DOI:10.2214/AJR.05.0606
AJR 2006; 187:812-817
© American Roentgen Ray Society
MDCT of the Elbow in Pediatric Patients with Posttraumatic Elbow Effusions
Vernon Chapman1,2,
Brian Grottkau3,
Maurice Albright3,
Ahmed Elaini3,
Elkan Halpern3 and
Diego Jaramillo4
1 The Children's Hospital of Denver, 7136 S Hudson Ct., Centennial, CO
80122.
2 Massachusetts General Hospital for Children, Boston, MA.
3 Radiology Imaging Associates, Englewood, CO.
4 Children's Hospital of Philadelphia, Philadelphia, PA.
Received April 8, 2005;
accepted after revision July 10, 2005.
Address correspondence to V. Chapman
(vernon.chapman{at}riaco.com).
Abstract
OBJECTIVE. The purpose of this study was to determine the
performance characteristics of MDCT in the detection of fractures in children
with posttraumatic elbow effusions and to assess the effect of MDCT findings
on clinical management.
SUBJECTS AND METHODS. Unenhanced MDCT of the elbow was prospectively
performed without sedation on 31 children 20 months to 16 years old who had
posttraumatic elbow effusions. Two blinded reviewers independently and in
consensus characterized all MDCT scans as positive or negative for the
presence of fracture. Level of interobserver agreement was determined with the
kappa statistic. Sensitivity, specificity, positive predictive value, and
negative predictive value of MDCT for fracture detection were determined for
the consensus MDCT interpretations with follow-up radiographs as the reference
standard. Patients were treated with casts and instructed to return in 2-3
weeks for clinical and radiographic follow-up unless a change in management
was indicated on the basis of MDCT findings. The frequency of alteration of
management was determined.
RESULTS. Both reviewers detected fractures in 15 (48%) of the
patients individually and in 16 (52%) of the patients by consensus.
Interobserver agreement for fracture detection with MDCT was excellent
(
= 0.85). The sensitivity, specificity, positive predictive value, and
negative predictive value of MDCT in the detection of fractures were 92%, 79%,
79%, and 92%, respectively. Four (13%) of the children had changes in
management based on the MDCT findings.
CONCLUSION. MDCT is a sensitive means of evaluating for
radiographically occult fractures in children with posttraumatic elbow
effusions. It has a high negative predictive value and a high level of
interobserver agreement. MDCT findings may lead to alteration of treatment of
children with nondisplaced lateral condylar and radial head fractures.
Keywords: elbow • fracture • MDCT • pediatric • trauma
Introduction
Accurate diagnosis of injuries to the elbow of a child can be difficult for
orthopedists and radiologists, because radiographs can be normal or show only
secondary findings in children with fractures. A common secondary finding
after elbow trauma is a joint effusion with a fat pad sign, which is elevation
of the normal anterior fat pad or visualization of the posterior fat pad on a
lateral radiograph of the elbow
[1]. In the setting of acute
trauma, studies of case series of the use of MRI and follow-up radiographs
have shown fractures in 17-76% of children with only the presence of a fat pad
sign on initial elbow radiographs
[2-5].
Sonography and arthrography are useful in examining children with
posttraumatic elbow effusions but can be painful and invasive
[6,
7]. Although it has proven
effective in examinations of these patients
[3,
8], MRI is time-consuming and
requires sedation of younger patients. In the past, CT also required sedation,
limiting its use in pediatric imaging. However, with the development of
multidetector technology, CT examinations take seconds to perform, eliminating
the need for sedation in most cases. In addition, MDCT studies can be
reformatted and evaluated in multiple planes, reducing the manipulation
necessary for optimal imaging of the area of interest. The sensitivity of MDCT
surpasses that of plain radiography in the evaluation of trauma to the face,
spine, and bony pelvis
[9-12].
Its use in the evaluation of acute extremity trauma is less well studied
[13]. To our knowledge, the
role of MDCT in evaluation of the elbows of children with posttraumatic
effusions has not been studied. The purpose of this study was to prospectively
determine the performance characteristics of MDCT in the detection of
radiographically occult fractures in pediatric patients with posttraumatic
elbow effusions and to assess the effect of MDCT findings on clinical
management.
