• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Quantitative Vibration Threshold Testing in Carpal Tunnel Syndrome: Analysis Strategies for Optimizing Reliability

Posted on: Wednesday, 25 February 2004, 06:00 CST

ABSTRACT: Tuning forks and electronic vibrometers have been used to quantify vibration sensation thresholds, which are thought to be affected early in carpal tunnel syndrome (CTS). The purpose of this study was to identify a reliable testing procedure for a newly designed, computer-controlled vibrometer (PCV50; Ztech, Salt Lake City, UT). Fifty-two patients (mean age 488 years) with electromyographically confirmed CTS were tested on one occasion. The computer-controlled vibrometer, with a fixed frequency of 50 Hz, used stepwise changes in amplitude to determine vibration sensation threshold. Each patient had three vibrometer measures (trials) taken on the pulp of the third digit of their right and left hands during the first test session and were retested by a single repetition 40 to 60 minutes later (retest). Intraclass correlation coefficients (ICCs) were used to examine several data analysis strategies. The strategy that generated the highest TCCs for both the right and left hands assumed that the first trial was a learning or practice attempt, and compared the average of the second and third trials with the score from the second session (ICC = 0.86 and 0.89, respectively). The computer-controlled vibrometer offered an easily administered, quantitative, and comfortable means to assess median nerve function. Using this reliable testing procedure will allow for additional investigations to determine its usefulness in the early detection and accurate quantification of CTS-related impairment.

J HAND THER. 2004;17:24-30.

Carpal tunnel syndrome (CTS) is a relatively common compression neuropathy affecting the median nerve at the wrist. The prevalence of CTS in the general population is thought to be approximately 5%1/ 2 and as high as 15% in certain high-risk occupational groups.3 Early identification of CTS is considered paramount to optimize outcomes of conservative and surgical treatment.4"6 Various measures of cutaneous sensibility, including vibration, have been used to aid in the diagnosis and grade the severity of CTS in both clinical and research settings.

Vibration sensation was the first cutaneous sense reported to be altered in several studies of nerve compression using human subjects.7"9 Touch sensation perception, including vibration sense, is carried by large-diameter myelinated afferent nerve fibers, which are the first to be affected by mechanical pressures and may also be the first to be affected by localized ischemia.10"12 For this reason, various measures of vibration sensibility have been examined and several quantitative vibratory instruments have been introduced.

Vibration sensation measures have been used for several decades to identify, localize, and categorize nerve injuries. A vibrating tuning fork touched against various locations on the patient and compared with the unaffected contralateral side has classically been used. This technique, although simple and effective, produces only a binary or dichotomous response (yes-no). Therefore, its use in grading treatment effects or disease progression is limited. The tuning fork also lacks a fixed or measurable amplitude and has limited use in bilateral conditions, which are common in CTS.

To address the deficiencies of the tuning fork, electronically controlled devices, which generate a modifiable vibratory stimulus, have been developed. Vibrometers, as they have become known, have adjustable amplitude, whereas the frequency of vibration can be either fixed or adjustable. The reliability of several of these vibrometers has been studied. Two studies by Gerr et al.13,14 compared the reliability of two threshold testing protocols using a fixed-frequency (120-Hz) vibrometer. They reported high correlations (r > 0.81) for both the method of limits and the forced-choice protocols in both studies. Their main conclusion in both studies was that the method of limits protocol required significantly less time to administer. In a study of a multifunction sensory testing tool, 50-Hz vibration was determined to be the most reliable measure of tactile sensibility when compared with 150-Hz vibration, light touch, pinprick, warming, and two- point discrimination.15

Several vibrometers have been shown to display concurrent validity,8,16,17 whereas other research has identified only a weak correlation with other measures of CTS.18,19 There may be variability in these findings for a number of reasons, including differences in equipment, testing protocol, vibratory frequency used, and the population studied. To accurately test the validity of a tool, or to make recommendations regarding its use clinically, a consistent protocol and testing procedure with the best possible reliability should be developed.

