Handbook of Clinical Neurology, Vol.

160 (3rd series)

Clinical Neurophysiology: Basis and Technical Aspects
K.H. Levin and P. Chauvel, Editors
https://doi.org/10.1016/B978-0-444-64032-1.00014-X
Copyright © 2019 Elsevier B.V. All rights reserved

Chapter 14

Nerve conduction studies: Basic concepts

JINNY TAVEE*
Neuromuscular Division, Department of Neurology, Northwestern Feinberg School of Medicine, Chicago, IL, United States

Abstract
Nerve conduction studies (NCSs) are an essential tool in the evaluation of the peripheral nervous system.
The sensory nerve action potential (SNAP) provides information on the sensory nerve axon and its path-
way from the distal receptors in the skin to the dorsal root ganglia, while the compound muscle action
potential (CMAP) is an assessment of the motor nerve fibers from their origins in the anterior horn cell
to their termination along muscle fibers. Various parameters of the SNAP and CMAP waveforms are used
to determine the number of functioning nerve fibers and the speed of conduction. Similarly, specific elec-
trodiagnostic patterns involving SNAP and CMAP amplitudes, latencies and other measurements can help
discern the underlying nerve pathophysiology as either axon loss or demyelinating in nature. Numerous
technical and environmental factors can affect the NCS and should be recognized and corrected if possible.
Finally, while basic NCSs are a noninvasive and low-risk procedure, safety issues for patients with
implanted electrical devices should be considered.

INTRODUCTION
For the patient presenting with sensory changes, weak-
ness, or other potential manifestations of peripheral ner-
vous system disease, nerve conduction studies (NCSs)
provide invaluable information regarding the integrity
of myelinated nerve fibers. In conjunction with the needle
electrode examination, which is the other main component
of the electrodiagnostic examination (EDX), the NCS can
be used to objectively confirm the presence of a lesion,
characterize the underlying pathophysiology, and deter-
mine degree of severity. They can also help with lesion
localization and prognostication. However, it is important
to keep in mind that an NCS or even a complete EDX
cannot supplant a careful history and neurologic examina-
tion, which should be performed prior to each study. In
addition, NCSs do not evaluate unmyelinated nerve fibers.
Thus they are not the ideal modality for studying pain or
autonomic dysfunction, which are mediated through
thinly myelinated A-delta and unmyelinated C fibers.
For patients presenting with these symptoms, specialized
small nerve fiber testing may be considered.
Routine NCSs are composed of motor and sensory
nerve responses. The mixed NCS, which simultaneously
assesses both motor and sensory components, is used in
the evaluation of specific diagnoses, such as carpal tun-
nel syndrome (CTS). Late responses and other special-
ized studies are discussed elsewhere. During the basic
procedure, an electrical stimulus is applied to a nerve
generating an action potential that is recorded from a
muscle (motor NCS) or separate site along the nerve
itself (sensory NCS.) The recorded responses or wave-
forms are evaluated for various parameters that indicate
the number of functioning axons and speed of the fastest
nerve fibers. The results are then compared to normative
values, which may be established in individual laborato-
ries or obtained from published reference data if similar
measurement protocols are used. When an NCS is per-
formed on the unaffected contralateral limb, side-to-side
comparisons can be made. This is especially helpful in
younger patients whose individual normative range for
nerve values (e.g., higher amplitudes) may exceed that
of the general population.

