| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Service de Radiologie, Groupe Hospitalier Pellegrin, CHRU Bordeaux, Place
Amélie Raba Léon, 33076 Bordeaux Cedex, France.
2 Service de Radiologie, Centre Hospitalier Général de Pau, 4
Blvd. Hauterive, 64000 Pau, France.
Received September 10, 2001;
accepted after revision November 19, 2001.
Address correspondence to O. Hauger.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Sonography of the scaphoid bone with a 12-MHz transducer was performed in 54 patients with clinicaly suspected scaphoid fracture and normal findings on initial radiographs, including specific scaphoid images. Three levels of clinical suspicion were considered: high (20%), moderate (30%), and low (50%). Attention was paid to the continuity of the scaphoid cortex and to the surrounding soft tissues (i.e., hemarthrosis or hematoma). Data from early sonograms were then compared with the results of radiography repeated 10-14 days after the initial trauma. In cases of persistent suspicion despite normal findings on follow-up radiographs, the presence of fracture was evaluated on CT (four patients), MR imaging (one patient), or bone scanning (one patient).
RESULTS. Follow-up examinations proved fracture of the scaphoid waist in five patients. In all patients, diagnosis of fracture was suspected on initial sonograms showing cortical disruption associated with soft-tissue abnormalities. There was one false-positive finding and no false-negative results. Using cortical disruption as a diagnostic criterion, we found the sensitivity, specificity, and accuracy of high-resolution sonography for the depiction of scaphoid fracture to be 100%, 98%, and 98%, respectively. Using soft-tissue abnormalities alone as a criterion, we found the sensitivity, specificity, and accuracy of high-resolution sonography to be 100%, 65%, and 68%, respectively. The overall prevalence of occult fracture was 9%, ranging from 3.7% for low suspicion to 27% for high suspicion of fracture.
CONCLUSION. High-resolution sonography is a reliable and accurate method of evaluating occult fractures of the scaphoid waist. Cortical disruption is the diagnostic key. Soft-tissue abnormalities alone lack specificity.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The diagnosis of scaphoid fracture is often challenging because of unspecific clinical presentation [4] and equivocal or negative imaging findings. Indeed, the rate of negative findings on initial conventional radiographs is high, ranging from 20% to 25% [5]. Radiography is usually repeated after 10-14 days in case of clinical suspicion of scaphoid fracture without initial abnormality. During this delay, depending on the degree of the clinical suspicion, the wrist can be immobilized [6]. Immobilization is usually prolonged if the clinical suspicion persists, even when findings on follow-up radiographs remain negative. The major limitation of this strategy is that the wrists of some patients without fracture may be immobilized, sometimes for a long period, with important consequences in terms of quality of life and health care costs [7]. On the other hand, the absence of early treatment (i.e., early immobilization) of an occult fracture increases the rate of complications and the risk of secondary displacement.
CT, MR imaging, and, to a lesser extent, radionuclide bone scanning have already been evaluated in the diagnosis of occult scaphoid fractures [8,9,10,11,12,13,14], offering a high sensitivity and specificity. However, these imaging modalities have limitations in terms of availability and cost-effectiveness. To our knowledge, few studies have examined the role of sonography in the analysis of the scaphoid bone in cases of wrist trauma [15,16,17], and none have specifically focused on occult fractures.
The purpose of our study was to evaluate the diagnostic accuracy of sonography in the detection of clinically suspected scaphoid fracture with normal findings on initial radiographs (occult scaphoid fractures).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
If the three criteria were present, the suspicion of scaphoid fracture was considered high (20%, 11/54). If only pain and tenderness were present, suspicion was considered moderate (30%, 16/54). If only pain was present, suspicion was considered low (50%, 27/54). Mechanisms of injury were a fall on an outstretched hand (94%, 51/54) or direct trauma (6%, 3/54).
Initial radiographs included posteroanterior, lateral, semipronated oblique scaphoid, radial oblique scaphoid, and posteroanterior images with the wrist in ulnar deviation. They were obtained within 7 days of the trauma (range, 0-5 days; mean, 1.3 days) and were evaluated for scaphoid fracture signs (i.e., radiolucent line through the bone, cortical disruption, or cortical avulsion) by the consensus of two radiologists experienced in musculoskeletal imaging. Patients with initial radiographic evidence of a scaphoid fracture were excluded from the study.
