WORK A Journal of Prevention,

Assessment & Rehabilitation

ELSEVIER

Work 8 (1997) 37-44

Inter-rater agreement and accuracy of clinical tests used in diagnosis of Carpal Tunnel Syndrome J.e. MacDermid a*, J.F. Kramer b, R.M. McFarlane a, J.H. Roth a a Hand

and Upper Limb Centre, St. Joseph's Health Centre, London, Ontario, N6A 4L6, Canada of Physical Therapy, University of Western Ontario, London, Ontario, Canada

b Department

Abstract

A sample of 42 subjects (84 hands) with complaints of pain, numbness and/or tingling were used to study the inter-rater agreement and accuracy of seven clinical tests which can be used to assist in the diagnosis of carpal tunnel syndrome (CTS). A 'gold standard diagnosis', defined as a clinical diagnosis of CTS rendered by one of two hand surgeons and supported by electrodiagnostic evidence of CTS, was used to classify the 84 hands as to the presence of CTS. Two therapists were blinded to patient history, electrophysiology, diagnosis and all evaluations performed by the other therapist, until clinical testing was completed. Tests performed included: wrist flexion, wrist extension, Tinel's, tethered median nerve (TMN), pinch, vibration and Semmes-Weinstein monofilament (SWMF) tests. Substantial inter-rater agreement was observed between the two therapists for five of the clinical tests (k> 0.71\ with SWMF and TMN having lower agreement. The most accurate test was Phalen's wrist flexion test. Good accuracy was demonstrated by pinch and vibration tests. Tinel's test was characterized by lower sensitivity, but false positives were rare. Wrist extension and TMN tests had poor sensitivity. SWMF testing was very sensitive, but a high number of false positives occurred when 'normal' was classified as 2.83. Reliability and accuracy of these tests supports their use as components of a clinical diagnosis of CTS. Copyright © 1997 Elsevier Science Ireland Ltd.

Keywords: Carpal tunnel syndrome; Diagnosis; Reliability; Validity; Sensitivity; Specificity

1. Introduction Results of provocative and sensory clinical tests are commonly used by a variety of rehabilitation

* Corresponding author. Tel.: + 1 5196466100 ext 5504; fax:

+ 1 5196466049

professionals to assist with the diagnosis of carpal tunnel syndrome (CTS). CTS diagnosis is based largely on the patient's clinical presentation (Phalen, 1966). However, increased utilization of electro diagnosis has been accompanied by controversy on the value of commonly used clinical examination techniques such as Phalen's and Tinel's tests. Some studies have suggested that

1051-9815/97/$17.00 Copyright © 1997 Elsevier Science Ireland Ltd. All rights reserved

Pll SI051-9815(96)00218-5

38

J.C. MacDennid et al. / Work 8 (1997) 37-44

the accuracy of clinical tests is no better than chance (Golding et al., 1986, Heller et aI., 1986, deKrom et aI., 1990). As clinical test interpretation involves a subjective component, disagreement on the usefulness of diagnostic test results could be related to differences in test protocols, instruments or skill levels of evaluators. However, it is also possible that the tests themselves are inherently inconsistent and as a result are characterized by low agreement on repeated testing of the same individual. A prerequisite in defining the capabilities of diagnostic tests is to determine agreement between testers on the results of the test. Perhaps of greater importance is the accuracy of the test, which can be determined by comparing the test result with the 'true' result. Systematic errors such as the failure to detect abnormality (low sensitivity) or frequent false positives (low specificity) contribute to reduced accuracy in diagnosis. Clinical diagnosis of CTS remains an essential element of management in CTS patients. Electrodiagnostic tests are valuable in assisting with diagnosis in the hospital setting. However, these generally require expensive equipment, specialized personnel and can be time consuming and uncomfortable. For this reason, they are less suitable for workplace screenings or on site evaluations. Clinical tests offer the advantagc of being simple, inexpensive and quick to perform. If they are reliable and accurate they could be a useful tool for therapists involved in workplace screening or assessments. A variety of clinical examinations and tests can be used to form a clinical diagnosis and a number of these have been studied for accuracy. However, no study has evaluated the retest agreement of commonly used clinical tests and their accuracy remains controversial. Therefore, the purposes of this prospective study were: (1) to examine the agreement between evaluators on the results of clinical diagnostic tests and (2) to examine the accuracy of selected clinical tests in diagnosis of CTS by comparing clinical test results with those of a predefined gold standard diagnosis.

