Ta

1

SCIENTIFIC ARTICLE

Scratch Collapse Test for Evaluation of Carpal and

Cubital Tunnel Syndrome

Christine J. Cheng, MD, MPH, Brendan Mackinnon-Patterson, MPH, John L. Beck, MD,Susan E. Mackinnon, MD

Purpose The purpose of this study was to evaluate the clinical usefulness of a new test, the scratchcollapse test, for the diagnosis of carpal tunnel syndrome and cubital tunnel syndrome.

Methods The scratch collapse test was prospectively compared with Tinel’s sign and flexion/nerve compression in 169 patients and 109 controls. One hundred nineteen patients were diagnosedwith carpal tunnel syndrome and 70 patients were diagnosed with cubital tunnel syndrome basedon history, examination, and positive electrodiagnostic test. For the new test, the patient resistedbilateral shoulder external rotation with elbows flexed. The area of suspected nerve compressionwas lightly “scratched,” and then resisted shoulder external rotation was immediately repeated.Momentary loss of shoulder external rotation resistance on the affected side was considered apositive test. The sensitivity, specificity, and predictive values were calculated.

Results For carpal tunnel syndrome, sensitivities were 64%, 32%, and 44% for the scratch collapsetest, Tinel’s test, and wrist flexion/compression test, respectively. For cubital tunnel syndrome,sensitivities were 69%, 54%, and 46% for the scratch collapse test, Tinel test, and elbow flexion/compression test, respectively. The scratch collapse test had the highest negative predictive value(73%) for carpal tunnel syndrome. Tinel’s test had the highest negative predictive value (98%) forcubital tunnel syndrome. Specificity and positive predictive values were high for all of the tests.

Conclusions The scratch collapse test had significantly higher sensitivity than Tinel’s test andthe flexion/nerve compression test for carpal tunnel and cubital tunnel syndromes. Accuracyfor this test was 82% for carpal tunnel syndrome and 89% for cubital tunnel syndrome.This novel test provides a useful addition to existing clinical maneuvers in the diagnosisof these common nerve compression syndromes. ( J Hand Surg 2008;33A:1518– 1524.Copyright © 2008 by the American Society for Surgery of the Hand. All rights reserved.)

Type of study/level of evidence Diagnostic II.

Key words Carpal tunnel syndrome, cubital tunnel syndrome, diagnosis, evaluation, scratchcollapse test.

nsod

HE DIAGNOSIS OF carpal tunnel syndrome andcubital tunnel syndrome remains primarily clin-ical, despite the wide availability of electrodi-

gnostic testing. Numerous studies have shown that

From the Kansas City Bone & Joint Clinic, Overland Park, KS; Division of Plastic and Reconstructive Surgery,Washington University School of Medicine, St. Louis, MO; and the Southern California Orthopedic/SportsMedicine Center, Santa Fe Springs, CA.

Received for publication September 21, 2007; accepted in revised form May 23, 2008.

The authors would like to thank Linda Schultz, RN, PhD, for her assistance in collecting data and

obtaining informed consent from the study and control subjects. d

518 � © ASSH � Published by Elsevier, Inc. All rights reserved.

erve conduction studies are not perfect,1 with reportedensitivities ranging from 49% to 84% in the evaluationf carpal tunnel syndrome.2 Thus, without a gold stan-ard, the clinician relies primarily on patient-reported

o benefits in any form have been received or will be received related directly or indirectly to theubject of this article.

orresponding author: Susan E. Mackinnon, MD, Division of Plastic and Reconstructive Surgery,ashington University School of Medicine, 4990 Children’s Place, Northwest Tower, Suite 1150, St.

ouis, MO 63110; e-mail: mackinnons@wudosis.wustl.edu.