The purpose of the current study was to investigate the potential use of a recently developed vibrometer (PCV50, Ztech, Salt Lake City, UT) and to examine strategies to optimize the reliability of test measures in a sample of patients with confirmed CTS. It was hypothesized that the staircase method of threshold determination programmed into the vibrometer would provide acceptable reliability (intraclass correlation coefficient [ICC] > 0.75), and that differences in reliability would be observed depending on the measurement strategies used for selecting the criterion score.20 These differences in reliability formed the basis for recommendations regarding the testing protocol for the prototype vibrometer.

METHODS

Testers

The second author (JCM) trained two testers in the use of the vibrometer. A set of written instructions was formulated to ensure that consistent instructions were provided to patients by both testers. Beyond instructions, the test application or recording of results is not examiner-dependent and thus no intertester reliability testing was conducted.

Subjects

The subjects were all patients diagnosed with CTS by one of three hand surgeons at the Hand and Upper Limb Centre (HULC) in London, Ontario. Prospective subjects were required to speak and read English and to be less than 65 years of age to minimize possible confounding effects as a result of a high prevalence of asymptomatic median neuropathy in older individuals.1'21 Patients with inflammatory joint disease, diabetes, concurrent injury of either hand, or previous wrist or hand surgery were also excluded due to the potential confounding effect these conditions may have on the symptoms reported. This protocol was approved by the university ethics review board.

The diagnosis of CTS was made by the surgeon based on the patient's history, physical examination findings, and nerve conduction study results. The electrodiagnostic criteria required a sensory latency of >3.5 ms or a median to ulnar conduction latency difference of >0.2 ms. all nerve conduction testing was completed by one of two electromyographers in the same laboratory using the same equipment. Neurophysiological testing has been supported as a valid and reliable diagnostic tool for CTS.22

Vibrometer

The vibrometer (Figure 1) had a 2-mm diameter aperture, in the center of which was positioned the 1-mm diameter vibrating post. A displacementcontrolled actuator modulated the amplitude of the vibration, whereas the frequency was constant at 50 Hz. A patient interface button was used to indicate when the subject perceived a sensation. The system was self-calibrated each time the computer program was initiated.

Each stimulus was provided for 1 second followed by a 3-second interval of no stimulus. The stimulus was interpreted as perceived if the subject activated the patient interface button during or within 1 second after the stimulus. The vibration threshold was determined using the staircase method of threshold determination, which has been reported to demonstrate adequate psychometric properties.23 The procedure began with a high-amplitude stimulus, which was decreased in large steps until the stimulus was no longer perceived. The amplitude was then gradually increased in smaller increments until it was again perceived. This process continued until a total of six changes in direction occurred. The computer software calculated the average of these "peaks" and "valleys" to determine the threshold amplitude in micrometers.

FIGURE 1. Vibrometer PCV50; Ztech, Salt Lake City, UT.

FIGURE 2. Patient positioning during vibration sensation threshold testing.

Vibration Threshold Testing

To facilitate temperature stabilization, subjects rested for 10 minutes in the laboratory before any testing.24 The subject was seated comfortably at an adjustable-height table and positioned so that the forearms and hands rested on the table surface (Figure 2). all testing was performed in a separate room from the main clinic in a quiet atmosphere with minimal distraction or background noise. The hand being tested was positioned palm down with the forearm resting on a foam pad, and the vibrometer was moved into position so that the center of the pulp of the third digit was situated over the vibrometer aperture. This arrangement placed the subject's wrist in slight extension (<10), the forearm pronated, the elbow flexed less than 90 and the shoulder elevated less than 45. The patient was instructed to rest his or her finger passively on the vibrometer. The patient interface \button, held in the patient's opposite hand, was activated by thumb pressure. After the explanation of the subject's role in this test, a single suprathreshold sample stimulus was given. The computer screen was positioned so that the subjects could not view it. Subjects were also requested to close their eyes during testing to avoid any visual cues from the test administrator and to allow for maximum concentration on the task. No detectable noise was produced by the vibrometer. Subjects were informed when each test session was initiated and completed. A rest period of approximately 1 minute between test sessions was provided.