2. Methods Both hands (n = 84) from 42 informed volunteers were assessed independently by two experienced hand therapists (one a certified hand therapist and one a charter member of the American Society of Hand Therapists) during a single test session. All subjects were new patients referred to the clinic with a history of a gradual onset of any of the following symptoms: pain, numbness or tingling. All consented to participate in the study as approved by the university ethics board. Patients with acute injuries, previous upper extremity surgery, complaints relatcd to congenital malformations, Dupuytren's disease, tumours or severe hand deformities were not included. The two hand therapists were separated from each other during testing and were blinded to the history, electrophysiology, diagnosis, and results from the other therapist's examination. The following tests were performed in random order: wrist flexion, wrist extension, tethered median nerve (TMN), Tinel's, pinch, vibration (tuning fork) and Semmes-Weinstein monofilament (SWMF) tests. Sensory testing was performed on the pulp of the thumb, index, long and small fingertips, using the small finger as a comparator (patients with ulnar neuropathy were excluded from accuracy of SWMF/vibration analysis). SWMF testing was performed as described by Bell-Krotoski (1995) using the 20 piece kit. A classification of abnormal was assigned if the SWMF threshold for any of the radial three digits was greater than 2.83 and greater than the threshold for the small finger. Vibration testing with the prong of a 256 cycle per second tuning fork was performed on the fingertip as described by Dellon (DeIlon, 1978). Five provocative tests were performed with reproduction of symptoms of numbness and/or tingling in the median nerve distribution taken as a positive test result. The wrist flexion test (Phalen's test) (Phalen, 1966) was performed by placing the dorsal aspects of the hands together and maintaining a position of maximum wrist flexion. Wrist extension (reversed Phalen's) was

J.e. MacDennid et al. / Work 8 (1997) 37-44

performed by placing the palms together and maintaining a position of full passive extension. The pinch test (Ditmars and Houin, 1986) was performed by having the patient actively pinch a piece of paper between the tips of the thumb, index and long fingers using MP flexion and i. p. extension. These tests were scored as positive if symptoms were reproduced within 60 s. The tethered median nerve test (LaBan et aI., 1986 and 1989) was performed by having the examiner push against the patients fingertip, pushing the index finger of the supinated forearm into maximum extension. This position was maintained for 15 s, and was scored positive if symptoms were reproduced. Tinel's test was performed using a reflex hammer to percuss over the palm from the proximal palmar crease to the distal wrist crease with electric shocks projected throughout the median nerve distribution indicating a positive test result (Mossman and Blau, 1986). The gold standard diagnosis of CTS was defined as a clinical diagnosis of CTS made by either of two experienced hand surgeons (based on history, motor and sensory inspection) plus independently obtained electrodiagnostic evidence of CTS. Physician examination included a detailed history of symptoms and aggravating factors, two point discrimination and light touch sensory evaluation and strength testing of abductor pollicis brevis by manual muscle testing. All

39

nerve conduction and electromyographic testing was performed by blinded staff neurologists. Evaluation of normality was considered in the context of the entire neurophysiologic examination, which included testing of ulnar, radial and proximal nerves as sources of pathology and examination of distal latencies, amplitudes and conduction times for motor and sensory nerves. Severity of nerve compression was rated as mild, moderate, or severe for Table 1 according to the extent of neurophysiologic abnormality. Inter-rater agreement results were determined by comparing test results between the two hand therapists using the kappa (k) statistic (Cohen, 1960), which adjusts for chance agreement. Confidence intervals were calculated around each k and interpreted with respect to a subjective scale proposed by Landis and Koch (1977), where a k of 0.00-0.20 indicates slight agreement, 0.21-0.40 fair, 0.41-0.60 moderate, 0.61-0.80 substantial and 0.81-1.00 almost perfect agreement. Accuracy was determined by comparing the conclusions of normal versus abnormal for each test with that of the independently obtained gold standard diagnosis. The only patients excluded from accuracy analysis where those whose pathology would interfere with test interpretation. Therefore, patients with ulnar neuropathy (6%) were excluded from sensory evaluations that incorporated an ulnar comparison in the decision

Table 1 Subjects CTS

47 (15) Numbness and/or tingling

Age Primary complaint Reported frequent numbness Reported frequent pain Symptoms worse at night Symptoms restricted to hand Primary diagnosis

n = sample, FCR = Flexor carpi radialis, RSI

en = 36 hands)

