Journal of the American Academy of Orthopaedic Surgeons 190 Electrodiagnostic evaluation can be useful in distinguishing among a variety of causes for numbness, weakness, and pain. Although most commonly used in diagnos- ing entrapment neuropathies, such as carpal tunnel syndrome and radiculopathies, electrodiagnostic evaluation often plays an impor- tant role in assessing more com- plex conditions. In individuals with severe traumatic neuropa- thies, electromyographic (EMG) and nerve conduction studies can establish a prognosis for signifi- cant functional recovery; those with severe or complete axon loss will have a less favorable outcome than those with evidence of neu- rapraxia. Radial and sciatic nerve lesions are two common examples of this. In patients who present with diffuse numbness and weak- ness, it may be difficult to clinical- ly differentiate central lesions (such as those of motor neuron dis- ease) from peripheral neuropathy or spinal stenosis (cervical and/or lumbar). In many cases, electrodi- agnostic evaluation can establish whether central or peripheral pro- cesses (or both) contribute to a pa- tientÕs symptoms. To provide useful information, the electrodiagnostic examination must include a clinical assessment as well as neurophysiologic testing. The electrodiagnostic medical con- sultation should always start with a directed history and physical examination and should utilize electrophysiologic testing to help answer the diagnostic questions posed by the differential diagnosis considered by the referring physi- cian and the consultant. Diagnoses should not be made solely on the basis of electrophysiologic Òabnor- malities,Ó but rather in the context of the patientÕs complaints. Neurophysiology of Impulse Transmission and Measurement The axon membrane is composed of a lipid bilayer, permeable to water but not to most ions or larger mole- cules. This selective permeability, coupled with the presence of the Na + /K + -ATPÐdependent electro- genic pump, allows for maintenance of a resting membrane potential of 60 to 90 mV, which is negative inside the axon membrane. Sodium ions are accumulated outside the membrane at a concentration about 12 times greater than inside, and potassium ions are concentrated inside the cell, at a concentration about 30 times greater than outside. There are also mechanisms pres- ent to allow the generation of an action potential. An action poten- tial is a traveling depolarization that allows transmission of infor- mation along the nerve. It is gener- ated by a specific set of mecha- Dr. Robinson is Professor of Rehabilitation Medicine, University of Washington School of Medicine, Seattle, and Chief of Rehabilitation Medicine and Director, Electrodiagnostic Medicine Laboratory, Harborview Medical Center, Seattle. Reprint requests: Dr. Robinson, Rehabilitation Medicine, Harborview Medical Center, Box 359740, 325 Ninth Avenue, Seattle, WA 98104. Copyright 2000 by the American Academy of Orthopaedic Surgeons Abstract The electrodiagnostic evaluation assesses the integrity of the lower-motor- neuron unit (i.e., peripheral nerves, neuromuscular junction, and muscle). Sensory- and motor-nerve conduction studies measure compound action potentials from nerve or muscle and are useful for assessing possible axon loss and/or demyelination. Needle electromyography measures electrical activity directly from muscle and provides information about the integrity of the motor unit; it can be used to detect loss of axons (denervation) as well as reinnerva- tion. The electrodiagnostic examination is a useful tool for first detecting abnormalities and then distinguishing problems that affect the peripheral ner- vous system. In evaluating the patient with extremity trauma, it can differen- tiate neurapraxia from axonal transection and can be helpful in following the clinical course. In patients with complex physical findings, it is a useful adjunct that can help discriminate motor neuron disease from polyneuropathy or myeloradiculopathy due to spondylosis. J Am Acad Orthop Surg 2000;8:190-199 Role of Neurophysiologic Evaluation in Diagnosis Lawrence R. Robinson, MD Lawrence R. Robinson, MD Vol 8, No 3, May/June 2000 191 nisms. Specifically, voltage-gated Na + channels based in the axon membrane are activated by partial membrane depolarization; opening of Na + channels allows inflow of Na + ions such that the membrane becomes further depolarized and even briefly hyperpolarized (i.e., relatively positive inside the mem- brane [30 to 40 mV]). Closing of sodium channels and opening of K + channels, with resultant K + efflux, then rapidly brings the membrane back to the resting state and ready for another wave of depolarization after an absolute refractory period (i.e., time during which the nerve cannot be depolar- ized again) of about 1 msec. 1 These sequential depolarizations proceed along the axon membrane. In the absence of myelin (e.g., on autonomic fibers and slow pain fibers), this is a slow process, with a conduction velocity of about 5 to 15 m/sec, depending on axon diame- ter. Myelin, however, allows for faster conduction, as currents jump from one node of Ranvier to the next; saltatory (node-to-node) con- duction speeds of 40 to 70 m/sec are achieved. Most motor and sen- sory fibers in human peripheral nerves are myelinated; the largest- diameter and