*Correspondence to: Jinny Tavee, M.D., Associate Professor, Medical Director, Neuromuscular Division, Director, EMG Neuro-
diagnostic Testing Center, Feinberg School of Medicine, Department of Neurology, Abbott Hall Suite 1114 710 N Lakeshore Drive,
Chicago, IL 60611, United States. Tel: +1-312-695-7950, E-mail: jinny.tavee@nm.org
218
The use of NCS protocols can also be helpful in
detecting a suspected lesion and excluding other diagno-
ses being considered in the differential. While there
should be a minimum set of sensory and motor NCSs
for each referring diagnosis, protocol deviations may
be necessary depending on the findings seen throughout
the procedure.
MOTOR NERVE CONDUCTION STUDIES
In most EDX laboratories, a standard motor NCS is per-
formed using an external handheld stimulator with two
prongs, consisting of a cathode (negatively charged)
and an anode (positively charged). The prongs are at a
fixed distance of about 3–4 cm, which is close enough
to direct the current flow down toward the nerve rather
than toward the surrounding tissues and far enough apart
so that the current does not pass between the stimulating
electrodes. The surface recording electrodes consist of an
active electrode (G1) and a reference electrode (G2),
which are often in the form of flat surface disks that
can be taped onto the skin. The active electrode is placed
over the muscle belly at the end-plate region, and the ref-
erence electrode is positioned more distally over the ten-
don. This is commonly known as the belly tendon
technique. A ground electrode is placed between the
stimulating and recording electrodes, where it serves to
reduce stimulus or shock artifact.
Once the skin is prepared, the stimulating electrodes
are pressed down against the skin directly over and
along the course of the motor nerve with the cathode
closest to the recording electrode (Fig. 1). This helps
to ensure that the action potential generated under the
cathode will propagate along the nerve axon and even-
tually reach the recording electrodes. If the stimulating
electrodes are reversed, the hyperpolarization that
Fig. 14.1. Motor nerve conduction studies. Median motor nerve stimulation is performed at a distal (A) and proximal site (B) with
recording electrodes placed over the belly of the abductor pollicis brevis muscle and distal tendon. G1 is the active recording
electrode and G2 is the reference recording electrode. Note that the cathode (darker stimulator prong) is positioned closest to
the recording electrode.
J. TAVEE
normally occurs under the anode results in blocked con-
duction of the nerve impulse, a phenomenon known as
anodal block.
Stimulation is initiated at 0 mA and slowly increased
(usually up to 50 mA) until a maximal response is
obtained. The resultant biphasic waveform is the com-
pound muscle action potential (CMAP). The CMAP
represents the summated muscle fiber action potentials
within the muscle being recorded. When the size
of the CMAP remains the same despite the use of
higher current, the intensity is increased an additional
20%–25% to confirm supramaximal stimulation of
the nerve.
For a routine motor NCS, stimulation is performed at
two different sites along the course of the nerve, resulting
in a distal and proximal waveform for each muscle. In
most laboratories, the upper extremity protocol consists
of median and ulnar nerve stimulation at the wrist and
elbow, recording from the abductor pollicis brevis and
abductor digiti minimi, respectively. In the lower extrem-
ity, peroneal and tibial stimulation is performed at the
ankle and knee while recording from the extensor digi-
torum brevis (peroneal) and abductor hallucis (tibial)
muscles.
The main CMAP parameters that are used for analysis
include the amplitude, latency, and conduction velocity
(Fig. 14.2). The CMAP amplitude (millivolts) is mea-
sured from baseline to negative peak and reflects the
number of motor nerve fibers responding to the stimulus.
The latency (milliseconds) is the time from the moment
of stimulation to the onset or “takeoff” of the negative
phase of the CMAP. Only the latency from stimulation
at the distal site, known as the distal or onset latency,
is reported in routine NCSs. The proximal latency is used
for calculating conduction velocity. While the distal
latency represents the time required for activation of
 NERVE CONDUCTION STUDIES: BASIC CONCEPTS 219