363-5023/08/33A09-0010$34.00/0

Ns

CWL

0

oi:10.1016/j.jhsa.2008.05.022

SCRATCH COLLAPSE TEST 1519

symptoms and a battery of clinical maneuvers to diag-nose these common conditions. Even the most fre-quently used tests for diagnosing carpal tunnel syn-drome, Tinel’s and Phalen’s, vary widely in theirreported sensitivities (44% to 75%) and specificities(48% to 100%).1,3–5 Attempts have been made to clar-ify the diagnostic utility of provocative tests for carpaltunnel syndrome through systematic reviews of theliterature, but the widespread variation in researchmethodology and incomplete reporting has led to in-conclusive recommendations.6,7

For a diagnostic test to be clinically useful, it must beeasily performed, reliable, reproducible, and have highsensitivity and specificity. If the sensitivity of a test islow, then patients who actually have the condition willbe missed. If the specificity is low, then patients who donot have the condition will not be eliminated. If avail-able tests have low to moderate sensitivity and speci-ficity, then a combination of tests will be more accurate.Ideally, the chosen tests should be independent, such assubjective versus objective measurements, to increasediagnostic validity.1 Nearly all of the provocative ma-neuvers for peripheral nerve compression rely on thepatient reporting whether symptoms are elicited, so acombination of these subjective tests would probablynot be independent.

Peripheral nerve injury, including compression neu-ropathy, can cause neuropathic pain characterized byhyperalgesia (increased response to painful stimuli) andallodynia (painful response from normally nonpainfulstimuli).8,9 Clinically, and in rat models, these occurwith skin stimulation both within and outside of theterritory of the compromised nerve.8–10 Painful cutane-ous stimulus has been noted to cause a period of inhi-bition in tonic voluntary muscle activity in humans.This period of electrical silence has been termed thecutaneous silent period.11,12 Although its exact mech-anism is poorly understood, it is generally thought to bean inhibitory spinal reflex that may play a protectiverole in facilitating withdrawal of a limb from potentiallyharmful stimuli.12–14

We introduce a new diagnostic test for carpal tunneland cubital tunnel syndrome, which we have termed thescratch collapse test, in which the examiner scratchesthe patient’s skin lightly over the area of nerve com-pression while the patient performs sustained resistedshoulder external rotation bilaterally. If the patient hasallodynia due to compression neuropathy, a brief loss ofmuscle resistance will be elicited. This test does not relyon patient report, providing a more objective evaluationmethod than most clinical tests for nerve compression.

We compared the scratch collapse test with other com-

JHS �Vol A, No

monly used tests for carpal and cubital tunnel syndromein a prospective clinical study of patients with compres-sion neuropathy symptoms and asymptomatic controls.We hypothesized that the scratch collapse test would bereliable, reproducible, and have comparable sensitivityand specificity to the other existing tests.

MATERIALS AND METHODSFrom January 1, 2004, to December 1, 2005, 169 adultpatients were referred to and evaluated for carpal tunnelsyndrome or cubital tunnel syndrome by a single sur-geon (S.E.M.). The clinical diagnosis of carpal tunnelsyndrome was made if the patient had symptoms ofnumbness, tingling, and/or pain in the median nervedistribution, nocturnal or activity-related symptoms inthe median nerve distribution, and the absence of clin-ical evidence of median nerve compression proximal tothe carpal tunnel. The clinical diagnosis of carpal tunnelsyndrome was supported by electrodiagnostic studies.Patients were diagnosed with cubital tunnel syndrome ifthey had numbness, tingling, and/or pain in the ulnarnerve distribution, weakness or wasting of the ulnar-innervated intrinsic hand muscles, and the absence ofclinical evidence of ulnar nerve compression atGuyon’s canal. The diagnosis of cubital tunnel syn-drome was supported by electrodiagnostic studies. Pa-tients with clinical or electrodiagnostic evidence of cer-vical disk or nerve root disease, those who hadpreviously had surgery for carpal or cubital tunnel syn-drome, and those with normal nerve conduction studieswere also excluded.

There were 79 men and 90 women in the studygroup, with a mean age of 52 years (range, 24–87years). One hundred nineteen patients (49 men, 70women) were diagnosed with carpal tunnel syndrome.Seventy-four patients had bilateral involvement. Sixty-four patients (35 men, 29 women) were diagnosed withcubital tunnel syndrome. Twenty-two patients had bi-lateral involvement. A control group of 109 adult con-trols were recruited from a community church group toparticipate in the study, based on their absence of symp-toms or signs of carpal and cubital syndrome. Electro-diagnostic studies were not obtained in these controlsubjects. All patients in the study group and all of thecontrol group had the following tests performed in thisorder: Tinel’s test over the median nerve at the wristand the ulnar nerve at the elbow; wrist flexion com-bined with direct compression over the median nerve atthe wrist for 60 seconds; elbow flexion combined withdirect compression over the ulnar nerve at the elbow;and the scratch collapse test. Positive results for Tinel’s

test and provocative flexion/compression tests were re-

vember

ered

1520 SCRATCH COLLAPSE TEST

corded if the patient experienced paresthesias in theappropriate nerve distribution. Demographic data forthe study and control groups are given in Table 1.