Each subject had the third digit of each hand tested four times. The first three vibration threshold measures (trials) were completed consecutively at the start of the session, alternating from side to side, the starting side determined randomly. The fourth measure (retest trial) was completed after 40 to 60 minutes. Between the third and fourth measures, the subject completed several short written questionnaires and was tested for hand strength, touch threshold, and hand dexterity. These tasks simulated time spent being assessed or undergoing treatment in a clinical setting.

Data Analysis

Reliability analyses were performed separately for the right and left hands of all subjects. Transcription from data sheets and computer output into a statistical software package (SPSS Inc., Chicago, IL) were independently rechecked by a research assistant. ICCs (2,1) and their associated 95% confidence intervals (CIs) were used to determine the reliability of repeated measures.25 The standard error of measurement (SEM) was also calculated for each of the comparison strategies.26

In cases when the subject could not perceive the maximum amplitude produced by the vibrometer (180 m), a threshold value was not produced. These subjects' data were removed for purposes of the primary analyses, because the subject was deemed not testable.

Several analysis strategies were used to compare the data from the four trials for each hand (Table 2). First, each of the individual trials from 1 to 3 was compared with the retest trial. The lowest value from trials 1 to 3 was then compared with the value from the retest. The median value of trials 1 to 3 was then compared with the retest. The high and the low scores of trials 1 to 4 were then removed and the two remaining values compared. Finally, the average of trials 1 to 3 and 2 to 3 were then calculated and compared with the retest trial.

RESULTS

Sixty patients volunteered to participate in this study (Figure 3). Of these, three cases were removed as a result of incomplete datasets (two subjects left the testing session before all trials were completed and the data on a third was lost as a result of computer failure). Five additional cases were omitted before the data analybes, because these patients were unable to perceive the maximum amplitude produced by the vibrometer during at least one of the trials. The resulting group of 52 subjects (13 males and 39 females) ranged in age from 30 to 63 years (mean age 48 8 years). Left hand-dominant individuals made up 10% of the test group. Eight individuals (15%) reported only their left side as affected, whereas 12 (23%) reported only right-sided symptoms. Bilateral symptoms of CTS were documented in 32 (62%) of the patients.

FIGURE 3. Methodology flow chart. ICCs = intraclass correlation coefficients; SEM = standard error of measurement.

The means and standard deviations of trials 1 to 3 and the retest trial are presented in Table 1. The ICC values for test-retest reliability are presented in Table 2. The data analysis strategy that generated the highest ICCs for both the right and left hands was that which assumed that the first trial was a learning or practice attempt and compared the average of trials 2 to 3 with the retest trial (ICC = 0.86 and 0.89, respectively); followed closely by comparison of the average of trials 1 to 3 with the retest (ICC = 0.85 and 0.80). The averaged trial comparisons also produced the lowest SEM values for both the right and left hands (Table 2). Slight to moderate reliability (<0.70) was generally noted for methods in which individual repetitions were compared.

DISCUSSION

The reliability of instruments is essential to making clinical judgments and research conclusions. This sentiment was echoed by Fleiss, who stated, "The most elegant design of a clinical study will not overcome the damage caused by unreliable or imprecise measurement."20 Despite this, many new tools are used somewhat prematurely in both clinical practice and research without adequate consideration to defining the tool's underlying reliability. The current study identified a highly reliable method of measurement using a newly designed vibrometer. Based on the evaluation of several analysis strategies, the highest degree of reliability was achieved by performing three trials, omitting the first as a practice trial and averaging the second two trials. In this way, ICCs of 0.86 and 0.89 were achieved for the right and left hands, respectively (Table 2).

Because this vibrometer was a new tool, the best way to use the data that it produced was not known. Whether one trial would be sufficient or if multiple trials would need to be grouped together to obtain reasonable reliability of the values generated was unclear. The stepwise method of threshold determination used by the vibrometer should have minimized the measurement error and theoretically produced reliable results from a single trial. In this case, the need for multiple trials, which are traditionally performed to improve a tool's reliability, would be eliminated. This was determined to not be the case. Only fair to moderate reliability (<0.75) was seen for the majority of these comparisons. Furthermore, the large SEM associated with the first trial reflects the clinical uncertainty attributable to this strategy. It is unclear why the stepwise method for threshold determination was ineffective in this application. The most likely explanation is that patients require some time to "learn" the response that they are detecting and become more skilled at sensing it.