100% 97% 75% 42% Mild CTS36% Moderate CTS 35% Severe CTS 28%

=

Repetitive strain injury

non-CTS

en = 48 hands)

37 (13) Pain 25% 76% 25% 56% DeQuervains tendonitis 31 % RSI or Diffuse tendonitis 45% Osteoarthritis 12% FeR Tendonitis 6% Ulnar neuropathy 6%

J.e. MacDermid et al. /

40

rule. Proximal nerve pathology or diffuse neuropathy was not considered to be a contraindication to the use of provocative tests. Sensitivity, the true positive rate, and specificity, the true negative rate, were determined (Sackett et aI., 1985) and compared with arbitrary subjective criteria as follows: > 85% excellent, 70-85% good, 50-70% fair and < 50% poor accuracy. However, the importance of early treatment in CTS and the minimal side effects associated with conservative treatments imply that more emphasis on sensitivity may be warranted. 3. Results The gold standard diagnosis classified 26 subjects as having CTS (10 bilateral, 16 unilateral) and 16 as not having CTS (Table 1). These nonCTS subjects had a variety of conditions such as tendonitis, osteoarthritis and ulnar neuropathy. While the primary complaints of CTS patients and non-CTS patients differed, symptomatology alone could not differentiate all CTS patients from those with other pathologies. Symptoms of pain, numbness and tingling occurred in both CTS and non-CTS subjects. CTS subjects presented most frequently with complaints of numbness and tingling, whereas non-CTS patients more frequently reported pain as the predominant symptom. However, pain was present in the majority of CTS patients and 25% of non-CTS patients reported experiencing numbness and tingling. The numbness or tingling of non-CTS

Work 8 (1997) 37-44

patients tended to be infrequent and diffuse, whereas CTS patients tended to have specific median nerve symptoms. Inter-rater agreement varied widely among the seven clinical tests, with five tests demonstrating at least substantial agreement (Table 2). Diagnostic test accuracy was similar for each of the two evaluators, but varied between clinical tests (Table 3). Except for SWMF, specificity tended to be better than sensitivity, that is false negatives were more common than false positives. The highest accuracy was observed for Phalen's test which had excellent sensitivity and specificity. Vibration perception demonstrated good accuracy. Sensitivity was excellent for SWMF testing, however specificity was poor to moderate, dependent on tester. The other clinical tests had lower accuracy. 4. Discussion This study provides the first data on the test-retest agreement of physical examination tests in diagnosis of CTS and additional data on the accuracy of these tests. Some studies have reported the incidence of positive clinical test results in a series of CTS patients, without including data on the incidence in patients who did not have CTS (Phalen, 1966, Ditmars and Houin, 1986, LaBan et aI., 1986, Novak et aI., 1992). While sensitivity of a test is important, without data on specificity, one cannot evaluate the usefulness of a diagnostic test. A high rate of false positives {low speci-

Table 2 Reliability of clinical tests Test

n

Kappa

Subjective rating

CI

Phalen's Tinet's Pinch Wrist extension Vibration TMN SWMF

84 82 82 82 84 78 78

0.88 0.81 0.76 0.72 0.71 0.49 0.22

Almost perfect Almost perfect Substantial Substantial Substantial Moderate Fair

0.77-0.98 0.66-0.98 0.62-0.91 0.55-0.88 0.56-0.86 0.26-0.71 0.26-0.42

n = number of hands tested; CI = 95% confidence interval around kappa

J.e. MacDennid et al. / Work 8 (1997) 37-44

41

Table 3 Accuracy of clinical tests Test

Tester

n

Sensitivity (%)

Specificity (%)

