Keywords
carpal tunnel syndrome - electrodiagnostic - nerve conduction studies - median nerve
- sensitivity
Introduction
Carpal tunnel syndrome (CTS) is an important cause of hand pain and impairment caused
by compression of the median nerve at the wrist.[1] It is the most common perihedral nerve entrapment of the upper limb encountered
worldwide.[2] The diagnosis is primarily based on clinical evaluation and is considered when patients
exhibit common symptoms like numbness, tingling, nighttime paresthesia, and/or neuritic
“pins-and-needles” sensation in the radial 3.5 digits.[3]
Electrophysiological testing enhances clinical evaluations by assessing the reduction
in nerve conduction speed through the carpal tunnel and subsequent axonal damage.
These tests confer additional diagnostic certainty beyond clinical impression by providing
quantitative data.[4] Nerve conduction studies (NCSs) are currently the only tool that can quantitatively
reveal the deterioration in median nerve function that occurs as CTS evolves from
grade 1 to grade 4, as patients in these grades are clinically indistinguishable.[5] The exact methods chosen are very important for determining early lesions of the
median nerve, which will affect the sensitivity and specificity of the results.
Electrodiagnostic (EDX) testing is performed using generally accepted standardized
techniques according to the American Association of Neuromuscular and Electrodiagnostic
Medicine (AANEM) summary statement.[6] Each EDX technique has a different recommendation level, which is influenced by
the quality and consistency of the evidence backing it.[7] The recommended EDX studies for patients suspected of having CTS include median
sensory or mixed nerve conduction studies, median motor conduction studies, needle
examination of the abductor pollicis brevis (APB), ulnar and/or radial motor and sensory
assessments, and needle electromyography (EMG) of the limb muscles innervated by the
C5 to T1 roots.[8]
[9]
Research has shown that evaluating sensory nerve responses is more effective than
relying on absolute median nerve latency for identifying median nerve issues linked
to CTS.[10]
[11] Evaluating the median sensory latency against the sensory latencies of the radial,
ulnar, and median nerves (segments external to the carpal tunnel) yielded the highest
precision in verifying a clinical diagnosis.[10]
[11]
In this study, several EDX techniques were selected and performed on participants.
This study involves a direct comparison of the diagnostic value in patients with clinical
CTS, with the goal of finding a combination of EDX techniques that offers high sensitivity
for early CTS diagnosis.
Materials and Methods
Design and Ethical Considerations
A case–control study was conducted. All the participants agreed to and comprehended
the detailed electrophysiological protocols. Ethical approval was obtained from the
ethics committee of our institution (approval number: KY24088).
Participants
The initial inclusion criteria for this study were to recruit 100 participants in
the CTS group and 70 participants in the healthy control group (HCs). After rigorous
screening, the following exclusions were made: from the CTS group, 3 participants
with other nerve injuries, 4 with a history of diabetes mellitus, and 2 with systemic
lupus erythematosus; from the HCs group, 1 participant with other nerve injuries and
1 who was unable to tolerate electrophysiological testing. Consequently, from January
2023 to December 2023, 68 HCs (136 hands) were recruited through our medical center,
and 91 adult patients(CTSs, 162 hands) clinically diagnosed with CTS were recruited
from the outpatient clinic of orthopaedics and neurology in our center. The sample
size was based on the number of observed cases within the study area throughout the
research period. Patients diagnosed with CTS in this study had at least one of the
following primary symptoms: (1) numbness, tingling pain, or paresthesia on the radial
side of the three fingers and the radial side of the ring finger; (2) nocturnal awakening
due to such sensory symptoms; and (3) a positive Tinel and/or Phalen sign. Patients
with previous wrist trauma or surgery, as well as those with diabetes mellitus, pregnancy,
or polyneuropathy were excluded. None of the HCs showed any symptoms of neurological
damage or a history of underlying conditions, such as diabetes.
Outcome Measures
We used the Vedi Keypoint EMG evoked potential instrument imported from Denmark to
conduct upper limb nerve conduction tests in the two groups. The patients were conscious
during the tests, which were conducted in an undisturbed and quiet environment with
an indoor temperature maintained at approximately 24°C. The nerve conduction test
sites included the median and ulnar nerve.[12] The participants were placed in the supine position with their forearms supinated.
The sensitivity was configured to 5 to 20 V/div, the low- and high-frequency filters
were set to 20 to 2 kHz, and the sweep speed was adjusted to 2 ms/div.