Another approach was to determine whether the lowest values of the multiple trials were more reliable. It has been suggested that a minimum value for perception threshold testing, or a maximum value for strength testing, may indicate the true maximum or minimum limit, and therefore be a more reliable measure.27 ICCs of 0.58 and 0.43 for the right and left hands, respectively, demonstrated that this approach did not improve the test-retest reliability associated with this instrument. This is similar to findings reported with shoulder strength evaluation showing that the maximum trial is not more reliable than an averaged score.28

The median value of the first three trials was compared with the retest (fourth) trial. A modest improvement in reliability was noted (0.71 and 0.72) with this approach. This may initially be viewed as selective data inclusion, as one trial is being selected and two are ignored, but can be justified as a result of the systematic nature of the selection. The median value may be interpreted as a reasonable estimate of the true threshold score. This strategy has been used in other threshold sensory evaluation tests, including one- and two-point light touch assessments.29

TABLE 2. Data Analysis Strategies and Reliability Measures*

TABLE 1. Means and Standard Deviations (SDs) for Grouped Data*

The final strategy used was to combine or average select trials together. This can be described as a method used to reduce the "noise" that may be present in the raw measures of a tool. This process is commonly used in statistics and measurement research. An example would be using the average of three goniometric measures rather than just one. In the current study, using the average of the second and third trials produced the highest reliability coefficients (Table 2). When the first trial was included in the average, the reliability was slightly decreased. Excluding the first trial can be easily justified by deeming it a practice trial. Use of a practice trial is common practice with vibrometers15'30'31 and other measurement instruments, including isokinetic testing.28

When the range of measurement of a tool is not large enough to include all potential measured values, this is termed a ceiling effect. For example, a grip dynamometer that measures only up to 50 kg will not provide a true grip strength value for a patient who grips greater than 50 kg. This is considered a limitation of a measurement tool. In the current study, the vibrometer was able to generate a vibration amplitude of up to 180 m. There were five subjects for which the vibrometer was unable to calculate a threshold value, because they did not perceive the highest amplitude that the vibrometer could produce. It was assumed that a valid estimate of vibration threshold could not be obtained in patients who "hit" the ceiling. This tool may still be deemed useful for patients who exhibit vibration sensory loss to this degree, but only from the standpoint of identifying that vibration perception is severely reduced. In this scenario, the tool is no longer quantitative, but rather is reduced to providing a yes-or-no response. Investigating the feasibility of increasing the maximum amplitude that this vibrometer is able to produce may be beneficial, because five of 57 subjects (9 %) could not perceive the maximum amplitude.

Using the reliable data analysis strategy for threshold determination identified in the current study, additional research is required to support its overall use. Although a strong case for construct validity has been presented in that vibration sense has been shown to be affected in experimental studies of nerve compression, confirmation of concurrent validity would serve to bolsterthis tool's acceptance. Vibration sensation thresholds, calculated by averaging the second and third trials, should be compared with other reliable measures of median nerve compression, CTS severity, and functional impact. These additional studies will assist users of this tool to properly interpret the results of individual measures and changes in measurement. As a continuation of this line of investigation, study of this vibrometer's sensitivity to change would allow clinicians to better evaluate treatment effects and to make treatment recommendations with greater confidence.

Development of normative vibration threshold values could prove useful in the application of this vibrometer as a screening tool for CTS. As stated previously, it appears that several other vibrometers have been tested in this manner without clearly defining a method for application and also without determining the tool's reliability in a population with CTS or in a mixed population. A large study of normative values may also provide information regarding effects of age, hand dominance, and gender on vibration thresholds specific to this tool. Additional investigation of this tool's reliability should also be performed on samples of patients drawn from a population of various upper extremity conditions.

