Jump to:

Overview and Description

Definition of the assessment/treatment procedures

The electrodiagnostic (EDX) assessment of cervical, thoracic, and lumbosacral radiculopathies will be reviewed.  Approximately 60% – 90% of all radiculopathies affect the lumbosacral nerve roots with the vast majority of those affecting the L5 and S1 roots.1-3  Cervical radiculopathies likely account for only 5%-10% of all spinal nerve root disorders affecting the C7, C6, C8 and C5 nerve roots in decreasing frequency.2-3  Thoracic radiculopathies occur, but are rare; representing only 0.15% – 4% of symptomatic disc herniations of the spine.4  They have also been reported in a subset of patients with diabetic polyradiculoneuropathy.5  

EDX techniques including nerve conduction studies (NCS) and needle electromyography (NEE) are routinely used in the clinical evaluation of patients with suspected radicular lesions.  Given the lack of a universal “gold standard” for evaluating radiculopathies, studies that establish physiological dysfunction of spinal nerve roots continue to be useful.5   

NCS are needed to rule-out the presence of other neurological conditions which may produce similar signs and symptoms as radiculopathies (i.e., peripheral entrapment neuropathies, plexopathies, or peripheral polyneuropathies).  As the pathology is at the root level, sensory NCS are expected to be normal given the proximal location of the dorsal root ganglion (DRG) within the neural foramen; however, abnormal superficial peroneal and sural sensory nerve action potentials (SNAPs) have been reported with L5 and S1 root lesions, respectively, suggesting a more distal location of the DRG within the foramen.5  Normal SNAPs argue for a radiculopathy versus a more peripheral lesion in patients with radiating extremity pain and/or paresthesias (i.e., an L5 or C8 radiculopathy rather than a peroneal or ulnar neuropathy, respectively).

Compound muscle action potential (CMAP) amplitudes may be reduced by severe axonal lesions affecting the primary nerve root innervation of the muscles used for pick-up (i.e., extensor digitorum brevis for L5 radiculopathies) during routine NCS. In cervical radiculopathies, however, distal CMAP amplitudes are not helpful in lesions outside a C8-T1 myotomal distribution as the more commonly affected nerve roots in the cervical spine (i.e., C6, C7) do not innervate the muscles commonly used for pick-up in the upper limb (i.e., abductor pollicis brevis and abductor digiti minimi). 

The value of F waves in the evaluation of radiculopathies remains limited in clinical practice.  F waves are often normal in less severe lesions, but may be abnormal in severe nerve root, plexus or peripheral nerve lesions.6 Despite reports that various F-wave parameters (i.e., minimal latency, maximal latency, chronodispersion) may improve the diagnostic yield for cervical radiculopathy when combined with needle EMG, abnormal F-wave findings cannot be used to localize a lesion to a specific nerve root level.7 This fact plus the supramaximal stimulation required to elicit F-waves, contributes to the low use of these studies to diagnose cervical radiculopathy.

H-reflexes can be helpful in the electrodiagnosis of S1 radiculopathy.  Asymmetrical prolongation or absence of H reflexes may be present in S1 nerve root injuries with a reported sensitivity of 50% and a specificity of 91%.8  It should be noted that H-reflexes become abnormal as soon as compression occurs and may remain abnormal indefinitely, making this test more useful than the other late responses.9  The upper extremity H-reflex study (mediated through C6 and C7 roots through the median nerve) has been shown to add utility to the electrodiagnosis of cervical radiculopathy, especially in patients whose clinical symptoms are not clear and needle EMG is normal.10

Somatosensory evoked potentials (SEPs) have the theoretical advantage of evaluating afferent nerve fiber dysfunction in radiculopathies.  Specifically, dermatomal SEPs should be useful as they assess the sensory fibers of a single nerve root.  Based on the available evidence, however, SEPs are still considered investigational for acute cervical and lumbosacral radiculopathies.   They may be useful in patients with suspected cervical myelopathy or other central causes of neurological symptoms (i.e., multiple sclerosis).5,11

The NEE remains the most widely accepted method for confirming the presence of a radiculopathy affecting the motor axons.1  Ideally, six muscles including the paraspinals should be screened for both cervical and lumbar radiculopathies as sensitivities of the NEE are relatively poor.12  In patients with cervical radiculopathy, sensitivities ranged from 50%-72% while sensitivities for lumbosacral radiculopathy ranged from 49% – 86%.12-13  In many patients, however, axonal loss is never found as overlapping muscle innervations and varying degrees of axonal injury likely obscure the detection of such lesions.  In other circumstances with a negative NEE, a demyelinating process with or without conduction block or a purely sensory lesion affecting the DRG must be considered.5

Ideally, the EDX examination should coincide with the development of denervation potentials (i.e., positive sharp waves and fibrillations) in the axial and limb muscles.  Abnormal spontaneous electrical activity usually first appears in the paraspinal muscles approximately 7-10 days following injury, followed by the proximal and then distal limb muscles several weeks later.3 Reinnervation follows this same proximal-to-distal pattern.  As such, the NEE should be performed approximately 3-4 weeks post-injury.