Sensory Ganglia Segment Method of Median Nerve
We segmented the median nerve sensory conduction using the orthodromic method. The
median sensory NCS were recorded from thumb (D1), index (D2), and middle finger (D3)
antidromically with standard distance of 10 cm (D1) and 14 cm (D2 and D3) between
stimulation and recording electrodes. The latency, sensory nerve conduction velocities
(SNCV) and amplitude (SNAP) of D1-W, D2-W, D3-W were calculated. The median nerve
was stimulated at the D3 and palm, with subsequent recordings taken at the palm and
wrist, with a fixed distance of 8 cm, to calculate and compare the SNCV difference
between D3 to the palm and palm to the wrist (P-W). We also recorded the SNCV from
wrist to elbow, which was stimulated at the wrist and recorded at the elbow (W-E).
Comparison of the Sensory Distal Latency of the Median to the Ulnar Nerve in the Ring
Finger (M-U ringdiff)[6]
Stimulation occurred 14 cm proximal to the recording electrode over either the median
or ulnar nerves at the wrist level. The responses were recorded using patch electrodes
placed on the ring finger. M-U ringdiff was used to calculate the latency difference
between the median-ring and ulnar-ring recordings.
Ulnar Sensory Nerve Conduction Studies
Ulnar sensory NCS was recorded from the little finger (D5), whose stimulation was
14 cm proximal to the recording electrode. Latency, SNCV, and SNAP from D5 to the
wrist (D5-W) were calculated. The difference between the median sensory distal latency
and ulnar sensory distal latency was determined by comparing these two measurements.[11]
Motor Nerve Conduction Studies
Motor studies of the median and ulnar nerves were conducted by recording compound
muscle action potentials (CMAP) from the APB and abductor digiti minimi muscles. The
recording electrodes (G1) were positioned over the muscle bellies, whereas the reference
electrodes (G2) were placed over the distal tendinous insertions. Stimulation of the
median and ulnar nerves occurred at the wrist, 8 cm proximal to G1, and at or below
the elbow level. Following the subtraction of the latency differences, the forearm
motor nerve conduction velocities were subsequently calculated. All responses were
confirmed to be supramaximal, defined as maximal when the amplitude of the CMAP ceased
to increase despite further increments in stimulus intensity. Distal motor latency
(DML) from onset and CMAP amplitude were measured.
Bias
Given the variability in detection techniques, proficiency levels, and expertise among
different electrophysiologists, this study invited a single physician to perform all
electrophysiological tests on the participants. The testing environment and instruments
were maintained consistently throughout the study.
Carpal Tunnel Syndrome Grading and Criteria
We used Bland's electrophysiological grading scale[13]
[14] ([Table 1]). The SNCV of W-P slowed down > 10 m/s than D3-P, M-U (D2-W compare D5-W) latency
delayed ≥ 0.5ms, M-U ringdiff latency delayed ≥ 0.5ms, were considered abnormal.[11]
Table 1
Neurophysiological grading scale for carpal tunnel syndromes
|
Grade of CTS
|
EDX abnormality
|
|
1-very mild
|
CTS detected only in two sensitive tests (e.g., SNCV slower of W-P, M-U, M-U ringdiff)
|
|
2-mild
|
Median DML < 4.5 ms with the SNCV of D3-W < 40 m/s
|
|
3-moderately severe
|
Median DML ≥ 4.5ms and < 6.5 ms with preserved SNAP
|
|
4-severe
|
Median DML ≥ 4.5 ms and <6.5 ms with absent SNAP
|
|
5-very severe
|
Median DML ≥ 6.5 ms with CMAP ≥ 0.2 mV
|
|
6-extremely severe
|
Median CMAP < 0.2mV
|
Abbreviations: CMAP, compound muscle action potential; CTS, carpal tunnel syndrome;
DML, distal motor latency; SNAP, sensory nerve conduction amplitude; SNCV, sensory
nerve conduction velocities.
Statistical Analysis
Statistical analyses were performed using STATA 11 for Windows. The chi-square test
was conducted to evaluate the consistency of the NCS outcome proportions across individual
digits and aggregated values. To examine intergroup differences in the NCS results,
an Independent Samples t-test was employed. Furthermore, an analysis was undertaken to assess the concurrent
criterion validity between different NCS methodologies, with the correlation strength
determined using the Pearson correlation coefficient for interval data or Spearman's
rank correlation for ordinal variables. Each test for detecting CTS used a receiver
operating characteristic curve, and the area under the curve (AUC) with a 95% confidence
interval (95% CI) was compared for each test.[15] Statistical significance was set at p < 0.05.