The new clinical test was performed with the patientfacing the examiner, with arms adducted, elbowsflexed, and both hands outstretched with wrists at neu-tral position (Fig. 1). The patient was asked to performsimultaneous resisted bilateral shoulder external rota-tion, keeping the arms abducted. The examiner gentlypushed against both of the patient’s forearms, askinghim or her to sustain steady resistance. With fingertips,the examiner then scratched or swiped the skin overly-

TABLE 1. Demographic Data of Study and Control

Controls

n 109

Men/women, n (%) 48/61 (44/56)

Mean age in years (range) 52 (19–92)

Mean BMI in kg/m2 (range) 26 (18–38)

No. affected hands, dominant/nondominant N/A

Mean symptom duration in months N/A

Diabetes mellitus, n (%) N/A

Tobacco use, n (%) N/A

Workers’ compensation, n (%) N/A

Litigation, n (%) N/A

BMI, body mass index; N/A, not applicable.*p � .05 compared with rest of experimental subjects.†p � .01 compared with control subjects.

FIGURE 1: The figure illustrates the scratch collapse test. Thewith arms adducted, elbows flexed, and hands outstretched wadduction/internal rotation to the forearms applied by the exfingertips over the course of the compressed nerve (ulnar nervetemporary loss of the patient’s external resistance tone is consid

ing the course of the potentially compressed nerve. The

JHS �Vol A, No

median nerve was scratched over the carpal tunnel atthe volar wrist. The ulnar nerve was scratched over thecubital tunnel at the medial elbow. A positive scratchcollapse test was recorded for the median or ulnar nerveif the patient demonstrated a momentary loss of exter-nal resistance tone on the affected side after “scratch-ing” over the carpal tunnel or cubital tunnel, respec-tively. This loss of muscle resistance was quite brief,with the patient regaining strength essentially immedi-ately with repeat resistance testing. However, the testcould be repeated successfully without evidence of fa-tigue or habituation (this video may be viewed at the

ups

ll SubjectsSubjects With Carpal

Tunnel SyndromeSubjects With Cubital

Tunnel Syndrome

169 119 64

/90 (47/53) 49/70* (41/59) 35/29 (55/45)

52 (24–87) 53 (26–87) 49* (24–86)

0† (17–58) 31† (17–58) 29† (18–45)

57/112 95/98 45/41

36 40 42

22 (13) 19 (16) 6 (9)

40 (24) 23 (19) 18 (28)

64 (38) 42 (35) 34 (53)

14 (8) 9 (8) 6 (9)

ient and examiner are labeled. The patient faces the examinerrists at neutral. Step A: The patient resists bilateral shoulderer. Step B: Next, the examiner “scratches” or swipes withbow illustrated). Step C: Step A is immediately repeated. Briefa positive scratch collapse test.

Gro

A

70

3

patith waminat el

Journal’s Web site, www.jhandsurg.org).

vember

SCRATCH COLLAPSE TEST 1521

Data analysis used 2 � 2 contingency tables tocalculate the sensitivity and specificity of each test. Thepresence or absence of each nerve compression syn-drome was recorded in the rows, and the results of eachtest were recorded in the columns. Sensitivity is thelikelihood that the test is positive if the condition ispresent. Specificity is the likelihood that a test is nega-tive if the condition is absent. The prevalence of carpaltunnel syndrome and cubital syndrome was determinedin the study population, and the associated positive andnegative predictive values were calculated for each testusing 2 � 2 contingency tables. Prevalence is the per-centage of people who are affected in the study group.Positive predictive value is the likelihood that the con-dition is present if the test is positive. Negative predic-tive value is the likelihood that the condition is absent ifthe test is negative. These reflect the usefulness of a testand are dependent on prevalence in the population.

