medical community became aware of the
radiologic manifestations of nonaccidental trauma. Since Caffey's landmark
article in 1946 [5], much has
been learned about the mechanisms of injury and skeletal manifestations of
child abuse. The characteristic locations and patterns of osseous injury seen
in victims of child abuse have been well established previously and are beyond
the scope of this article [2,
6,
7]. The main radiologic
contribution to the evaluation of child abuse continues to be the
documentation of radiologic features of osseous injuries that are specific for
abuse. Performing a high-detail skeletal survey is the most common way of
evaluating whether skeletal injuries specific for abuse are present in
infants. Although high-detail skeletal surveys are routinely performed on
living children in cases of suspected abuse, the routine use of radiography in
deceased children in whom abuse is suspected has not yet been established. If
postmortem radiography is performed, then it is often done by those with no
formal training in radiography, and the images are interpreted by
nonradiologists. Prior reports have documented the utility of performing
high-quality postmortem skeletal surveys for both the documentation of child
abuse and the prosecution of offenders
[3,
4,
7,8,9].
This study sought to expand the focus of prior studies to evaluate the utility
of high-detail postmortem radiography in neonates, infants, and children not
only in cases of suspected abuse, but also in cases of unexpected death.
Abuse is one of the four categories of causes of unexpected death in
children. The other three categories are natural, accidental, and undetermined
causes. In our study, the most common cause of death in cases of abuse was
blunt force injury of the head. This finding corresponds to findings reported
in another study [8]. Blunt
force injury of the head was readily detectable at pathologic examination:
findings most often consisted of contusions, skull fractures, cerebral
cortical contusions, and subdural hemorrhages. Although the acute cause of
death in these cases could be determined at routine autopsy examination, this
examination did not address the issue of the presence of acute or remote
coexisting extremity trauma because the routine autopsy protocol includes only
visual examination of the extremities.
Previous reports have confirmed that routine gross inspection of the
extremities at autopsy is insensitive for the documentation of metaphyseal
injury; however, documentation of this type of injury is paramount in cases of
suspected abuse [3,
8]. Information about the full
extent and chronicity of trauma is critical to the investigation of child
abuse and to the prosecution of offenders. The prosecution of abuse cases is
often difficult because injuries sustained from an accidental fall can often
mimic those of abuse. Similarly, the lack of findings in an asphyxiated infant
at autopsy can make a case of suspected child abuse hard to substantiate. As
Kleinman et al. [8] found,
"The process ultimately rests on the perceived likelihood that a jury of
the defendant's peers will view the strength of the evidence as sufficient to
warrant a guilty verdict." Therefore, the medical community should take
measures to identify all injuries that may support the suspicion that death
resulted from inflicted trauma.
Postmortem radiographic finding of fractures that carry a high index of
suspicion for abuse have proved extremely useful in the correct classification
and prosecution of cases of suspected child abuse. At the time of this
writing, five of the seven homicide cases that had positive findings on the
skeletal survey were awaiting trial. The death of one patient, a 5-week-old
female infant, was tried as a homicide and resulted in the conviction of the
victim's father (Fig.
4A,4B,4C).
Subsequent neglect charges against the victim's mother are also being pursued.
The skeletal survey findings in the other patient were helpful in obtaining a
confession from the victim's father, who was convicted of neglect of a
dependent (Class C felony), involuntary manslaughter, and battery resulting in
bodily injury. He received a 20-year prison sentence and 4 years of probation.
Both the law enforcement personnel and the prosecutors involved in these cases
unanimously supported the use of high-detail postmortem skeletal surveys in
the evaluation of unexpected childhood death (Hibbard RA, personal
communication). This finding also supports those of another study
[8]. Of particular importance
to the prosecution of the offenders was the severity of the charges that could
be filed on the basis of the skeletal survey findings. The presence of both
acute and healing injuries on skeletal surveys provided substantial evidence
of the chronicity of physical abuse. According to several prosecutors, this
evidence allowed the offenders to be charged with murder rather than
aggravated battery, thus allowing much longer sentences if the defendants were
convicted.
View larger version (154K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A. —5-week-old female infant with multiple blunt force injuries
of chest and abdomen and shaking injury of head (recent and remote subdural
hemorrhage found at autopsy). These extremity fractures were detected on
postmortem skeletal survey. Victim's father was convicted of murder and her
mother was charged with neglect. Posteroanterior radiograph of left wrist
reveals transverse fracture of distal left radial diaphysis and corner
fracture of distal left ulnar metaphysis (arrows).
|
|
View larger version (164K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4B. —5-week-old female infant with multiple blunt force injuries
of chest and abdomen and shaking injury of head (recent and remote subdural
hemorrhage found at autopsy). These extremity fractures were detected on
postmortem skeletal survey. Victim's father was convicted of murder and her
mother was charged with neglect. Posteroanterior radiograph of right wrist
shows classic metaphyseal fracture of distal right ulna (arrow).
|
|
View larger version (93K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4C. —5-week-old female infant with multiple blunt force injuries
of chest and abdomen and shaking injury of head (recent and remote subdural
hemorrhage found at autopsy). These extremity fractures were detected on
postmortem skeletal survey. Victim's father was convicted of murder and her
mother was charged with neglect. Anteroposterior radiograph of left tibia and
fibula show classic metaphyseal fractures of proximal and distal left tibia
(arrows).
|
|
Although the justification for the use of postmortem skeletal surveys
appears obvious in cases of homicide and suspected child abuse, it is also
useful in cases of accidental, natural, and undetermined causes of death.