most heavily myelin- ated fibers are spindle afferents and alpha motor neurons. When a nerve is electrically stimulated, the propagation of these action potentials can be re- corded by using surface electrodes. The voltage at the skin surface for these action potentials ranges from a few microvolts to a few hundred microvolts. A recording from nerve is usually referred to as a compound nerve action potential (CNAP). If the action potential is recorded from a pure sensory nerve, it is referred to as a sensory nerve action potential (SNAP). When motor nerves are stimulated, potentials can be recorded directly from muscle. Because each axon synapses with many muscle fibers, a much larger response is usually produced at the muscle level. The amplitude of the resulting com- pound muscle action potential (CMAP) is typically a few millivolts. If mixed axons are involved (e.g., motor and sensory), the response is best referred to as a CNAP. Principles of Nerve Conduction Studies Sensory- and Mixed-Nerve Conduction Studies Typically, CNAPs and SNAPs are measured by electrically stimu- lating a peripheral nerve and re- cording the response a known dis- tance away. Recording that reflects propagation along the nerve in a physiologic direction (e.g., after stimulating a digital sensory nerve and recording from the wrist) is referred to as Òorthodromic record- ing.Ó However, stimulation of a nerve usually activates the nerve in both directions from the point of stimulation. If recordings are from a nonphysiologic direction (e.g., stimulation of the median sensory nerve at the wrist and recording from a digital nerve), this is re- ferred to as Òantidromic recording.Ó The speed of conduction is the same in either direction. For clinical purposes, there are, in broad terms, usually two measures one makes of CNAPs or SNAPs: (1) speed of conduction (i.e., latency or velocity) and (2) size of the response (i.e., amplitude) (Fig. 1). Traditionally, the speed of conduc- tion for CNAPs and SNAPs has been measured in terms of latency (i.e., the time between the onset of stimulation and either the onset or the peak of the potential). Peak latency is easier to measure, particu- larly when the potential is small or the baseline is noisy. Onset latency, although more difficult to measure, has the physiologic significance of representing the arrival of the impulse via the fastest-conducting nerve fibers at the recording elec- trode. Conduction velocity for CNAPs can be derived by dividing the distance between the stimulation site and the active (G1) electrode by the onset latency, represented by the equation CV = d/t, where CV = con- duction velocity in meters per sec- ond, d = distance between stimula- tion site and recording electrode in millimeters, and t = onset latency in milliseconds. Latency and conduction velocity can be affected by a number of physiologic and pathologic factors. In healthy control subjects, slowed conduction can be a result of fac- tors such as the temperature of the extremity or even normal aging. Pathologically, demyelination pro- duces slowing. Conditions that result in loss of axons, particularly faster-conducting axons, also pro- duce slowing of nerve conduction or prolongation of latency. The amplitude of the CNAP can be measured from baseline to peak or from peak to peak. In general, the size of the CNAP and the SNAP is roughly proportional to the num- ber of axons depolarizing under the active electrode. It can be affected Figure 1 Measures of the SNAP or CNAP. Latency is the time between stimu- lus and the onset or peak of the potential. Amplitude is measured from peak to peak. Conduction velocity (CV) is calculated as distance divided by onset latency. peak latency amplitude onset latency distance onset latency CV = Neurophysiologic Evaluation Journal of the American Academy of Orthopaedic Surgeons 192 by a number of physiologic and pathologic factors. Cold increases the amplitude of both the CNAP and the SNAP. Aging produces smaller-amplitude SNAPs, probably as a result of gradual loss of large myelinated axons. Pathologically, loss of axons will reduce the amplitude of the CNAP. Distal lesions between the sites of stimulation and recording will de- crease the amplitude of the CNAP immediately, as conduction cannot traverse the lesion. Proximal le- sions (e.g., brachial plexus lesions) that separate sensory axons from their cell bodies (in the dorsal root ganglion) will produce distal axon loss due to axonal (wallerian) de- generation over time (usually 7 to 10 days after injury). 2 Thus, a reduced-amplitude SNAP can be due to an axonal lesion anywhere distal to the dorsal root ganglion. Motor-Nerve Conduction Studies The principles of stimulation and recording for motor-nerve conduc- tion studies are similar to those used for sensory-nerve conduction studies with several exceptions. The primary difference is that motor-nerve conduction studies involve recording a CMAP over muscle rather than recording direct- ly from nerve. Therefore, the distal latency involves not only conduc- tion along the nerve from the point of stimulation (proceeding at about 50 m/sec), but also includes neuro- muscular junction transmission time (which takes about 1 msec) and conduction along muscle fibers (about 3 to 5 m/sec). Although la- tency from a distal stimulation site can be measured, it cannot be con- verted into a nerve conduction velocity in the same way as a SNAP can be, because of this additional time for neuromuscular junction transmission and muscle fiber con- duction. Therefore, to evaluate con- duction velocities, motor nerves are typically stimulated in two places, and the distance between the two stimulation sites is divided by the dif- ference in latency; neuromuscular- junction transmission time and muscle-fiber conduction velocity are canceled out in the process (Fig. 2). Many of the same factors affect motor-nerve conduction studies as affect sensory-nerve conduction studies. 