ORIGINAL ARTICLE
Characteristics of Rib Fractures in Child Abuse—The Role of
Low-Dose Chest Computed Tomography
Thomas R. Sanchez, MD,* Angelo D. Grasparil, MD,† Ruchir Chaudhari, MD,‡
Kevin P. Coulter, MD,§ and Sandra L. Wootton-Gorges, MD*
Objectives: Our aim is to describe the radiologic characteristics of rib
fractures in clinically diagnosed cases of child abuse and suggest a comple-
mentary imaging for radiographically occult injuries in highly suspicious
cases of child abuse.
Methods: Retrospective analysis of initial and follow-up skeletal surveys
and computed tomography (CT) scans of 16 patients younger than
12 months were reviewed after obtaining approval from our institutional re-
view board. The number, location, displacement, and age of the rib frac-
tures were recorded.
Results: Out of a total 105 rib fractures, 84% (87/105) were detected on
the initial skeletal survey. Seventeen percent (18/105) were seen only after
follow-up imaging, more than half of which (11/18) were detected on a
subsequent CT. Majority of the fractures were posterior (43%) and anterior
(30%) in location. An overwhelming majority (96%) of the fractures
are nondisplaced.
Conclusions: Seventeen percent of rib fractures analyzed in the study
were not documented on the initial skeletal survey. Majority of fractures
are nondisplaced and located posteriorly or anteriorly, areas that are often
difficult to assess especially in the acute stage. The CT scan is more sensi-
tive in evaluating these types of fractures. Low-dose chest CT can be an im-
portant imaging modality for suspicious cases of child abuse when initial
radiographic findings are inconclusive.
Key Words: rib fractures, nonaccidental trauma, child abuse,
low-dose chest CT
(Pediatr Emer Care 2016;00: 00–00)
R ib fractures are injuries strongly associated with child abuse.
In the absence of a history of massive accidental trauma, un-
derlying bone fragility or cardiopulmonary resuscitation, rib frac-
tures are likely be due to nonaccidental injury.1
A properly performed skeletal survey in accordance with the
American College of Radiology and Society for Pediatric Radiol-
ogy guidelines remains the primary imaging modality for the eval-
uation of suspected child abuse. However, even with the addition
of oblique views of the chest, plain radiographs can miss acute
and nondisplaced fractures of the ribs and computed tomography
(CT) has been shown to be more sensitive in evaluating these
types of fractures.2
From the *Division of Pediatric Radiology, University of California, Davis
Medical Center Children's Hospital, Sacramento CA; †Cardinal Santos Med-
ical Center, Philippines; ‡Department of Radiology, and §Department of Pedi-
atrics, University of California, Davis Medical Center and Children's Hospital,
Sacramento, CA.
Disclosure: The authors declare no conflict of interest.
Reprints: Thomas R. Sanchez, MD, Division of Pediatric Radiology, University
of California, Davis Medical Center Children's Hospital. 4860 Y St. Suite
3100 ACC, Sacramento CA 95817 (e‐mail: trsanchez@ucdavis.edu).
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
ISSN: 0749-5161
Pediatric Emergency Care • Volume 00, Number 00, Month 2016
Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
METHODS
A retrospective analysis of 16 clinically diagnosed cases of
child abuse between January 2008 and January 2012 was per-
formed after obtaining approval from the institutional review
board. Initial skeletal surveys and follow-up skeletal surveys were
analyzed for the number, presence, and location of acute and
healing rib fractures. All 16 patients had a follow-up skeletal sur-
veys between 1 and 2 weeks after the initial imaging.
The number, location, displacement, and age of the rib
fractures were recorded as well as the clinical data, including
age and sex of the infants. Acute fractures were diagnosed
when lucency was noted with or without a cortical step-off on
initial radiographs without periosteal reaction. Subacute frac-
tures were identified when periosteal reaction or callous forma-
tion are already present. We also recorded the number of acute
fractures that were not visible on the initial radiographs and
were only seen on follow-up imaging after the formation of cal-
lus. In identifying the location of the fractures the ossified ribs
were equally divided into five segments: anterior, anterolateral,
lateral, posterolateral, and posterior. Displacement was judged
based on the severity of cortical step-off: 0, no cortical step-off;
1, less than 50% cortical width displacement; 2, more than 50%
cortical width displacement; and 3, completely displaced or
overriding fracture.
In 5 patients where the initial skeletal survey was negative
and the emergency physician or pediatrician maintained a high
suspicion for child abuse, a chest CT scan was performed within
24 hours of the initial skeletal survey to evaluate the presence of
rib fractures. This was then compared with the original and
follow-up skeletal surveys.
RESULTS
A total of 16 patients (7 men and 9 women) with ages rang-
ing from 1 to 11 months were included in the study. Of 105 frac-
tures evaluated, there was almost equal distribution of fractures
involving the right (46%) and left (54%) ribs.
A total of 105 acute and healing rib fractures were noted
on both the initial and follow-up skeletal surveys. On the initial
survey, only 87 of 105 fractures (83%) were seen. The signifi-
cant majority of these fractures, 71 of 87 (82%) fractures was
already in the healing or subacute stage, whereas the remaining
16 of 87 (18%) fractures was in the acute stage (Table 1). On
follow-up imaging (either the follow-up skeletal survey after
1 to 2 weeks or the chest CT performed within 24 hours of
the first skeletal survey), an additional 18 (17%) of a total of
105 rib fractures were detected. Eleven (11/18) of these frac-
tures were identified on the chest CT scans of 5 patients,
whereas 7 (7/18) fractures were identified on the follow-up
skeletal surveys in patients who did not have chest CT. The
CT correctly identified all the fractures not visible on the initial
skeletal surveys and the follow-up skeletal survey confirmed
that the chest CT missed no fracture.
The majority of fractures were posterior (43%) and anterior
(29.5%) in location (Table 2). The overwhelming majority of the
www.pec-online.com 1
Sanchez et al Pediatric Emergency Care • Volume 00, Number 00, Month 2016

