| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Emergency Radiology |
1 From the Program for the Assessment of Radiological Technology (ART Program), Departments of Radiology (J.J.N., E.H.G.O., A.Z.G., G.P.K., M.G.M.H.), Epidemiology and Biostatistics (J.J.N., E.H.G.O., M.G.M.H.), Orthopaedic Surgery (J.A.N.V.), and Traumatologic Surgery (A.B.v.V.), Erasmus MC, University Medical Center Rotterdam, Dr Molewaterplein 40, 3015 GD Rotterdam, the Netherlands. Received July 8, 2003; revision requested September 24; revision received April 5, 2004; accepted May 19. Supported in part by the Revolving Fund from Erasmus University Medical Center Rotterdam and by an unrestricted grant from Esaote, Genoa, Italy. Address correspondence to M.G.M.H. (e-mail: m.hunink@erasmusmc.nl).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Informed consent was obtained from all participating patients; the institutional review board approved the randomized controlled trial and use of data to create prediction models. Of 90 patients (37 female, 53 male; mean age, 40.0 years), 87 with acute wrist trauma were randomized to undergo radiography (n = 43) or radiography and a short MR imaging examination with low-field-strength dedicated extremity MR system (n = 44). Age, sex, trauma mechanism, presence of tenderness of the anatomic snuffbox, radiographic results, MR imaging results, and treatment data were collected. Univariable and multivariable logistic regression analysis was used to create four models for prediction of treatment need.
RESULTS: Thirty-six patients had one or more fractures; one patient had a marked soft-tissue lesion. In univariable analysis, age (odds ratio, 1.02; 95% confidence interval: 1.00, 1.05), anatomic snuffbox tenderness (odds ratio, 2.31; 95% confidence interval: 0.90, 5.96), radiographic results (odds ratio, 31.2; 95% confidence interval: 8.90, 109), and positive MR imaging results versus MR imaging not performed (odds ratio, 1.86; 95% confidence interval: 0.57, 6.06) were significantly predictive of treatment need. In multivariable analysis, radiographic results (odds ratio, 24.7; 95% confidence interval: 6.59, 93.1) and positive MR imaging results (odds ratio, 6.28; 95% confidence interval: 1.27, 31.0) were significantly predictive of treatment need. Negative MR imaging results were not significantly predictive (odds ratio, 0.87; 95% confidence interval: 0.20, 3.82).
CONCLUSION: A short MR imaging examination with a low-field-strength MR imaging system following radiography in initial evaluation of patients with acute wrist trauma has additional value in prediction of treatment need; it does not have value in identification of patients who can be discharged without further follow-up.
© RSNA, 2005
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
When tenderness of the anatomic snuffbox is present, a scaphoid bone fracture may be clinically suspected; however, a tear of the triangular fibrocartilage complex or a rupture of interosseous carpal ligaments cannot be detected at the initial physical examination. An initial MR imaging examination may play a role in the detection of these lesions, as well. In a meta-analysis, Hobby et al (9) found MR imaging to be an accurate means of diagnosing tears of the triangular fibrocartilage complex; however, its role will likely be more limited in the detection of rupture of wrist ligaments. Although the specificity of MR imaging for the detection of rupture of the scapholunate ligament may be high, its sensitivity is low (9–12).
The high costs of high-field-strength MR imaging systems have prevented a widespread use of MR imaging in acute wrist injury. However, the gradual ongoing development of and increase in use of low-field-strength dedicated extremity MR imaging systems may have changed this situation, since these systems are considerably less expensive than high-field-strength MR imaging systems, both in acquisition and in utilization, and their development is progressing steadily.
We were interested in determining whether a short MR imaging examination in the acute setting could possibly guide treatment more adequately and lead to better patient outcomes or earlier recovery. Although the costs of an additional examination with MR imaging would increase the costs of the diagnostic work-up, costs could potentially be reduced by a decrease in use of medical resources and a decrease in productivity losses, provided that the earlier diagnosis would lead to earlier treatment and recovery.
The purpose of this study was to assess the predictive value of a short MR imaging examination, in addition to or instead of radiography, in patients who present with acute wrist trauma to identify patients who require additional treatment versus those who do not and can be discharged without further follow-up.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
The patients were randomized into groups who underwent radiography alone and radiography combined with an MR imaging examination. Randomization was conducted with the drawing of computer-generated random assignments from consecutively numbered sealed envelopes by research staff and radiology technologists on service. Exclusion criteria were as follows: additional injury of the head, back, thorax, or abdomen; compound fracture of the wrist; preexisting symptoms in the same wrist; and intoxication with alcohol or ingestion of nonmedicinal drugs such as heroin and cocaine. Patients were included from 8:00 AM to 11:00 PM, 7 days a week, by research staff or radiology technologists on service.