NEE abnormalities found in two or more muscles innervated by the same nerve root but by different peripheral nerves confirms the presence of a radiculopathy.2  However, paraspinal fibrillations and/or positive sharp waves may be the only EDX abnormality found in about 20% of patients with lumbosacral root injuries and in about 40% of patients with cervical root injuries.14-15   In contrast to most limb muscles, abnormal spontaneous electrical activity may be seen in the cervical and lumbar paraspinal muscles of asymptomatic, older individuals.16-18  The close similarity in appearance of atypical appearing endplate spikes with fibrillation potentials makes diagnosing cervical or lumbosacral radiculopathy contentious when relying on electrical abnormalities in the paravertebral muscles alone.19 

Ultrasound guidance for needle placement can enhance the accuracy of muscle selection for the diagnosis of cervical and lumbar radiculopathy, especially in deeper muscles with multisegmental root innervation.   Examples include the iliacus and psoas muscles to help differentiate an upper lumbar radiculopathy (i.e., L2/3) from a femoral mononeuropathy, the posterior tibialis muscle to help differentiate an L5 radiculopathy from a sciatic mononeuropathy or lumbosacral plexopathy and the supinator muscle to help differentiate a C5/6 radiculopathy from a radial mononeuropathy.20-21

The density of fibrillation potentials depends upon the innervation ratio (i.e., ratio between the total number of extrafusal muscle fibers to the number of innervating motor axons) of the muscle being sampled.22  The innervation ratio may be quite high in larger muscles such as the gastrocnemius where one motor axon innervates roughly 1,934 muscle fibers.22  Therefore, losing a small percentage of motor axons to root injury will generate a relatively large number of fibrillating muscle fibers.  It should be noted, however, that fibrillations and positive sharp waves will disappear with time due to motor unit reinnervation and/or loss of muscle membrane reactivity.5

The NEE in patients with long-standing cervical or lumbar radiculopathy may detect changes consistent with chronic neurogenic injury.  Relying solely though on motor unit action potential (MUAP) morphology to diagnose radiculopathy is not recommended as there are rarely enough motor axons injured innervating a particular muscle to produce discernible alterations in the recruitment pattern and, even rarer, are they found in a definable myotomal distribution5.  Moreover, the time constraints and sampling size imposed by a proper motor unit analysis limits the evaluation of motor unit duration, amplitude, and polyphasia necessary to adequately characterize a chronic radiculopathy.5 If reinnervation cannot keep pace with ongoing axon loss, both low- and high-voltage fibrillation potentials may be seen suggesting an acute-on-chronic radiculopathy.

In contrast to neurotemetic lesions such as radial nerve palsies resulting from gunshot or stab wounds where the EDX examination can help prognosticate return of motor function, no such information can be reliably inferred from the EDX examination of patients with cervical or lumbosacral radiculopathy.  The inherent difficulty correlating fibrillation potential density with the severity of nerve root injury and the lack of available distal CMAP amplitudes makes prognostication difficult.  Furthermore, the multisegmental innervation of most peripheral nerves helps preserve the distal CMAP amplitude even with severe injury to a single nerve root.  Identifying those patients with demyelinating or purely sensory lesions is also limited as the NEE is expected to be normal with the possible exception of reduced recruitment.

Relevance to Clinical Practice

EDX testing is an extension of the physician’s history taking and physical examination.  Perhaps the best rationale to support its use is establishing whether nerve root injury is present and which roots are affected.  In that context, performing EDX studies when root injury has previously been established or disproved through complimentary investigations to the patient’s physical examination such as magnetic resonance imaging (MRI) or computed tomography (CT) myelography is unwarranted.   