fractures (96.2%) were nondisplaced. Only 3.8% of the fractures

TABLE 1. Age of Rib Fractures Seen on Initial Skeletal Survey demonstrated grade 1 displacement. No fractures demonstrated
grade 2 or grade 3 displacement (Table 3).
No. Fractures Age of the Fracture Percentage, %
16/87 Acute 18 DISCUSSION
71/87 Subacute 82 Rib fractures, particularly if multiple and posterior or
posteromedial in location, have been strongly associated with
child abuse. In a child 3 years or younger, rib fractures have a
TABLE 2. Location of Rib Fractures 95% positive predictive value for child abuse.3 Multiple rib frac-
tures in the absence of overt trauma are also strongly associated
No. Fractures Distribution/Location Percentage, %
with child abuse.4
57/105 Left 54 In reviewing skeletal surveys for nonaccidental trauma,
48/105 Right 46 high suspicion is necessary especially when evaluating rib
45/105 Posterior 43 fractures. Fractures that are already in the subacute or healing
3/105 Posterolateral 3 stages can be easily detected on initial radiographs of the chest.
However, even with additional oblique views of the chest,
31/105 Anterior 29.5
acute and nondisplaced rib fractures may not be detected on
10/105 Anterolateral 9.5 initial imaging.
16/105 Lateral 15 Our retrospective study showed that 17% of rib fractures
were missed on initial radiographic imaging. Factors that may
contribute to nondetection include incomplete or nondisplaced
TABLE 3. Displacement of Rib Fractures fractures and superimposition with other bony structures5 (Fig. 1).
The formation of callous, which makes the fractures more con-
No. Fractures Displacement Percentage, % spicuous, will only be seen on follow-up imaging several weeks
101/105 Grade 0 (nondisplaced) 96.2 later (Fig. 2).
The CT has been shown to be more sensitive in demonstrat-
4/105 Grade 1 (less than 50% 3.8
displacement) ing even acute rib fractures that are occult on plain radiographs.2
In our retrospective study, 11 rib fractures missed by the initial
0/105 Grade 2 (more than 50% 0
displacement) skeletal survey were all successfully demonstrated on the chest
CT. However, the greater radiation dose of a routine chest CT
0/105 Grade 3 (completely displaced 0
or overriding) scan has made it a less attractive alternative. Recent changes in
CT scan protocol by adjusting the kVp and mA settings, and

FIGURE 1. A, Left oblique chest radiograph appears normal. B, Corresponding CT showing a subtle fracture of the anterior costochondral
junction of the right sixth rib. C, Normal contralateral costochondral junction for comparison.