Amplitude

Stimulus

Distal
latency
Duration
Fig. 14.2. Compound muscle action potential (CMAP). The
commonly measured components of the CMAP include the
onset or distal latency, which is the time between the stimulus Fig. 14.3. Sensory nerve conduction studies (NCS). Median
artifact and the onset of the negative phase; the amplitude, sensory NCS is performed antidromically, stimulating the
which is measured from baseline to negative peak; and the median nerve at the wrist, while recording distally from ring
duration of the negative phase. electrodes placed over the second digit. G1 is the active record-
ing electrode and G2 is the reference recording electrode. Like
the motor NCS, the cathode is closest to the recording
the fastest conducting nerve fibers, it also includes the electrodes.
time interval for nerve conduction through the neuro-
muscular junction and then muscle fiber activation. Peak
The conduction velocity (meters per second) is the latency
speed of the fastest conducting nerve fibers and is calcu-
lated by dividing the distance between the two stimula-
tion sites by the difference between the proximal and Baseline-to-peak
distal latencies. Because the latencies (and thus the time amplitude
for neuromuscular junction transmission and muscle
fiber activation) do not factor into the final calculation,
the conduction velocity more closely reflects the conduc-
tion time of the fastest motor nerve fibers. In the normal
adult, conduction velocity in the upper extremity is typ-
ically 50 m/s, while that of the lower extremity is 40 m/s. Fig. 14.4. Sensory nerve action potential (SNAP). The two
Other measurements include the CMAP area, which is commonly reported SNAP parameters are the amplitude mea-
sured from baseline to peak, and the peak latency, which is the
calculated by the computer and is actually more accurate
time from stimulation to the maximum negative peak.
than the amplitude in its representation of the total num-
ber of activated motor fibers, because it includes all of
The two primary SNAP measurements are the ampli-
the nerve fibers rather than just the more synchronized
tude and peak latency (Fig. 14.4). The amplitude (micro-
ones. However, it is not as commonly reported as
volts) is measured from baseline to negative peak and
amplitude. Waveform duration, which is a measure of
reflects the number of sensory nerve fiber action poten-
synchrony, can be helpful in the diagnosis of a number
tials. The peak latency (milliseconds) is the time from the
of conditions, particularly demyelinating neuropathies
moment of nerve stimulation to the negative peak of the
and critical illness myopathy (Goodman et al., 2009).
response. Other parameters include the onset latency,
which is measured from the time of nerve stimulation
to the waveform’s departure from baseline and represents
SENSORY NERVE CONDUCTION
the fastest conducting sensory fibers. Duration of the
STUDIES
SNAP is measured from the waveform’s departure from
For a sensory NCS, both stimulation and recording baseline to its final return to baseline, and is shorter than
occur along the nerve (Fig. 14.3). As with motor NCS, that of the CMAP, which can help distinguish the two in
the cathode is positioned closest to the recording elec- cases where one may be absent. If the onset latency is
trode, and stimulation intensity is gradually increased used rather than the peak latency, the conduction velocity
until there has been supramaximal stimulation (usually can be calculated with stimulation at only one site. Prox-
only 30–40 mA is required). The resultant waveform is imal stimulation is not routinely performed with sensory
the sensory nerve action potential (SNAP), a biphasic NCSs due to the effects of temporal dispersion and phase
or triphasic waveform that represents summated nerve cancellation, which result in a SNAP of reduced ampli-
action potentials. tude and prolonged duration.
220 J. TAVEE
Unlike the magnification effect seen with the CMAP,
which represents activated muscle fibers as opposed to
motor nerve fibers, the SNAP is a direct measurement
of the nerve action potential and is a hundred times smal-
ler. This leads to an increased susceptibility to technical
factors and background noise. Special consideration is
therefore warranted in the choice of technique and
parameters used for analysis when performing a sensory
NCS. For instance, because the SNAP amplitude is so
small, it is often challenging to identify the exact onset,
which affects the accuracy of the latency measurement.
This is why the peak latency is more commonly used than
the onset latency. Furthermore, the sensory nerve con-
duction velocity may also be inaccurate given that it is
calculated from the onset latency. The result is that in
many laboratories technical factors preclude the routine
reporting of onset latency and conduction velocity with
sensory NCSs.
A common protocol for routine sensory NCSs in the
upper extremity consists of median, ulnar, and superficial
radial sensory stimulation at the wrist while recording
from the second digit (median), fifth digit (ulnar), and ana-
tomic snuffbox along the dorsal aspect of the wrist and
hand (radial), respectively. In the lower extremity, the sural
and superficial peroneal sensory nerves are stimulated at
the calf and recorded at the ankle or dorsum foot.
Antidromic vs orthodromic technique
As depolarization results in bidirectional conduction of
the nerve impulse, a sensory NCS can be performed
