| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Musculoskeletal Imaging |
1 From the Departments of Medicine (S.M.W., A.C.F.H., M.F., K.S.W.) and Radiology and Diagnostic Imaging (J.F.G.), Prince of Wales Hospital, Shatin, Hong Kong; and Institute for International Health, University of Sydney, Australia (S.K.L.). Received January 15, 2003; revision requested March 26; final revision received November 19; accepted January 2, 2004. Address correspondence to S.M.W. (e-mail: jsmwong@hkstar.com).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: A prospective cohort of 133 patients suspected of having CTS were referred to a teaching hospital between October 2001 and June 2002 for electrodiagnostic study. One hundred twenty patients (98 women, 22 men; mean age, 49 years; range, 19–83 years) underwent sonography within 1 week after electrodiagnostic study. Radiologist was blinded to electrodiagnostic study results. Seventy-five patients had bilateral symptoms; 23 patients, right-hand symptoms; and 22 patients, left-hand symptoms (total, 195 symptomatic hands). Cross-sectional area of median nerve was measured at three levels: immediately proximal to carpal tunnel inlet, at carpal tunnel inlet, and at carpal tunnel outlet. Flexor retinaculum was used as a landmark to margins of carpal tunnel. Optimal threshold levels (determined with classification and regression tree analysis) for areas proximal to and at tunnel inlet and at tunnel outlet were used to discriminate between patients with and patients without disease. Sensitivity, specificity, and false-positive and false-negative rates were derived on the basis of final diagnosis, which was determined with clinical history and electrodiagnostic study results as reference standard.
RESULTS: For right hands, sonography had sensitivity of 94% (66 of 70); specificity, 65% (17 of 26); false-positive rate, 12% (nine of 75); and false-negative rate, 19% (four of 21) (cutoff, 0.09 cm2 proximal to tunnel inlet and 0.12 cm2 at tunnel outlet). For left hands, sensitivity was 83% (53 of 64); specificity, 73% (24 of 33); false-positive rate, 15% (nine of 62); and false-negative rate, 31% (11 of 35) (cutoff, 0.10 cm2 proximal to tunnel inlet).
CONCLUSION: Sonography is comparable to electrodiagnostic study in diagnosis of CTS and should be considered as initial test of choice for patients suspected of having CTS.
© RSNA, 2004
Index terms: Nerves, diseases • Wrist, abnormalities • Wrist, injuries, 43.419 • Wrist, US, 43.419, 43.1298
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
With the Equivalence of Correlated Proportion module within a sample size software (PASS2002; J. Hintz, Kaysville, Utah), we estimated that a sample of 120 patients would be needed to obtain a power of .8 with an 
of .05. To be specific, the module calculates sample size for equivalence tests in which two dichotomous responses are measured in each subject. The parameters used in the calculations included the following: a one-sided (noninferiority) alternative, the standard proportion (ie, the proportion of patients among those who were eligible to be recruited with a diagnosis of CTS set at 0.8), an 89% sensitivity of sonography (4), and 0.088 [0.8 – (0.8 x 0.89)] as the maximum allowable difference between the electrodiagnostic study and sonographic proportions that would still result in the conclusion of equivalence.
We then invited consecutive patients who were suspected of having CTS and were referred to the electrodiagnostic unit at Prince of Wales Hospital, Shatin, Hong Kong, for confirmation of the diagnosis at an electrodiagnostic study to participate in this study. The study protocol was approved by the university research ethics committee; written informed consent was obtained in all patients. All patients were referred because their physicians suspected that they had a primary diagnosis of CTS. Patients were excluded from the study who had diabetes mellitus because there was a lack of accepted diagnostic criteria for this group of patients and who were suspected of having other coexisting disorders, such as polyneuropathy or thoracic outlet obstruction, in addition to CTS at referral because they required a more detailed electrodiagnostic assessment.
