J Forensic Sci, 2018

doi: 10.1111/1556-4029.13919
PAPER Available online at: onlinelibrary.wiley.com

PATHOLOGY/BIOLOGY

Kimberly Wells ,1 B.S.; Rudy J. Castellani,1 M.D.; and Joyce L. deJong,1 D.O.

Can Neck Contusions in Unexplained Pediatric

Deaths be Explained by Cardiopulmonary
Resuscitation?*

ABSTRACT: The distribution of cutaneous contusions in infants may raise the possibility of maltreatment. Neck contusions are particularly
problematic since they seldom occur outside the setting of abuse, while cardiopulmonary resuscitation (CPR)-related maneuvers may involve
the neck. To address the role of CPR in neck findings, we examined 260 consecutive pediatric autopsies in which CPR was attempted. No neck
contusions were identified in manners of death classified as natural, undetermined, or suicide. Contusions were present in two of 80 deaths clas-
sified as accident and had obvious accidental causes. About 26% of deaths classified as homicide had neck contusions with no explanation pro-
vided by the caregiver (p < 0.0001). In conclusion, neck contusions in deceased children with no apparent explanation should be regarded as
suspicious for abuse and investigated accordingly. CPR is not a plausible explanation for neck contusions in children.

KEYWORDS: forensic science, forensic pathology, cardiopulmonary resuscitation, child abuse, injury, contusion

Location and pattern of contusions involving skin and soft tis-
sues may provide insight into potential maltreatment of children.
In a case–control study, Pierce et al. noted that bruising of the
torso, ear, or neck in a child under the age of 4 years and any
bruising of an infant <4 months of age were characteristics
strongly predictive of abuse. The number of bruises was also
significant, in that no children in the accident group had more
than four bruises (n = 53), compared to 43% in the abuse group
(18/42). Regarding the neck specifically, bruising of the neck
was present in 18 out of 103 abused children, and in none of the
53 control children who sustained accidental trauma. Pascoe
et al. (1) noted neck bruises in 12% of children with suspected
child abuse and neglect, and no neck bruises in children with
accidental injuries (p < 0.001). Dunstan et al. (2) quantitated
skin contusions according to maximum diameter and used logis-
tic regression to discriminate accidental bruises from maltreat-
ment. Bruises to the head and neck were common in abused
children, rare in nonabused children, and were a better discrimi-
nator of maltreatment than bruising of the limbs. Only six out of
189 control children had bruises to the face and “other head and
neck.” Neck contusions per se were not described in accidentally
injured children.
Given the suspicious nature of neck contusions in young chil-
dren, it is not surprising that the issue of cardiopulmonary resus-
citation (CPR) and its role in producing cutaneous injuries
1
Department of Pathology, Western Michigan University Homer Stryker
MD School of Medicine, Kalamazoo 49007, MI.
Corresponding author: Joyce L. deJong, D.O.
E-mail: joyce.dejong@med.wmich.edu
*Presented at the International Association of Forensic Sciences (IAFS)
Meeting, August 21–25, 2017, in Toronto, Ontario, Canada.
Received 24 July 2018; and in revised form 4 Sept. 2018; accepted 12
Sept. 2018.
sometimes arise in cases of homicidal physical abuse. Since it is
customary to attempt CPR even in obviously deceased infants
and young children, CPR-related contusions to the head and
neck, if such a phenomenon exists with any frequency, should
be well characterized in the literature and well recognized by
pediatricians and forensic pathologists.
An unbiased, evidence-based analysis of the neck region is
therefore needed to better understand the spectrum of changes
that may be ascribed purely to CPR. We sought to test the the-
ory of CPR as a plausible alternative explanation to maltreatment
in young children presenting at autopsy with neck contusions by
performing a retrospective review of autopsy reports and pho-
tographs of the neck in 260 consecutive pediatric autopsies in
which CPR had been performed.
Materials and Methods
A computer search of all deaths reported to the Office of the
Medical Examiner was performed in 2016 of deaths from 2008
to 2016 with the following criteria: (i) age 0–10 years; (ii) CPR
was attempted; (iii) a complete autopsy was performed. Two
hundred and sixty unselected deaths with these criteria were
identified. Each case included photographs taken during autopsy
using a single lens reflex camera. The autopsies were complete
and included external photographs of all surfaces of the body
after the body had been undressed and cleaned, additional pho-
tographs of injuries, and an internal examination of the head,
neck, thoracic, and abdominal cavities. Photographs were avail-
able to review in all cases. In all autopsies, the external descrip-
tion of the neck, the description of the evidence of injury section
of the autopsy, and photographs obtained as a matter of routine
of the neck were evaluated, specifically for the presence or
absence of contusions in the vicinity of the neck (see diagram).
The pericraniocervical line, which demarcates the head from the

