Sensory Conduction Studies

In early and mild CTS, mild sensory nerve conduction slowing across the CT is often the only abnormal finding. The peak latency of the median sensory nerve action potential (SNAP) is typically delayed or prolonged. The use of peak latency for detecting CTS, instead of the onset latency, has become increasingly embraced by many electrodiagnosticians because of the frequent challenge in identifying the onset of the SNAP in the presence of a large stimulus artifact. Sensory amplitudes vary greatly among individuals and offer limited diagnostic value until the amplitude becomes smaller than the range of normal reference values. As CTS progresses, sensory peak latency is progressively delayed and the amplitude becomes smaller. In more advanced CTS, there is often no recordable SNAP even with signal averaging and enhancement.

There are many different protocols for median nerve sensory conduction study to help diagnose CTS. Using antidromic techniques with stimulation applied to the wrist, median SNAPs can be recorded at the thumb, the index, the long, and the ring fingers. Median sensory fiber electrical signals can also be recorded in the long finger with the application of two stimuli, one at the wrist and one at the palm. When present, relative sensory nerve conduction slowing in the proximal segment can reveal focal median sensory nerve injury localized to the CT. Conduction block is suspected when the SNAP elicited with proximal stimulation is 50% or less in amplitude than that recorded with stimulation applied distal to the wrist. Proximal segmental conduction slowing relative to a more distal segment helps in differentiating CTS from length-dependent peripheral neuropathy. The method of sensory “inching” along the median nerve between the wrist and palm is not regularly used, but could potentially help localize conduction slowing to a short segment of the median nerve.

Motor Conduction Studies

Median motor nerve conduction is usually obtained recording over the abductor pollicis brevis (APB) muscle. Delay of the distal motor latency (DML) supports the diagnosis of CTS. However, by itself, mild prolongation of the median DML does not absolutely suggest focal demyelination. The preferential loss of the large fast-conducting myelinated nerve fibers could also result in slower nerve conduction and delayed DML. It has been shown that mild proximal conduction slowing can occur in CTS and, therefore, a decreased median motor nerve conduction velocity in the forearm does not exclude the diagnosis of CTS. Comparing the conduction velocity and DML of the median nerve to that of the ulnar nerve, which travels outside the CT, helps to confirm CTS and differentiate it from length-dependent peripheral polyneuropathy. One should also be aware of the possibility of anatomic variants such as a Martin-Gruber anastomosis in the forearm, but its presence generally does not cause confusion in the diagnosis of CTS in contrast to EDX for ulnar neuropathy at the elbow.

Short segment studies testing the recurrent median motor branch distal to the compression site have the potential for more precisely localizing a median neuropathy to the CT. In performing such testing, distal motor conduction block can sometimes be demonstrated as consequence of significant focal motor nerve demyelination. However in practice, the short distance between stimulation and recording sites and the close proximity of the recurrent branch of the median motor nerve to the nearby ulnar motor nerve make such short segment testing challenging.

Needle EMG Studies

Needle EMG is often helpful in further characterizing the neuropathic insult, especially when the compound muscle action potential amplitude is reduced because this can be a consequence of either distal demyelination or denervation. Needle EMG can provide evidence of denervation by the presence of membrane instability, such as fibrillation potentials and positive sharp waves, and altered motor unit morphology. With severe CTS, electrodiagnostic evidence of axonal loss of motor nerves and subsequent motor unit reorganization can be seen in the intrinsic hand muscles innervated by the median nerve, including the APB, the superficial head of the flexor pollicis brevis, opponens pollicis, and the first and second lumbricals. However, axon loss of motor nerves is generally uncommon in CTS. Also, fibrillation potentials and positive sharp waves cannot be used to quantify the extent of motor axon loss. For most CTS cases, the use of needle EMG is debated.