Imaging
Radiographs were obtained in posteroanterior and lateral projections. If a scaphoid bone fracture was suspected, a scaphoid bone radiographic series was obtained that consisted of a posteroanterior view with the hand in ulnar deviation, a lateral view, and two oblique views. Immediately after radiography, an MR imaging examination was performed with a 0.2-T dedicated extremity MR imaging system (Artoscan M; Esaote, Genoa, Italy). The time for our standard MR imaging protocol for the wrist is about 30 minutes. This time would be both too time-consuming and too costly for application in all patients with acute wrist injury. Therefore, we used a shortened MR imaging protocol (Table 1). This protocol was achieved by using only one signal acquired for each sequence and a rectangular field of view and by limiting the phase-encoding direction of the matrix. Shortening of the duration of the examination comes at the cost of some loss in quality, but in a pilot study, we found that the subjective image quality was acceptable (15). The average acquisition time was 6 minutes 30 seconds, with a total examination time, including start-up of the MR imaging system and patient positioning, of approximately 15 minutes. A dual phased-array wrist coil was used.
|
Outcome
We used additional treatment after initial presentation as the outcome measure. We did not use the initial treatment in the emergency department as the outcome measure, since we were only interested in the need for treatment during follow-up visits. By analyzing the need for treatment during these follow-up visits, we were able to assess whether initial MR imaging could add information in the discrimination between patients who did not need follow-up versus those who needed additional treatment and follow-up. Patients with obvious fractures were included to analyze the possibility of replacing radiography with MR imaging.
Data Collection
Data were collected about age, sex, side of trauma, tenderness of the anatomic snuffbox, radiographic results, MR imaging results (whether a traumatic lesion was visible, not visible, or uncertain), and therapeutic procedures (J.J.N.). Data were obtained with direct assessment and with review of all information available from emergency department records, hospital records, outpatient clinic records, and the electronic hospital information system. Questionnaires were sent to all patients at 1, 6, and 12 weeks and at 6 months after the initial visit to obtain information about treatment performed outside our hospital. In case questionnaires were not returned, we interviewed the patients by means of telephone (J.J.N., E.H.G.O.), provided they could be reached within two attempts of calling them. The final diagnosis was obtained with review of all information from diagnostic imaging, follow-up, and questionnaires. The follow-up period was 6 months. If the patients indicated on the questionnaire that they did not have daily symptoms anymore, however, the follow-up was terminated, since no additional effects of the injury were to be expected.
Patients
From August 1999 to May 2001, 90 patients with a mean age of 40.0 years (37 female patients [mean age, 40.8 years; range, 16.4–80.0 years] and 53 male patients [mean age, 38.8 years; range, 16.7–71.1 years]) were included. The age difference between male and female patients was not statistically significant (P = .58). A flow diagram that shows the progress of patients as they passed through the trial is presented in Figure 1. Although inclusion was intended to be consecutive, only about half of all eligible patients were randomized. The reasons were that some patients refused to participate, others had language problems, and still others were inadvertently missed. After enrollment, three patients appeared to have preexisting symptoms of the injured wrist, and they were subsequently excluded. Of the remaining 87 patients, 44 were allocated to the MR imaging strategy and 43 were allocated to the reference group. In four patients who were allocated to the MR imaging strategy, MR imaging was not performed or MR imaging results were not interpretable because of technical failure (one case), because the examination was interrupted due to pain (one case), because the MR imaging system was not available (one case), and because the patient was not able to extend the elbow so that the joint could not reach the middle of the magnet bore (one case). After their first visit to the emergency department, 37 patients required additional therapy and 50 patients did not.
|
Statistical Analysis
The predictive value of age, sex, side of trauma, trauma mechanism, tenderness of the anatomic snuffbox, radiographic results, and MR imaging results was analyzed by using univariable logistic regression. For the MR imaging results, the "MR imaging not performed" group to which patients were randomized was used as a reference group to analyze the incremental value of the MR imaging information compared with the absence of this information. Therefore, the variable "MR imaging results" was recoded and had three different outcomes: positive or uncertain, negative, or MR imaging not performed. In the regression analysis, the MR imaging results were analyzed, with the outcome of MR imaging not performed as the reference group. In this way, the whole patient population could be included in the analysis of the MR imaging results.
To account for the fact that the radiographs and the MR images contain overlapping information, we created an additional combined-imaging variable containing both the radiographic and MR imaging results. This variable had five possible outcomes: (a) negative radiographic results and MR imaging not performed, (b) positive radiographic results and MR imaging not performed, (c) negative radiographic and MR imaging results, (d) discrepant radiographic and MR imaging results, and (e) positive radiographic and MR imaging results. The outcome of negative radiographic results and MR imaging not performed was considered the reference group for the analysis of the combined-imaging variable. With this combined-imaging variable, we took into account the diagnostic interaction between radiography and MR imaging.