View larger version (123K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A —10-year-old boy with left elbow pain after falling down
stairs. Lateral radiograph of elbow shows fat pad sign with elevation of
anterior and posterior fat pads (arrows). No fracture was identified
on frontal or lateral radiographs of elbow.
|
|
Subjects and Methods
Patients
We prospectively performed unenhanced MDCT of the elbow on 31 children, 12
girls and 19 boys. The patients' ages ranged from 20 months to 16 years, with
an average age of 7.2 years. Patients were included in the study if there was
evidence of posttraumatic elbow effusion as determined by the presence of the
fat pad sign on a lateral radiograph of the elbow and if frontal and lateral
radiographs obtained immediately after the trauma did not show a fracture. All
patients were examined and underwent scanning within 48 hours of injury. None
of the patients had previous significant musculoskeletal abnormalities. The
institutional review board at our hospital approved this study, and informed
consent was obtained from the parents of all children participating in the
study.
MDCT Technique
Scans were obtained with a 16-MDCT scanner (LightSpeed 16; GE Healthcare)
without sedation or administration of IV contrast material. Scanning was
performed with the patient in the prone position with the affected arm held
above the head in approximately 90 degrees of flexion. Frontal and lateral
localizer radiographs were obtained with the minimum technique allowed by the
scanner (80 kVp, 10 mA). Coverage included the distal part of the humerus and
the proximal portions of the radius and ulna for a total of approximately 80
mm. MDCT scan parameters were as follows: 100 kVp, z-axis automatic
tube current modulation (noise index, 20; minimum, 25 mA; maximum, 200 mA),
rotation speed, 0.5 s/rotation; table speed, 13.75 mm/rotation; beam pitch,
1.375:1; detector configuration, 16 x 1.25 mm. Reconstructions included
slice thickness of 2.5 mm, image spacing of 2.5 mm, full reconstruction mode,
and bone reconstruction algorithm as well as slice thickness of 1.25 mm, image
spacing of 1.0 mm, full reconstruction mode, and standard reconstruction
algorithm. Sagittal and coronal reformations were obtained from the latter
reconstructions.
Follow-up Radiographs
Frontal and lateral radiographs of the elbow were obtained 2-3 weeks after
the initial evaluation.
View larger version (148K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A —3-year-old girl with left elbow pain after wrestling. Lateral
radiograph of elbow shows fat pad sign with elevation of anterior and
posterior fat pads (arrows). No fracture was identified on frontal or
lateral radiographs of elbow.
|
|
View larger version (100K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —3-year-old girl with left elbow pain after wrestling. Coronal
reformatted MDCT image shows fracture (arrow) of lateral condylar of
distal humerus with less than 2-mm displacement of fracture fragments. Patient
was treated nonoperatively, and fracture healed without complication.
|
|
Image Analysis
Two reviewers interpreted all of the images in this study. One reviewer was
a fourth-year radiology resident, and the other was a senior pediatric
musculoskeletal radiologist. Both reviewers were blinded to clinical and
demographic data. All MDCT studies and follow-up radiographs were reviewed on
a PACS workstation (Impax RS 3000 1K review station, AGFA Technical Imaging
Systems). MDCT studies were reviewed with bone (window level, 500 H; window
width, 3,000 H) and soft-tissue (window level, 40 H; window width, 400 H)
settings.
View larger version (148K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A —15-year-old boy with left elbow pain after falling off
bicycle. Lateral radiograph of elbow shows fat pad sign with elevation of
anterior and posterior fat pads (arrows). No fracture was identified
on frontal or lateral radiographs of elbow.
|
|
The two reviewers first examined MDCT images independently. They
characterized MDCT scans as either positive or negative for the presence of
fracture. An MDCT scan was considered positive only if there was direct
evidence of fracture (cortical discontinuity or bone deformity). Indirect
evidence of fracture (joint effusion or soft-tissue swelling) was not
considered sufficient for characterization of a scan as positive. Reviewers
recorded the site of the fracture as supracondylar, lateral condylar, medial
epicondylar, radial head, radial neck, or proximal ulnar. Both reviewers in
consensus then reviewed MDCT images. Together the reviewers characterized each
MDCT scan as either positive or negative and recorded the site of
fracture.