3 There are, however, two important differences. First, be- cause motor-neuron cell bodies reside in the anterior horn of the spinal cord rather than in the dor- sal root ganglion, the amplitude of the response is diminished by axon loss at the anterior horn cell or dis- tally (i.e., not at the dorsal root gan- glion). A root lesion proximal to the dorsal root ganglion, for exam- ple, would diminish the amplitude of the CMAP but not that of the SNAP. Second, because recording is from muscle, neuromuscular- junction transmission defects or primary myopathies may reduce the amplitude of the CMAP. Late Responses There are two ÒlateÓ responses (i.e., occurring late after the CMAP or M wave), which sometimes pro- vide useful information: the F wave and the H wave. 4 The F wave (so named because it was first recorded in foot muscles) is a late response usually recorded from distal mus- cles. Physiologically, when a motor nerve is stimulated distally, axons are depolarized in both directionsÑ distally (orthodromically) and proxi- mally (antidromically). The ortho- dromic volley activates the muscle distally, and the antidromic volley proceeds proximally to the anterior horn cell. It is thought that the F wave occurs when a small percent- age (3% to 5%) of antidromically activated motor cell bodies dis- charge and produce orthodromic activation of their motor axons. This is noted as a small-amplitude (about 100 to 200 µV) late (about 30 msec in the distal upper limb) potential. F-wave measurements usually find their greatest applicability in the assessment of multifocal or dif- fuse processes, especially those af- fecting proximal areas of the periph- eral nervous system. Acquired or inherited demyelinating polyneu- ropathies that produce multifocal or diffuse slowing are clinical set- tings in which F waves can provide additional useful information. Although it would seem appealing to use F waves for the diagnosis of brachial plexopathy or some en- trapment neuropathies, they are usually not of significant help in these applications, nor do they offer unique information not ob- tained by conventional nerve con- duction studies. Because the F wave is produced by only a small percentage of the motor axons, the presence of just a few normally conducting fibers will result in nor- mal latencies. Moreover, the F- wave volley traverses such a long distance of peripheral nerve that a focal lesion, unless there is severe demyelination, would not be ex- pected to produce marked abnor- malities in F-wave latencies. The H wave (named after Hoff- man) involves synaptic transmis- amplitude amplitude latency1 latency2 Wrist Elbow ∆ distance lat2 − lat1 CV = Figure 2 Measures of the CMAP. Latency is the time between stimulus and the onset of the potential. Amplitude is measured from baseline to peak. Conduction velocity (CV) can be calculated as the distance be- tween two points divided by the latency difference between two points. Lawrence R. Robinson, MD Vol 8, No 3, May/June 2000 193 sion at the spinal cord level and is in many ways analogous to the muscle stretch reflex. However, instead of activating stretch recep- tors within the muscle mechanical- ly, the large-diameter afferent nerve fibers are activated electrical- ly. After the afferent volley reaches the spinal cord, a monosynaptic reflex excites alpha motor neurons, and a late response is produced in the muscle. The H reflex can usual- ly be elicited only in the soleus muscle in adults. The most useful application of the H wave is in the detection of S1 radiculopathy. 5 It has been shown that the H wave is more sensitive than needle electromyography in the assessment of S1 radiculopathy, probably related to the fact that the H wave can depict conduction block and demyelination, whereas needle electromyography can be used to detect only motor axon loss. Principles of Needle Electromyography Needle electromyography assesses the function of the motor unitÑthe combination of an anterior horn cell, an axon, and all the muscle fibers supplied by the single axon. It is very sensitive for detection of axon loss at any level along the lower motor neuron once sufficient time has elapsed for fibrillations and other abnormalities to develop (usually 2 to 3 weeks). 