Latency Differences in CTS

There are two methods to decide whether the median nerve latency is normal or prolonged, both of which require normal reference values obtained from healthy individuals without CTS. The first is the absolute latency method. Using this method, the median sensory or motor latency is compared against the normal reference value. Conduction slowing is determined if the latency exceeds the upper limit of normal. This absolute latency method has several drawbacks, especially for sensory conduction because the skin temperature effects and other factors. The recommendation is that each EDX laboratory establishes its own normal data. Unfortunately, this is not always achievable.

An increasing awareness of variation among normal reference values from laboratory to laboratory has led to the comparison reference value method. This newer method has been well described and was recommended for use in diagnosing CTS in a recent 2011 American Association of Neuromuscular and Electrodiagnostic Medicine (AANEM) monograph. In this method, conduction study of the median nerve is compared with that of another nerve, usually in the same limb and typically the ulnar. Comparison reference values from different investigators show less variation, because of better built-in controls for differences in subjects, techniques, and equipments.

For instance, for the assessment of CTS, latency measurements can be compared using median–thenar and ulnar–hypothenar pairs. With this protocol, a DML difference of greater than 1.2 to 1.8 ms is said to be significant. This technique is especially useful when studying patients who have generalized peripheral polyneuropathy, most commonly from diabetes. As supporting evidence for probable CTS, this author uses a difference of greater than 2.0 ms comparing the median DML with the ipsilateral ulnar DML. When there are no sensory responses in the upper extremities of a diabetic patient who has additional unilateral hand paresthesias and weakness, such a motor comparison can be quite helpful to demonstrate evidence for CTS in the presence of diabetic peripheral neuropathy. Another useful technique in this setting of possible peripheral neuropathy and superimposed CTS is a comparison of the difference in DML between the median-innervated second lumbrical muscle and the ulnar-innervated first palmar interosseous muscle. The author applies this approach in her laboratory but not routinely.

The Combined Sensory Index

Though the best electrodiagnostic test for CTS has yet to be determined, many studies for CTS have been designed over the last few years. For instance, using the absolute method, one can assess for CTS by obtaining median sensory latency recording from the thumb, and the index, long, and ring fingers. Using the comparison method, CTS can be diagnosed by determining median-radial sensory latency differences to the thumb, median-ulnar sensory differences to the ring finger, and transpalmar median-ulnar latency differences. An important question is whether one normal test is sensitive enough to refute the diagnosis of CTS. Because CT compression can affect some median nerve fascicles earlier or preferentially, one electrodiagnostic test may demonstrate CTS when another does not. Performing limited studies could reduce the sensitivity of EDX considerably, causing a false negative error. However, increasing the number of tests also likely increases the probability of reducing the specificity of testing, causing a false positive error in proportion to the number of tests performed. Although a false negative error may limit early detection and treatment, a false positive error could result in overtreatment, which has the potential to cause harm to the patient.

The combined sensory index (CSI) is an attempt to maximize sensitivity for detecting CTS without reducing specificity by using a single score derived from multiple sensory tests. The CSI, sometimes called the Robinson index, is easy to obtain. The CSI is the sum of comparisons of sensory latencies collected with three established sensory tests for the study of CTS using the following formula: CSI = ringdiff + thumbdiff + palmdiff. In this formula, ringdiff is the peak latency difference of the median and ulnar antidromic sensory nerve conduction to the ring finger stimulating 14 cm proximally, the thumbdiff is the peak latency difference of the median and radial antidromic sensory nerve conduction to the thumb stimulating 10 cm proximally, and the palmdiff is the transpalmar peak latency difference of the median and ulnar orthodromic conduction using a distance of 8 cm. For example, if the latencies are 3.8 ms for median nerve conduction to the ring finger and 3.4 ms for ulnar nerve conduction to the ring finger, then the ringdiff is 0.4. If the median latency is 3.4 ms and the ulnar latency is 3.8 ms, then the ringdiff would be a negative number (−0.4). Using negative numbers helps to cancel random errors such as distance measurement and take into account of ulnar or radial sensory neuropathy if present.