either orthodromically (physiologic direction of nerve
conduction) or antidromically. In most laboratories, the
antidromic technique is preferred, mainly because it
results in larger SNAP amplitudes. With antidromic sen-
sory NCSs, the nerve is stimulated proximally and the
response recorded distally where the nerve is closer to
the surface. The shorter distance between the nerve
and recording electrode results in a higher SNAP ampli-
tude. The reverse is true with the orthodromic method, in
which the nerve is stimulated distally and recorded prox-
imally, where the nerve may be deeper underneath the
surface. One disadvantage of the antidromic technique
is that CMAPs and SNAPs may be evoked simulta-
neously by proximal nerve stimulation, corrupting the
measured distal response. Careful evaluation of the
waveform features of morphology and duration can help
identify the response as being motor or sensory.
MIXED SENSORIMOTOR NCS
Like sensory NCS, mixed NCSs are direct studies in
which the mixed nerve action potential (MNAP) that is
generated with stimulation represents the summated
action potentials of individual sensory and motor nerve
fibers. While it is also expressed in microvolts, the
MNAP amplitude is usually higher than that of the SNAP
obtained from the same nerve segment, as both the sen-
sory and motor components of the nerve are included in
the MNAP measurement. Its use in most laboratories is
limited to palmar studies for the evaluation of CTS and
plantar studies for tarsal tunnel syndrome and polyneuro-
pathy. An advantage of the mixed NCS is inclusion of Ia
muscle afferents, which are heavily myelinated and are
among the first fibers to be affected in demyelinating
nerve lesions (Preston and Shapiro, 2013).
COMMON FACTORS AFFECTING NCS
In everyday practice, a number of technical and physio-
logic factors can potentially lead to altered waveforms
and spurious findings. Constant vigilance and trouble-
shooting in addition to a standardized approach to the
procedure are needed to ensure optimal results. The fol-
lowing factors can influence the NCS.
Temperature
The cool limb is one of the most common problems seen
in the EDX laboratory and can affect both the NCS and
needle electrode examination. Larger amplitudes, longer
durations, prolonged latencies, and slowed conduction
velocities are changes seen with reduced limb tempera-
tures (Fig. 14.5). Physiologically, cold temperatures lead
P
25.6°C
33.2°C
P
20 µV
2 ms
Fig. 14.5. Temperature effect. Median sensory nerve action
potential obtained at 25.6°C (dashed lines) and 33.2°C (solid
line). At cooler temperatures, delayed sodium channel activa-
tion results in slowed conduction times (note the prolonged
SNAP peak latency), while delayed inactivation results in
increased amplitudes.
 NERVE CONDUCTION STUDIES: BASIC CONCEPTS
to slowed opening and delayed inactivation of sodium
channels. The former affects the rate of nerve conduc-
tion, while the latter results in larger amplitudes (Louis
and Hotson, 1986; Kimura, 2001). Limb temperature
should be monitored during the procedure and main-
tained at or near the ideal temperature of 33°C. Warming
the limb may be accomplished by soaking the limb in
warm water, wrapping it in a warm towel or using an
infrared heat lamp prior to the study. While mathematical
corrections exist for latency and conduction velocity, the
result is an estimation that lacks true accuracy.
Age
Both amplitude and conduction velocity are affected by
age due to factors including nerve fiber attrition, loss of
myelin, and age-related changes of the nerve and muscle
membrane (Lederman, 2000). Nerve conduction veloci-
ties in the full-term infant are half the adult value and
reach the adult range by 3–5 years, correlating with the
degree of myelination (Thomas and Lambert, 1960;
Kimura, 2001). After the age of 30, conduction velocity
slows each decade, with up to a 10% decrease by 60 years
of age (Mayer, 1963). Similarly, there is a reduction in
recorded amplitude with each decade, especially after
age 60. This is particularly apparent for sensory responses
in the distal lower extremity (Falco et al., 1992). In one
study, the sural nerve response was absent in healthy
adults over the age of 74 years (Tavee et al., 2014).
221
Recording electrode placement
When performing a routine motor NCS using the belly
tendon technique, an initial positive wave or “dip” pre-
ceding the CMAP elicited at all sites of stimulation indi-
cates that the electrode may not be placed directly over
the belly of the muscle. The position of the recording
electrode should be adjusted until the positive wave dis-
appears or is minimized.
Submaximal/supramaximal stimulation
Supramaximal nerve stimulation is essential to perform-
ing quality NCSs, as this leads to synchronized activa-
tion of all nerve fibers and an accurate waveform for
analysis. During a routine NCS, the gradual increase
in stimulation intensity results in a corresponding rise
in SNAP or CMAP amplitude as more nerve fibers are
activated, until the nerve is supramaximally stimulated
(Fig. 14.6). To a lesser extent, the latency time also
shortens with increasing intensity. Submaximal stimula-
tion can thus result in falsely reduced amplitudes and
mildly prolonged latencies. Increasing stimulus intensity
beyond just supramaximal or slippage of the stimulating
electrodes off the nerve can lead to current spread to
neighboring nerve trunks, affecting waveform configu-
ration and amplitude.