EDX testing can help clarify a patient’s diagnosis when imaging studies are equivocal.  MRI scans of the lumbar spine are known to have high false-positive rates with one study reporting 27% of normal subjects having a disc protrusion.23   The false-positive rate of cervical spine MRI scans is lower with only 10% of asymptomatic subjects reported having a herniated or bulging disc.24 

EDX testing may be helpful in predicting patients’ response to both non-operative and operative management of cervical and lumbosacral radiculopathy.  Savage et al. found greater improvement in low back-related disability outcomes after physical therapy in those patients with sciatica who had positive EMGs compared to those patients with normal EDX testing.  A notable weakness of the study was that criteria for a positive EMG was denervation in the limb muscles and/or paraspinal muscles alone.25  Annaswamy et al. found that those patients with EMG-positive lumbar radiculopathy had better pain scores and functional improvement after interlaminar epidural steroid injection (ESI) than those with EMG-negative radiculopathy.26  In a larger study of patients with chronic lumbar or cervical radicular pain, McCormick et al. found that those with abnormal EDX were more likely to have better intermediate and long-term outcomes compared to those patients with normal EDX after lumbar transforaminal ESI.27  EDX did not predict outcomes in patients with cervical radicular pain as the investigators felt that the numbers in the cervical spine group were too small to draw clear conclusions. 

As surgical success rates vary, EDX testing can provide objective data to help identify patients who may benefit from surgical decompression. The NEE, for example, was reported to correlate well with nerve root injury at surgery in 78.6% of patients based on findings of fibrillations and positive sharp waves in limb muscles in one early study.28 A later study by Falck et al. compared the outcome of patients with lumbar radiculopathy from disc herniation treated surgically to those patients with lumbar radiculopathy treated conservatively but found the initial EMG had no prognostic ability to predict outcome at five years.29 Tullberg et al. found that those patients undergoing surgery for lumbosacral radicular pain from a herniated disc fared worse with normal pre-operative EDX as compared to patients with an abnormal EDX.30  In a small study of patients with cervical radiculopathy, Alrawi et al., found that those patients with positive pre-operative EMGs fared better after discectomy and fusion than patients with normal EMGs.31

Complications arising from EDX testing are exceedingly rare in the general population.  Imaging studies performed on patients after undergoing NEE demonstrated an aggregated risk of bleeding of approximately 1% for all muscles with no significantly higher incidence seen in patients on systemic anticoagulation.32 As such, there is no compelling reason to defer the NEE in patients suspected of having a radiculopathy who are taking commonly prescribed anticoagulants. Patients on novel oral anticoagulants were found to have minimal risk of clinically relevant hemorrhagic complications, not significantly different from those patients on warfarin in a 2019 study.33 Concern over puncturing the pleural and abdominal cavities, especially in obese patients, limits the use of the NEE of the intercostals and anterior abdominal wall muscles in diagnosing thoracic radiculopathies. Ultrasound with Doppler imaging may be used to help avoid blood vessels when localizing deep muscles targeted for NEE. 

Routine EDX testing should not pose a safety hazard to patients with cardiac pacemakers or defibrillators, but NCS should be avoided in patients with external cardiac pacemakers as the conductive lead threaded into the heart is electrically sensitive.34 In patients with significant anxiety disorders, skin infections, grafts, or varicosities, proceeding with the EDX examination remains at the discretion of the examining physician. 

The ability of patients to provide informed consent underlies most medical procedures and tests, including EDX testing.  Hospitalized patients, however, especially those in intensive care settings, may not be able to provide informed consent. In these cases, family members may be approached for consent if the requested testing is thought to clarify the patient’s diagnosis or change management.

Guidelines for ethical behavior in EDX medicine have been developed by the American Association of Neuromuscular and Electrodiagnostic Medicine (AANEM).35 Central to the EDX consultation is the physician’s fiduciary duty to safeguard the patient’s physical, psychological, and emotional interests as it relates to the prescribed testing. The EDX examination should be planned to minimize patient discomfort by stimulating the least number of nerves and needling the fewest muscles necessary to confirm or refute the presence of a radiculopathy.  Should the patient experience significant discomfort, the examination should be terminated and the patient advised that no lasting harm is expected. As patients have a tendency to overestimate the benefits of medical intervention and underestimate its harms, shared decision making in EDX medicine, especially as it relates to the diagnosis of cervical and lumbar radiculopathies, may help reduce the number of unnecessary tests and the procedures resulting from their use.36