One hundred thirty-three consecutive patients who were suspected of having CTS as the principal diagnosis and who underwent an electrodiagnostic study for confirmation of it from October 2001 to June 2002 were invited to join the study. Of this number, five patients failed to return for the subsequent sonographic examination after they underwent the initial nerve conduction study. Three patients were excluded from the study because they had concurrent diabetes mellitus. Four patients had an initial referral to rule out concurrent neurologic diseases (two who were suspected of having thoracic outlet obstruction and two who were suspected of having double crush syndrome). One patient was excluded from the final analysis after sonographic findings revealed a bifid median nerve, which resulted in technical difficulty in interpretation of the results (Fig 1).
|
In all patients, the medical history was reviewed by one of the authors (M.F.) at the electrodiagnostic study. Each patient was asked not to divulge the results of the electrodiagnostic study to the subsequent examiner at sonography to minimize bias.
Electrodiagnostic Technique
All patients underwent a standardized nerve conduction study that was based on the American Academy of Neurology summary statement about the desired protocol for patients who are suspected of having CTS (1). This nerve conduction study consisted of 8-cm transcarpal orthodromic median and ulnar sensory peak latencies, median forearm conduction velocities, 8-cm median motor compound muscle action potential, and distal motor latencies. These parameters were measured by using standard techniques of supramaximal stimulation and surface electrodes, with adjustment for skin temperature (Nicolet; Viking III, Madison, Wis). The nerve conduction studies were performed with the guidance of a neurologist (M.F.) who had a special interest in electrodiagnostic studies, along with an experienced electrodiagnostic study technician. Patients referred to our laboratory who were suspected of having CTS underwent an electrodiagnostic study that was performed in both hands regardless of the hand in which symptoms were manifested. Mild discomfort relating to electrical shock during the procedure was reported by 11 (9.5%) patients.
Sonographic Technique
Sonographic examinations were performed within 1 week after the electrodiagnostic study by a radiologist (J.F.G.) experienced in musculoskeletal sonography. Examinations were performed by using a 13–5-MHz linear array transducer (Multi-D; Sonoline Elegra Advanced; Siemens, Munich, Germany). Subjects were seated facing the examiner. The arms were extended; wrists were rested on a hard flat surface, forearms were supinated, and the fingers were semiextended. Transverse images of the median nerve were obtained at three levels: immediately proximal to the carpal tunnel inlet (Fig 2), at the carpal tunnel inlet (Fig 3), and at the carpal tunnel outlet (Fig 4). The flexor retinaculum was used as a landmark, in preference to the bone landmarks, to the margins of the carpal tunnel. The flexor retinaculum is seen at high-resolution sonography as a variably bowed echogenic band that spans the carpus.
|
|
|
|
|
|
Reference Standard
The diagnosis of CTS was confirmed with findings of an electrodiagnostic study. In addition to these findings, a compatible clinical history (sensory symptoms over the distribution of the median nerve with or without positive results with the Phalen maneuver and/or a positive Tinel sign) was used to confirm the diagnosis. The clinical history was elicited by the neurologist during the electrodiagnostic study before the data were reported. A difference of more than 0.4 msec between the median and ulnar sensory peak latencies or a prolonged median distal motor latency of more than 4 msec was defined as confirmatory electrophysiologic evidence of CTS. The examiner had no knowledge of the sonographic results at the time he or she reported the results, since sonography was performed after the electrodiagnostic study.
Statistical Analysis
Reliability testing was performed in the symptomatic hands only, and the reliability value was estimated with the intraclass correlation coefficient. The intraclass correlation coefficient (2.1) was used for interreader reliability, while the intraclass correlation coefficient (3.1) was used for intrareader reliability. The two intraclass correlation coefficient models used have been described previously by Shrout and Fleiss (7). Briefly, the intraclass correlation coefficient (2.1) is useful for assessment of interreader reliability, whereas the intraclass correlation coefficient (3.1) is appropriate for intrareader reliability.