Results
A total of 91 patients with CTS (male/female: 24/67, mean age: 51 years) and 68 HCs
(male/female: 19/49, mean age: 45 years) were enrolled in this study. Of the 91 CTS
patients,71 (71/91, 78.02%) had bilateral CTS and 20 (20/91, 21.98%) had unilateral
CTS. The EDX results of 162 hands (right, 86; left, 76) in the CTSs and 136 hands
in the HCs were recorded. In the cohort of 162 hands, 79 hands were categorized as
very mild and 83 as moderately severe. The demographic characteristics and clinical
manifestations of all the patients with CTS are summarized in [Table 2].
Table 2
Demographics and clinical characteristics of carpal tunnel syndromes
|
Variables
|
N = 91, hands = 162
|
|
Age, median (range) years
|
50.62 (23–81)
|
|
Gender
|
|
|
Males
|
24 (26.37)
|
|
Females
|
67 (73.63)
|
|
Unilateral vs. bilateral
|
|
|
Unilateral
|
20 (21.98)
|
|
Bilateral
|
71 (78.02)
|
|
Side
|
|
|
Right
|
86 (53.09)
|
|
Left
|
76 (46.91)
|
|
Grade
|
|
|
Mild
|
79 (48.77)
|
|
Moderately severe
|
83 (51.23)
|
Electrophysiological Results
SNCV from the D1/D2/D3/palm to the wrist of the CTSs was significantly slower than
that of the HCs. The DML of the median nerve in the CTSs group was longer than that
of the HCs. Median (D2)-ulnar(D5) and median-ulnar (both recorded at ring finger)
sensory latency difference of CTSs were longer than 0.5 ms, and significantly longer
than that of HCs (p < 0.05). SNCV from D3 to the palm in HCs was slower than that from the palm to the
wrist, while it was the opposite in CTSs ([Table 3]).
Table 3
Electrophysiological results in healthy controls and carpal tunnel syndromes
|
Parameters
|
HCs
|
CTSs
|
t/Z
|
p-Value
|
|
DML (ms)
|
3.44 (3.25, 3.67)
|
4.50 (3.94, 5.09)
|
−12.54
|
0.000[a]
|
|
D1-W (m/s)
|
62.47 ± 6.61
|
43.08 ± 8.69
|
21.33
|
0.000[a]
|
|
D2-W (m/s)
|
66 (62.65, 69.30)
|
48.20 (42.70, 53.40)
|
12.94
|
0.000[a]
|
|
D3-W (m/s)
|
64.14 ± 5.76
|
45.99 ± 7.84
|
22.40
|
0.000[a]
|
|
W-E (m/s)
|
73.10 (71.40, 76.15)
|
72.60 (71.10, 74.93)
|
1.42
|
0.16
|
|
M-U (ms)
|
0.27 (0.16, 0.37)
|
1.05 (0.79, 1.59)
|
−14.64
|
0.000[a]
|
|
M-U ringdiff (ms)
|
0.12 (0.07, 0.29)
|
1.05 (0.65, 1.75)
|
−14.55
|
0.000[a]
|
|
P-W (m/s)
|
74.1 (67.45, 79.65)
|
48.10 (43.20, 52.20)
|
14.35
|
0.000[a]
|
|
D3-P vs P-W (m/s)
|
−9.15 (−15.4, −3.6)
|
11.05 (4.50, 15.12)
|
−13.37
|
0.000[a]
|
Abbreviations: CTS, carpal tunnel syndrome; DML, the distal motor latency of median;
D1-W/D2-W/D3-W, SNCV from D1/D2/D3 to wrist; W-E, SNCV from wrist to elbow; M-U, median
(D2)-ulnar (D5) sensory latency difference; M-U ringdiff, median-ulnar (both recorded
at ring finger) sensory latency difference; P-W, SNCV from palm to wrist; D3-P versus
P-W: the SNCV difference between D3 to the palm and palm to the wrist (P–W);
a Significantly different when comparing CTSs and HCs, p < 0.05.