2

81 77

87 86

90 86

1 2

77 73

77 77

80 72

2

77 73

72 70

88 78

2

80 76

65 75

96 85

78 74

59 41

92

2 2

80 76

52 36

92 95

2

79 70

86 85

60 32

Phalen's Vibration Pinch Wrist Extention Tinet's TMN SWMF

n

=

94

number of hands tested; TMN = tethered median nerve

ficity) would negate the utility of a highly sensitive test in distinguishing patients with CTS from those with other hand pathologies. Other studies which incorporate non-CTS subjects in the study design (Golding et aI., 1986, Heller et aI., 1986, deKrom et aI., 1990) have not reported the type of testers or methods that were used to perform the clinical tests and tended to rely solely on electro diagnostic criteria as the gold standard diagnosis of CTS. These studies found poor sensitivity and specificity of a variety of commonly used clinical tests. However, the generalizability of these results to experienced therapists is questionable. Studies which have used experienced hand surgeons and therapists as testers have tended to report favourable sensitivity and specificity of selected clinical tests such as Phalen's test and SWMF testing (Szabo et aI., 1984, Gellman et aI., 1986). However, these studies have used asymptomatic subjects rather than patients with other pathologies as controls. This creates an artificial simplicity in diagnosis that does not challenge the diagnostic test nor reflect clinical practice. As the role of a clinical test is to distinguish CTS

patients from those with other pathologies that might potentially be confused with CTS, it is essential that this process be replicated in studies evaluating the accuracy of clinical diagnostic tests. The inclusion of asymptomatic controls could be expected to over estimate the true accuracy of a test used in clinical evaluation (Sackett et aI., 1985). The proper 'control' group for evaluation of diagnostic test accuracy has been described by Sackett et ai. and includes patients with the same pathology in areas other than that of the target disorder (i.e. ulnar nerve pathology) and other pathologies in the same area (i.e. OA or tendonitis). To establish clinical relevance, controls must be those that would be encountered in the clinic and methods must be similar to those used in the clinical/workplace setting. For this reason, all testing was performed in the clinic with procedures readily available in the typical hand clinic. Patients tended to be ones that ascribed their symptoms to factors in the workplace and the majority of diagnoses were ones frequently related to workplace disorders. Accuracy can only be determined by comparing

42

J.e. MacDennid et al. /

the test result with the 'true' result. The 'true' result must be represented by the most rigorous form of a diagnosis, the 'gold standard diagnosis'. The acceptability of this gold standard diagnosis is typically the most difficult and controversial element of studies that evaluate the accuracy of a diagnostic test. This is particularly true in CTS where no single diagnostic method has been accepted as a gold standard. Previous work has illustrated the imperfect sensitivity of electrodiagnosis. Redmond and Rivner (1988) demonstrated 46% of asymptomatic individuals may have at least one indicator suggestive of CTS and 14% may have abnormal distal sensory latencies. These results suggest that electrodiagnosis, while valuable in diagnosis of CTS, can not, in isolation, be considered either as a substitute for clinical examination or as a gold standard diagnosis. However, for the purpose of studying clinical diagnostic methods it provides objective evidence to which one can compare the subjective results of clinical tests and was, therefore, used as a component of the 'gold standard diagnosis'. The effect of this lack of a universally acceptable gold standard diagnosis is demonstrated by the four patients in which the electrophysiologic and clinical diagnoses did not agree. Two subjects (three hands) with suspected carpal tunnel syndrome determined by both the surgeon and the clinical test results, had normal electrodiagnostic test results. Thus, according to the pre-defined gold standard diagnosis they did not have CTS. Although adherence to study protocol required that these patients be classified as non-CTS, both responded to night splinting, suggesting falsely negative electrophysiology. This 3.5% of the study sample would have slightly lowered the specificity results for the clinical tests. Two subjects (two hands) without CTS symptoms but electrodiagnostic test results consistent with CTS were also classified as non-CTS according to study protocol. Six months later these two patients were still asymptomatic, suggesting that while the study gold standard diagnosis was correct, the electrodiagnos tic tests were falsely positive. As the gold standard diagnosis was not questionable in these cases these two patients would not alter study results. The controversies generated in the diag-