A variable was included in the multivariable logistic regression analysis if the P value (
2 test) was less than .1 in the univariable analysis. This lenient P value was chosen to avoid rejection of a variable that could potentially contribute in the prediction of outcome (treatment) in the multivariable analysis. We created four multivariable models for the prediction of the need for additional treatment after the initial emergency department visit. In model 1, only radiographic results were considered; in model 2, radiographic results were omitted and MR imaging results were used instead; in model 3, both radiographic and MR imaging results were included as additive variables; and in model 4, the combined-imaging variable was used.
The likelihood ratio test was used in a stepwise backward approach to assess whether the variable contributed significantly to the model. A P value less than .05 indicated a significant contribution of the variable in the model. The Hosmer-Lemeshow goodness-of-fit test was used to assess calibration of the models (16). A threshold P value of .05 was used and indicated good calibration of the model. The area under the receiver operating characteristic (ROC) curve was used to compare predictive performance across the models (17). The difference between the areas under the ROC curve of the models was analyzed by using the z statistic for ROC curves and was corrected for the fact that they were derived from the same patients (18). The correction factor was calculated by using the Pearson product moment correlation method. The difference was considered significant if the z statistic P value was less than .05.
The Akaike information criterion (AIC) was used to compare the predictive power of the models (19). The AIC is derived from the 2 statistic and takes into account the complexity and the accuracy of the model, with correction for the degrees of freedom. In this way, direct comparison of models is possible. A value for AIC of zero or smaller indicates no predictive value of the model. The significance of a difference in AIC of nested models (eg, the model with radiographic results vs the model with radiographic and MR imaging results) was tested by using the likelihood ratio test. In nonnested models (eg, the model with radiographic results vs the model with the combined-imaging variable), this test of significance cannot be used. For these models, a difference in AIC larger than 2.0 was considered significant. All analyses were performed with software (SPSS for Windows, release 10.0.7; SPSS, Chicago, Ill).
Sample Size
For each variable analyzed in the multivariable logistic regression analysis, at least 10 patients with an event and at least 10 patients without an event are necessary (20). Because the data for this study were obtained from a randomized controlled trial that addressed the costs and effectiveness of early MR imaging of wrist trauma, the number of variables analyzed in the multivariable regression analysis was determined by the number of patients included in the randomized controlled trial. Since 37 of the included patients needed additional therapy and 50 patients did not, we could analyze four variables ([3 + 5]/2) in the multivariable regression analysis.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
Of 12 patients with a scaphoid bone fracture, nine had tenderness of the anatomic snuffbox, and two did not; in one case, the presence or absence of tenderness of the anatomic snuffbox was not reported. All patients with a scaphoid bone fracture that was occult at radiography but visible at MR imaging (Fig 2) had tenderness of the anatomic snuffbox. No patients without a scaphoid bone fracture at MR imaging had a fracture during follow-up. In seven patients with any type of fracture, immobilization was followed by physical therapy. Furthermore, physical therapy was prescribed in four patients with negative radiographic and MR imaging results. One patient with a fracture of the distal radius and ulna (Smith type) underwent surgery 6 weeks after trauma because conservative therapy had failed.
|
|
|
Regression Analysis
Univariable analysis of the variables (Table 3) showed that age, tenderness of the anatomic snuffbox, and radiographic and MR imaging results were significantly predictive of the need for additional treatment, with a threshold P value of .1. Sex and side of trauma were not significantly predictive of treatment.
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
The only additional contribution of MR imaging results in our population was in the identification of radiographically occult fractures of the scaphoid bone and the distal radius. Of these radiographically occult fractures, all the scaphoid bone fractures and one of the distal radius fractures were treated with prolonged cast immobilization. The reason to prolong immobilization of this distal radius fracture was persistent pain. In contrast to scaphoid bone fractures, radiographically occult distal radius fractures are not known to have potential complications if they are not diagnosed. Our results suggest that the potential value of initial MR imaging for the acutely injured wrist is only to be expected in clinically suspected but radiographically occult scaphoid bone fractures. The number of scaphoid bone fractures in our study was too small for us to make a definitive statement in regard to the value of MR imaging for exclusion of scaphoid bone fractures when anatomic snuffbox tenderness is present. The value of MR imaging for the detection of scaphoid bone fractures has been shown by others (4,8,22,23). Results of some studies indicate that early MR imaging in these patients can be cost-effective from a hospital perspective (4,23) and, potentially, also from a societal perspective (22).
The addition of a short dedicated extremity MR imaging examination to the initial diagnostic strategy did not lead to the diagnosis of soft-tissue injury that required treatment. MR imaging diagnosis of triangular fibrocartilage complex lesions and scapholunate ligament rupture did not lead to treatment, and persistent symptoms did not occur.