Electrophysiological Diagnostic Profiles in Carpal Tunnel Syndromes
All techniques had a high specificity for the diagnosis of CTSs. The M-U and M-U ringdiff
showed higher sensitivity and specificity than the other techniques. The AUC of the
M-U study was 0.992 (95% CI: 0.983–1.00), which was significantly higher than that
of DML (p = 0.000), D1-W (p < 0.05), D2-W (p < 0.05), D3-W (p = 0.005), and D3-PvsP-W (p < 0.001). The AUC of the M-U ringdiff was 0.989 (95% CI: 0.979–0.999), which was
significantly higher than that of DML (p = 0.000), D1-W (p < 0.05), D2-W (p < 0.05), and D3-PvsP-W (p = 0.0013). The electrophysiological diagnostic profiles of all the patients with
CTS are summarized in [Table 4].
Table 4
Electrophysiological diagnostic profiles in carpal tunnel syndromes
|
Parameters
|
Sensitivity (%)
|
Specificity (%)
|
PPV (%)
|
NPV (%)
|
AUC
|
|
DML (ms)
|
50
|
100
|
100
|
62.7
|
0.922
|
|
D1-W (m/s)
|
79.6
|
97.1
|
97
|
80
|
0.963
|
|
D2-W (m/s)
|
69.1
|
100
|
100
|
73.1
|
0.969
|
|
D3-W (m/s)
|
73.5
|
99.3
|
99.2
|
75.8
|
0.971
|
|
M-U (ms)
|
98.1
|
97.1
|
97.5
|
97.8
|
0.992[a]
|
|
M-U ringdiff (ms)
|
98.1
|
98.1
|
98.1
|
98.1
|
0.989[b]
|
|
P-W (m/s)
|
62.3
|
98.5
|
98.1
|
68.7
|
0.983
|
|
D3-P vs P-W (m/s)
|
61.1
|
96.3
|
95.2
|
67.5
|
0.950
|
Abbreviations: AUC, area under the curve; NPV, negative predictive value; PPV, positive
predictive value.
a AUC differs significantly when M-U is compared against these parameters: DML (p = 0.000), D1-W (p < 0.05), D2-W (p < 0.05), D3-W (p = 0.005), D3-PvsP-W (p < 0.001).
b AUC shows a significant variation when M-U is compared to the subsequent parameters:
DML (p = 0.000), D1-W (p < 0.05), D2-W (p < 0.05), D3-PvsP-W (p = 0.0013).
Electrophysiological Diagnostic Profiles in Very Mild Carpal Tunnel Syndromes
As at least two sensitivity tests were used to diagnose very mild CTS, we selected
four sensitivity tests and studied them in pairs. The combinations were as follows:
method 1: combined D3-W and M-U; method 2: D3-W and M-U ringdiff; method 3: D3-W and
D3-PvsP-W; method 4: M-U and M-U ringdiff; method 5:M-U and D3-PvsP-W; and method
6: M-U ringdiff and D3-PvsP-W. Finally, Method 4 showed higher sensitivity than the
other methods, and the AUC area was significantly larger than that of any other method
(p = 0.000) ([Table 5]).
Table 5
Electrophysiological diagnostic profiles of methods in very mild carpal tunnel syndromes
|
Parameters
|
Sensitivity (%)
|
Specificity (%)
|
PPV (%)
|
NPV (%)
|
AUC
|
95% CI
|
|
Method 1
|
50.6
|
100
|
100
|
77.7
|
0.753
|
0.698–0.809
|
|
Method 2
|
50.6
|
100
|
100
|
77.7
|
0.753
|
0.698–0.809
|
|
Method 3
|
32.9
|
100
|
100
|
72.0
|
0.665
|
0.612–0.717
|
|
Method 4
|
94.9
|
100
|
100
|
97.1
|
0.975[a]
|
0.950–0.999
|
|
Method 5
|
53.2
|
100
|
100
|
78.6
|
0.766
|
0.710–0.821
|
|
Method 6
|
53.2
|
100
|
100
|
78.6
|
0.766
|
0.710–0.821
|
Abbreviations: AUC, area under the curve; CI, confidence interval; NPV, negative predictive
value; PPV, positive predictive value.
a Comparing Method 1 with Methods 2 to 6 reveals a significant difference in AUC (p = 0.000).