Work 8 (1997) 37-44

nosis of these patients exemplify the difficulty in choosing a comparator diagnosis when studying clinical tests for CTS diagnosis. Previous studies have used either electrodiagnosis alone or a clinical diagnosis alone as the 'gold standard'. We elected to define a gold standard diagnosis that incorporated electrodiagnostic and clinical criteria in an attempt to minimize methodologic error. The most accurate test was the wrist flexion test. This finding is consistent with previous studies (Phalen, 1966, Gellman et aI., 1986, Novak et aI., 1992). The next most accurate test was vibration testing with a tuning fork, with the 77% sensitivity found in this study being identical to that reported by Dellon (1978). New information from this study is that vibration with a tuning fork is 80% specific in differentiating CTS patients from patients with other hand pathologies. While more sophisticated methods are available for testing vibration sensibility, the tuning fork remains useful for screening where these more expensive devices are inaccessible or impractical. Of the tests evaluated, wrist flexion and vibration testing with a tuning fork provided the best overall accuracy and are of sufficient accuracy to assist with a clinical diagnosis of CTS. The maintained pinch test was proposed to reproduce CTS symptoms through dynamic compression of the carpal canal via the lumbricals (Ditmars and Houin, 1986). Although not as accurate as wrist flexion or vibration, it may be helpful in clinical examination. In general the less frequently reported clinical tests such as wrist extension and TMN were less accurate and reliable. As a result, they may not offer any major advantage over more traditionally used tests. In agreement with previous studies, SWMF had very high sensitivity in detection of CTS (Szabo et aI., 1984). However, unlike these previous reports, specificity was low. The critical difference between our results and those reported previously is that others have used asymptomatic subjects rather than non-CTS hand patients in their study design. Our lower specificity values may reflect the fact that SWMF are quite able to distinguish CTS patients from patients without any pathology, but are less able to differentiate between CTS patients and patients with hand complaints due to

J.e. MacDennid et al. / Work 8 (1997) 37-44

other pathologies. Further investigation is required to determine the role of selected cutoff for normality (filament 2.83), and the effect of other hand pathologies or disuse on sensibility, as possible explanations for low specificity in the present study. Inter-rater agreement of clinical testing reflects variation in test results due to variations in patient response or examiner technique and interpretation. Our results suggest that different testers are likely to agree on the results of Phalen's, Tinel's, pinch, vibration with a tuning fork and wrist extension tests used in diagnosis of CTS. The finding that five of the seven clinical tests were at least substantially reliable (k > 0.71) is superior to the agreement on physical examination techniques used by specialist physicians in diagnosis of other pathologies (Waddell et aI., 1982, Solari et aI., 1989, Rudy et aI., 1990). The TMN and SWMF tests demonstrated less than substantial agreement (k < 0.49). Because stretching into full composite extension is typically uncomfortable, there may be difficulty in distinguishing CTS complaints from those experienced by patients with other pathologies when using the TMN test. The low inter-rater agreement of the SWMF testing using the decision rule that abnormality was indicated by a SWMF threshold greater than 2.83 and greater than that in the small finger was an unexpected finding. We have demonstrated that this more complex decision rule lowers the reliability of decisions on normality of SWMF (MacDermid et aI., 1994). However, agreement was high for vibration testing which used a similar decision rule. Of more importance in explaining the relatively low reliability for SWMF is the fine gradients between successive filaments in the 20 piece kit. Most disagreement on SWMF threshold occurred due to variation by one filament in the fine gradients between successive filaments in the range 2.83-3.61. Use of the five piece kit or the recently available enhanced filaments might reduce this source of error, however, further investigation is required to address this question. Previous studies on physical diagnosis indicate that agreement on clinical findings improves as tests are made more objective, when evaluators

43

have similar training and when training sessions are used to increase agreement (Waddell et aI., 1982, Solari et aI., 1989, Rudy et aI., 1990). The effect of variations in these factors on the results of reliability and accuracy remains open to future investigation. Clinical testing can be both reliable and accurate in diagnosis of CTS, but varies with the test and the tester. Because of the high prevalence of carpal tunnel syndrome in occupational disorders, techniques which are used to assist in diagnosis of CTS must be evaluated for their diagnostic performance. Of the methods tested in this study, wrist flexion and vibration perception demonstrated the best overall agreement and accuracy. Variations in test methods, patient population and evaluators will impact on clinical test performance and require further study. However, the reliability and accuracy of clinical tests in differentiating CTS patients from patients with other hand pathologies as described in this study were sufficiently high that specific clinical tests can be considered as useful adjuncts in the diagnosis of CTS. While these tests may be used in the clinic or workplace to evaluate patients, diagnosis of carpal tunnel requires piecing together the clinical history and results of various tests to make a judgement whether the complete picture is one characteristic of carpal tunnel syndrome. No single test provides a completely accurate diagnosis. References Bell-Krotski, J.A. (1990) Light touch-deep pressure using Semmes-Weinstein monofilaments. In: Hunter, J.M., Schneider, L.H., Mackin, E.J. and Callahan, A.D. (Eds.), Rehabilitation of the Hand (third edition), Mosby, St. Louis, pp. 585-593. Cohen, J. (J 960) A coefficient of agreement for nominal scales. Educ. Psych. Meas. 20, 37-46. DeKrom, M.C.T.F.M., Knipchild P.G., Kester A.D.M. and Spaans F. (1990) Efficacy of provocative tests for diagnosis of carpal tunnel syndrome. The Lancet 335,393-395. Dellon, A.L. (1978) Clinical use of vibratory stimuli to evaluate peripheral nerve injury and compression neuropathy. Plas. and Reconstruct. Surg. 62, 466-476. Ditmars, D.M. and Houin, H.P. (1986) Carpal tunnel syndrome. Hand Clinics 2, 525-532. Gellman, H., Gelberman, R.H., Tan, A.M. and Botte, M.J. (1986) Carpal tunnel syndrome: an evaluation of the provocative diagnostic tests. J. Bone Joint Surg. 68, 735-737.