6 There are usually four distinct steps in the needle EMG examination for each muscle: (1) insertional activity, (2) spontaneous activity, (3) examina- tion of motor-unit potentials, and (4) assessment of recruitment. Insertional Activity Insertional activity is examined by moving the needle through the muscle briefly and observing the amount and duration of the electri- cal potentials produced. Insertional activity may be decreased or may be prolonged in duration. Decreased insertional activity can result if the needle is not positioned in muscle or is in a muscle that has marginal viability. Muscles that have become atrophied and fibrotic will have reduced insertional activity, as will muscles that have become necrotic due to compartment syndrome. Prolonged or increased insertional activity, as an isolated finding, is a very ÒsoftÓ abnormality. No diag- nosis should be made on the basis of this ÒabnormalityÓ when it is an isolated finding, as it may be seen in some asymptomatic individuals. In- creased insertional activity can also be seen in association with fibril- lations or positive sharp waves and thus may be an indicator of either denervation or a primary muscle disorder. Spontaneous Activity Spontaneous activity consists of electrical discharges that are seen without needle movement or vol- untary contraction. Fibrillation potentials represent abnormal spontaneous single muscle-fiber discharges. Fibrillation potentials are essentially always abnormal, but they are a nonspecific finding. Fibrillation potentials are often seen in denervated muscles. Myopathies may be associated with fibrillation potentials. Disorders characterized by upper-motor-neuron lesions, such as stroke and spinal cord injury, have been shown to produce fibrillation potentials; these are usu- ally seen early after onset of the lesion and can be confusing when trying to diagnose a peripheral- nerve lesion superimposed on an upper-motor-neuron lesion. Fibrillation potentials are usually graded on a scale from 1+ to 4+, with 1+ representing a repro- ducibly observed fibrillation in an isolated area and 4+ representing sustained fibrillation potentials (which often obscure the baseline) throughout the muscle. The size of fibrillation potentials has been cor- related with the time since onset of denervation. Large-amplitude fi- brillation potentials (>100 µV) are seen within the first year after onset of denervation; smaller amplitudes (<100 µV) are seen later. 7 It has been postulated that this relation- ship reflects muscle fiber atrophy over time, with smaller-diameter fibers producing smaller-amplitude fibrillations. Consequently, large- amplitude fibrillations in the pres- ence of a neuropathic lesion suggest recent denervation. Positive sharp waves can be thought of in much the same way as fibrillation potentials. They also represent abnormal spontaneous single-muscle-fiber discharges. Positive sharp waves can be seen in essentially all the same disorders as fibrillation potentials. In some cases of muscle trauma, positive sharp waves may be seen in isola- tion without associated fibrilla- tions. Positive sharp waves are thought to have the same patho- physiologic characteristics as fibril- lation potentials and can be graded by using the same scheme. Complex repetitive discharges, formally known as Òbizarre high- frequency discharges,Ó probably represent groups of muscle fibers firing in near synchrony. They are usually seen in chronic neuropathic and myopathic conditions. When seen in isolation, they are a nonspe- cific but usually abnormal finding, similar to positive sharp waves and fibrillations. Fasciculation potentials repre- sent spontaneous discharges of a single motor unit. As opposed to a fibrillation potential (in which only a single muscle fiber fires), a fascic- ulation potential involves the entire motor unit (the axon and all the muscle fibers that it supplies). Unlike fibrillation potentials, fasci- culations produce enough force that they can be seen on the skin Neurophysiologic Evaluation Journal of the American Academy of Orthopaedic Surgeons 194 clinically. Fasciculation potentials are often generated at the anterior horn cell, as in motor neuron dis- eases, but they may also be ectopi- cally generated distally along the axon, possibly even in intramuscu- lar axons. Fasciculation potentials can be seen in a variety of neuromuscular disorders. In addition to motor neuron disease and the syndrome of benign fasciculations, fasciculation potentials can be seen in chronic radiculopathies, peripheral polyneu- ropathies, thyrotoxicosis, and over- dosage of anticholinesterase med- ications. Motor-Unit Analysis A great deal of information can be obtained from analysis of voluntarily activated motor-unit action poten- tials (MUAPs) (Fig. 3). The MUAP represents the electrical potential cre- ated by the synchronous discharge of all the muscle fibers supplied by a single motor axon. Theoretically, in neuropathic conditions in which there has been partial denervation and reinnerva- tion, one will see changes represen- tative of the underlying process of axonal sprouting (Fig. 4). Within days after partial denervation, intra- muscular axons that remain unaf- fected will send sprouts, usually emanating from distal nodes of Ranvier, to reinnervate nearby denervated muscle fibers. These sprouts, particularly early on, are not yet well myelinated and, there- fore, conduct slowly. Consequently, in the early phases of reinnervation, one will note increased polyphasici- ty and increased duration of the MUAP as a result of temporal dis- persion in newly formed sprouts and poor synchronization of muscle- fiber discharges. As these sprouts mature, synchronization of muscle- fiber discharges improves; the polyphasicity tends to be reduced, and one is left with large-amplitude, long-duration