Follow-up radiographs were interpreted by both reviewers in consensus and
categorized as positive or negative on the basis of presence or absence of
healing (periosteal reaction or sclerosis). Reviewers were blinded to the
findings of the MDCT studies.
Patient Treatment
The consensus interpretation of the MDCT study was communicated to the
referring pediatric orthopedist at the time of initial evaluation. All
patients were treated with an elbow cast in 90 degrees of flexion and
instructed to return in 2-3 weeks for clinical and radiographic follow-up,
unless the orthopedist believed the patient would benefit from a change in
management. Such changes in management were recorded.
View larger version (141K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A —7-year-old girl with left elbow pain after falling out of
chair. Lateral radiograph of elbow shows fat pad sign with elevation of
anterior and posterior fat pads (arrows). No fracture was identified
on frontal or lateral radiographs of elbow.
|
|
View larger version (83K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4B —7-year-old girl with left elbow pain after falling out of
chair. Sagittal reformatted MDCT image shows nondisplaced Salter type 2
fracture (arrow) of radial neck. Follow-up radiographs obtained 2
weeks after injury showed no evidence of healing fracture.
|
|
Statistical Analysis
The sensitivity, specificity, positive predictive value, and negative
predictive value of elbow MDCT in fracture detection were determined for the
individual and consensus interpretations of the follow-up radiographs, which
were used as the reference standard. The 95% CI was calculated for
sensitivity, specificity, positive predictive value, and negative predictive
value of the consensus interpretations. Interobserver variability for the
interpretation of all MDCT studies was determined with the kappa statistic.
The percentage of patients with changes in management as a result of the MDCT
findings was determined.
View larger version (144K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5A —3-year-old boy with right elbow swelling after fall. Lateral
radiograph of elbow shows fat pad sign with elevation of anterior and
posterior fat pads (arrows). No fracture was identified on frontal or
lateral radiographs of elbow.
|
|
View larger version (84K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5B —3-year-old boy with right elbow swelling after fall. Axial
(B) and coronal oblique (C) reformatted MDCT images show
nondisplaced fracture (arrows) of proximal ulna. Follow-up
radiographs obtained 3 weeks after injury showed no evidence of healing
fracture.
|
|
View larger version (90K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5C —3-year-old boy with right elbow swelling after fall. Axial
(B) and coronal oblique (C) reformatted MDCT images show
nondisplaced fracture (arrows) of proximal ulna. Follow-up
radiographs obtained 3 weeks after injury showed no evidence of healing
fracture.
|
|
Results
Both reviewers identified fractures in 15 (48%) of the 31 children and had
identical interpretations in 14 cases. Fractures identifiedby both reviewers
included five supracondylar (Figs.
1A,
1B,
1C, and
1D), three radial neck, two
lateral condylar (Figs. 2A and
2B), two radial head (Figs.
3A,
3B, and
3C), and two proximal ulnar
fractures. Each reviewer identified an additional supracondylar fracture not
identified by the other reviewer. In consensus, both reviewers identified
fractures in 16 (52%) of the children. Interobserver agreement for detection
of fracture with MDCT was excellent with a kappa value of 0.85 (95% CI,
0.65-1.0).
Twenty-six patients underwent follow-up radiography. Of the five patients
with no radiographic follow-up, one patient with MDCT evidence of a proximal
ulnar fracture had sustained multiple trauma, was transferred to another
hospital, and was lost to follow-up. One patient with evidence of a radial
head fracture on MDCT had the cast removed at a 1-week follow-up visit and did
not return for the scheduled clinical and radiographic follow-up 2 weeks
later. Three patients with no evidence of fracture on MDCT had normal findings
at clinical examinations 2-3 weeks after trauma and did not undergo repeated
radiography.
In consensus, both reviewers found evidence of fracture healing on
follow-up radiographs of 12 (46%) of the 26 patients. The reviewers
independently and in consensus identified 10 of these fractures on MDCT,
including five supracondylar, two radial neck, and two lateral condylar
fractures and one radial head fracture. One reviewer identified an additional
supracondylar fracture that exhibited evidence of healing on follow-up
radiographs. Both reviewers identified no fracture on MDCT scans of a patient
found to have evidence of a healing lateral condylar fracture. Among the 14
patients with no evidence of healing on follow-up radiographs, both reviewers
identified an additional two fractures with MDCT, including a radial neck
fracture (Figs. 4A and
4B) and a proximal ulnar
fracture (Figs. 5A,
5B, and
5C), and one reviewer
identified an additional supracondylar fracture.