Limitations of this device and study should be noted. The ceiling effect, as discussed above, limits the quantitative application of this tool to individuals with very high vibration thresholds. This study assessed vibration with a frequency of 50 Hz, as this was preprogrammed into the vibrometer. This limits the generalizability of the results to this frequency of vibration. Furthermore, the results are applicable only to this model of vibrometer. The various aspects of testing apparatus design, including size of the vibrating post and damping control, make this type of study instrument- dependent. Finally, this study examined the reliability of this tool in a sample of patients with confirmed CTS. These individuals had confirmed electromyographic (EMG) evidence of median nerve dysfunction. These results should not be applied to individuals with early or mild CTS (i.e. patients without EMG evidence) without additional investigation.

CONCLUSION

The vibrometer tested in this study provided a rapid, quantitative, and comfortable means to assess vibration sensibility, which required little examiner skill to administer and was able to provide stable results after administration of a practice trial. Additional testing is required to determine its usefulness in the detection and accurate quantification of CTS-related impairment and other neurologic pathologies.

REFERENCES

1. Atroshi I, Gummcsson C, Johnsson R, Ornstein E, Ranstam J, Rosen I. Prevalence of carpal tunnel syndrome in a general population. JAMA. 1999;282:153-8.

2. de Krom MC, Knipschild PG, Rester AD, Thijs CT, Boekkooi PF, Spaans F. Carpal tunnel syndrome: prevalence in the general population. J Clin Epidemiol. 1992;45:373-6.

3. Stevens JC, Sun S, Beard CM, O'Fallon WM, Kurland LT. Carpal tunnel syndrome in Rochester, Minnesota, 1961 to 1980. Neurology. 1988;38:134-8.

4. Gelberman RH, Rydevik BL, Pcss GM, Szabo RM, Lundborg G. Carpal tunnel syndrome. A scientific basis for clinical care. Orthop Clin North Am. 1988;19:115-24.

5. Lundborg G, lie -Stenstrom AK, Sollerman C, Stromberg T, Pyykko I. Digital vibrogram: a new diagnostic tool for sensory testing in compression neuropathy. J Hand Surg [Am]. 1986;ll:693-9.

6. Rosen B, Lundborg G, Abrahamsson SO, Hagberg L, Rosen I. Sensory function after median nerve decompression in carpal tunnel syndrome. Preoperative vs postoperative findings. J Hand Surg [Br]. 1997;22:602-6.

7. Gelberman RH, Szabo RM, Williamson RV, Dimick MP. Sensibility testing in peripheral-nerve compression syndromes. An experimental study in humans. J Bone Jt Surg [Am]. 1983;65:632-8.

8. Szabo RM, Gelberman RH, Dimick MP. Sensibility testing in patients with carpal tunnel syndrome. J Bone Jt Surg [Am]. 1984;66:60-4.

9. Szabo RM, Gelberman RH, Williamson RV, Dellon AL, Yaru NC, Dimick MP. Vibratory sensory testing in acute peripheral nerve compression. J Hand Surg [Am]. 1984;9:104-9.

10. Fullerton PM. The effect of ischemia on nerve conduction in the carpal tunnel syndrome. J Neurol Neurosurg Psychiatry. 1963;385- 7.

11. Rydevik B, Lundborg G, Bagge U. Effects of graded compression on intraneural blood flow. An in vivo study on rabbit tibial nerve. J Hand Surg [Am]. 1981;6:3-12.

12. Sunderland S. The nerve lesion in the carpal tunnel syndrome. J Neurol Neuro.surg Psychiatry. 1976;39: 615-626.

13. Gerr F, Hershman D, Letz R. Vibrotactile threshold measurement for detecting neurotoxicity: reliability and determination of age- and height-standardized normative values. Arch Environ Health. 1990;45:148-54.

14. Gerr FE, Letz R. Reliability of a widely used test of peripheral cutaneous vibration sensitivity and a comparison of two testing protocols. Br J lnd Med. 1988;45:635-9.

15. Horch K, Hardy M, Jimenez S, Jabaley M. An automated tactile tester for evaluation of cutaneous sensibility.] Hand Surg [Am]. 1992;17:829-37.