MUAPs. The in- crease in amplitude is a result of the increased number of muscle fibers belonging to the same motor unit within the recording area of the tip of the EMG needle. Myopathic changes in the MUAP result from loss of individual mus- cle fibers. In myopathic conditions, the MUAPs are typically small in amplitude and short in duration. Furthermore, fewer muscle fibers from the same motor unit fire with- in the recording area of the needle electrode. Recruitment Evaluation of motor unit recruit- ment can assess whether reduced strength is due to a reduction in the lower-motor-neuron pool or to poor central effort. In distinguish- ing between these two possibilities, the primary feature that is mea- sured is the motor-unit firing rate. Central recruitment implies that there are reduced numbers of mo- tor units firing but that they are fir- ing at normal or slow speed. This 1 2 3 duration amplitude Figure 3 Measures of the MUAP include duration (from onset to termination), amplitude (from peak to peak), and num- ber of phases (numbered, as shown). Figure 4 Top, Normal MUAP, recorded by a needle electrode from muscle fibers within its recording area. Middle, After denervation, single muscle fibers spontaneously dis- charge, producing fibrillations and positive sharp waves. Bottom, When reinnervation by axon sprouting has occurred, the newly formed sprouts will conduct slowly, producing temporal dispersion (i.e., prolonged MUAP duration) and MUAP polyphasicity. The high- er density of muscle fibers within the recording area of the needle belonging to the enlarg- ing second motor unit results in an increased-amplitude MUAP. Normal Denervation Reinnervation Lawrence R. Robinson, MD Vol 8, No 3, May/June 2000 195 is by far the most common Òabnor- malityÓ in recruitment, but in isola- tion it is completely nondiagnostic. Central recruitment can be reflec- tive of upper-motor-neuron le- sions, pain, or poor voluntary ef- fort. Reduced recruitment (noted in less severe conditions) and dis- crete recruitment (noted in more severe conditions) are pathologically significant and imply that there are reduced numbers of motor units firing rapidly. Interpretation of the Electrodiagnostic Examination Principles of Localization Needle electromyography is con- ventionally used for evaluation of lesions that are primarily axonal or so proximal that it is not possible to stimulate both proximal and distal to an entrapment site. Muscles that are supplied by multiple peripheral nerves, roots, or areas of the plexus are examined, and a localization is made on the basis of the distribu- tion of abnormalities. A sciatic nerve lesion in the thigh can be dis- tinguished from L5 radiculopathy, for example, if there is evidence of denervation in muscles supplied by the superficial and deep branches of the peroneal nerve but not the tensor fasciae latae or paraspinal muscles. Thus, localization is based on finding abnormalities distal to a branch point but normal findings proximally. 8,9 Nerve conduction studies are best at localizing the site of patho- logic change when there is demye- lination. As mentioned previously, demyelination causes focal slowing and conduction block; the presence of these findings can precisely lo- calize a focal entrapment. Conduc- tion block and slowing is observed only in demyelination and neura- praxia. It is not present in lesions with axon loss once wallerian de- generation has occurred (about 7 days after onset); therefore, localiza- tion of purely axonal lesions de- pends primarily on EMG findings. Deducing the Pathophysiology Neurapraxia and demyelination are best demonstrated when there is focal conduction block and slow- ing on nerve conduction studies but a large-amplitude CMAP or SNAP is elicited distal to the site of the lesion. Purely neurapraxic injuries have no electrophysiologic evidence of axon loss (fibrillation potentials or positive sharp waves) or reinnervation. Axon-loss lesions (e.g., axonot- mesis and neurotmesis 10 ) are usually demonstrated by evidence of de- nervation on needle EMG examina- tion as well as small-amplitude CMAP and SNAP responses with stimulation and recording distal to the site of the lesion. While needle electromyography is a more sensi- tive indicator for motor-axon loss, measurement of CMAP or SNAP amplitude is a better measure of the degree of axon loss and of prognosis. Axonotmesis and neurotmesis can- not usually be distinguished on elec- trodiagnostic studies, because the primary difference between the two conditions is integrity of the support- ing structures (which have no elec- trophysiologic function) (Table 1). Timing of Electrophysiologic Changes The time course of electrodiag- nostic changes after the onset of a neuropathic lesion is an important consideration that influences the interpretation of the electrophysio- logic examination. Neurapraxia, demyelination, and severe axon loss produce electrophysiologic changes immediately at onset if the nerve can be stimulated both proxi- mal and distal to the lesion. How- ever, proximal lesions, in which it is not possible to get proximal and distal to the lesion, do not immedi- ately produce changes on distal nerve conduction studies or elec- tromyography. Moreover, distinc- tion between neurapraxia and ax- onotmesis cannot be made until 7 days have passed, allowing time for wallerian degeneration to have progressed to the point that stimu- lation of motor axons elicits no Table 1 Electrodiagnostic Findings in Various Peripheral Nerve Disorders Root Plexus Focal Axonal Demyelinating Finding Lesion Lesion Entrapment Polyneuropathy Polyneuropathy Motor nerve amplitude +/− (focal) +/− (diffuse) +/− Sensory nerve amplitude Normal (focal) +/− (diffuse) +/− Distal latency Normal Normal (focal) Normal (diffuse) Conduction velocity Normal Normal (focal) Normal (diffuse) Fibrillations + (acute) + (acute) +/− (severe) + +/− Large polyphasic MUAPs + (chronic) + (chronic) +/− (severe) + +/− Neurophysiologic Evaluation Journal of the American Academy of Orthopaedic Surgeons 196 motor responses. 