| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, University of Michigan, 1500 E Medical Center Dr.,
TC-2910L, Ann Arbor, MI 48109-0326.
2 Department of Orthopaedic Surgery, University of Michigan, Ann Arbor,
MI.
Received June 7, 2007;
accepted after revision July 11, 2007.
Address correspondence to J. A. Jacobson
(jjacobsn{at}umich.edu).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Through a search of radiology records from January 1 to December 31, 2005, 103 consecutive patients were identified as having radiographic and CT examinations of the wrist. After excluding incomplete or nondiagnostic examinations and those with a greater than 6-week interval between imaging studies, the final study group consisted of 61 wrist examinations in 60 patients. Two musculoskeletal radiologists and one emergency radiologist blindly reviewed CT examinations, and each bone (scaphoid, lunate, triquetrum, pisiform, trapezium, trapezoid, capitate, hamate, metacarpals, distal radius, distal ulna) was categorized as normal or fractured, with agreement reached by consensus. Each prospective radiographic report was categorized as either normal or fracture/equivocal for each osseous structure. Results were compared using the chi-square and Fisher's exact tests.
RESULTS. In the proximal carpal row, lunate and triquetrum fractures were often radiographically occult (0% and 20%, respectively, detected at radiography); whereas in the distal carpal row, trapezoid, capitate, and hamate fractures were often occult (0%, 0%, and 40% detected at radiography, respectively). Hamate fractures were significantly associated with metacarpal fractures, and distal radius fractures were associated with scaphoid and ulna fractures.
CONCLUSION. Thirty percent of wrist fractures were not prospectively diagnosed on radiography, suggesting that CT should be considered after a negative radiographic finding if clinically warranted. The location of a dorsal scaphoid avulsion fracture emphasizes the need for specific radiographic views or cross-sectional imaging for diagnosis.
Keywords: CT • fracture • radiography • trauma • wrist
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Radiographic examination is routinely used in the initial evaluation of a suspected acute wrist fracture. However, because of overlapping structures, possible suboptimal positioning and technique, lack of dedicated special radiographic views, and other problems inherent to radiographic analysis, missed wrist fractures do occur [4]. Because of this, other imaging methods such as CT are also used both in the primary evaluation of suspected wrist fractures [5-8] and in the assessment of bone healing after fracture [9]. Although radiography has been shown to be limited compared with CT in the evaluation of acute wrist trauma, further studies with larger populations and MDCT with multiplanar reformatting are needed.
In our clinical experience, we have noted a spectrum of wrist fractures that were not originally detected on radiographs but were later confirmed on CT. In addition, we hypothesized that recognizing a pattern of fractures may improve radiographic detection. Therefore, the purpose of our study was to determine which wrist fractures are not prospectively diagnosed at radiography using CT as the gold standard and to identify fracture patterns of the wrist that may improve radiographic detection.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Patient records were reviewed in the electronic patient database by one of the authors, including age, mechanism of injury, and time between imaging studies. All patients were included who had a documented wrist injury that was evaluated initially on radiography and, at some point, on CT as well. Radiographic images obtained closest to the time of trauma were identified. Radiographs for individual patients were excluded if they were obtained after the CT examination. Patients were excluded when corresponding CT examinations of the wrist injury in question were not available. No exclusions were made on the basis of sex, age, or mechanism of wrist injury.
Once patient information had been acquired, a data sheet was prepared for each individual patient to record prospective radiographic findings as either normal or fracture/equivocal for each osseous structure as dictated by the staff radiologist at the time of patient presentation. Prospective radiographic interpretations were completed as part of patient care by one of 16 radiologists in the general call pool, of whom eight were fellowship-trained musculoskeletal radiologists, four were musculoskeletal fellows, and four had subspecialty training but not in musculoskeletal radiology.