TABLE 2. Clinical Tests Results Summary

Clinical Test

Study Group

Extremities WithPositive Result

n, (%)

ExtremNega

n

Tinel sign at wrist 70 (29) 14

Median nerve flexion/compressionat wrist

90 (38) 11

Scratch collapse test at wrist 140 (59) 7

Tinel sign at elbow 61 (48) 5

Ulnar nerve flexion/compressionat elbow

51 (40) 6

Scratch collapse test at elbow 76 (59) 3

TABLE 3. Sensitivity, Specificity, and Predictive Va

Sensitivity(%)

Specificity(%)

Wrist Tinel 32* 99

Wrist provocative 44* 99

Wrist scratch collapse 64 99

Elbow Tinel 54* 99

Elbow provocative 46* 99

Elbow scratch collapse 69 99

Prevalence of carpal tunnel syndrome � 70%; prevalence of cubital t*p � .001 compared with scratch collapse test.

Screening tests are more useful when the condition is

JHS �Vol A, No

more common. The McNemar test of correlated pro-portions was used to compare the sensitivity and spec-ificity of the diagnostic tests.15 Chi-square analysis wasused to evaluate all categorical variables, and indepen-dent t-test was used to evaluate all continuous variablesbetween the study and control groups. Kappa statisticwas calculated for agreement between clinical tests andfor interrater reliability of the scratch collapse test. Datawere analyzed with statistical software (SPSS; SPSS,Inc., Chicago, IL).

RESULTSDetails of the study and control groups are shown inTables 1, 2, and 3. The study subjects had significantlyhigher body mass index than the control subjects (p �.01). All other demographic parameters were similarbetween the 2 groups. The patients diagnosed withcarpal tunnel syndrome were more likely to be female

ExtremitiesUnknown/

Not Recorded

Control Group

s Withesult

)

Extremities WithPositive Result

n, (%)

Extremities WithNegative Result

n, (%)

2) 20 3 (1) 215 (99)

9) 32 2 (�1) 216 (�99)

3) 20 2 (�1) 216 (�99)

0) 16 2 (�1) 216 (�99)

7) 17 0 218

7) 18 1 (�1) 217 (�99)

of Clinical Tests

PositivePredictive Value

(%)

NegativePredictive Value

(%)Accuracy

(%)

96 59 65

98 65 72

99 73 82

97 98 84

96 78 81

99 86 89

syndrome � 32%.

itietive R, (%

8 (6

6 (4

8 (3

1 (4

0 (4

4 (2

lues

unnel

(59%, p � .03), have bilateral involvement (62%, p �

vember

1522 SCRATCH COLLAPSE TEST

.002), and were less likely to report an injury associatedwith their diagnosis (26%, p � .02). The patients diag-nosed with cubital tunnel syndrome were younger thanthe rest of the study group (mean age 49 years, p � .05),more likely to report an injury associated with theirdiagnosis (50%, p � .001), and were more likely tohave workers’ compensation insurance (53%, p �.005).

The scratch collapse test was reproducible, withexcellent interrater reliability (� � .98, p � .001).For patients with carpal tunnel syndrome, the resultsof the 3 clinical tests showed only slight to fairagreement, with the least agreement between thescratch collapse and wrist flexion/nerve compressiontests (� � .18, p � .001). Correlation between theclinical tests was much better in patients with cubitaltunnel syndrome. The results of the 3 clinical testsshowed moderate to substantial correlation, with theleast agreement between the scratch collapse andelbow flexion/nerve compression tests (� � .458, p� .001). Chi-square for trend analysis of the clinicaltests showed no relationship between duration ofsymptoms and frequency of positive test results inany of the study subgroups.

The scratch collapse test was more sensitive than anyof the other clinical tests in both carpal tunnel syndromeand cubital tunnel syndrome patients, as shown in Table3. For carpal tunnel syndrome, the scratch collapse testhad better sensitivity (64%) compared with that of wristflexion/nerve compression test (44%; p � .001) andTinel test (32%; p � .001). For cubital tunnel syn-drome, the scratch collapse test was more sensitive(69%) than elbow flexion/nerve compression test (46%;p � .001) and Tinel test (54%; p � .002). All of theclinical tests were highly specific (99%), without nota-ble differences between tests.