2 Ten days after the onset of a complete lesion, SNAPs will be absent as well. Therefore, 7 to 10 days after onset, a neurapraxic injury (in which the distal amplitudes will be normal) can be differentiated by nerve con- duction studies from an axonot- metic lesion (in which the distal amplitudes will be reduced). Two to three weeks after the onset of injury, the needle EMG study starts to show fibrillation potentials and positive sharp waves. 6 Proximal muscles demon- strate these abnormalities first; more distal muscles, later. Radicu- lopathies, for example, may show paraspinal abnormalities at day 10 to 14 after onset, but distal-limb muscle changes may not be appar- ent for 3 to 4 weeks after onset. Fibrillations and positive sharp waves may persist for several months or even many years after a single injury, depending on the extent of reinnervation. The timing and type of electro- physiologic changes consequent to reinnervation will depend in part on the mechanism of reinnervation. When reinnervation is a result of axonal regrowth from the site of the lesion (usually in complete injuries), the appearance of new MUAPs will not occur until motor axons have had sufficient time to regenerate across the distance between the lesion site and the muscle (usually proceeding at a rate of a few mil- limeters a day). When these new axons first reach the muscle, they will innervate only a few muscle fibers, producing short-duration, small-amplitude potentials, some- times referred to as Ònascent poten- tials.Ó With time, as more muscle fibers join the motor unit, the MUAPs will become larger, more polyphasic, and longer in duration. Motor-unit potential changes will also develop when reinnerva- tion occurs by axonal sprouting. Polyphasicity and increased dura- tion develop first as newly formed, poorly demyelinated sprouts sup- ply the recently denervated muscle fibers. As the sprouts mature, large- amplitude, long-duration MUAPs develop and persist indefinitely. Evaluation of Common Clinical Entities Hand Numbness (Case 1) A 50-year-old woman presents with a 3-month history of progres- sive right-hand numbness. The numbness involves all digits of the hand but is restricted to the palmar aspect. She reports mild chronic neck pain but denies symptoms in the feet. Physical examination demonstrates normal strength and muscle stretch reflexes; sensation is normal to pin prick and light touch. There is a positive Tinel sign over the median nerve at the wrist and at the ulnar groove bilaterally, but no Phalen sign. The differential diagnosis in this case includes median neuropathy at the wrist (e.g., carpal tunnel syn- drome), cervical radiculopathy, and ulnar neuropathy. Electrodiag- nostic studies are therefore oriented toward looking for evidence of slowing in peripheral nerves or evi- dence of denervation in the mus- cles of the upper limb. A notable finding is slowing in the median nerve at the wrist, with prolonged latencies compared with both radial and ulnar nerves (Fig. 5). It has recently been shown that it is better (in terms of sensitivity, specificity, and reliability) to perform the three comparisons of median and ulnar nerves illustrated and then to add the median-ulnar and median-radial nerve latency differences, rather than looking at individual tests alone (Fig. 6). 11 There is no evi- dence of slowing in the ulnar nerve, nor is there evidence of de- nervation in the C5 to T1 myotomes of the upper limb; thus, the find- Nerve Conduction Studies Stimulate Record Latency (msec) Amplitude Velocity (m/sec) Median nerve (sensory) Wrist Ring finger 4.8 12 µV Ulnar nerve (sensory) Wrist Ring finger 3.5 8 µV Median nerve (sensory) Wrist Thumb 4.1 21 µV Radial nerve (sensory) Wrist Thumb 2.8 11 µV Median nerve (sensory) Palm Wrist 3.1 20 µV Ulnar nerve (sensory) Palm Wrist 2.1 22 µV Median nerve (motor) Wrist APB 4.5 (<4.3) 6.7 (³5.0) mV Elbow APB 6.1 (³5.0) mV 51 (³50) Ulnar nerve (motor) Wrist ADM 3.6 (<3.8) 8.3 (³5.0) mV Below elbow ADM 8.1 (³5.0) mV 57 (³50) Above elbow ADM 7.7 (³5.0) mV 61 (³50) Needle EMG Spontaneous Activity Motor Unit Action Potentials Muscle Myotome Ins. Act. Fibs/PSWs Amplitude Duration Phasicity Recruitment Deltoid C5,6 Normal None Normal Normal Normal Full Biceps C5,6 Normal None Normal Normal Normal Full Pronator teres C6,7 Normal None Normal Normal Normal Full ECR C6,7 Normal None Normal Normal Normal Full FCR C6-8 Normal None Normal Normal Normal Full Triceps C7,8 Normal None Normal Normal Normal Full APB C8,T1 Normal None Normal Normal Normal Full FDI C8,T1 Normal None Normal Normal Normal Full Cervical paraspinals C5-T1 Normal None Figure 5 Findings