Two musculoskeletal radiologists and one emergency radiologist, each fellowship trained, then blindly reviewed CT examinations by consensus to classify each bone (scaphoid, lunate, triquetrum, pisiform, trapezium, trapezoid, capitate, hamate, proximal metacarpals, distal radius, distal ulna) as either positive or negative for fracture. Results from prospective radiographic and consensus CT findings were compared using the chi-square and Fisher's exact tests.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CT showed a total of 69 fractures in 45 of 61 (74%) wrists; 27 wrists had one fracture, 13 had two fractures, four had three fractures, and one had four fractures (Table 1). Of these cases with fracture at CT, radiography reports indicated a total of 48 fractures in 39 of 61 (64%) examined wrists; 32 patients had one fracture, and eight patients had two fractures. There was a fracture of the scaphoid in 16 wrists at CT (with 13/16 or 81% diagnosed at prospective radiography; nine waist, three dorsal avulsion, one distal pole, and three proximal pole) (Fig. 1A, 1B, 1C, 1D), lunate in three (0/3 or 0% seen at radiography) (Fig. 1A, 1B, 1C, 1D), triquetrum in five (1/5 or 20% seen at radiography; four dorsal avulsion and one body) (Fig. 2A, 2B), pisiform in one (1/1 or 100% seen at radiography) (Fig. 3A, 3B), trapezium in three (2/3 or 67% at radiography; one body and two volar trapezial ridge) (Fig. 4A, 4B, 4C, 4D), trapezoid in three (0/3 or 0% at radiography) (Fig. 5A, 5B), capitate in two (0/2 or 0% at radiography) (Fig. 6A, 6B), hamate in five (2/5 or 40% at radiography; body in three and hook in two) (Fig. 7A, 7B), distal radius in 17 (17/17 or 100% at radiography), distal ulna in six (6/6 or 100% at radiography), and proximal metacarpal in eight (6/8 or 75% at radiography). Overall, 30% of the fractures seen on CT were not prospectively diagnosed at radiography.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The following are additional details with regard to the prospective radiography reports. The scaphoid was equivocal for fracture in 15 cases (13 negative and two positive for fracture at CT) and positive for fracture in 14 (three negative and 11 positive for fracture at CT). In cases in which a positive or equivocal radiograph for scaphoid fracture was followed by CT, the indication for CT was confirmation of fracture in six, evaluation for fracture displacement in five, and evaluation for nonunion in two. The lunate was equivocal for fracture in one (negative for fracture at CT). The triquetrum was equivocal for fracture in one (negative for fracture at CT) and positive for fracture in two (one negative and one positive for fracture at CT). The pisiform was positive for fracture in one (also positive for fracture at CT). The trapezium was equivocal for fracture in one (negative at CT) and positive for fracture in two (both positive for fracture at CT). The trapezoid was not equivocal or positive for fracture in any case. The capitate was equivocal for fracture in three (all negative for fracture at CT). The hamate was equivocal for fracture in five cases (four negative and one positive for fracture at CT) and positive for fracture in one (also positive for fracture at CT). The radius was equivocal for fracture in one (negative at CT) and positive for fracture in 17 (all positive for fracture at CT). The ulna was equivocal for fracture in two (one negative and one positive for fracture at CT) and positive for fracture in six (one negative and five positive fractures at CT). The proximal metacarpals were equivocal for fracture in one (positive for fracture at CT) and positive for fracture in five (all positive for fracture at CT).
A significant relationship was seen between radiography reports and CT findings for scaphoid (p = 0.003), pisiform (p = 0.02), trapezium (p = 0.005), radius (p < 0.001), ulna (p < 0.05), and metacarpal (p < 0.05) fractures. A significant relationship was also seen between scaphoid and radius fractures (p = 0.003), hamate and metacarpal fractures (p < 0.05), and radius and ulna fractures (p = 0.005).