FIGURE 2. A, Patient suspected of non-accidental trauma, the initial chest radiograph was normal. B, Two weeks later, the follow-up
radiograph now shows healing seventh and eighth posterior rib fractures.

2 www.pec-online.com © 2016 Wolters Kluwer Health, Inc. All rights reserved.

Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Pediatric Emergency Care • Volume 00, Number 00, Month 2016 Rib Fractures in Child Abuse

with the additional use of iterative reconstruction to reduce im-

age noise, has now made it possible to perform the study with
a significantly lower radiation dose.6 This makes CT scan of
the chest a viable option in highly suspicious cases of non-
accidental trauma when initial skeletal surveys fail to show high
specificity rib fractures. Not only can it show subtle acute frac-
tures (especially anterior costochondral fractures which are com-
mon but difficult to see on plain radiographs), it can also identify
fractures hidden by other bony structures (Fig. 3) and can also
evaluate parenchymal and pleural injuries. In our institution,
we already have a low-dose chest CT (LDCT) protocol that
can be used for patients with high suspicion for abuse if the skel-
etal survey and chest radiographs are negative or inconclusive.
We use the following parameters: 100 Kvp, 15 mA, rotation time
of 0.5 seconds, pitch of 1.0, and adaptive iterative reconstruction
FIGURE 4. Low-dose CT of the chest. The subtle non-displaced
blend of 20%. The effective radiation dose comparisons are acute fracture of the right seventh posterior rib was not evident on
measured using the latest software programs, PCXMC (STUK, the 4-view chest radiograph. Estimated dose of the CT study is only
Helsinki, Finland) for radiography and Radimetrics eXposure 0.48 mSv compared to the 4-view chest radiograph which
(Bayer, Inc, Toronto, Canada) for the CT scan with final dose es- is 0.35 mSv.
timates based on the ICRP 103 weighting factors.7 Our latest re-
sults showed that the effective radiation dose of an LDCT using of the spine and scapula,6 and in the detection of lung parenchy-
our protocol is nearly the same as a 4-view chest radiograph mal injuries, pleural effusion, or small pneumothoraces.
(Fig. 4). The estimated dose of our LDCT study for a particular
3-month-old patient is only 0.48 mSv and approaches the dose CONCLUSIONS
of the 4-view chest radiographs (anteroposterior, lateral and bi-
Although our study sample is small, we found that 17% of
lateral oblique as recommended by the joint American College
rib fractures are missed on initial skeletal survey. The LDCT is a
of Radiology and Society of Pediatric Radiology committees),
more sensitive examination that can be an important imaging mo-
which has an estimated dose of 0.35 mSv. This adheres to the as
dality for suspicious cases of child abuse when initial radiographic
low as reasonably achievable principle of minimizing radiation dose
findings are negative or inconclusive.
while at the same time using a more sensitive imaging modality.
Another advantage of performing CT is that it can also eval-
uate the rest of the bony thorax including high specificity fractures REFERENCES
1. Bulloch B, Schubert CJ, Brophy PD, et al. Cause and clinical
characteristics of rib fractures in infants. Pediatrics. 2000;105:E48.
2. Wootton-Gorges SL, Stein-Wexler R, Walton JW, et al. Comparison of
computed tomography and chest radiography in the detection of rib fractures
in abused infants. Child Abuse Negl. 2008;32:659–663.
3. Dwek JR. The radiographic approach to child abuse. Clin Orthop Relat Res.
2011;469:776–789.
4. Kemp AM, Dunstan F, Harrison S, et al. Patterns of skeletal fractures in
child abuse: systematic review. BMJ. 2008;337:1518.
5. Kleinman PK, Marks SC, Adams VI, et al. Factors affecting visualization
of posterior rib fractures in abused infants. AJR Am J Roentgenol.
1988;150:635–638.
6. Sanchez TR, Lee JS, Coulter KP, et al. CT of the chest in suspected child
abuse using submillisievert radiation dose. Pediatr Radiol. 2015;45:1072–1076.
FIGURE 3. Subacute left first rib fracture (arrow) not seen on the 7. Valentin J. (2007) Recommendations of the International Commission
chest radiograph due to superimposition of the clavicle. on Radiation Protection, ICRP Publication, 103.

© 2016 Wolters Kluwer Health, Inc. All rights reserved. www.pec-online.com 3

Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.