from nerve conduction and needle EMG studies in case 1. Normal val- ues are shown in parentheses. Abbreviations: ADM = abductor digiti minimi; APB = abduc- tor pollicis brevis; ECR = extensor carpi radialis; FCR = flexor carpi radialis; FDI = first dorsal interosseous; Fibs/PSWs = fibrillations/positive sharp waves; Ins. Act. = insertional activity. Lawrence R. Robinson, MD Vol 8, No 3, May/June 2000 197 ings are consistent with carpal tun- nel syndrome but are not sugges- tive of ulnar neuropathy or cervical radiculopathy. Pain in the Low Back and Lower Limb (Case 2) A 45-year-old man reports low back pain extending into the left lower limb, with pain and numb- ness in the posterolateral thigh and leg and the lateral aspect of the foot. This started after an injury at work when he was lifting and rotating a heavy object. He had a similar episode 4 years pre- viously, which resolved with con- servative management. Physical examination demonstrates normal strength and sensation but a de- creased left ankle jerk. The diag- nostic questions in this case are whether a radiculopathy is present and, if so, at what level and of what duration. Needle electromyography was performed on the muscles of the left lower limb, evaluating com- monly affected myotomes (L3 to S2) to look for evidence of either acute denervation or prior dener- vation and reinnervation. The findings shown in Figure 7 indicate both recent denervation (fibrilla- tions and positive sharp waves) and reinnervation (large, long- duration MUAPs) in the left S1 dis- tribution. These findings allow one to infer that there is both a new- onset S1 radiculopathy and a pre- existing radiculopathy at the same level. Asymmetry of the H waves (smaller amplitude and longer latency on the left) confirms the presence of an abnormality at the S1 level. Combined Upper- and Lower- Motor-Neuron Findings (Case 3) A 70-year-old retired cardiac sur- geon complains of progressive weakness in the upper and lower limbs and muscle atrophy in the upper limbs. He has only vague sensory symptoms of numbness in the upper limbs. He denies bowel or bladder dysfunction. There is a history of chronic mild neck pain with no difficulty speaking or swal- lowing. He reports intermittent muscle twitching in the pectoral muscles, worse with cold (he is not sure if this is shivering). On physi- cal examination, there is marked muscle atrophy in the upper limbs but normal muscle bulk in the lower limbs. Strength is diffusely weak (4/5 on MRC scale) in the upper and lower limbs. Sensation is nor- mal. Muscle stretch reflexes are hyperactive in the upper and lower limbs. Cervical spine radiographs reveal marked degenerative changes (spondylosis). The diagnostic question in this case is whether cervical myelopathy or motor neuron disease is the cause of the patientÕs symptoms. Although the clinical features could be consistent with either diagnosis, the electrodiagnostic features are usually different. Cervical spondy- Figure 6 Nerve conduction studies in case 1. Note prolongation of peak latencies (values in parentheses) in median nerves compared with ulnar and radial nerves. The combined sensory index is calculated by adding the peak latency differences between median and ulnar nerves to the ring finger (4.8 Ð 3.5 = 1.3 msec), the median and radial latency differences to the thumb (4.1 Ð 2.8 = 1.3 msec), and the median and ulnar latencies with stimulation in the palm and recording over the wrist (3.1 - 2.1 = 1.0 msec); this difference totals 3.6 msec. Values of 1.0 msec or above are considered abnormal and consistent with median neu- ropathy at the wrist. Median ring (4.8) Ulnar ring (3.5) Median thumb (4.1) Radial thumb (2.8) Median palm (3.1) Ulnar palm (2.1) Nerve Conduction Studies Stimulate Record Latency, msec Amplitude, mV Left H wave Knee Soleus 35.1 1.7 Right H wave Knee Soleus 32.8 4.9 (Normal side-to-side difference for latency is 1.2 msec, with normal amplitude difference up to 40%.) Needle EMG Spontaneous Activity Motor Unit Action Potentials Muscle Myotome Ins. Act. Fibs/PSWs Amplitude Duration Phasicity Recruitment Vastus medialis L3,4 Normal None Normal Normal Normal Full Adductor longus L3,4 Normal None Normal Normal Normal Full Tibialis anterior L4,5 Normal None Normal Normal Normal Full Tensor fasciae latae L4-S1 Normal None Normal Normal Normal Full Biceps femoris L5,S1 Increased 1+/2+ Increased Increased Normal Full Peroneus longus L5,S1 Increased 1+/1+ Increased Increased Normal Full Soleus S1,2 Increased 2+/2+ Increased Increased Normal Lumbar paraspinals L3-S1 Normal None Figure 7 Findings from nerve conduction and needle EMG studies in case 2. Abbreviations: Fibs/PSWs = fibrillations/positive short waves; Ins. Act. = insertional activity. Neurophysiologic Evaluation Journal of the American Academy of Orthopaedic Surgeons 198 losis may produce lower-motor- neuron loss in the upper limbs due to root or anterior horn cell involve- ment, but it should not cause lower- motor-neuron loss in other regions of the body. In contrast, motor neu- ron disease produces widespread evidence of upper- and lower- motor-neuron loss and fascicula- tions. Electromyographic diagnosis of amyotrophic lateral sclerosis re- quires evidence of denervation in three of the following four ÒregionsÓ: bulbar, cervical, thoracic, and lum- bosacral. The needle EMG findings in