44

J.e. MacDermid et al. / Work 8 (1997) 37-44

Golding. D.N., Rose, D.M. and Selvarajah, K. (1986) Clinical tests for carpal tunnel syndrome: an evaluation. Brit. J. Rheum. 25, 388-390. Heller, L., Ring, H., Costeff H. and Solzi (1986) Evaluation of Tinel's and Phalen's signs in diagnosis of the carpal tunnel syndrome. Eur. Neurol. 25, 40-42. LaBan, M.M., Friedman, N.A and Zemenick, G.A (1986) 'Tethered' median nerve stress test in chronic carpal tunnel syndrome. Arch. Phys. Med. Rehabil. 67, 803-804. laBan, M.M., MacKenzie, J.R. and Zemenick, G.A (1989) Anatomic observations in carpal tunnel syndrome as they relate to the tethered median nerve stress test. Arch. Phys. Med. Rehabil. 70, 44-46. Landis, J.R and Koch, G.G. (1977) The measurement of observer agreement for categorical data. Biometrics 33, 159-174. MacDermid, J.e., Kramer, J.F. and Roth, J.H. (1994) Decision making in detecting abnormal Semmes-Weinstein Monofilament thresholds in carpal tunnel syndrome. J. Hand Ther. 7, 158-162. Mossman, S.S. and Blau, J .N. (1987) Tinel's sign and the carpal tunnel syndrome. Br. Med. J. 294, 680. Novak, e.B., MacKinnon, S.E., Brownlee, R and Kelly, L. (1992) Provocative sensory testing in carpal tunnel syndrome. J. Hand Surg. 17B, 204-208. Phalen, G.S. (1966) The carpal tunnel syndrome: seventeen

years experience in diagnosis and treatment of six hundred and fifty-four. J. Bone Joint Surg. 48A, 211-228. Redmond, M.D. and Rivner, D. (1988) False positive electrodiagnostic tests in carpal tunnel syndrome. Muscle Nerve 11, 511-517. Rudy, T.E., Turk, D.e., Brena, S.F., Sterg, RL. and Brody, M.e. (1990) Quantification of biomedical findings of chronic pain patients: development of an index of pathology. Pain 42, 167-182. Sackett, D.L., Haynes, RB. and Tugwell, P., (1985) Clinical Epidemiology. A Basic Science for Clinical Medicine, Little, Brown and Company, Toronto. Solari, A, Filippini, G., Gagliardi, L., Bevilacqua, L., Amantini, A, Giuliani, G., Messina, e., Rossi, G., Savettieri, G. and Tredici, G. et al. (1989) Interobserver agreement in the diagnosis of multiple sclerosis. Arch. Neurol. 46, 289-292. Szabo, RM., Gelberman, RH. and Dimick, M.D. (1984) Sensibility testing in patients with carpal tunnel syndrome. J. Bone Joint Surg. 66A, 60-64, 1984. Waddell, G., Main, C.J., Morris, E.W., Venner, RM., Rae, P., Sharmy, S.H. and Galloway, H. (1982) Normality and reliability in the clinical assessment of backache. Br. Med. J. 284,1519-1522. Weinstein, S. (1993) Fifty years of somatosensory research: from the Semmes-Weinstein Monofilaments to the Weinstein Enhanced Sensory Test. J. Hand Ther. 6,11-22.

Inter-rater agreement and accuracy of clinical tests used in diagnosis of Carpal Tunnel Syndrome.

A sample of 42 subjects (84 hands) with complaints of pain, numbness and/or tingling were used to study the inter-rater agreement and accuracy of seve...
1MB Sizes 1 Downloads 0 Views