Normal Regional Anatomy on Radiographs and Sonograms
The scaphoid occupies a vulnerable position, bridging the proximal and
distal rows of the carpus. On both radiographs and sonograms, the morphology
of the scaphoid depends on the position of the wrist. The ulnar deviation
elongates the scaphoid and affords the best view of the waist, the site of 70%
of fractures. The radial deviation foreshortens the scaphoid and tends to
obscure the presence of a fracture but affords visualization of the distal
pole of the scaphoid [8]. The
radial margin of the waist of the scaphoid is the site of a small angular
protuberance, the tubercle, corresponding to a frequent normal variant. An
irregular pattern of this small protuberance can be mistaken for a fracture.
On sonography, the cortex of the scaphoid is seen as a thin continuous and
usually regular echogenic line, except for the tubercle area in which the
cortex may appear irregular. A particularity of the scaphoid bone is the
absence of periosteum [19],
which explains why fracture healing occurs without periosteal callus formation
[18]. Seventy-five percent of
the surface of the scaphoid is covered with cartilage, which facilitates
mobilization with adjacent bones
[20]. The remaining 25% allows
attachment of ligaments and tendons.
Sonography
After informed consent of the patient was obtained, sonography of the
scaphoid bone was performed with a sonographic system (HDI 3000 and 5000;
Advanced Technology Laboratories, Bothell, WA) by using a
high-spatial-resolution 12-MHz transducer. Comparative sonography of both
normal and painful wrists was performed within 7 days of trauma (range, 0-7
days; mean, 1.6 days) by one of two investigators experienced in performing
musculoskeletal sonography. Patients were seated upright in front of the
radiologist. The scaphoid bone was scanned in the longitudinal and transverse
planes, from the dorsal and lateral directions, in both normal and ulnar
deviation positions to elongate the scaphoid. The ulnar deviation in the
lateral direction was obtained by placing a support on the ulnar side of the
wrist (Fig.
1A,1B).
Attention was paid to the continuity of the anterior echogenic margin of the
scaphoid corresponding to the cortex and the surrounding soft tissues.
Scaphoid fracture was considered if the cortex was discontinuous. Two types of
soft-tissue abnormalities were reported: hemarthrosis, defined as a diffuse
and compressible collection, and hematoma, defined as a focal and
incompressible collection located at the level of the scaphoid (Figs.
2 and
3). Sonographic examinations
were interpreted immediately without knowledge of the follow-up radiographic
findings.
|
|
|
|
Clinical and Radiologic Follow-Up
The same protocol of conventional radiography was repeated 10-14 days after
the initial trauma for all patients. Diagnosis of initially occult scaphoid
fracture was based on visualization of follow-up radiographs of a fracture
line through the bone, cortical disruption, or cortical avulsion. Absence of a
fracture was confirmed by the resolution of the clinical signs associated with
the absence of a fracture line on follow-up radiographs. If suspicion of a
fracture still persisted despite normal findings on follow-up radiographs,
other diagnostic techniques were performed to obtain a diagnostic clue. These
examinations consisted of CT (four patients), MR imaging (one patient), or
bone scanning (one patient). Among these six patients, three benefited from
repeated sonography. Results from the follow-up examinations were finally
compared with the initial sonographic findings.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
|
Of the six patients with evidence of cortical disruption, five (83%) had proven scaphoid fractures on follow-up radiographs (Fig. 4A,4B,4C), whereas one (17%) had normal findings on both radiographs and CT scans. In this false-positive case, CT showed a degenerative wrist with cortical irregularity of the tubercle. No false-negative result was observed.