The optimal threshold values for levels proximal to the tunnel inlet, at the tunnel inlet, and at the tunnel outlet for discrimination between patients who received a diagnosis of the disease and those who did not were found by using a classification and regression tree analysis (8). Classification and regression tree analysis is a statistical procedure that was developed for estimation of prediction errors in a classification model. In our application, patients who were suspected of having CTS and were classified on the basis of findings at an electrodiagnostic study were entered into the model as the target variables, whereas scores for levels proximal to the tunnel inlet, at the tunnel inlet, and at the tunnel outlet were entered as predictors. The aim was to find the optimal cutoffs for the three-predictor variables that would produce a classification system that could be used to discriminate between patients with CTS and those without CTS, with the minimum misclassification rate. Classification and regression tree analysis was performed separately for each hand (with appropriate data in patients with unilateral symptoms), as well as for the corresponding hand (with data in those with bilateral symptoms).
A separate analysis also was conducted only for patients with bilateral symptoms with a combination of data in both hands. The sensitivity, specificity, and false-positive and false-negative rates were then calculated. All the data analyses were performed with statistical software (Answer Tree, version 2.1; SPSS, Chicago, Ill).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Interreader reliability for sonographic examination was established in eight patients with CTS in whom assessment was performed by the same examiner and another radiologist during one examination. The reliability coefficients were 0.87, 0.71, and 0.90 for measurements at levels proximal to the tunnel inlet, at the tunnel inlet, and at the tunnel outlet, respectively.
One hundred twenty patients who were suspected of having primary CTS entered the study. At sonography, none had concurrent ganglia, nerve sheath tumors, or other secondary causes of CTS.
When only right-hand data were used, the best cutoffs were a median nerve cross-sectional area of 0.09 cm2 at the level proximal to the tunnel inlet and 0.12 cm2 at the tunnel outlet. As shown in Figure 5, patients with an area larger than 0.09 cm2 at the level proximal to the tunnel inlet were classified as having CTS. Those with an area of 0.09 cm2 or smaller at the level proximal to the tunnel inlet required further measurements and were classified as having CTS if the tunnel outlet area was larger than 0.12 cm2. Measurement at the tunnel inlet was not a clinically important predictor. The sensitivity, specificity, and false-positive and false-negative rates were 94% (66 of 70), 65% (17 of 26), 12% (nine of 75), and 19% (four of 21), respectively.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
The conclusion derived from this study is that a combination of measurements at the level proximal to the tunnel inlet and at the tunnel outlet provided the best sensitivity and specificity. The threshold values derived may seem complex, but this result is to be expected because of the uncertainty surrounding normal values for median nerve dimension and side-to-side variation, along with the fact that multiple-level swelling of the median nerve in the carpal tunnel seems to be the rule in patients with CTS. Nerve caliber and swelling is a continuum, and attempts to define single universal threshold values will always have limited success.
The benefit of a single cutoff is obvious for daily practice. We computed a single cutoff for both left and right hands in our study, and radiologists in other centers may do likewise for diagnostic ease. We stress, however, that it is not easy to justify independence of the two hands in the same patient; as a result, such a single cutoff might violate the primary assumption of independent observations. Because of this consideration, we decided against supplying the combined left- and right-hand cutoffs in Results in this article. Nevertheless, when data in both hands were combined in our study, the best cutoffs were 0.10 cm2 at the level proximal to the tunnel inlet and larger than 0.12 cm2 at the tunnel outlet. The sensitivity, specificity, and false-positive and false-negative rates were 86% (93 of 108), 74% (31 of 42), 11% (11 of 104), and 33% (15 of 46), respectively.
The false-positive rate was 11%–15% with sonography. It is generally accepted that approximately 10% of patients with classic symptoms of CTS who are referred for an electrodiagnostic study will have normal results at that study (12). No consensus exists for this group, but some clinicians have successfully treated these patients with surgical decompression, which is the treatment for patients with true CTS (13,14). Since these patients who were suspected of having CTS were referred for confirmation of the diagnosis, one could argue that normal results of an electrodiagnostic study would be used to classify them into the 10% false-negative category according to traditional diagnostic criteria (though now recognized as abnormal findings at sonography). Previously, paresthesia in patients with CTS has been shown to occur prior to failure of conduction in unmyelinated sensory fibers, as measured with nerve conduction tests (15).