Height
Slowed conduction velocities and prolonged latencies
are seen in taller individuals, likely a factor of limb length
and cooler temperatures in more distal portions of the
limbs (Kimura, 2001).

Sex
Women have higher amplitudes and faster conduction
velocities than men. While the latter may be attributed
to height, it is unclear why the amplitude is higher.

Body habitus
Nerve response amplitudes are directly related to the dis-
20 µV

tance between the nerve and recording electrodes. This is

most commonly encountered in patients with peripheral
edema or large limb size, in whom low amplitude or
absent sensory responses can be misconstrued as patho-
logical. In such situations, increasing the duration of the
stimulus or increasing the separation between the stimu-
lating electrodes will increase the penetration of current
and possibly improve the response.
2 ms
Fig. 14.6. Effect of stimulation intensity. The effect of
increasing stimulus strength on the recorded SNAP, from sub-
threshold to supramaximal. Reproduced with permission from
Lederman, R.J. Nerve conduction studies. In: Levin, K.H. and
L€uders, H., 2000. Comprehensive clinical neurophysiology.
Saunders, Philadelphia, pp. 89–111.
222 J. TAVEE
Nerve anomalies
Anomalous innervations are anatomic variants that can
result in spurious waveforms that can be misinterpreted as
pathological if not recognized. One of the most commonly
encountered nerve anomalies is the Martin-Gruber anasto-
mosis, which has been reported in 15%–30% of anatomic
studies (Erdem et al., 2002; Rodriguez-Niedenf€ uhr et al.,
2002). With this anomaly, nerve fibers destined for ulnar
innervated hand muscles are misdirected at the brachial
plexus level into the median nerve trunk until the forearm,
where they cross over to the ulnar nerve trunk. If crossing-
over fibers innervate the abductor digiti minimi muscle,
routine ulnar studies demonstrate a higher ulnar CMAP
with wrist stimulation compared with stimulation above
the elbow, because crossing-over fibers are stimulated at
the wrist but not at the elbow. This pattern can be misinter-
preted as a pathological conduction block along ulnar nerve
fibers at the elbow. A “cross-over study” in which the
median nerve is stimulated and the abductor digiti minimi
is recorded will confirm the cross over and rule out ulnar
Fig. 14.7. Electrodiagnostic patterns. (A) Normal motor nerve
neuropathy at the elbow. conduction studies with distal (S1) and proximal
(S2) stimulation. (B) Demyelinating conduction block lesion
Other considerations results in a normal amplitude with stimulation distal to the
Other sources of artifact and their results include mea- lesion and reduced amplitude with proximal stimulation.
(C) Demyelinating conduction block with nonuniform slowing
surement errors (inaccurate latency and conduction
results in a similar pattern as (B) with temporal dispersion.
velocity), cathode–anode reversal (reduced amplitude (D) Axon loss lesion results in diffusely reduced amplitudes.
and prolonged latency with normal conduction velocity),
filter settings (amplitude and latency changes), and
excess stimulation (increased shock artifact, costimula- loss is sufficiently severe to involve the fastest nerve
tion of adjacent nerves, and shortened latency due to cur- fibers. However, when it does occur, slowing is mild
rent jump distal to the site of stimulation). in proportion to the degree of amplitude reduction.
In the case of a traumatic or sudden axon loss lesion in
PATHOPHYSIOLOGY which the nerve is transected, infarcted, or severely
Nerve pathophysiology may be characterized as either injured, wallerian degeneration ensues and the segment
axon loss or demyelination. Axon loss causes failure of the nerve distal to the lesion undergoes rapid degener-