The CSI has a high specificity (ie, few false positives), a high sensitivity (ie, few false negatives), and excellent test-retest reliability. When the upper limit of the normal CSI is set at 0.9 ms, the test sensitivity is 0.83 and the specificity 0.95. If the normal upper limit for the CSI is raised from 0.9 ms to 1.1 ms, the sensitivity remains essentially the same at 0.82; however, the specificity increases to 1.00 indicating very low if any probability of false positive error. Generally, a CSI of 1.0 ms or greater would be consistent with CTS. The test and retest reliability for the CSI has been shown to be superior to that of other techniques in identifying CTS because of improved control for variables such as hand size, height, age, and temperature.

Study Approaches and Reporting of CTS

In the author’s laboratory, CTS is considered with any clinical presentation of pain, numbness, paresthesias, and weakness in the upper extremities. Other potential diagnoses, such as arthritis, tendinopathy, polyneuropathy, brachial plexopathy, or cervical radiculopathy, are also entertained. Clinical impression of the likelihood of CTS from history and physical examination is formed before EDX study. The author usually begins with median motor nerve conduction, followed by ulnar motor nerve conduction, and then the three sensory tests that form the CSI study. When bilateral CTS is suspected, the more symptomatic hand is studied first. If the median nerve is normal, the less symptomatic side is not always studied. The author does not perform NCS in the asymptomatic limb for the purpose of contralateral comparison of the median nerves given the significant incidence of bilateral EDX abnormalities in some asymptomatic population. This is in accordance with AANEM guidelines.

For most patients referred to our EDX laboratory with possible CTS, the author performs all three CSI sensory tests and reports the calculated CSI. The added time, cost, and discomfort to the patient of these studies are quite reasonable. The benefit of obtaining a complete CSI is its excellent reproducibility. Repeat CTS referrals on the same patients are fairly common in the author’s institution. These patients have recurrent CTS after a period of satisfactory symptom relief using conservative treatments. Comparison of two CSI scores could provide information in regard to progression of CTS and is potentially of help to the patient and the treating physician in making treatment choices. A change of the CSI in such repeat studies of greater than 0.3 ms is considered to be statistically significant. The author is occasionally asked to study the patient who has persistent hand pain and numbness after CT release surgery. The comparison of two CSI scores obtained before and after the operation, when available, is useful in differentiating residual median neuropathy versus incomplete release or recurrent CTS, which is rare.

For nerve conduction studies, the author prefers warming the upper limb rather than extrapolating using a latency or conduction velocity corrected for temperature. Individual thermostats in the study rooms help regulate room temperature to minimize skin temperature drop during the study, which is usually between 1°C and 2°C in our laboratory. Warming pads are used to maintain skin temperature, though a heat lamp or a hand-held hair dryer can also work. The author does not routinely perform needle EMG for CTS evaluation.

Needle EMG is performed if there is concern for the possibility of ulnar neuropathy, brachial plexopathy, cervical radiculopathy, or if there is a history of trauma. For example, non–median innervated C8 muscles, such as the first dorsal interosseus, abductor digiti minimus, or extensor indicis proprius, are examined in addition to the APB muscle if C8 radiculopathy is suspected. Needle EMG study of median-innervated forearm muscles, such as the pronator quadratus and pronator teres, help to differentiate CTS from less common median nerve entrapment syndromes that occur proximal to the wrist.

When providing the patient with CTS and the referring clinician with our EDX impression, the author confirms the localization of a median neuropathy at the wrist and comments on the electrodiagnostic evidence or sensory nerve involvement only or if there is focal demyelination of the motor nerve. The severity of the demyelination is also commented on in reference to the degree of conduction slowing and the presence or absence of motor conduction block. Denervation from CTS is rare, but when present, the extent and chronicity are also discussed in the report.

No matter the EDX protocol used, some patients with the clinical symptoms of CTS have negative findings. Conservative treatment (see later discussion) is a satisfactory approach for many of these patients. Conversely, EDX CTS in patients without clinical CTS symptoms should be reported with caution to referring clinicians to avoid the potential risk of overtreatment. A prevalence study of the Swedish general population found 18.4% of a control group to have abnormal EDX findings characteristic for CTS.