Sonography measures a different parameter (structural pathologic abnormalities of nerve swelling) from that (physiological malfunctions of the median nerve) measured at an electrodiagnostic study. It is possible that some patients who have CTS that is detectable at sonography may have swelling of the median nerve that is not severe enough to cause impairment in conduction. Therefore, one could also argue that the patients identified in this study do not have sonographically determined false-positive findings but rather false-negative findings suggestive of CTS at the electrodiagnostic study. One must understand the limitations of an electrodiagnostic study concerning its sensitivity and specificity (16). It would be interesting to follow the group of patients with sonographically determined false-positive cases of CTS longitudinally to see if they eventually develop abnormalities that are compatible with CTS at an electrodiagnostic study.
If one were willing to retest those with negative results for diagnosis of CTS at an electrodiagnostic study, then the false-positive rate with a cutoff of 0.10 cm2 at the level proximal to the tunnel inlet in the right hand would only be 2%. This makes the diagnosis of CTS on the basis of the measurements at the levels proximal to the tunnel inlet and at the tunnel outlet in the right hand almost certain in those with the disease; retesting at an electrodiagnostic study can be performed in the event of an initial negative sonogram. If sonography is chosen as the first diagnostic test, with the assumption that one in five patients will require retesting after an initial negative sonogram, considerable cost reduction could still be achieved. An electrodiagnostic study costs $2,500 Hong Kong dollars ($320 U.S. dollars) in our center, with the procedure lasting as long as 30 minutes, whereas sonography costs $500 Hong Kong dollars ($60 U.S. dollars) and requires a performance time of as long as 15 minutes. Professional societies are urging the performance of cost-effectiveness studies to determine the usefulness of an electrodiagnostic study (17).
Each investigation yields different information, but sonography seems to have an advantage in this sense for confirmation of CTS in selected patients because of its cost and easy accessibility. Not all sonographic machines, however, have the capacity for high-resolution and high-frequency imaging. Both tests were found to be highly reproducible, as all the reliability coefficients were greater than 0.70. The overall reliability of sonography seems compatible with that of an electrodiagnostic study when one considers that multiple measurements were made at sonography.
Our study had limitations. First, our electrodiagnostic unit is a tertiary referral center for the surrounding population of 800,000 patients from various specialties including general medicine, orthopedics, and neurology, with the study group limited to patients with CTS as a primary diagnosis. Spectrum and selection bias are inherent in this type of hospital-based study because our patients represented those with severe enough symptoms to warrant a referral in the first place. Generalization of our results regarding accuracy of both an electrodiagnostic study and sonography might not be applicable in primary care, as was demonstrated in previously published community-based research in regard to the use of the electrodiagnostic study (18).
Second, electrodiagnostic study results combined with clinical symptoms as determined by a neurologist were regarded as the reference standard in this study; other centers may perform electrodiagnostic study techniques in addition to those we performed in our study to confirm the diagnosis of CTS. The American Academy of Neurology issued new practice parameters that were published toward the end of this study (17). The performance of sonography compared with such alternative electrodiagnostic strategies is uncertain. It is likely that additional electrodiagnostic studies may produce better sensitivity but reduced specificity (19).
Third, sonography is an operator-dependent test, and appropriate experience is required to ensure reliability and reproducibility. With appropriate training of operators, however, this issue can be resolved readily.
Although the data in this study suggest that sonography is effective in the diagnosis of CTS, the extended role of sonography in that diagnosis awaits further definition. In the past, the electrodiagnostic study has had other roles in the treatment of CTS. Grading systems for the severity of CTS that are based on electrodiagnostic study results have been devised (20–23). In addition, electrodiagnostic study results can be used to identify other conditions that mimic CTS, such as cervical radiculopathy, polyneuropathy, or other median nerve entrapment syndromes (24,25). Whether sonography plays a role as a predictor of treatment response or a role in the differentiation of CTS from conditions that simulate CTS remains unknown.
Finally, the role of a combination of sonography and an electrodiagnostic study in the diagnosis of CTS remains to be determined, provided that the issue about standard criteria for diagnosis is resolved (26). An electrodiagnostic study was included as part of the reference standard in the present study and in most previous studies; thus, the assessment of the diagnostic accuracy of sonography as a stand-alone diagnostic test of choice independent of the electrodiagnostic study will remain difficult.