With follow-up radiographs as the reference standard, the sensitivity of
MDCT for fracture detection was 83% and 92% for the individual reviewers and
92% (95% CI, 62-100%) for both reviewers in consensus. The specificity was 79%
and 86% for the individual reviewers and 79% (95% CI, 49-95%) for both
reviewers in consensus. The positive predictive value was 77% and 85% for the
individual reviewers and 79% (95% CI, 49-95%) for both reviewers in consensus.
The negative predictive value was 85% and 92% for the individual reviewers and
92% (95% CI, 62-100%) for both reviewers in consensus.
Four (13%) of the patients had alterations in management based on the MDCT
findings. Two patients with radial head fractures had the cast removed after 1
week to allow early mobilization and prevent restricted range of motion. Two
children with nondisplaced lateral condylar fractures were treated
nonoperatively with weekly follow-up radiographs for close monitoring for
delayed displacement.
Discussion
The results of our study suggest that MDCT is a sensitive (92%) and
specific (79%) means of evaluating for radiographically occult fractures of
the elbow in children and has a high negative predictive value (92%). MDCT
depicted occult injuries in 52% of children with joint effusion and normal
radiographs. Excellent interobserver agreement ( = 0.85) suggests that
MDCT is reproducible, and the short examination time and lack of need for
sedation suggest that it is practical.
There has been controversy regarding the prevalence of fractures in cases
in which a child sustains acute elbow trauma and radiographs show only a fat
pad sign but no identifiable fracture. The variability appears to stem from
the reference standard used to determine the presence of occult injuries. For
example, using follow-up radiographs as a standard, Donnelly et al.
[2] found that only 17% of
posttraumatic effusions were associated with fractures, whereas studies in
which MRI was used as a standard have shown that more than one half of
patients have bone injury [3,
4]. Consequently, the
specificity (79%) and positive predictive value (79%) observed in our study
likely reflect the use of a reference standard, follow-up radiographs, that is
less accurate than MDCT. Specifically, three fractures—a supracondylar
fracture, a radial neck fracture (Figs.
4A and
4B), and a proximal ulnar
fracture (Figs. 5A,
5B, and
5C)—were identified on
MDCT, and there was no evidence of healing on the follow-up radiographs.
In our experience, MDCT of the elbow in children is a relatively easy and
painless examination. All of the patients in the current study, which included
two 4-year-olds, one 2-year-old, and a 20-month-old, underwent scanning
without incident. This success was likely due to the minimal manipulation
necessary to perform MDCT and the relative speed with which the examination is
performed. In contrast, a series of radiographs necessitates positioning the
injured elbow as many as four times, an often time-consuming process that is
painful for the child. Furthermore, MDCT can be performed with no image
degradation when the elbow is in a cast
[14]. The additional views
that may be needed after a radiographic series are not necessary with MDCT,
because the data can be reformatted in any plane or examined on a 3D
workstation to help solve diagnostic dilemmas. MDCT of the elbow performed
with an automated tube current modulation technique entails a much lower
radiation dose than a fixed tube current technique
[14]. Furthermore, when
properly performed, the radiation dose from MDCT of the elbow is limited to
the elbow and does not include the closest critical organs, the corneas.
It is difficult to assess the extent to which added information about
fractures obtained with MDCT affects management or improves outcome. Examples
of changes in management in the current study include early mobilization of
two radial head fractures to prevent restricted range of motion and more
frequent radiographic follow-up of two nondisplaced lateral condylar fractures
for close monitoring for delayed displacement. The latter of these examples is
particularly noteworthy because some orthopedic surgeons pin all lateral
condylar fractures and would consider prospective knowledge of the presence of
this type of fracture particularly useful.
Limitations of the current study were the relatively small sample size,
which led to the wide confidence intervals observed, and the use of only
frontal and lateral radiographs for the initial evaluation of fractures.