In our study population, the prevalence of carpaltunnel syndrome was 70%. The positive predictive val-ues for the clinical tests were high, with the scratchcollapse test having the highest positive predictivevalue (99%), followed by the wrist flexion/nerve com-pression test (98%) and Tinel test (96%). Negativepredictive values for the tests for carpal tunnel syn-drome were also highest for the scratch collapse test(73%), followed by the wrist flexion/compression test(65%) and Tinel test (59%). The prevalence of cubitaltunnel syndrome was much lower in our study group(32%). Nevertheless, the positive predictive values forthe clinical tests remained high, with the scratch col-lapse test having the highest positive predictive value(99%), followed by Tinel test (97%) and elbow flexion/

nerve compression test (96%). Negative predictive

JHS �Vol A, No

value, however, was highest for Tinel test (98%), fol-lowed by the scratch collapse test (86%) and elbowflexion/nerve compression test (78%).

DISCUSSIONThere is no perfect gold standard for the diagnosis ofcompression neuropathy. For carpal tunnel syndrome, acombination of positive electrodiagnostic findings andclinical symptoms combined with positive clinical test-ing is believed to be the most accurate method.16 Ac-curacy of diagnosis can be improved by combiningtests, especially when the tests have limited sensitivityand/or specificity. Systematic review of the literature onthe diagnosis of carpal tunnel syndrome found reportedsensitivities for Tinel’s test to range from 45% to 75%and wrist flexion/compression test to range from 49% to89%.6,7 The analogous tests have performed better indiagnosing cubital tunnel syndrome, with reported sen-sitivities for Tinel’s test and elbow flexion/compressiontest of 70% and 91%, respectively,17 although 20% to30% false-positive rates have been reported in asymp-tomatic individuals.18,19 In our study, the sensitivities ofthe Tinel’s and flexion/compression tests for both car-pal and cubital tunnel syndrome were in the lower rangeof previously reported values. Because we only in-cluded study subjects who had positive electrodiagnos-tic findings, some subjects who had mild nerve com-pression with normal electrodiagnostic findings mighthave been excluded, potentially increasing the false-negative rate and decreasing the computed sensitivity.On the other hand, the specificities of the same testswere high, which reflects our use of an asymptomaticcontrol group that did not undergo electrodiagnostictesting. These patients would be expected to have neg-ative results for most diagnostic tests, which wouldfalsely elevate specificity.7 Electrodiagnostic testingwas not performed in the control group because carpaland cubital tunnel syndromes are believed to be clinicaldiagnoses. The diagnosis would be unclear for individ-uals without signs and symptoms of compression neu-ropathy but with abnormal electrodiagnostic studies.This combination of findings has been considered byexpert panel consensus for carpal tunnel syndrome tohave poor positive predictive value.16

The high prevalence (70%) of carpal tunnel syn-drome in our subjects reflects their referral to a handsurgeon’s practice and is much higher than the re-ported prevalence of 5% to 15% in population-basedstudies.3,20 Positive and negative predictive values ofscreening tests are dependent on disease prevalenceand can markedly affect the utility of a test. All of the

clinical tests in this study had positive predictive

vember

SCRATCH COLLAPSE TEST 1523

values greater than 95%, which is high comparedwith previously reported values of 79% to 91%.21

These values would be expected to be lower in apopulation with lower disease prevalence, such aspatients who might present to a general practitioner’soffice for evaluation.

The scratch collapse test had significantly (p � .001)higher sensitivity than that of Tinel test and of theflexion/nerve compression test for both carpal tunneland cubital tunnel syndrome patients. The accuracy ratefor this test was 82% for diagnosing carpal tunnelsyndrome and 89% for diagnosing cubital tunnel syn-drome. There is definitely a learning curve associatedwith this test. The patient is instructed to only match theforce of the examiner’s resistance and not try to push ashard as they can. The patient’s shoulders should remainadducted, with their elbows at their sides. The examinergently pushes against the patient’s forearms and not thehands or wrists. The test seems to be quite anatomicallyexact and will be negative if stimulation does not occurover the nerve. Thus, the “scratch” needs to be along thecourse of the nerve. If the surface landmarks for a nerveare unclear to the examiner, then all of the fingersshould be used to “swipe” over the general location ofthe nerve, rather than only “scratching” over the nerve.