In the proximal carpal row, lunate and triquetrum fractures were often not diagnosed on radiography (0% and 20% correctly diagnosed, respectively), whereas in the distal carpal row, trapezoid, capitate, and hamate fractures were not diagnosed on radiography (0%, 0%, and 40%, respectively).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Carpal fractures account for 18% of hand and wrist fractures and 6% of all fractures overall [11, 12]. The scaphoid is the most common carpal fracture, accounting for 10% of all hand fractures and 60-70% of all carpal fractures [13]. Although our study confirmed that the scaphoid is the most commonly fractured carpal bone, the incidence was not as high (16/38 = 42%) as previously reported. The incidence of fractures of the capitate, lunate, and hamate bones in the literature accounts for only 1-3% of all carpal fractures [14], but our study showed an incidence of 15% (10/69). Fractures of the pisiform, triquetral, trapezium, and trapezoid bones are reportedly much less common [15]; however, our study showed an incidence of 17% (12/69) collectively for these four bones. The differences between the incidences of various fractures in this study and those reported in the literature may be explained by selection bias; more symptomatic or significant injuries are more likely to have further diagnostic evaluation with CT.
When evaluating for carpal fractures, several techniques are commonly used, including radiography, CT, and MRI. A three-view radiographic examination (lateral, posteroanterior, and semipronated oblique) is often initially performed [16]. However, other specialized projections are sometimes used, including semisupinated oblique, dedicated scaphoid, and carpal tunnel projections, depending on the clinical scenario [10]. When using CT in the evaluation of wrist trauma, it is important to optimize technique, such as obtaining thin slices and performing multiplanar reformation imaging, to obtain a high diagnostic yield. Current MDCT scanners can obtain slice thicknesses of less than 1 mm, which allows isotropic imaging and high-resolution reformation in any plane [17]. The use of CT in the setting of an acute wrist injury allows visualization of anatomic structures without the overlap of other structures that confounds radiographic interpretation. Indeed, our results show that fractures were identified on CT that were not prospectively seen at radiography. Radiation exposure must always be a consideration when imaging with CT, especially with children, as well as expense.
In addition to CT, some authors have promoted MRI as the technique of choice for occult wrist fracture examination [18-20]. However, CT has been shown to overlook trabecular fractures of the scaphoid, and MRI may not identify cortical fractures [21]. Evidence has suggested that the early use of MRI in diagnosing scaphoid fractures is considered a cost-effective strategy [22, 23]. Regardless, many major university hospitals opt for CT or scintigraphy in the evaluation of scaphoid trauma, even when there is access to MRI [24]. At this time, no diagnostic imaging protocol has been established and universally accepted for the evaluation of suspected scaphoid fractures, much less for the evaluation of abnormalities involving other carpal bones [24]. This is true at our institution as well. In the emergency department setting, an equivocal radiograph is typically followed by CT, in part because of the accessibility of CT at our institution. A patient with a negative radiograph and a high clinical suspicion for fracture is typically referred to an orthopedic surgeon, who will then determine the need for advanced imaging studies such as CT, MRI, or bone scan.
Although our study has shown that some carpal fractures (such as the lunate, triquetrum, capitate, and hamate) are often occult on radiographic examination, we also have found that associations between specific fractures are common. For example, although fractures of the hamate were seen only 40% (2/5) of the time radiographically, hamate fractures had a statistically significant (p < 0.05) association with proximal metacarpal fractures, which were prospectively detected by radiography at a rate of 75% (6/8). This would suggest that although radiography may not be sufficiently sensitive for the diagnosis of hamate fractures, the occurrence of metacarpal fractures should make one suspicious for coexisting hamate fracture. Another fracture association that our study found was between scaphoid and radius fractures (p = 0.003). Although scaphoid fractures were generally detected on radiography (sensitivity of 81%), the association with distal radial fractures (sensitivity of 100%) may allow an even higher rate of scaphoid fracture identification. Given the high morbidity associated with missed scaphoid fractures, this is an important association. Finally, our study shows a statistically significant association between distal ulna and distal radius fractures (p = 0.005), although both were uniformly detected at radiography (sensitivities of 100%).
