Research paperImpact of technical variations on the ring-finger test for carpal tunnelsyndromeDaniel Gregor Schulzea,b,g,⇑, Karl-Christian Nordbyc, Milada Cvancarova Småstuend,e, Thomas Clemmc,Margreth Grotled,g, John Anker Zwarta,b,g, Kristian Bernhard Nilsena,f,gaDepartment of Neurology, Oslo University Hospital, Oslo, NorwaybInstitute of Clinical Medicine, Faculty of Medicine, University of Oslo, Oslo, NorwaycNational Institute of Occupational Health, Oslo, NorwaydOslo and Akershus University College, Oslo, NorwayeInstitute of Basic Medical Sciences, University of Oslo, Oslo, NorwayfDepartment of Neuromedicine and Movement Science, Norwegian University of Science and Technology, Trondheim, NorwaygResearch and Communication Unit for Musculoskeletal Health, Oslo University Hospital, Norwayarticle infoArticle history:Received 4 February 2019Received in revised form 19 November 2019Accepted 23 November 2019Available online 11 December 2019Keywords:Nerve conduction studyRing finger testMedian-ulnar nerves comparison testsabstractObjective:To assess if recording the sensory latencies of the median and ulnar nerves one-by-one (con-secutive) or at the same time (simultaneous) in the ring-finger test for carpal tunnel syndrome (CTS) willshow equivalent results or if it will lead to a different clinical classification of patients.Methods:We assessed the limits of agreement between the simultaneous and the consecutive methodbased on the median- ulnar sensory latency difference derived by both methods in 80 subjects and com-pared the number of minimal CTS cases identified by the two methods.Results:Limits of agreement ranged from0.23 to 0.29 ms. A significantly higher proportion of subjectswith minimal CTS (only detectable by using the comparison test) was found using the simultaneousmethod (n = 8 and 2, respectively; p = 0.03).Conclusion:The two methods have a poor to moderate agreement as indicated by the range of the limitsof agreement (0.5 ms).Significance:Even small methodological changes to the ring-finger test can lead to results with differentclinical meaning in the same individual and one should be aware of which method was used when inter-preting results.Ó2019 International Federation of Clinical Neurophysiology. Published by Elsevier B.V. This is an openaccess article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).1. IntroductionIn patients with clinically suspected carpal tunnel syndrome(CTS), nerve conduction studies (NCS) can demonstrate reducedmedian nerve conduction velocity (Bland, 2007; Werner andAndary, 2011; Deniz et al., 2012). However, in a considerable pro-portion of patients with clinically suspected CTS the sensory nerveconduction velocity in the median nerve is within the normalrange, but reduced function of the median nerve can be demon-strated if the median and the ulnar nerve are directly comparedand a latency difference above a defined cut-off value is found(Padua et al., 1997). A frequently used test to compare the ulnarand the median nerve in these patients is the ring finger test(Uncini et al., 1989), which compares directly the sensory latenciesof the ulnar and the median nerve from the 4th finger. This test hasthe advantage that the patient serves as their own control and it istherefore less dependent on age, sex and hand temperature thantests which compare conduction velocities to normal values. Thetest s diagnostic precision varies considerably (Wang and Yan,2013; Kouyoumdjian et al., 2002; Capone et al., 1998). Differentpractical approaches to the test are in use, and this might partlyexplain the varying diagnostic precision. The orthodromic ring fin-ger test may be performed in a consecutive manner where onenerve is recorded first, then the setup is rearranged and the othernerve recorded. The test can also be performed by recording thesensory potential of the median and the ulnar nerves simultane-ously, thus saving time and potentially making the examinationless unpleasant for the patient. It is unclear, if these two methodscan be used interchangeably.https://doi.org/10.1016/j.cnp.2019.11.0052467-981X/Ó2019 International Federation of Clinical Neurophysiology. Published by Elsevier B.V.This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).Abbreviations:CTS, carpal tunnel syndrome; EMG, electromyography; HAVS,hand arm vibration syndrome; NCS, nerve conduction study.⇑Corresponding author at: Department of Neurophysiology, Oslo UniversityHospital, Postboks 4950 Nydalen, 0424 Oslo, Norway.E-mail address:gresch@ous-hf.no(D.G. Schulze).Clinical Neurophysiology Practice 5 (2020) 23–29Contents lists available atScienceDirectClinical Neurophysiology Practicejournal homepage: www.elsevier.com/locate/cnp