Because it is a test that does not rely on the patient’ssubjective report of whether symptoms are elicited, thescratch collapse test provides a more objective additionto other clinical maneuvers. Combining tests that tapinto different aspects of a condition helps meet theassumption of conditional independence, which in-creases the diagnostic validity of the full complement oftests when a gold standard is not available.1 Becausethis is a novel test, the patient has not likely had itadministered before, making it potentially more diffi-cult to feign a positive result. We have found this testuseful in patients with potential secondary gain as acomplicating factor. Those patients who exhibit a pos-itive scratch collapse test when other tests have beenequivocal or negative can feel validated that a diagnosishas been established. The test can also be repeated in apatient in rapid succession, without any observed fa-tigue, so several trials can be performed for verification.In the setting of multiple sites of peripheral nerve com-pression, the most severely affected site will collapse,whereas the less severe ones may not. This can behelpful when trying to assign the appropriate treatmentoptions for each site.

Rat models have been developed for the study ofneuropathic pain, using chronic constriction of eitherthe sciatic nerve9,10 or spinal nerves.8 Mechanical allo-

dynia (painful response from normally nonpainful stim-

JHS �Vol A, No

uli) and heat hyperalgesia (increased response to painfulstimuli) within the cutaneous territory of the injurednerves were produced after nerve injury.8–10 The spon-taneous and elicited electrical axonal activity disap-peared with chemical or physical inactivation of thedorsal root ganglia.9,10 Abnormal electrical responsewas also noted with stimulation of adjacent nerve ter-ritories or even in the contralateral limb, which suggestsa component of central sensitization.10 This experimen-tal animal response to chronic constriction neuropathyseems similar to the exaggerated neurogenic pain re-sponse that can occur in humans.

In response to painful cutaneous stimulation, humansalso exhibit a temporary inhibition of tonic voluntarymuscle activity, with a brief period of electromyo-graphic silence. This cutaneous silent period was firstdescribed by Hoffmann in 192211 and has been re-corded in healthy volunteers as well as in patients withcentral nervous system disorders such as Friedreich’sataxia22 and complete spinal cord injury.12 MaximalEMG suppression occurs only with painful stimulus,begins within 100 milliseconds after the stimulus, lasts50 to 100 milliseconds, does not show habituation, andhas been recorded in various upper- and lower-extremitymuscles after stimulation of digital nerves.11–14,22 Theexact mechanism is poorly understood, with synergis-tic13 and antagonistic11 muscle groups being simulta-neously affected and the inhibitory period is more rapidthan the voluntary muscle relaxation.13 Most investiga-tors feel that it is mediated primarily by slow-conducting A-� fibers and may be a protective inhibi-tory spinal reflex.11-14,22

We believe that the scratch collapse test uses thephenomenon of the cutaneous silent period to brieflyinhibit tonic shoulder external rotation as a response tonoxious stimulus of the skin overlying a chronicallyconstricted nerve. We have tested and demonstrated itsusefulness in the clinical diagnosis of carpal tunnelsyndrome and cubital tunnel syndrome. It has shownsimilar or improved diagnostic ability when comparedwith other widely used clinical tests. Although it is easyto perform and has excellent reproducibility betweenexaminers, a learning curve is involved in using thescratch collapse test. The “scratch” stimulus needs to bedirectly over the course of the compromised nerve toelicit a consistent positive response. Further studies totry to elucidate the underlying physiologic mechanismof this test would be useful. Whereas the patients in thisstudy had positive electrodiagnostic findings, this test ismost useful as an additional diagnostic tool in individ-uals with positive symptoms but negative electrodiag-

nostic studies. Future investigations may address the

vember

1524 SCRATCH COLLAPSE TEST

utility of this test for compression neuropathies in otheranatomic sites or for other painful conditions.

REFERENCES1. LaJoie AS, McCabe SJ, Thomas B, Edgell SE. Determining the

sensitivity and specificity of common diagnostic tests for carpaltunnel syndrome using latent class analysis. Plast Reconstr Surg2005;116:502–507.