© 2018 American Academy of Forensic Sciences 1

2 JOURNAL OF FORENSIC SCIENCES
FIG. 1––The neck is defined as extending from the base of the cranium
and the inferior border of the mandible to the thoracic inlet. The pericranio-
cervical line, which demarcates the head from the neck, runs from the mid-
point of the chin anteriorly to the external occipital protuberance
posteriorly.
FIG. 2––Autopsy photographs demonstrate neck contusions in four deceased children. The decedents were between 10 and 20 months of age. All decedents
had acute subdural hemorrhages and extensive retinal hemorrhages. Trauma history reported by caregivers varied from no trauma to minor trauma, which
changed with repeated interviews. Manner of death was classified as homicide in each case. Note the rounded contour of the contusions, particularly in A and
C, suggesting finger marks or grip marks.
neck, runs from the midpoint of the chin anteriorly to the exter-
nal occipital protuberance posteriorly (Fig. 1). The neck is
defined as extending from the base of the cranium and the infe-
rior border of the mandible to the thoracic inlet. All available
photographs were reviewed and at a minimum, and photos of
the anterior and posterior surfaces of the neck were available.
Some cases also included photographs of the side of the neck.
Two reviewers were assigned to each case. The reviewer was
provided the list of deaths to review with the photographs and
the autopsy report to allow them to read the description of the
neck examinations and evidence of injury section. Evaluating the
written description was essential in the event a contusion was
described in the report, but not captured by a photograph. The
reviewers found complete correlation between the photographs
and the written descriptions of the neck in the photographs; there
were no photographs that displayed what may have been a con-
tusion that lacked a description in the written report. Similarly,
all written documentation of contusions was documented with
photography. The cause and manner of death were also recorded,
and the major anatomical findings assessed (Fig. 2).
Results
The ages of the 260 decedents were: 5 (<1 h); 1 (<24 h); 21
(1–30 days); 28 (31–60 days); 30 (61–90 days); 52 (3–
5 months); 34 (6–11 months); 27 (12–23 months); 12 (24–