Discussion
CTS continues to be one of the most recognized and common types of median nerve compression,
accounting for approximately 90% of all entrapment neuropathies.[2] Prolonged exposure to vibrations or forceful repetitive motions is believed to be
the primary cause of CTS. Additionally, certain conditions such as diabetes, pregnancy,
and severe obesity may increase the risk of developing this syndrome.[16]
[17] According to several previous studies, early-stage CTS will disturb the patient's
sleep by a pronounced sensation of hand numbness accompanied by a feeling similar
to swelling, despite the absence of visible swelling.[4] Therefore, timely identification and intervention are crucial to alleviate patient
discomfort and enhance prognosis. In this study, we aimed to explore a highly sensitive
and specific electrophysiological diagnostic combination for CTS.
Of the 91 CTS patients included in this study, 67(73.63%) were female patients, which
affirmed that CTS was more frequent in women.[18] This may be related to the fact that women undertake more handwork (such as housekeeping
and textile work). In this study, 71 (71/91, 78.02%) patients had bilateral CTS, which
was much more than unilateral CTS. This has also been mentioned in a previous review.[2]
In this study, most electrophysiological test results were significantly different
between the CTSs and HCs. In previous studies, the NCS demonstrated high validity
and reliability in confirming the clinical diagnosis of CTS, with a sensitivity exceeding
85% and a specificity of 95%.[19] In this study, all EDX techniques showed high specificity (>96.3%), positive predictive
value (>95.2%), and large AUC (0.922 as the smallest) in the diagnosis of CTSs. Among
these tests, D1-W, D3-W, M-U, and M-U ringdiff had a higher sensitivity (>70%). In
particular, for the M-U and M-U ringdiff, the sensitivity reached 98.1%. Prior research
has confirmed that when comparing median nerve distal sensory latencies with those
of the radial or ulnar nerves, sensory studies exhibit greater sensitivity than motor
studies.[20] In this study, the DWL with 50% sensitivity was much lower than that of all sensory
conduction studies. We considered that EDX techniques of sensory nerves were still
the most sensitive EDX tests for diagnosing CTS, which is consistent with the results
of a previous study.[21] Due to its high sensitivity and specificity, comparing the NCS of the median with
that of the ulnar nerve in the same hand (M-U and M-U ringdiff) represents the most
precise EDX method for diagnosing CTS. Comparing distal segment sensory NCS of the
median nerve to median sensory nerve conduction through the carpal tunnel in the same
limb proved ineffective for diagnosing patients with clinically suspected CTS. Although
the AUC area of D3-P vs. P-W was large (0.950), the observed sensitivity was notably
below the anticipated level (only 61.1%). These findings align with those of previous
investigations.[7] Conversely, some studies have indicated that segmental studies of the median nerve
are more sensitive than comparative studies.[21]
[22] Variations in outcomes might stem from the differing levels of disease severity
among the study participants and distinct methodologies, including varying abnormal
test cutoff points and statistical procedures.
According to Bland's electrophysiological grading scale,[13] at least two sensitive abnormal EDX tests can diagnose very mild CTS. Several sensitive
EDX tests we selected from the AANEM guidelines.[6] As is well understood, employing two complementary comparison methods that concur
reduces the likelihood of false positives or false negatives.[11] By pairing these sensitive techniques, we confirmed that the combination of M-U
and M-U ringdiff is the best EDX test combination for CTS diagnosis. This combination
had a higher sensitivity (94.9%) and negative predictive value (97.1%) than other
combinations. It also showed a significantly larger AUC than the other methods (p = 0.000). The diagnostic effectiveness of the M-U and M-U ringdiff in measuring comparative
values in the same hand is attributed to patients being their own controls, with nerve
conduction variations affecting both nerves equally.
Conclusion
Based on the evidence of this study, EDX tests, including sensory and motor conduction
studies, can be used to confirm clinical CTS. If clinical CTS patients exhibit normal
median motor distal latency or SNCV from D1 to D3 to the wrist, a comparison of median
and ulnar nerve conduction through the wrist is recommended, including the M-U and
M-U ringdiff. Although the segmental comparison of the median nerve was more sensitive
than that of the comparative studies in many studies, including the AANEM guidelines,
in this study, the sensitivity was notably below the anticipated level. The scope
of this study was limited to comparing sensory conduction latency between the median
and ulnar nerves. In future studies, we will expand this investigation to include
a comparison with the radial nerve to further enhance our understanding.