In our study population, no patient was treated because of soft-tissue injury during the follow-up period; therefore, we could not demonstrate any predictive value of initial MR imaging in this respect. Our results suggest that the incidence of soft-tissue injury in wrist trauma is relatively low; however, the observed low incidence of soft-tissue injury in our study may have been influenced by the relatively short follow-up period of 6 months. Ligamentous instability, especially, can potentially take months to years to be recognized clinically. A low incidence of soft-tissue injury after wrist trauma was also found by Raby (23) in a study about MR imaging in patients who were suspected of having scaphoid bone fractures. In many studies (9–12), a low sensitivity of MR imaging for the detection of scapholunate ligament rupture was reported, although the specificity was generally high.
Another limitation of this study was that a considerable number of eligible patients were inadvertently missed. The main reason for missing patients was that inclusion took place from 8:00 AM to 11:00 PM, 7 days a week. Consequently, a large number of radiology technologists were involved in recruitment, and although they had been instructed, not all were equally aware of the importance of consecutive recruitment and many were forgetful. Since it was often impossible to retrace whether missed patients would have been eligible, exact numbers of missed eligible patients could not be obtained. In a few instances, foreign patients could not be included because they did not understand the informed consent form. We do not expect that patients who were forgotten to be asked or patients who did not understand the native language will have caused a selection bias.
We assumed that the clinical decision about whether or not to treat a patient was always correct. It is possible, however, that some patients were treated unnecessarily or that patients were inadvertently not treated. It was difficult to estimate how many patients were treated unnecessarily; in regard to patients who were inadvertently not treated, we found that four patients without a demonstrated lesion at radiography or MR imaging still had symptoms of the wrist injury 3 months afterward. One of these patients was suspected of having a borderline personality disorder. One patient received physical therapy, but without a clear diagnosis, and two patients received physical therapy but did not seek medical attention for their persistent symptoms. It may have been possible that these last three patients in fact had a substantial but undetected lesion. This potentially missed detection may have been caused by the fact that we used an MR imaging protocol with a limited matrix size and limited number of signals acquired to reduce imaging time. If we missed a lesion in these patients, however, that lesion apparently was not one that led to persistent symptoms without treatment, since all three patients were symptom free within 6 months.
The strategy of a short MR imaging examination in the initial evaluation of acute wrist trauma should only be implemented in the daily routine if it has been shown to be cost-effective. Although we could not show that the MR imaging examination would be helpful in this respect for identification of patients who would not need additional therapy, there may be a potential role for it in expediting the time to final diagnosis and, thereby, the time to necessary treatment. This could result in earlier recovery and, thereby, in a reduction of time off work, which is an important cost to society. A reduction of this cost may outweigh the costs of a short MR imaging examination. Further work should focus on this issue.
In conclusion, a short MR imaging examination with a low-field-strength MR imaging system, in addition to radiography, in the initial work-up of all patients with acute wrist trauma has additional value in the prediction of the need for treatment, but it does not have value in the identification of patients who can be discharged without follow-up. A short MR imaging examination in all patients with acute wrist trauma is, therefore, not recommended. Early MR imaging may, however, be valuable when performed in patients who are clinically suspected of having a scaphoid bone fracture but in whom radiographs are normal.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
See also the articles by Nikken et al and Oei et al in this issue.
Author contributions: Guarantor of integrity of entire study, M.G.M.H.; study concepts and design, all authors; literature research, J.J.N.; clinical studies, J.J.N., E.H.G.O.; data acquisition and analysis/interpretation, all authors; statistical analysis, J.J.N.; manuscript preparation, definition of intellectual content, revision/review, and final version approval, all authors; manuscript editing, J.J.N.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Related Articles
This article has been cited by other articles:
|
|
 |
|
  F. J. P. Beeres, S. J. Rhemrev, P. den Hollander, L. M. Kingma, S. A. G. Meylaerts, S. le Cessie, K. A. Bartlema, J. F. Hamming, and M. Hogervorst Early magnetic resonance imaging compared with bone scintigraphy in suspected scaphoid fractures J Bone Joint Surg Br, September 1, 2008; 90-B(9): 1205 - 1209. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Memarsadeghi, M. J. Breitenseher, C. Schaefer-Prokop, M. Weber, S. Aldrian, C. Gabler, and M. Prokop Occult Scaphoid Fractures: Comparison of Multidetector CT and MR Imaging--Initial Experience Radiology, July 1, 2006; 240(1): 169 - 176. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. J. Nikken, E. H. G. Oei, A. Z. Ginai, G. P. Krestin, J. A. N. Verhaar, A. B. van Vugt, and M. G. M. Hunink Acute Peripheral Joint Injury: Cost and Effectiveness of Low-Field-Strength MR Imaging--Results of Randomized Controlled Trial Radiology, September 1, 2005; 236(3): 958 - 967. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