this case (Fig. 8) demonstrate evidence of denervation in the upper limbs, con- sistent with two processes. There is denervation of C6-innervated mus- cles, consistent with a C6 radicu- lopathy. Additionally, the distal muscles of the upper and lower limbs demonstrate denervation, suggesting a distal peripheral poly- neuropathy. However, extensive evaluation of other body regions (including the tongue, thoracic paraspinal muscles, and proximal lower limbs) did not show evidence of denervation. Fasciculations were limited to two distal hand muscles and were not widespread. Nerve conduction studies dem- onstrate slowing of conduction dif- fusely (in the sural, peroneal, and ulnar nerves) but more severe ab- normalities in the median nerve (with absent sensory response and very prolonged motor latency). These findings confirm the pres- ence of a peripheral polyneuropa- thy and also suggest a superim- posed median neuropathy at the wrist. Thus, the findings are more con- sistent with cervical spondylosis and myeloradiculopathy than with motor neuron disease. A peripheral polyneuropathy with focal median neuropathy is also present. Surgi- cal decompression of the cervical spine resulted in rapid improve- ment. Summary The electrodiagnostic examination is a useful tool for detecting problems affecting the peripheral nervous sys- tem. Clinical assessment and defini- tion of the questions to be answered are essential to tailor the electrodiag- nostic examination for each patient. Potential pitfalls include performing tests in a standardized manner with- out examining the patient, not form- ing a differential diagnosis, technical errors, examining too few areas, and overinterpretation of minor devia- tions from Ònormal.Ó However, when performed appropriately, elec- trodiagnostic studies contribute sig- nificantly to the evaluation of patients with peripheral nervous system com- plaints. Nerve Conduction Studies Stimulate Record Latency, msec Amplitude Velocity, msec Median nerve (sensory) Wrist Thumb Absent response Radial nerve (sensory) Wrist Thumb 3.7 (²2.7) 3 µV (³5) Sural nerve (sensory) Leg Ankle 6.0 (²4.0) 1 µV (³5) Median nerve (motor) Wrist APB 5.1 (<4.3) 8.8 mV (³5.0) Elbow APB 7.5 mV (³5.0) 49 (³50) Ulnar nerve (motor) Wrist ADM 3.6 (<3.8) 8.3 mV (³5.0) Below elbow ADM 8.1 mV (³5.0) 50 (³50) Above elbow ADM 7.7 mV (³5.0) 49 (³50) Peroneal nerve (motor) Ankle EDB 8.6 (²6.0) 2.5 mV (³2.0) Knee EDB 2.5 mV (³2.0) 35 (³40) Needle EMG Spontaneous Activity Motor Unit Action Potentials Muscle Myotome Ins. Act. Fibs/PSWs Fasc Amplitude Duration Phasicity Recruitment Deltoid C5,6 Normal None None Normal Normal Normal Full Biceps C5,6 Normal None None Normal Normal Normal Full Extensor carpi radialis C6,7 Increased 2+/2+ None Normal Normal Normal Central Pronator teres C6,7 Increased 1+/1+ None Normal Normal Normal Full Triceps C7,8 Normal None None Increased Increased Normal Reduced APB C8,T1 Increased 1+/1+ 1+ Increased Increased Normal Reduced FDI C8,T1 Increased 1+/1+ 1+ Increased Increased Normal Reduced Pectoralis major C5-T1 Normal None None Normal Normal Normal Full Cervical paraspinals C5-T1 Normal None None Vastus medialis L3,4 Normal None None Normal Normal Normal Full Adductor longus L3,4 Normal None None Normal Normal Normal Full Tibialis anterior L4,5 Normal None None Normal Normal Normal Full Tensor fasciae latae L4-S1 Normal None None Normal Normal Normal Full Biceps femoris L5,S1 Normal None None Normal Normal Normal Full Soleus S1,2 Increased 2+/2+ None Increased Increased Normal Lumbar paraspinals L3-S1 Normal None None Tongue XII Normal None None Figure 8 Findings from nerve conduction and needle EMG studies in case 3. Normal val- ues are shown in parentheses. Abbreviations: ADM = abductor digiti minimi; APB = abductor pollicis brevis; EDB = extensor digitorum brevis; Fasc = fasciculations; FDI = first dorsal interosseous; Fibs/PSWs = fibrillations/positive short waves; Ins. Act. = insertional activity. Lawrence R. Robinson, MD Vol 8, No 3, May/June 2000 199 References 1. Dumitru D: Electrodiagnostic Medicine. Philadelphia: Hanley & Belfus, 1995, pp 341-384. 2. Chaudhry V, Cornblath DR: Wallerian degeneration in human nerves: Serial electrophysiological studies. Muscle Nerve 1992;15:687-693. 3. Miller RG: Injury to peripheral motor nerves. Muscle Nerve 1987;10:698-710. 4. Fisher MA: H reflexes and F waves: Physiology and clinical applications. Muscle Nerve 1992;15:1223-1233. 5. Braddom RL, Johnson EW: Standardi- zation of H reflex and diagnostic use in S1 radiculopathy. Arch Phys Med Rehabil 1974;55:161-166. 6. Daube JR: Needle examination in clin- ical electromyography. Muscle Nerve 1991;14:685-700. 7. Kraft GH: Fibrillation potential ampli- tude and muscle atrophy following peripheral nerve injury. Muscle Nerve 1990;13:814-821. 8. Sunderland S: Nerves and Nerve Inju- ries, 2nd ed. Edinburgh: Churchill- Livingstone, 1968. 9. Campbell WW, Pridgeon RM, Riaz G, Astruc J, Leahy M, Crostic EG: Sparing of the flexor carpi ulnaris in ulnar neu- ropathy at the elbow. Muscle Nerve 1989;12:965-967. 10. Seddon H: Surgical Disorders of the Peripheral Nerves, 2nd ed. New York: Churchill-Livingstone, 1975, pp 21-23. 11. Robinson LR, Micklesen PJ, Wang L: Strategies for analyzing nerve conduc- tion data: Superiority of a summary index over single tests. Muscle Nerve 1998;21:1166-1171.