We thus aimed to assess if the two methods showed acceptableagreement and hypothesized that the number of correctly identi-fied minimal carpal tunnel syndrome cases (defined as clinicalsymptoms and NCS findings indicative of median nerve entrap-ment present) would not differ between the methods.2. Methods2.1. SubjectsThe study population is that of a Norwegian health survey ofrock drillers and road workers in a Norwegian construction com-pany. The NCS were performed between November 2015 and June2016. The data were collected with the purpose of assessingperipheral nerve damage in road workers and rock drillers withexposure to hand-arm vibration. The study was approved by theRegional Ethics Committee (REK 2013/1031). All participants pro-vided a written consent. Data were collected assuming that thesimultaneous and the consecutive method would produce equiva-lent results. We performed both methods as a quality check.2.2. Nerve conduction studiesAll NCS were performed using Focus Keypoint.Net EMG equip-ment (Natus medical incorporated, Pleasanton, USA). Skin temper-ature was measured with an Exergen dermaTemp handheldinfrared thermographic scanner (Exergen corporation, Watertown,MA, USA) and maintained at over 30 degrees C in all cases. If skintemperature was below 30 degrees, the participants held theirhands in warm water for 2–5 min and temperature was measuredagain. We used the same type of electrodes for the consecutive andthe simultaneous recording. For recording, we used pre- gelled dis-posable surface electrodes (Alpine biomed, Skovlunde, Denmark).For stimulation, we used a hand-held stimulation bar with fixedinter-electrode distances. The stimulation electrodes consist of felttips which had a diameter of 7.5 mm and were soaked in salinesolution. When being placed on the proximal phalanx and usingsupramaximal stimulation, they allow simultaneous stimulationof the median and ulnar nerves. In addition, an E0 ‘‘ground elec-trode” was placed on the hand being tested. We performed the fol-lowing protocol: We started with the motor study of the ulnar andmedian nerves. Then we performed sensory study of ulnar nerve(palm, 4th and 5th finger, starting with the 4th) and then sensorystudy of the median nerve (palm, 2nd, 3rd, 4th finger, starting withthe 4th). The measurements we hereby obtained from the 4th fin-ger constitute the consecutive method. We then performed thesimultaneous method by adding a pair of electrodes for the ulnarnerve and placed these at the previously marked and used record-ing spot. The extra pair of electrodes used for recording the ulnarnerve in the simultaneous method was connected to a separateamplifier. In summary, the simultaneous method consists of sim-ply adding a pair of recording electrodes for the ulnar nerve afterhaving performed the consecutive median nerve sensory testing,so that we ended up with one pair for the median and one pairfor the ulnar nerve (Fig. 1).2.3. Procedures common to both methodsWe performed both methods orthodromically. A fixed distanceof 14 cm was used from the stimulation point (cathode position) atthe proximal phalanx of the ring finger to the active registrationelectrode over the median nerve and the ulnar nerve at the wrist.We measured the distances, marked all points for electrode place-ment and used the same recording and stimulation points for bothmethods. We used the same type of recording and stimulationelectrodes for both nerves and in both methods. Recording elec-trodes were placed with an inter-electrode distance of 3 cm. Weachieved supramaximal stimulation by increasing stimulus inten-sity by 20% after the amplitude reached its maximum. We thengave a series of 10 stimuli with pulse duration of 0.1 ms each.Motor amplitudes were automatically measured from baseline topeak. Sensory amplitudes were measured from the negative peakto the intersection of a line drawn from the first to the last positivepeak. Latencies were calculated based on the peak of the negativedeflection (seeFig. 2a and b for examples of the simultaneousrecording). Conduction velocities were measured based on theonset of the response. Low (at 20 Hz) and high frequency filter(at 10 kHz) settings were employed.2.4. Description of the two methodsWe always performed the consecutive method first: we placedthe recording electrodes at the measured distance for the ulnarcommon stimulation site median nerve recording siteulnar nerve recording siteequal distanceFig. 1.Recording set-up for the simultaneous method.1 ms3μVulnar nervemedian nerveulnar nervemedian nerve1 ms3μVABFig. 2.Examples of recordings with the simultaneous method in two differentpatients (a, b).24D.G. Schulze et al./Clinical Neurophysiology Practice 5 (2020) 23–29