Another finding of interest in our study involves a particular fracture of the scaphoid bone. Of the 16 scaphoid fractures detected on CT by retrospective consensus, 25% (4/16) were dorsal avulsion fractures of the scaphoid (Figs. 2A, 2B and 4A, 4B, 4C, 4D). A search of the literature revealed only two prior studies that have described such a fracture. Compson et al. [25] described three cases of acute dorsal avulsion fracture of the scaphoid wherein the fractures arose from a ridge on the dorsum of the scaphoid. Mayfield et al. [26] referred to a "scaphoid waist avulsion fracture" that probably results from tension of the radial collateral ligament, caused by forced ulnar deviation. As indicated by Compson et al. and verified by our results, the scaphoid dorsal avulsion fracture is visible only on the 45° anteroposterior semisupinated oblique radiographic view of the scaphoid. Mayfield et al. similarly note that this unusual fracture pattern was seen only on the internal oblique projection and was not visualized on a straight posteroanterior projection. Compson [27], in a follow-up article, noted that only the 45° oblique anteroposterior (semisupinated) view adequately identified this dorsal avulsion fracture. Our review of the most commonly used classification scheme for scaphoid fractures reveals that this unusual fracture pattern is not included [28]. An isolated dorsal avulsion fracture of the scaphoid appears to have a good prognosis, but such fractures may mimic a fracture through the body of the scaphoid, particularly those of the proximal pole that have a poor prognosis [25]. Thus, the question arises as to whether the semisupinated oblique view should be routinely used in the evaluation of the patient with snuffbox tenderness or a suspected scaphoid fracture, particularly because a dorsal avulsion fracture may be confused with another scaphoid fracture pattern. Indeed, the semisupinated oblique view has been shown to be the most sensitive radiographic view for the detection of distal radius fractures [17].
We acknowledge that our study has several limitations. First, a broad range of radiologists (some of whom did not have subspecialty training in bone radiology) performed the prospective radiographic interpretations. However, this is probably similar to most radiology practices. The study results were also limited by using CT as the gold standard. However, two musculoskeletal radiologists and one emergency radiologist, each fellowship-trained and board-certified, performed a consensus reading, thus strengthening the gold standard. In addition, CT is not a perfect gold standard for the diagnosis of carpal bone fractures because false-negatives have been shown [29]. However, by using thin slices and multiplanar reformation, the incidence of such false-negatives is likely minimized. Finally, because this study was retrospective, the results may be limited by selection bias because those patients who underwent CT in addition to radiography either had a high clinical suspicion of wrist fracture or underwent CT as the result of the prospective radiographic interpretation. In addition, the clinical significance of the fractures was not assessed.
In conclusion, radiography failed to detect 30% of wrist fractures identified at CT, which suggests the importance of further advanced or cross-sectional imaging for fracture detection in the correct clinical scenario. The location of a dorsal scaphoid avulsion fracture emphasizes the need for specific radiographic views or cross-sectional imaging for fracture detection.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  Y.-Y. Liao and Y.-M. Lin McNemar Test Is Preferred for Comparison of Diagnostic Techniques Am. J. Roentgenol., October 1, 2008; 191(4): W190 - W190. [Full Text] [PDF]
|
|
|
|
|
|
E. M. Caoili and J. A. Jacobson Reply Am. J. Roentgenol., October 1, 2008; 191(4): W191 - W191. [Full Text] [PDF]
|
|
|
|
 |
|
 R. Kaewlai, L. L. Avery, A. V. Asrani, H. H. Abujudeh, R. Sacknoff, and R. A. Novelline Multidetector CT of Carpal Injuries: Anatomy, Fractures, and Fracture-Dislocations1 RadioGraphics, October 1, 2008; 28(6): 1771 - 1784. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