Although negative findings on a skeletal survey do not exclude abuse, the
absence of skeletal injury—when combined with the autopsy and death
scene findings—may suggest sudden infant death syndrome as the cause of
death. This information relieves the family and caretakers of unwarranted
suspicion by medical and legal authorities and provides reassurance that the
death of their infant could not have been prevented
[10].
Similarly, postmortem skeletal surveys should be performed in cases in
which the circumstances surrounding the child's death are unknown or not
entirely clear in the perimorbid period. Although our data revealed no
skeletal injuries in those who died of accidental or undetermined causes, it
is important to remember that the manner of death is determined
retrospectively by a pathologist after review of autopsy and death scene
information. Often, the death scene investigation is only in its preliminary
stages at the time of autopsy, and, thus, many facts may remain undiscovered.
The skeletal survey findings may substantiate or contradict the findings at
the death scene investigation by law enforcement personnel and allow the cause
of death to be categorized as accidental or undetermined.
Although our study reveals the utility of high-detail postmortem
radiography, limitations should be addressed. Two involve the materials and
methods portion of the study. First, no fracture specimens were removed at
autopsy to correlate with the radiographic findings. Thus, the findings of
this study lack correlation with a gold standard. Second, not all radiographic
studies were interpreted by the same pediatric radiologist. Although
undesirable, this practice was unavoidable because of changes in personnel
during the 2-year study period.
The third limitation of this study is one of logistics and expense.
Although some centers provide funding for routine postmortem radiography, this
was not the case in our study. Performing a postmortem skeletal survey is
time-consuming, often requiring 30-45 min of a technologist's time and 30-45
min of radiography suite time. It also requires the commitment of a
radiologist to evaluate all radiographs for adequate quality before the
decedent is removed from the department. Not only is the study itself
nonreimbursable from a technical and professional standpoint, additional
revenue is lost because a substantial amount of time in a radiography suite is
used, during which reimbursable studies could otherwise be performed. Further
investigation with a true cost-to-benefit analysis may prove an interesting
topic for future research about this subject.
Despite the aforementioned limitations, we support the routine use of
high-detail postmortem radiography of neonates, infants, and children in the
evaluation of unexpected death. Postmortem skeletal surveys provide additional
information about the extent and chronicity of extremity trauma that is not
routinely documented by autopsy and should be performed in all cases of
suspected child abuse in neonates, infants, and children. Normal findings on
postmortem radiography may also contribute to the accurate classification of
natural, accidental, and undetermined causes of death and ensure the proper
handling of these cases by medical and law enforcement personnel.
References
-
United States Department of Health and Human Services.
Child maltreatment 1999: reports from the States to the National
Child Abuse and Neglect Data System. Washington, DC: United
States Government Printing Office, 2001
-
Kleinman PK. Diagnostic imaging of child
abuse, 2nd ed. St. Louis: Mosby, 1998:8
-167, 242-247
-
Kleinman PK, Marks SC Jr, Richmond JM, Blackbourne BD. Inflicted
skeletal injury: postmortem radiologic—histopathologic study in 31
infants. AJR
1995;165:647
-650[Abstract/Free Full Text]
-
Kleinman PK, Marks SC, Nimkin K, Rayder SM, Kessler SC. Rib
fractures in 31 abused infants: post-mortem radiologic-histopathologic study.
Radiology
1996;200:807
-810[Abstract/Free Full Text]
-
Caffey J. Multiple fractures in the long bones of infants suffering
from chronic subdural hematoma. AJR
1946;56:163
-173
-
Kleinman PK, Marks SC, Blackbourne B. The metaphyseal lesion in
abused infants: a radiologic—histopathologic study.
AJR
1986;146:895
-905[Abstract/Free Full Text]
-
Kleinman PK. Diagnostic imaging in infant abuse.
AJR
1990;155:703
-712[Free Full Text]
-
Kleinman PK, Blackbourne BD, Marks SC, Karellas A, Belanger PL.
Radiologic contributions to the investigation and prosecution of cases of
fatal infant abuse. N Engl J Med
1989;320:507
-511[Abstract]
-
Thomsen TK, Elle B, Thomsen JL. Post-mortem radiological
examination in infants: evidence of child abuse? Forensic Sci
Int 1997;90:223
-230[Medline]
-
American Academy of Pediatrics. Committee on Child Abuse and
Neglect. Distinguishing sudden infant death syndrome from child abuse
fatalities. Pediatrics
1994;94:124
-126[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