nerve, and performed the sensory ulnar nerve stimulation and reg-istration. We then removed the recording electrodes, and repeatedthe procedure for the sensory median nerve. For the simultaneousmethod, we placed the recording electrodes at the given distancesfor both nerves at the wrist. We then stimulated at the phalanx ofthe ring finger and recorded the sensory responses of the medianand the ulnar nerves simultaneously.2.5. Analysis of the NCS dataFor each subject, the ulnar and median nerve sensory latencies,amplitudes and conduction velocities were extracted using bothmethods. We graded the NCS findings according to Padua et al.(Padua et al., 1997) as either normal, minimal median nerveentrapment, mild median nerve entrapment, moderate mediannerve entrapment or severe median nerve entrapment. Minimalmedian nerve entrapment was defined as a sensory latency differ-ence0.5 ms between the median and the ulnar nerve in theabsence of other findings; mild median nerve entrapment wasdefined as reduced sensory conduction velocity in the mediannerve, moderate median nerve entrapment as additional increasedmotor distal latency, severe median nerve entrapment as theabsence of sensory responses. For the minimal grade, we usedthe absolute latency difference, for the other grades (mild, moder-ate and severe), results were considered abnormal according tonormal values for the laboratory. These normal values are derivedfrom an unpublished northern joint-effort and are integrated intothe software. They are adjusted for age, sex and height.3. CTS diagnosisClinical criteria for the diagnosis of CTS were based on the pres-ence of at least two of the following symptoms: nocturnal episodesof paresthesia in the median nerve distribution, numbness in thefingers innervated by the median nerve, alleviation of symptomsby shaking of the limb. Positive family history of CTS (parents orsiblings) and weakness in the hand as supportive criteria couldreplace one of the criteria above if only one of the obligatory crite-ria were met. Subjects were diagnosed with CTS if the clinical cri-teria were met and if at the same time NCS findings indicative of atleast minimal median nerve entrapment were present.4. Statistical analysisFor all participating subjects, a one sample Student‘st-test wasused to test whether the difference between the two methods isequal to zero. The normality assumption was tested by means ofvisual inspection using a histogram and Q-Q plots and was foundto be satisfied. We used a Bland Altman plot to compare the tworecording methods. Using the difference between latency of theulnar and of the median nerve, we calculated both the differencebetween the simultaneous recording and the consecutive record-ing and the mean of the simultaneous and the consecutive methodfor each subject. We then plotted these two values against oneanother. The limits of agreement were calculated as 1.96 * standarddeviation of the measurement differences on either side of themean. The mean and the upper and lower limit of agreement wereincluded as horizontal lines in the figures (Figs. 3 and 4). P-values < 0.05 were considered statistically significant.When assessing the ability of the two methods to identify caseswith minimal CTS, we used the clinical criteria for CTS as a goldstandard. We used a McNemar test to compare the number of sub-jects with a minimal CTS diagnosis (defined as both clinical criteriamet and median-ulnar latency difference0.5 ms) derived by thetwo methods. All analyses were performed using IBM SPSS v 24software.5. Results5.1. Sample descriptionWe examined the left and right hands of 80 subjects aged 22 to68 years (median 43 years). All subjects were men. 72 subjects hadtechnically satisfying NCS data on at least one hand.5.2. Nerve conduction findingsThe results of the nerve conduction studies are summed up