16. Borg K, Lindblom U. Increase of vibration threshold during wrist flexion in patients with carpal tunnel syndrome. Pain. 1986;26:211-9.

17. Hardy M, Jimenez S, Jabaley M, Horch K. Evaluation of nerve compression with the automated tactile tester. J Hand Surg [Am]. 1992;17:838-42.

18. Gerr F, Letz R, Harris-Abbott D, Hopkins LC. Sensitivity and specificity of vibrometry for detection of carpal tunnel syndrome. J Occup Environ Med. 1995;37:1108-15.

19. Werner RA, Franzblau A, Johnston E. Quantitative vibrometry and electrophysiological assessment in screening for carpal tunnel syndrome among industrial workers: a comparison. Arch Phys Med Rehabil. 1994;75:1228-32.

20. Fleiss JL. The Design and Analysis of Clinical Experiments. New York-NY: John Wiley and Sons, Inc, 1986.

21. Gescheider GA, Bolanowski SJ, Hall KL, Huffman KE, Verrillo RT. The effects of aging on information-processing channels in the sense of touch: absolute sensitivity. Somatosens Mot Res. 1994;ll:345-57.

22. Jablccki CK, Andary MT, So YT, Wilkins DE, Williams FH. Literature review of the usefulness of nerve conduction studies and electromyography for the evaluation of patients with carpal tunnel syndrome. AAEM Quality Assurance Committee. Muscle Nerve. 1993;16:1392-414.

23. Yarnitsky D. Quantitative sensory testing. Muscle Nerve. 1997;20:198-204.

24. Virrillo RT, Bolanowski SJ. The effect of skin temperature on the psychophysical responses to vibration on glabrous skin. J Acoust Soc Am. 1986;80:528-32.

25. Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull. 1979;86:420-8.

26. Streiner DL, Norman GR. Health Measurement Scales. A (Practical Guide to Their Development and use) 2nd ed. Oxford: Oxford University Press, 1995.

27. Kuhlman JR, Iannotti JP, Kelly MJ, Ricgler FX, Gevaert ML, Ergin TM. Isokinetic and isometric measurement of strength of external rotation and abduction of the shoulder. J Bone Joint Surg [Am]. 1992;74:1320-33.

28. Magnusson SP, Gleim GW, Nicholas JA. Subject variability of shoulder abduction strength testing. Am J Sports Med. 1990;18:349- 53.

29. Dellon AL, Keller KM. Computer-assisted quantitative sensorimotor testing in patients with carpal and cubital tunnel syndromes. Ann Plast Surg. 1997;38:493-502.

30. Goldberg JM, Lindblom U. Standardised method of determining vibratory perception thresholds for diagnosis and screening in neurological investigation. J Neurol Neurosurg Psychiatry. 1979;42:793-803.

31. Grunert BK, Wertsch JJ, Matloub HS, McCallum-Burke S. Reliability of sensory threshold measurement using a digital vibrogram. J Occup Med. 1990;32:100-2.

Mark C. Hubbard, BSc(PT), MSc

Orthopaedic & Rehab Department

Penn State Hershey Medical Center

Hershey, Pennsylvania

Joy C. MacDermid, BSc(PT), PhD

Clinical Research Laboratory

Hand and Upper Limb Centre

St. Joseph's Health Care London

London, Ontario, Canada

School of Rehabilitation Science

McMaster University

Hamilton, Ontario, Canada

Ontario Ministry of Health and Long-Term Care

Health Research Personnel Development Program

Ontario, Canada

John F. Kramer, PhD

Trevor B. Birmingham, MSc(PT), PhD

School of Physical Therapy

University of Western Ontario

Ontario, Canada

Correspondence and reprint requests to Joy C. MacDermid, BSc(PT), PhDx Hand and Upper Limb Centre, St. Joseph's Health Care London, PO Box 5777, London, Ontario, Canada N6A 4L6. E-mail: .

doi:10.1197/j.jht.2003.10.004

Copyright Hanley & Belfus, Inc. Jan-Mar 2004

More News in this Category