|
|
|
Among the eight patients with evidence of hematoma alone, none had proven scaphoid fracture on follow-up. In these patients, CT (three patients), MR imaging (one patient), and bone scanning (one patient) were performed in addition to radiography because of the persistence of a clinical suspicion of fracture. Repeated sonographic examinations performed in three of these patients showed resolution of the hematoma (Fig. 5A,5B,5C,5D). Of the eight patients with hemarthrosis alone, none had proven scaphoid fracture on follow-up. In total, none of the 16 patients with soft-tissue abnormalities alone (i.e., without cortical disruption) had a confirmed scaphoid fracture on follow-up. Among the 32 patients without cortical disruption and soft-tissue abnormality on initial examination, follow-up radiographic findings were considered normal in all patients.
|
|
|
|
Using cortical disruption as the criterion of fracture, we found the sensitivity, specificity, positive predictive value, and negative predictive value of sonography for early detection of occult scaphoid fractures to be 100%, 98%, 83%, and 100%, respectively. If soft-tissue abnormalities alone (i.e., hematoma or hemarthrosis) were used as the criterion of fracture, the sensitivity, specificity, positive predictive value, and negative predictive value of sonography for early detection of occult fractures were 100%, 65%, 23%, and 100%, respectively. The overall prevalence of occult fracture was 9.3% (5/54), ranging from 3.7% (1/27) for low suspicion to 6.3% (1/16) for moderate suspicion and to 27% (3/11) for high suspicion of fracture.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
To improve the diagnosis of what has been called the "clinical scaphoid" [5, 22], we evaluated the diagnostic accuracy of high-resolution sonography in the depiction of occult scaphoid fractures.
Our results show that sonography can depict occult scaphoid fractures at an early stage. When we considered cortical disruption as the diagnostic key, all five occult fractures in this group were depicted on sonography without false-negative results. The false-positive result observed corresponded to an irregular tubercle in an elderly patient with marked degenerative changes of the thumb and a history of wrist trauma. As on radiography [18], the tubercle area appears to be a critical zone to analyze on sonography. In all our patients, the cortical disruption was associated with either hemarthrosis or hematoma. However, in our experience, soft-tissue abnormalities alone cannot be considered as a strong indicator of scaphoid fractures because they can be related to other internal lesions such as intrinsic ligament tears.
Previous studies focusing on the depiction of bone fracture on sonography, particularly the tibia and the forearm [23, 24], have emphasized the importance of the subperiosteal hematoma as an indirect sign of fracture. This sign, a collection adjacent to the bone limited by a thin hyperechogenic line corresponding to the periosteum, was also suggested as an indirect sign of fracture in a previous study concerning the scaphoid [17]. However, the hypothesis seems unlikely because the scaphoid, which is almost entirely covered by cartilage [20], has the particularity of not being surrounded by periosteum [19]. We observed such a hyperechogenic line parallel to the cortex in two patients, but no fracture was found on follow-up examination (Fig. 6A,6B,6C). We believe that this finding corresponds to a positive echo signal of the anterior interface of the cartilage induced by a high gain level.
|
|
|
Sonography has already been suggested as an alternative diagnostic tool in patients with scaphoid fractures, but none of the previous studies were specifically focused on occult fractures. Using various direct and indirect criteria, the authors found a sensitivity ranging from 50% to 100%; a specificity, from 74% to 100%; and an accuracy, from 84% to 87% for the depiction of all types of fractures (occult or not) [15,16,17]. Our results appear superior with a sensitivity, specificity, and accuracy of 100%, 98% and 98%, respectively, which may be explained by three main reasons: first, technical improvements in sonography leading to higher spatial resolution, especially with the 12-MHz transducer used in this study allowing depiction of subtle posttraumatic changes; second, the use of stringent criteria (i.e., disruption of the cortex) to confirm the diagnosis of fracture. Most previous studies used less stringent criteria such as soft-tissue changes, which may also occur after a wrist sprain. Third, the five occult fractures depicted in this study involved the waist of the scaphoid, which is the most common site of fracture but also the easiest area to analyze on sonography. Fractures of the distal and especially the proximal poles of the scaphoid would likely be diagnostically more challenging.