inTables 1 and 2. The number of subjects with minimal median nerveentrapment (defined as median- ulnar sensory latency differences>0.5 ms without further NCS findings) was significantly differentbetween the two methods in the right hand (11 and 4, respectively,p = 0.01) and borderline statistically different in the left hand (11and 6, respectively, p = 0.06). A minimal median nerve entrapmentwas found in 16 subjects with at least one method in at least onehand (Table 3).In the right hand, the mean median-ulnar latency differencewas borderline statistically different between the two methods(p = 0.06; latency difference of 0.36 ms and 0.32 ms respectively);in the left hand, no statistically significant difference was found(p = 0.14; latency difference of 0.32 ms and and 0.31 ms, respec-tively). SeeFig. 5. We have not received negative feedback fromthe patients regarding unpleasantness of the simultaneousstimulation.5.3. CTS casesThe ability of the two methods to identify cases with minimalCTS is given inTables 3–5. The number of subjects diagnosed withminimal CTS (defined as median-ulnar sensory latency >0.5 ms asthe only NCS finding and clinical criteria for CTS fulfilled) was sig-nificantly different (8 and 2, respectively, p = 0.03) for the twomethods in both hands combined (n = 16) and in the right handalone (n = 11). In the left hand (n = 11), the simultaneous methodidentified 2, the consecutive 1 subject with minimal CTS. All sub-jects with minimal CTS in the left hand had bilateral CTS with morepronounced symptoms in the right hand.5.4. Bland Altman analysisIn the right hand, the mean difference in latency between themedian and the ulnar nerve with the consecutive method was0.32 ms (SD 0.34 ms), and the mean difference between the twomethods was 0.03 ms with a standard deviation of 0.13 ms.Ninety-five percent of measurements were within the limits ofagreement from0.23 ms to 0.29 ms (Fig. 3).In the left hand the mean difference between the two methodswas 0.023 ms, the mean latency difference with the consecutivemethod was 0.31 ms (SD 0.51 ms). Ninety-five percent of measure-ments were within the limits of agreement from0.23 ms to0.27 ms (Fig. 4).6. DiscussionOur data reveal that the agreement between the simultaneousand the consecutive method of conducting the ring-finger test isclinically not acceptable and that the two methods produce resultswith different clinical implications.D.G. Schulze et al./Clinical Neurophysiology Practice 5 (2020) 23–2925
We assessed agreement based on the limits of agreement andthe mean difference between the two methods. The limits of agree-ment were relatively wide. These ranged from0.23 ms to 0.29 msin the right hand and from0.23 ms to 0.27 ms in the left hand.Assessing the limits of agreement between the two methods isthe recommended approach when comparing two instrumentswhich measure the same quantitative value and is regarded assuperior to a straight forward computation of a correlation coeffi-cient (Bland and Altman, 1986; Watson and Petrie, 2010). Limits ofagreement show to which degree measurements of the same quan-titative value differ between two different instruments in the sameindividual. The limits of agreement illustrate the differencebetween the methods as a range in the same unit as the measuredquality, and allow to judge if this difference is clinically acceptable,or if it will lead to a different diagnosis simply because a differentinstrument was chosen. The limits of agreement indicate that ameasurement from one method might differ by0.23 ms to0.29 ms from the measurement obtained by the other method(for the right hand; and between0.23 ms and 0.27 ms for the lefthand). To put these values into clinical perspective, the most oftenused cut-off value to demonstrate a minimal median nerve entrap-ment and thus a CTS diagnosis with the ring-finger test is 0.5 ms(Uncini et al., 1989). We argue that the cut-off value is so closeto the limits of agreement, that patients might receive a CTS diag-nosis with one method but not the other. The cut-off value hasbeen a subject of discussion before (Logigian et al., 2014; Noderaet al., 2003; Preston et al., 1994; Wang and Yan, 2013;Kouyoumdjian et al., 2002; Capone et al., 1998), especially inpatients with diabetes or with ambiguous clinical symptoms(Salerno et al., 1998; Rivner et al., 2001). The mean differencebetween the two methods was small (0.03 ms) considering theclear