2. American Academy of Neurology, American Association of Elect-rodiagnostic Medicine, and American Academy of Physical Medi-cine and Rehabilitation. Practice parameter for electrodiagnosticstudies in carpal tunnel syndrome. Neurology 1993;43:2404–2405.

3. Katz JN, Larson MG, Sabra A, Krarup C, Stirrat CR, Sethi R, et al.The carpal tunnel syndrome: diagnostic utility of the history andphysical examination findings. Ann Intern Med 1990;112:321–327.

4. Durkan JA. A new diagnostic test for carpal tunnel syndrome. J BoneJoint Surg 1991;73A:535–538.

5. Williams TM, Mackinnon SE, Novak CB, McCabe S, Kelly L.Verification of the pressure provocative test in carpal tunnel syn-drome. Ann Plast Surg 1992;29:8–11.

6. Massy-Westropp N, Grimmer K, Bain G. A systematic review of theclinical diagnostic tests for carpal tunnel syndrome. J Hand Surg2000;25A:120–127.

7. MacDermid JC, Wessel J. Clinical diagnosis of carpal tunnel syn-drome: a systematic review. J Hand Ther 2004;17:309–319.

8. Sheen K, Chung JM. Signs of neuropathic pain depend on signalsfrom injured nerve fibers in a rat model. Brain Res 1993;610:62–68.

9. Tal M, Eliav E. Abnormal discharge originates at the site of nerveinjury in experimental constriction neuropathy (CCI) in the rat. Pain1996;64:511–518.

10. Sotgiu ML, Biella G. Contribution of central sensitization to thepain-related abnormal activity in neuropathic rats. Somatosens Mot

Res 2000;17:32–38.

11. Uncini A, Kujirai T, Gluck B, Pullman S. Silent period induced by

JHS �Vol A, No

cutaneous stimulation. Electroencephalogr Clin Neurophysiol 1991;81:344–352.

12. Logigian EL, Plotkin GM, Shefner JM. The cutaneous silent periodis mediated by spinal inhibitory reflex. Muscle Nerve 1999;22:467–472.

13. Leis AA, Stokic DS, Fuhr P, Kofler M, Kronenberg MF, Wissel J, etal. Nociceptive fingertip stimulation inhibits synergistic motoneuronpools in the human upper limb. Neurology 2000;55:1305–1309.

14. Kofler M. Functional organization of exteroceptive inhibition fol-lowing nociceptive electrical fingertip stimulation in humans. ClinNeurophysiol 2003;114:973–980.

15. Fleiss JL. Statistical methods for rates and proportions. 3rd ed.Hoboken, NJ: John Wiley & Sons, 2003:374–380.

16. Rempel D, Evanoff B, Amadio PC, deKrom M, Franklin G, Franz-blau A, et al. Consensus criteria for the classification of carpal tunnelsyndrome in epidemiologic studies. Am J Public Health 1998;88:1447–1451.

17. Novak CB, Lee GW, Mackinnon SE, Lay L. Provocative testing forcubital tunnel syndrome. J Hand Surg 1994;19A:817–820.

18. Kuschner SH, Ebramzadeh E, Mitchell S. Evaluation of elbowflexion and Tinel tests for cubital tunnel syndrome in asymptomaticindividuals. Orthopedics 2006;29:305–308.

19. Rayan GM, Jensen C, Duke J. Elbow flexion test in the normalpopulation. J Hand Surg 1992;17A:86–89.

20. de Krom MC, Knipschild PG, Kester AD, Thijs CT, Boekkooi PF,Spaans F. Carpal tunnel syndrome; prevalence in the general popu-lation. J Clin Epidemiol 1992;45:373–376.

21. Wiesman IM, Novak CB, Mackinnon SE, Winograd JM. Sensitivityand specificity of clinical testing for carpal tunnel syndrome. Can JPlast Surg 2003;11:70–72.

22. Serrao M, Parisi L, Pierelli F, Rossi P. Cutaneous afferents mediat-ing the cutaneous silent period in the upper limbs: evidences for arole of low-threshold sensory fibres. Clin Neurophys 2001;112:

2007–2014.

vember