Arguably, inclusion of at least one oblique view may have decreased the
false-negative rate of the initial radiographs and ultimately affected the
sensitivity, specificity, positive predictive value, and negative predictive
values observed for MDCT. An additional potential limitation is the
cost-effectiveness of elbow MDCT in children with posttraumatic effusions,
particularly given the relative costs of MDCT and radiography.
A factor limiting routine use of MDCT in children with posttraumatic elbow
effusions is the false-negative results observed in this study. Although MDCT
was very sensitive in the detection of fractures and has a high negative
predictive value, both reviewers did not identify a lateral condylar fracture,
and one reviewer did not identify a supracondylar fracture with evidence of
healing on followup radiographs. This observation suggests that the
combination of conservative management and radiographic follow-up is
appropriate for most patients and that MDCT may be most useful in cases in
which radiographs are normal but there is clinical suspicion of a lateral
condylar or radial head fracture.
In conclusion, MDCT of the elbow is a sensitive means of evaluating for
radiographically occult fractures in children with only joint effusion on
initial radiographs of the elbow after acute trauma. MDCT has a high negative
predictive value and a high level of interobserver agreement among
radiologists of varied experience. MDCT findings may lead to changes in
management for a subset of these patients, including those with nondisplaced
fractures of the lateral humeral condyle and radial head.
References
- Goswami G. The fat pad sign. Radiology2002; 222:419
-420[Free Full Text]
- Donnelly L, Klostermeier T, Klosterman L. Traumatic elbow effusions
in pediatric patients: are occult fractures the rule?
AJR 1998; 171:243
-245[Abstract/Free Full Text]
- Beltran J, Rosenberg Z, Kawelblum M, et al. Pediatric elbow
fractures: MRI evaluation. Skeletal Radiol1994; 23:277
-281[CrossRef][Medline]
- Griffith J, Roebuck D, Cheng J, et al. Acute elbow trauma in
children: spectrum of injury revealed by MR imaging not apparent on
radiographs. AJR 2001;176
: 53-60[Abstract/Free Full Text]
- Skaggs D, Mirzayan R. The posterior fat pad sign in association
with occult fractures of the elbow in children. J Bone Joint Surg
Am 1999; 81:1429
-1433[Abstract/Free Full Text]
- Blane CE, Kling TF Jr, Andrews JC, DiPietro MA, Hensinger RN.
Arthrography in the posttraumatic elbow in children.
AJR 1984; 143:17
-21[Abstract/Free Full Text]
- Davidson RS, Markowitz RI, Dormans J, Drummond DS. Ultrasonographic
evaluation of the elbow in infants and young children after suspected trauma.
J Bone Joint Surg Am 1994;76
: 1804-1813[Abstract/Free Full Text]
- Badelon O, Bensahel H, Mazda K. Lateral humeral condylar fractures
in children: a report of 47 cases. J Pediatr Orthop1988; 8:31
-34[Medline]
- Rhea J, Rao P, Novelline R. Helical CT and three-dimensional CT of
facial and orbital injury. Radiol Clin North Am1999; 37:489
-513[CrossRef][Medline]
- Sheridan R, Peralta R, Rhea J, Ptak T, Novelline R. Reformatted
visceral protocol helical computed tomographic scanning allows conventional
radiographs of the thoracic and lumbar spine to be eliminated in the
evaluation of blunt trauma patients. J Trauma2003; 55:665
-669[Medline]
- Guillamondegui O, Pryor J, Gracias V, Gupta R, Reilly P, Schwab C.
Pelvic radiography in blunt trauma resuscitation: a diminishing role.
J Trauma 2002; 53:1043
-1047[Medline]
- Haapamaki VV, Kiuru MJ, Koskinen SK. Multidetector computed
tomography diagnosis of adult elbow fractures. Acta
Radiol 2004; 45:65
-70[CrossRef][Medline]
- Pretorius E, Fishman E. Volume-rendered three-dimensional spiral
CT: musculoskeletal applications. RadioGraphics1999; 19:1143
-1160[Abstract/Free Full Text]
- Blickman JG, Dunlop RW, Sanzone CF, et al. Is CT useful in the
traumatized pediatric elbow? Pediatr Radiol1990; 20:184
-185[CrossRef][Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