difference in how many CTS cases (defined as NCS findingsand clinical symptoms present) the two methods identified. In apopulation like the present, increased sensory latencies of themedian nerve, as high as the cut-off value are more prevalent thanin the general population (Armstrong et al., 2008; Salerno et al.,mean of the latency difference measured by the simultaneous and consecutive method2,001,501,00,50,00-,50difference between the latency difference measured by the simultaneous and consecutive method,40,20,00-,20-,40Fig. 3.Bland Altman plot for the right hand showing on the x-axis the mean of the simultaneous and the consecutive measurement, and on the y-axis the difference betweenthe simultaneous and consecutive measurement. The average difference is indicated as a horizontal line at 0.03. The upper and the lower limits of agreement are indicated ashorizontal lines at 0.29 ms and0.23 ms, respectively.mean of the latency differences measured by the simultaneous and consecutive method2,001,501,00,50,00-,50 difference between the latency differences measuredby the simultaneous and consecutive method ,40,20,00-,20-,40-,60Fig. 4.Bland Altman plot for the left hand showing on the x-axis the mean of the simultaneous and the consecutive measurement, and on the y-axis the differencebetweenthe simultaneous and consecutive measurement. The average difference is indicated as a horizontal line at 0.02. The upper and the lower limits of agreement are indicated ashorizontal lines at 0.27 ms and0.23 ms, respectively.26D.G. Schulze et al./Clinical Neurophysiology Practice 5 (2020) 23–29
1998). It can be argued that in this population already a smallmean difference between the methods is enough to produce thedifferent sensitivities of the methods. In addition, one must con-sider that individual differences of up to +0.29 ms (or0.23 ms)are to be expected. This probably explains why we found a highernumber of individuals with a latency difference of0.5 ms withthe simultaneous method, and at the same time a small, not signif-icant average difference between the methods. However, this dif-ference trended towards being significant (p = 0.06). In the lefthand, we did not find a significant difference regarding the numberof CTS cases identified by the two methods. We assume that this isdue to the low number of cases with CTS in the left hand in thisgroup and that the analysis thus lacked the power necessary todemonstrate a significant difference. We rationalize this assump-tion by the limits of agreement, which are very similar to the righthand, especially considering the range (0.5 ms). Further, the num-ber of measurements0.5 ms was borderline statistically signifi-cantly different between the methods (p = 0.06).We found good specificity for both methods (Table 5), which issurprising in the light of previous reports (Redmond and Rivner,1988). A reason might be the higher prevalence of CTS in our pop-ulation compared to the general population (Franklin andFriedman, 2015; Barcenilla et al., 2012), which simply gives a falseimpression of high specificity.There are several technical factors that might have influencedour measurements. We cannot rule out that the variation betweenthe methods is due to error in measurement of the distances andplacement of the electrodes. We tried to minimize measurementerror as we measured and marked the stimulation and registrationsites before testing, used the same measurements for both meth-ods and kept the recording electrodes for the median nerve in placewhen switching over to the simultaneous method. It can be arguedthat the simultaneous method might be less prone to measure-ment and placement errors, as placement of stimulation electrodesis only performed once and not twice as in the consecutive method.Another possible explanation for the difference between the meth-ods is that the simultaneous method might be more self-controlling than the consecutive, as it measures both nerves withthe hand in the same posture and with the same stimulus intensityand background noise level and thus potentially reducing mea-surement errors. Not stimulating the ulnar and median nerves inthe 4th finger directly one after another might have led to a certainTable 1Summary of nerve conduction study data (NCS).Latency parameterHand Recording methodSimultaneousConsecutiveMedian nerve latencymean in ms (SD)Righthand3.35 (0.42)3.32(0.40)Lefthand3.27 (0.45)3.30 (0.51)Ulnar