35 months); 11 (36–47 months); and 39 (4–10 years). Neck con-
tusions were absent in 252 of the 260 decedents (97%) who had
CPR attempts. No neck contusions (0%) were identified in man-
ners of death classified as natural (37; 14%), undetermined (119;
46%), or suicide (1; <1%). Contusions were present in two of
80 (2.5%) deaths classified as accident (motor vehicle accident
with obvious major trauma related to the contusion and dog
mauling with patterned contusions consistent with the attack). In
both accident cases, the cause of the contusion was unambigu-
ously related to the injuries and not caused by CPR. Six of 23
(26%) deaths classified as homicide had neck contusions with
no explanation provided by the caregiver (p < 0.0001, chi
square). Neck contusions were significantly more common in
cases classified as homicide compared to cases classified as
other than homicide (p < 0.00001, chi square). In the homicide
cases, contusions tend to be small and rounded, suggesting fin-
ger marks or grip marks. The contusions in the six homicides
were consistent with physical abuse. In none of the cases with
neck, contusions were there specific evidence to support CPR as
the cause. In three of the eight deaths (the dog attack and two
homicides), the neck contusions were noted by first responders
before CPR was attempted.
Discussion
In this study, the association between neck contusions at
autopsy and homicidal abuse was highly statistically significant.
The precise origin of the neck contusions was not elucidated.
Importantly, the robust association between neck contusions and
homicide is noncircular, since neck contusions were not relied
upon or specifically considered for manner of death certification
in any of the homicides. The determination of the manner of
death relied on other findings in the investigation and autopsy,
not the presence or absence of neck contusions. Also important
is that neck contusions were not linked to CPR in any of the
260 unselected autopsy cases in this series.
The above findings are compatible with the literature. Kaplan
and Fossum evaluated 25 consecutive unexpected deaths of
infants (from 1–1/2 to 6 months) who had received CPR and
found fresh abrasions over the nasal bridge, undersurface of
the nose, lip, cheeks, and anterior chin surface (3). There were
no neck contusions. Price et al. (4) studied 324 children with
fatal inflicted abdominal injuries and cases of death due to nat-
ural causes in children up to 10 years old with no neck contu-
sions noted. Ryan et al. (5) evaluated 204 children between 0
WELLS ET AL. . NECK CONTUSIONS IN PEDIATRIC DEATHS 3
and 14 years of age for injuries related to CPR and noted inju-
ries in 65 out of 153 children, 63 of which consistent of super-
ficial bruises and abrasions. The authors noted increased
incidence of injuries with more prolonged resuscitative efforts,
but no neck contusions. Matshes and Lew studied a fixed
cohort of 546 children who underwent forensic autopsies over
a 10-year period, for evidence of CPR-related injuries (6).
Twenty-two lesions potentially attributable to CPR were identi-
fied in 19 cases: 15 cases of orofacial injuries compatible with
attempted endotracheal intubation; four cases with focal pul-
monary parenchymal hemorrhage; one case with prominent
anterior mediastinal emphysema; and two cases with anterior
chest abrasions. No neck injuries were reported. Bush et al. (7)
found iatrogenic injuries related to CPR in 3% of cases, but no
neck injuries. As noted above, bruising of the torso, ear, or
neck of a child <4 years of age was characteristic of abuse
according to Pierce et al. (8), and head/neck bruises were
highly statistically significantly associated with abuse in the
study by Dunstan et al. (2).
In conclusion, the association between neck contusions and
homicidal abuse was highly statistically significant in our study
of 260 consecutive pediatric autopsies with CPR attempts. The
identification of such lesions in the course of pediatric autopsy
and in the appropriate clinical context warrants investigation for
homicidal abuse. In contrast, CPR is not a plausible explanation
for neck contusions in young children.
References
1. Pascoe JM, Hildebrandt HM, Tarrier A, Murphy M. Patterns of skin
injury in nonaccidental and accidental injury. Pediatrics 1979;64(2):245–7.
2. Dunstan FD, Guildea ZE, Kontos K, Kemp AM, Sibert JR. A scoring sys-
tem for bruise patterns: a tool for identifying abuse. Arch Dis Child
2002;86(5):330–3.
3. Kaplan JA, Fossum RM. Patterns of facial resuscitation injury in infancy.
Am J Forensic Med Pathol 1994;15(3):187–91.
4. Price EA, Rush LR, Perper JA, Bell MD. Cardiopulmonary resuscitation-
related injuries and homicidal blunt abdominal trauma in children. Am J
Forensic Med Pathol 2000;21(4):307–10.
5. Ryan MP, Young SJ, Wells DL. Do resuscitation attempts in children
who die, cause injury? Emerg Med J 2003;20(1):10–2.
6. Matshes EW, Lew EO. Do resuscitation-related injuries kill infants and
children? Am J Forensic Med Pathol 2010;31(2):178–85.
7. Bush CM, Jones JS, Cohle SD, Johnson H. Pediatric injuries from car-
diopulmonary resuscitation. Ann Emerg Med 1996;28(1):40–4.
8. Pierce MC, Kaczor K, Aldridge S, O’Flynn J, Lorenz DJ. Bruising char-
acteristics discriminating physical child abuse from accidental trauma.
Pediatrics 2010;125(1):67–74.