To The Editor:
I read with interest, and then with dismay, "The Posterior
Fat Pad Sign in Association with Occult Fracture of the Elbow in
Children" (81-A: 1429-33, Oct. 1999), by Skaggs and Mirzayan, because
a radiograph in the article (Fig. 2) clearly demonstrates a plastic
bowing or buckle fracture of the distal humerus. The authors obviously
did not recognize this fracture, which casts considerable doubt
on the final conclusions of their study. In addition, in their second
case, in which they claim that a radial fracture was not visible
on the initial studies, the initial radiographs presented in their
article (Figs. 3-A through 3-D) do not include the area of the radius
where the fracture was finally demonstrated. The latter could be
an oversight, but the first problem—that is, presenting
a case in which the authors suggest that no fracture is present,
although a fracture is clearly present—substantially detracts
from the article.
Most likely the authors do not use comparison radiographs of
the uninjured limb in the evaluation of skeletal injuries sustained
in children. Comparison radiographs are essential, in my opinion,
for the evaluation of childhood fractures, because without them
physicians will continue to miss subtle fractures such as the one
present, but not detected, in Figure 2. Complicating this problem
is that the article was abstracted for the Yearbook of Diagnostic
Radiology1. Experienced
and talented radiologists, but, nonetheless, those who probably do
not believe in the value of comparison radiographs, also missed
the fracture; they simply went on to support the findings of the
original article. It is a little disappointing that so many well-educated
medical professionals would not be able to realize that the fracture was
present on the initial study.
I am including references for two articles that would have helped
the authors. The first, by Rogers et al., describes the anterior
humeral line, which should be used as a reference when diagnosing subtle,
supracondylar, buckle fractures in children2.
Also, in a review article by John et al.3,
attention was specifically paid to the subtlety of distal humeral
fractures. In this regard, it was stated that buckle fractures often
manifest as an increased angulation of the cortex of the distal humerus
on the ulnar side. This was clearly present in Figure 2-A in the
article by Skaggs and Mirzayan.
I think something should be done about this. In terms of the
outcome alone, clearly these patients would have enough pain that
the extremity would be immobilized. On the other hand, if diagnostic
accuracy is a consideration, then the article, as it now stands,
needs to be addressed. Perhaps the authors should say that without
the use of comparison radiographs, these subtle fractures will probably
be missed, but the presence of the displaced fat pad sign should
suggest that the extremity be immobilized and treatment be prescribed
as if the patient had a fracture. It is only a measure of how defined
you want to be in your initial diagnosis, and, to this end, despite
the 76% incidence of so-called missed fractures demonstrated
by the authors, there are at least two articles in the literature
that suggest that this incidence should be in the neighborhood of
15%4,5 .
D.L. Skaggs and R. Mirzayan reply:
We would like to respond to Dr. Swischuk’s letter point
by point.
Regarding the unappreciated fracture in Figure 2, a pediatric
radiologist blinded to the study, the pediatric emergency-medicine
attending physician, and the treating orthopaedic resident did not appreciate
a fracture on three radiographs of the elbow. While playing Monday-morning
quarterback, both Dr. Swischuk and the authors appreciated a fracture.
This in no way detracts from the message of the paper: if an elevated fat
pad is present but a fracture is not appreciated by the initial
treating physicians, there is a 76% chance that a fracture
is present. If Dr. Swischuk were to expeditiously proceed to the
emergency department every time there was a child with an elevated
fat pad and no obvious fracture, we have no doubt that additional
fractures would be appreciated at presentation. However, we do not believe
that this would change the treatment of children in whom a fracture
is identified, or that of children in whom a fracture is still not
identified.
Figure 3 may not include the area of fracture. Dr. Swischuk is
correct in stating that the radiographs made at the time of injury
may not include the area of the radius that was fractured. In our minds,
this adds further support to the clinical utility of an elevated
fat pad in identifying children who are likely to have a fracture
that is not appreciated, or possibly even present, on the initial radiographs.
In reference to the fact that the article was abstracted for
the Yearbook of Diagnostic Radiology, Dr. Swischuk
views it as a problem that "experienced and talented radiologists, but,
nonetheless, those who probably do not believe in the value of comparison
radiographs, also missed the fracture" and "went
on to support the findings of the original article." We would
encourage Dr. Swischuk to consider another possibility—that
many recognized experts, including the editors and reviewers of The
Journal of Bone and Joint Surgery and of the Yearbook
of Diagnostic Radiology, recognized the fracture, as well
as the inherent value of the study.
In regard to the use of comparison radiographs, Dr. Swischuk
suggests that we should state that "without the use of comparison
radiographs, these subtle fractures will probably be missed." We refer
Dr. Swischuk to two previous reports evaluating the usefulness of comparison
radiographs of the uninjured elbow. One report concludes that use
of comparison radiographs did not improve diagnostic accuracy in
elbow trauma assessed in a pediatric emergency department by residents,
emergency-department attending physicians, or pediatric radiologists6. The second report concludes that
use of comparison radiographs did not improve the diagnostic accuracy
of orthopaedic surgeons or that of orthopaedic residents7. In addition, use of comparison radiographs
increases costs and radiation exposure. To the best of our knowledge, Dr.
Swischuk’s opinion that "comparison radiographs
are essential . . . for the evaluation of childhood fractures" is not
supported by the literature.
In summary, the finding that 76% of children with elevated
fat pads and no fracture appreciated at presentation (by three doctors)
had evidence of fracture-healing at an average of three weeks after
the injury is practical information for clinicians treating children
who have sustained acute elbow trauma. If further retrospective
analysis of radiographs by experts discovers that some fractures
were initially missed, in the context of our findings, this discovery would
have no effect on treatment.