nerve latencymean in ms (SD)Righthand3.00 (0.28)3.00 (0.28)Lefthand2.97 (0.27)2.98 (0.27)Median- ulnar nervelatency differencemean in ms (SD)Righthand0.35 (0.36)0.32 (0.34)Lefthand0.31 (0.47)0.31 (0.51)Median-ulnar nervelatency differencerange in ms(minimum-maximum)Righthand2.03 (0.10–1.93) 1.85 (0.19– 1.66)Lefthand2.46 (0.49–1.97) 2.32 (0.38–1.94)Table 2Distribution of nerve conduction studies severity grades (N = 72).NCSseveritygradeLeft handsimultaneousLeft handconsecutiveRight handsimultaneousRight handconsecutive050 53 45 4911161142–411131619Table 3Agreement between the consecutive and simultaneous methodsin identifying minimal carpal tunnel syndrome (CTS) cases in bothhands.ConsecutiveNo CTSCTSTotalSimultaneousNo CTS808CTS628Total14216P = 0.03-0,250,000,250,500,751,001,251,501,752,000,000,250,500,751,001,251,501,752,00median-ulnar latency difference measured by the simultaneous method (ms)median-ulnar latency difference measured by the consecutive method (ms)Fig. 5.Raw median-ulnar sensory latency differences for both methods. The x-axis shows the latency difference between the median and ulnar nerves as measured by thesimultaneous method. The y-axis shows the latency difference between the median and ulnar nerves as measured by the consecutive method. The diagonalline representsequal values for x and y (y = 1x + 0).D.G. Schulze et al./Clinical Neurophysiology Practice 5 (2020) 23–2927
bias. This was due to a change in our clinical standard protocol bythe additional testing of the remaining median and ulnar fingerswith the objective of assessing peripheral nerve damage in thestudy population. However, we were aware of this limitation andmade an effort to keep the hand in the exact same position duringboth measurements. Placing an extra pair of electrodes at the med-ian nerve while stimulating the ulnar nerve might reduce measure-ment error. However, to achieve a realistic comparison betweenthe two methods, it is necessary to reposition the recording elec-trodes in the consecutive approach, as one uses only one pair ofelectrodes in this method. This is due to economic factors, as thecomparison test is not needed in patients with reduced conductionvelocity in the median nerve and the routine use of two sets ofelectrodes in all patients with suspected CTS would produce a lotof wasted electrodes. Also, we cannot exclude an order effect inwhich the investigator was biased by the result of the first methodwhen performing the second, as we have not randomized the orderof the two methods. However, we consider it unlikely that theorder in which the methods were performed was a source of biasin the present study, as the average latencies and the differencebetween the latencies were not different between the methods.Furthermore, we did not measure the skin temperature continu-ously, and thus it cannot be ruled out that sweating or cooling afterwarming might have influenced the measurements. However, wemeasured temperature again if we found significantly reducedconduction velocities.Even though we tried to minimize the impact of each of thesefactors, we cannot rule out that they influenced the measurementsand we argue that our study illustrates how important standard-ization of methods is.Further limitations need to be addressed. On a global basis,stimulation with a hand-held bar is less common than with ringelectrodes. However, at least in northern Europe, the orthodromicmethod is more commonly performed with a stimulation bar.Since the normative values used in our lab were collected in thismanner, we used this approach also in this study. We find ourapproach less time consuming, as we can use the stimulation barfor both motoric NCS and sensory NCS and find the re-positioning of the bar more time efficient. In this study, we didnot see a far-field potential from the median or ulnar nerves whenrecording the respective other nerve. This might be due to that wetried to place the electrodes as precisely as possible over thenerves. Further, this sample population had relatively large handsand wrists and thus possibly a wide distance between the medianand ulnar nerves. This, in addition to that all subjects were men,might reduce the external validity of our findings. Further, all sub-jects were exposed to risk factors for carpal tunnel syndrome suchas heavy manual work and use of vibrating tools (Kozak et al.,2015; Barcenilla et al., 2012), which explains the higher