of nerve transmission and eventually results in degener- ation in a proximal to distal (anterograde) fashion. In the
ation of the entire portion of the nerve distal to the lesion. first few days following acute nerve injury, normal sen-
In contrast, demyelination results in either conduction sory and motor nerve amplitudes will be elicited distal to
block or conduction slowing at the site of the lesion the site of the lesion. This is because the distal nerve seg-
and remains focal with no involvement of the segments ment has not yet fully degenerated and can still conduct
proximal or distal to the lesion. Both types have specific nerve impulses. After acute nerve transection, SNAP
manifestations on NCSs, although some lesions may amplitude begins to decline after day 5 and is complete
demonstrate combined pathophysiologic features, with by day 9–11. CMAP amplitude begins to decline after
one usually predominating over the other. day 3 and is complete by day 5–8, faster than SNAP
decline owing to loss of axon transport and neuromuscu-
Axon loss lar junction transmission, which fail before nerve action
potential conductivity (Chaudhry and Cornblath, 1992).
Axon loss lesions are the most common type of nerve
pathophysiology encountered in the EDX laboratory.
Demyelination
Axon loss is characterized by reduced or absent CMAP
and/or SNAP amplitudes (Fig. 14.7). Slowing of the dis- Demyelinating lesions have more varied manifestations
tal latency or conduction velocity only occurs when axon on NCSs and can affect any aspect of the nerve response.
 NERVE CONDUCTION STUDIES: BASIC CONCEPTS
In lesions characterized by nonfocal, uniform demyelin-
ation, conduction velocities will be slowed equally
without change in amplitude. With nonuniform demye-
lination some nerve fiber action potentials arrive at the
recording site before others, causing dispersion of the
CMAP or SNAP, with effects on amplitude and duration
(Fig. 14.7).
Research criteria for defining peripheral nerve demy-
elination have been suggested. These criteria include
latency prolongation >125% and conduction velocity
slowing to <80% of normal values (Dimachkie and
Barohn, 2013). When CMAP amplitude is less than
80% of normal, then the latency and conduction velocity
requirements are >150% and <70%, respectively
(Dimachkie and Barohn, 2013).
Like axon loss lesions, demyelinating conduction
block prevents nerve impulse transmission at the site
of the lesion. On NCS, this is traditionally defined along
motor nerve fibers and is manifested as reduced ampli-
tude on stimulation of the nerve proximal to the lesion,
and normal amplitude distal to the lesion (Fig. 14.7).
Definite conduction block is generally defined as a
50% or greater difference in amplitude between distal
and proximal stimulation sites.
SAFETY
Basic NCSs are noninvasive and of relatively low risk, as
nerve stimulation and recording are carried out on the
surface of the skin. However, there are a few absolute
and relative contraindications to the routine NCS. In
patients with external cardiac pacing wires, surface nerve
stimulation poses a significant risk of electrical injury to
the heart and should not be performed (AANEM, 2014).
Additionally, clinical judgment is recommended in
patients with recent acute myocardial infarction due to
the potential risk of inducing or exacerbating arrhythmia
(AANEM, 2014).
For patients with an internal cardiac pacemaker pro-
grammed to a bipolar sensing configuration or an
implanted cardiac defibrillator, routine NCSs are consid-
ered relatively safe (Schoeck et al., 2007; Gechev et al.,
2016). However, for those with an internal pacemaker
programmed to a unipolar sensing configuration, the
potential risks of electromagnetic interference should