In conclusion, we found the following diagnostic accuracy values of sonography: In the right hand, with a cutoff of 0.09 cm2 at the level proximal to the tunnel inlet and 0.12 cm2 at the tunnel outlet, the sensitivity, specificity, and false-positive and false-negative rates were 94%, 65%, 12%, and 19%, respectively. In the left hand, with a cutoff of 0.10 cm2 at the level proximal to the tunnel inlet only, the sensitivity, specificity, and false-positive and false-negative rates were 83%, 73%, 15%, and 31%, respectively.
We propose an algorithm for the evaluation of patients who are suspected of having CTS (Fig 7). We believe that all patients who are suspected of having CTS but do not have diabetes mellitus should undergo sonography as the initial diagnostic test. If sonography yields positive results, CTS is confirmed; if the results are negative, one could then refer the patient for an electrodiagnostic study. Performance of an electrodiagnostic study should be reserved for confirmation of CTS if the initial sonogram is negative or the initial history of the patient suggests that the pathologic diagnosis might include more than CTS alone.
|
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, S.M.W., J.F.G.; study concepts, S.M.W., J.F.G., A.C.F.H.; study design, S.M.W.; literature research, A.C.F.H.; clinical studies, J.F.G., A.C.F.H., M.F.; data acquisition, A.C.F.H., J.F.G., M.F.; data analysis/ interpretation, S.M.W., S.K.L.; statistical analysis, S.K.L.; manuscript preparation, S.M.W., A.C.F.H., J.F.G.; manuscript definition of intellectual content, S.M.W., K.S.W.; manuscript editing, K.S.W.; manuscript revision/review and final version approval, S.M.W., J.F.G.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  I. PINILLA, C. MARTIN-HERVAS, G. SORDO, and S. SANTIAGO The Usefulness of Ultrasonography in the Diagnosis of Carpal Tunnel Syndrome J Hand Surg Eur Vol., August 1, 2008; 33(4): 435 - 439. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Aleman, J. D. Berna, M. Reus, F. Martinez, G. Domenech-Ratto, and M. Campos Reproducibility of Sonographic Measurements of the Median Nerve J. Ultrasound Med., February 1, 2008; 27(2): 193 - 197. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L H Visser, M H Smidt, and M L Lee High-resolution sonography versus EMG in the diagnosis of carpal tunnel syndrome J. Neurol. Neurosurg. Psychiatry, January 1, 2008; 79(1): 63 - 67. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D P Bland Carpal tunnel syndrome BMJ, August 18, 2007; 335(7615): 343 - 346. [Full Text] [PDF]
|
|
|
|
 |
|
 H. Berner Hammer, E. A Haavardsholm, and T. Kristian Kvien Ultrasonographic measurement of the median nerve in patients with rheumatoid arthritis without symptoms or signs of carpal tunnel syndrome Ann Rheum Dis, June 1, 2007; 66(6): 825 - 827. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Oh, M. Belohlavek, C. Zhao, N. Osamura, M. E. Zobitz, K.-N. An, and P. C. Amadio Detection of Differential Gliding Characteristics of the Flexor Digitorum Superficialis Tendon and Subsynovial Connective Tissue Using Color Doppler Sonographic Imaging J. Ultrasound Med., February 1, 2007; 26(2): 149 - 155. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. B. Hammer, I. A. H. Hovden, E. A. Haavardsholm, and T. K. Kvien Ultrasonography shows increased cross-sectional area of the median nerve in patients with arthritis and carpal tunnel syndrome Rheumatology, May 1, 2006; 45(5): 584 - 588. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y.-M. Lin, T.-S. Lee, S. M. Wong, S. K. Lo, and N. A. Obuchowski Cutoff Point of 0.10 cm2 Appropriate for Both Hands * Drs Wong and Lo respond: * Dr Obuchowski responds: Radiology, February 1, 2005; 234(2): 642 - 643. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