prevalenceof CTS in our study as compared to the general population. Thesame is true for other conditions associated with high exposureto vibrating tools such as hand-arm-vibration syndrome (HAVS)which have a higher prevalence in populations such as the present.In patients with HAVS, nerve damage, especially to the mediannerve has been found (Rolke et al., 2013). However, we do notbelieve this can have affected our findings, as this should not causea higher proportion of median entrapment cases in the simultane-ous method alone. We did not systematically rule out other mus-culoskeletal conditions in the hands or forearms. Likewise, we didnot perform clinical provocation methods such as Tinels maneuveror Phalens test. However, these tests may be misleading as theymay have a higher specificity and sensitivity for tenosynovitis inthe hand than for carpal tunnel syndrome (El Miedany et al.,2008). This could introduce a bias when performing the tests inour study population with its high manual workload. As this studywas part of a health survey, we did not perform a priori calcula-tions of sample size. This might have influenced the significancelevel of the mean difference and it can be conceived that a largersample size might show a significant difference between themethods. Moreover, we had no information on the inherent mea-surement error of the orthodromic ring finger test. We could notuse ROC analysis to compare how precise the two methods arewhen measuring the latency, as this would require knowledge ofthe true latency, which is not known. Likewise, the number of car-pal tunnel cases in the material was too small to use receiver oper-ating curves with the clinical definition as a gold standard in orderto determine cut-off values. We used a clinical definition of CTS asthe reference standard. This might be problematic, as previousstudies have demonstrated a varying degree of correlationbetween clinical symptoms and NCS (Gomes et al., 2006) and dif-ferent clinical symptoms have different degrees of correlation toNCS findings (Schrijver et al., 2005). NCS has a false positive rate(Redmond and Rivner, 1988). Further, the clinical presentation ofCTS varies to a significant degree (Nora et al., 2004). We performedevery measurement only once in the same individual. In order tocompare test-retest reliability and thus measurement errorbetween and within the methods, it would have been necessaryto perform measurements twice in random order in the same indi-vidual. We can therefore not quantify to which degree the simul-taneous test is ‘‘self-controlling”.Our results suggest that technical variations in the way thering-finger comparison test is performed might lead to changesin the diagnostic accuracy of this test.Neurophysiologic labs should be aware of how easily theresults of ring-finger can be influenced by methodologicalvariations.Table 4Relationship between the clinical criteria for carpal tunnel syndrome (CTS) and theconsecutive and simultaneous methods in the minimal median entrapment subgroup(N = 16).Clinical criteriafor CTS metClinical criteriafor CTS not metConsecutivemethod0.5 ms median-ulnar latencydifference220.5 ms median-ulnar latencydifference66Simultaneousmethod0.5 ms median-ulnar latencydifference830.5 ms median-ulnar latencydifference05Table 5Relationship between the clinical criteria for carpal tunnel syndrome (CTS) and theconsecutive and simultaneous methods in the whole population (N = 72).Clinical criteriafor CTS metClinical criteriafor CTS not metConsecutivemethod0.5 ms median-ulnar latencydifference970.5 ms median-ulnar latencydifference1244Simultaneousmethod0.5 ms median-ulnar latencydifference1680.5 ms median-ulnar latencydifference543The consecutive method has a sensitivity of 42% and a specificity of 86%, thesimultaneous method a sensitivity of 76% and a specificity of 84%.28D.G. Schulze et al./Clinical Neurophysiology Practice 5 (2020) 23–29