be carefully considered, especially with more specialized
studies, which include proximal stimulation at the supra-
clavicular fossa or repetitive nerve stimulation studies
(Schoeck et al., 2007; Cronin et al., 2013; Gechev
et al., 2016). Caution is also advised in patients with
an implanted deep brain or spinal cord stimulator,
although NCSs have been safely performed in both cases
(AANEM, 2014; Gechev et al., 2016). A device check by
223
a trained specialist before and after the procedure is help-
ful as electromagnetic interference may inadvertently
turn the implanted device on or off.
SUMMARY
Routine NCSs are a safe and noninvasive means by
which the peripheral nervous system may be examined.
An understanding of the various electrophysiologic
patterns and technical factors that can affect NCS is inte-
gral to accurate interpretation of the findings and ulti-
mately formulation of the diagnosis. The determination
of nerve pathophysiology in particular is a key compo-
nent of the EDX that should be included in the final
report, as this information has implications for prognosis
and treatment.
REFERENCES
American Association of Neuromuscular and Electrodiagnostic
Medicine (AANEM) (2014). Risks of electrodiagnostic
medicine 2014, pp 1–8, http://www.aanem.org/getmedia/
50f4dd83-835c-46cb-a832-930851440e9e/risksinEDX.pdf.
aspx. Accessed 20 January 2019.
Chaudhry V, Cornblath DR (1992). Wallerian degeneration in
human nerves: serial electrophysiological studies. Muscle
Nerve 15 (6): 687–693.
Cronin EM, Gray J, Abi-Saleh B et al. (2013). Safety of
repetitive nerve stimulation in patients with cardiac implant-
able electronic devices. Muscle Nerve 47 (6): 840–844.
Dimachkie MM, Barohn RJ (2013). Chronic inflammatory
demyelinating polyneuropathy. Curr Treat Options
Neurol 15 (3): 350–366.
Erdem HR, Ergun S, Erturk C et al. (2002). Electrophysiological
evaluation of the incidence of Martin-Gruber anastomosis
in healthy subjects. Yonsei Med J 43 (3): 291–295.
Falco FJ, Hennessey WJ, Braddom RL et al. (1992).
Standardized nerve conduction studies in the upper limb
of the healthy elderly. Am J Phys Med Rehabil 71 (5):
263–271.
Gechev A, Kane NM, Koltzenburg M et al. (2016). Potential
risks of iatrogenic complications of nerve conduction
studies (NCS) and electromyography (EMG). Clin
Neurophysiol Pract 1: 62–66.
Goodman BP, Harper CM, Boon AJ (2009). Prolonged com-
pound muscle action potential duration in critical illness
myopathy. Muscle Nerve 40 (6): 1040–1042.
Kimura J (2001). Electrodiagnosis in diseases of nerve and
muscle: principles and practice, Oxford university press.
Lederman RJ (2000). Nerve conduction studies. In: KH Levin,
H L€uders (Eds.), Comprehensive clinical neurophysiology,
Saunders, Philadelphia, pp. 89–111.
Louis AA, Hotson JR (1986). Regional cooling of human
nerve and slowed Na+ inactivation. Electroencephalogr
Clin Neurophysiol 63 (4): 371–375.
Mayer RF (1963). Nerve conduction studies in man.
Neurology 13 (12): 1021–1030.
224 J. TAVEE
Preston DC, Shapiro BE (2013). Electromyography and neu-
romuscular disorders: clinical-electrophysiologic correla-
tions, Elsevier Health Sciences.
Rodriguez-Niedenf€uhr M, Vazquez T, Parkin I et al. (2002).
Martin-Gruber anastomosis revisited. Clin Anat 15 (2):
129–134.
Schoeck AP, Mellion ML, Gilchrist JM et al. (2007). Safety of
nerve conduction studies in patients with implanted cardiac
devices. Muscle Nerve 35 (4): 521–524.
Tavee JO, Polston D, Zhou L et al. (2014). Sural sensory
nerve action potential, epidermal nerve fiber density, and
quantitative sudomotor axon reflex in the healthy elderly.
Muscle Nerve 49 (4): 564–569.
Thomas JE, Lambert EH (1960). Ulnar nerve conduction
velocity and H-reflex in infants and children. J Appl
Physiol 15 (1): 1–9.