Other diagnostic tools have been evaluated for their accuracy in the early diagnosis of scaphoid fractures. Bone scintigraphy offers high sensitivity but lacks specificity because of a potential increase in focal uptake in cases of bone contusion or ligament injuries [25, 26]. CT and MR imaging offer sensitivity and specificity greater than 95% and a rate of detection ranging from 18% to 27% [8,9,10,11, 27,28,29,30], figures comparable to the results obtained in our study with high-resolution sonography. In our study, the 9.3% prevalence of occult scaphoid fractures detected on sonography is equivalent to the rate of detection in some bone scan studies [5, 31] but lower than that on MR imaging. This discrepancy may be explained by the high incidence of low clinical suspicion of scaphoid fracture in our study (50%). If only high clinical suspicion is considered (equivalent to levels of suspicion in previous MR studies), the rate of detection rises to 27%, which is better than that of bone scintigraphy and CT and equivalent to that of MR imaging.
Diagnosis of occult scaphoid fracture is of paramount importance because of its therapeutic implications. Because of a high rate of complications (i.e., delayed union, nonunion, avascular necrosis, and secondary osteoarthritis), treatment of patients with displaced fractures, usually with surgery, is well codified, [32]. On the contrary, treatment of patients with an undisplaced fracture, which is by definition an occult scaphoid fracture, may vary. It usually consists of cast immobilization of the wrist for a 4- to 8-week period. However, an alternative surgical treatment by percutaneous fixation can be proposed [33, 34]. The advantage of the latter method is a significant reduction in the immobilization time with a quick return to normal activities (4 days for sedentary workers and 5 weeks for manual workers). A delay in the diagnosis increases the risk of secondary displacement, which jeopardizes such a therapeutic approach. Another benefit of the early diagnosis of an occult fracture is that overtreatment of patients without significant lesions is avoided.
Compared with other diagnostic tools, sonography should become a diagnostic technique of choice in cases of clinical scaphoid. It is a reliable, nonionizing, available, and cost-effective technique. Moreover, it is not time-consuming because the mean time of the procedures ranged from 10 to 15 min in our study.
However, sonography has shortcomings. First, because all fractures in our study occurred at the waist, to our knowledge, it is unknown whether sonography will be efficient for the detection of fractures of the distal and especially proximal poles of the scaphoid. Fractures of the proximal pole, although less frequent (20% of scaphoid fractures), have a high rate of complications and must be treated promptly. Second, ulnar deviation, which is essential for complete visualization of the scaphoid, may be difficult to obtain if there is marked wrist pain. However, even with a limited ulnar deviation, the most common site of fracture (the waist) can be analyzed. Third, the extent and direction of the fracture line cannot be extensively depicted, but this finding has no significant consequence in the treatment of these undisplaced fractures. Lastly, the tubercle area may appear quite irregular, mimicking cortical disruption, as in one patient in our study. However, such irregularities are usually seen in the elderly population, which is less subject to scaphoid fractures. Considering these limitations, CT and MR imaging are still clearly of interest in certain circumstances, and the diagnostic strategy in Figure 7 can be proposed in patients in whom scaphoid fracture in suspected.
|
In conclusion, high-resolution sonography appears to be a reliable, accurate, available, and cost-effective method for evaluating occult fractures of the waist of the scaphoid. Diagnosis is based on cortical disruption associated with soft-tissue abnormalities, which alone are not a specific sign of fracture. Early evaluation of the scaphoid bone after wrist trauma is important to initiate prompt treatment in cases of occult fracture, because early treatment substantially decreases the rate of complication and avoids overtreatment of patients without fracture.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  D Barron Imaging trauma of the appendicular skeleton Imaging, September 1, 2007; 19(3): 323 - 335. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A M Groves, H K Cheow, K K Balan, P W P Bearcroft, and A K Dixon 16 detector multislice CT versus skeletal scintigraphy in the diagnosis of wrist fractures: value of quantification of 99Tcm-MDP uptake Br. J. Radiol., September 1, 2005; 78(933): 791 - 795. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. Berna, G. Chavarria, F. Albaladejo, L. Meseguer, A. Pellicer, M. A. Sanchez-Canizares, and D. Perez-Flores Panoramic Versus Conventional Radiography of Scaphoid Fractures Am. J. Roentgenol., January 1, 2004; 182(1): 155 - 159. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D Barron and T Branfoot Imaging trauma of the appendicular skeleton Imaging, December 1, 2003; 15(4): 324 - 340. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