Declaration of Competing InterestThe authors declare that they have no known competing finan-cial interests or personal relationships that could have appearedto influence the work reported in this paper.ReferencesArmstrong, T., Dale, A.M., Franzblau, A., Evanoff, B.A., 2008. Risk factors for carpaltunnel syndrome and median neuropathy in a working population. J. Occup.Environ. Med. 50, 1355–1364.Barcenilla, A., March, L.M., Chen, J.S., Sambrook, P.N., 2012. Carpal tunnel syndromeand its relationship to occupation: A meta-analysis. Rheumatology (Oxford) 51,250–261.Bland, J.D., 2007. Carpal tunnel syndrome. BMJ 335, 343–346.Bland, J.M., Altman, D.G., 1986. Statistical methods for assessing agreementbetween two methods of clinical measurement. Lancet 1, 307–310.Capone, L., Pentore, R., Lunazzi, C., Schonhuber, R., 1998. Pitfalls in using the ringfinger test alone for the diagnosis of carpal tunnel syndrome. Ital. J. Neurol. Sci.19, 387–390.Deniz, F.E., Oksuz, E., Sarikaya, B., Kurt, S., Erkorkmaz, U., Ulusoy, H., Arslan, S., 2012.Comparison of the diagnostic utility of electromyography, ultrasonography,computed tomography, and magnetic resonance imaging in idiopathiccarpal tunnel syndrome determined by clinical findings. Neurosurgery 70,610–616.El Miedany, Y., Ashour, S., Youssef, S., Mehanna, A., Meky, F.A., 2008. Clinicaldiagnosis of carpal tunnel syndrome: Old tests-new concepts. Joint Bone Spine75, 451–457.Franklin, G.M., Friedman, A.S., 2015. Work-related carpal tunnel syndrome:Diagnosis and treatment guideline. Phys. Med. Rehabil. Clin. N. Am. 26, 523-+.Gomes, I., Becker, J., Ehlers, J.A., Nora, D.B., 2006. Prediction of theneurophysiological diagnosis of carpal tunnel syndrome from thedemographic and clinical data. Clin. Neurophysiol. 117, 964–971.Kouyoumdjian, J.A., Morita, M.P., Molina, A.F., 2002. Usefulness of additional nerveconduction techniques in mild carpal tunnel syndrome. Arq. Neuropsiquiatr. 60,923–927.Kozak, A., Schedlbauer, G., Wirth, T., Euler, U., Westermann, C., Nienhaus, A., 2015.Association between work-related biomechanical risk factors and theoccurrence of carpal tunnel syndrome: An overview of systematicreviews and a meta-analysis of current research. BMC Musculoskelet Disord.16, 231.Logigian, E.L., Villanueva, R., Twydell, P.T., Myers, B., Downs, M., Preston, D.C.,Kothari, M.J., Herrmann, D.N., 2014. Electrodiagnosis of ulnar neuropathy at theelbow (une): A bayesian approach. Muscle Nerve 49, 337–344.Nodera, H., Herrmann, D.N., Holloway, R.G., Logigian, E.L., 2003. A bayesianargument against rigid cut-offs in electrodiagnosis of median neuropathy atthe wrist. Neurology 60, 458–464.Nora, D.B., Becker, J., Ehlers, J.A., Gomes, I., 2004. Clinical features of 1039 patientswith neurophysiological diagnosis of carpal tunnel syndrome. Clin. Neurol.Neurosurg. 107, 64–69.Padua, L., Lomonaco, M., Gregori, B., Valente, E.M., Padua, R., Tonali, P., 1997.Neurophysiological classification and sensitivity in 500 carpal tunnel syndromehands. Acta Neurol. Scand. 96, 211–217.Preston, D.C., Ross, M.H., Kothari, M.J., Plotkin, G.M., Venkatesh, S., Logigian, E.L.,1994. The median–ulnar latency difference studies are comparable in mildcarpal tunnel syndrome. Muscle Nerve 17, 1469–1471.Redmond, M. David, Rivner, Michael H., 1988. False positive electrodiagnostic testsin carpal tunnel syndrome. Muscle Nerve 11, 511–518.Rivner, Michael H., Swift, Thomas R., Malik, Khalid, 2001. Influence of age andheight on nerve conduction. Muscle Nerve 24, 1134–1141.Rolke, R., Rolke, S., Vogt, T., Birklein, F., Geber, C., Treede, R.D., Letzel, S., Voelter-Mahlknecht, S., 2013. Hand-arm vibration syndrome: Clinical characteristics,conventional electrophysiology and quantitative sensory testing. Clin.Neurophysiol. 124, 1680–1688.Salerno, D.F., Franzblau, A., Werner, R.A., Bromberg, M.B., Armstrong, T.J., Albers, J.W., 1998. Median and ulnar nerve conduction studies among workers:Normative values. Muscle Nerve 21, 999–1005.Schrijver, H.M., Gerritsen, A.A., Strijers, R.L., Uitdehaag, B.M., Scholten, R.J., De Vet, H.C., Bouter, L.M., 2005. Correlating nerve conduction studies and clinicaloutcome measures on carpal tunnel syndrome: Lessons from a randomizedcontrolled trial. J. Clin. Neurophysiol. 22, 216–221.Uncini, A., Lange, D.J., Solomon, M., Soliven, B., Meer, J., Lovelace, R.E., 1989. Ringfinger testing in carpal tunnel syndrome: A comparative study of diagnosticutility. Muscle Nerve 12, 735–741.Wang, Y.J., Yan, S.H., 2013. Improvement of diagnostic rate of carpal tunnelsyndrome with additional median-to-ulnar comparative nerve conductionstudies. Acta Neurol. Taiwan 22, 152–157.Watson, P.F., Petrie, A., 2010. Method agreement analysis: A review of correctmethodology. Theriogenology 73, 1167–1179.Werner, R.A., Andary, M., 2011. Electrodiagnostic evaluation of carpal tunnelsyndrome. Muscle Nerve 44, 597–607.D.G. Schulze et al./Clinical Neurophysiology Practice 5 (2020) 23–2929