  1. Dumitru D. Electrodiagnostic medicine.  Philadelphia, PA: Hanley & Belfus; 1995:523–584.
  2. Wilbourn AJ, Aminoff MJ. American Association of Electrodiagnostic Medicine minimonograph 32: The electrodiagnostic examination in patients with radiculopathies. Muscle Nerve. 1998; 21:1612-1631.
  3. Wilbourn AJ, Aminoff MJ. Radiculopathies. In: Brown WF, Bolton CF, editors. Clinical electromyography, 2nd ed. Boston, MA: Butterworth-Heinemann; 1993:177–209.
  4. Alvarez O, Roque CT, Pampeti M.  Multilevel thoracic disc herniations: CT and MR studies.  J Comp Assist Tomog 1988;12:649-652.
  5. Fischer MA. Electrophysiology of radiculopathies. Clinical Neurophysiology. 2002; 113:317–335.
  6. Hakimi K, Spanier D. Electrodiagnosis of Cervical Radiculopathy.  Phys Med Rehabil Clin N Am 24(2013);1-12.
  7. Lo YL, Chan LL, Leoh T, et al.  Diagnostic utility in F waves in cervical radiculopathy; electrophysiological and magnetic resonance imaging correlation.  Clin Neruol Neurosurg 2008;110:58-61.
  8. Marin R, Dillingham TR, Chang A, et al. Extensor digitorum brevis reflex in normals and patients with radiculopathies. Muscle Nerve 1995; 18:52–59.
  9. Barr K.  Electrodiagnosis of Lumbar Radiculopathy.  Phys Med Rehabil Clin N Am; 24(2013):79-91.
  10. Eliaspour D, Sanati E, Moqadam M, et al.  Utility of flexor carpi radialis H-reflex in diagnosis of cervical radiculopathy. J Clin Neurophysiol 2009;26:458-60.
  11. Aminoff MJ, Eisen AA.  AAEM minimonograph 19: somatosensory evoked potentials.  Muscle Nerve 1998;21:277-90.
  12. Dillingham TR. Evaluating the patient with suspected radiculopathy. Physical Medicine & Rehabilitation 2013; 5(5S):S41-49.
  13. American Association of Electrodiagnostic Medicine and American Academy of Physical Medicine and Rehabilitation. The electrodiagnostic evaluation of patients with suspected cervical radiculopathy: literature review on the usefulness of needle electromyography. Muscle Nerve 1999; 22(Suppl 8):S213–S221.
  14. Kuruoglu R, Oh SJ, Thompson B. Clinical and electromyographic correlations of lumbosacral radiculopathy. Muscle Nerve 1994; 17:250–251.
  15. Czyrny JJ, Lawrence J. Importance of paraspinal muscle electromyography in cervical and lumbosacral radiculopathies. Am J Phys Med Rehabil 1995; (74):458–459.
  16. Gilad R, Dabby M, Boaz M, et al.  Cervical paraspinal electromyography: normal values in 100 control subjects. J Clin Neurophysiol 2006;23:573-6.
  17. Dates ES, Mar EY, Bugola MR, et al. The prevalence of lumbar paraspinal spontaneous activity in asymptomatic subjects.  Muscle Nerve 1996;19(3):350-4.
  18. Nardin RA, Raynor EM, Rutkove SB.  Fibrillations in lumbosacral paraspinal muscles of normal subjects.  Muscle Nerve 1998;21(10):1347-9.
  19. Dumitru D, Diaz CA, King JC.  Prevalence of denervation in paraspinal and foot intrinsic musculature.  Am J Phys Med Rehabil 2001;80(7):482-90.
  20. Boon AJ, Oney-Marlow TM, Murthy NS, et al.  Accuracy of electromyography needle placement in cadavers: non-guided vs. ultrasound guided.  Muscle Nerve 2011; 44: 45-49.
  21. Albin SR, Hoffman LR, MacDonald CW, et al.  Ultrasonographic validation for needle placement in the tibialis posterior muscle.  International Journal of Sports Physical Therapy 2021; 16 (6).
  22. Dillingham TR. Electrodiagnostic medicine II: Clinical evaluation and findings.  In: Braddom RL, Chan L, Harrast MA, et al., editors.  Physical Medicine and Rehabilitation, 4th ed.  Philadelphia, PA: Elsevier; 2011: 211.
  23. Jensen MC, Brant-Zawadzki MN, Obuchowski N, et al.  Magnetic resonance imaging of the lumbar spine in people without back pain. N Engl J Med 1994; 331:69-73.
  24. Boden SD, McCowin PR, Davis DO, et al.  Abnormal magnetic-resonance scans of the cervical spine in asymptomatic subjects. A prospective investigation. J Bone Joint Surg 1990; 72A:1178-1184.
  25. Savage NJ, Fritz JM, Kircher JC, et al.  The prognostic value of electrodiagnostic testing in patients with sciatica receiving physical therapy.  Eur Spine J 2015; 24(3):434-43.
  26. Annaswamy TM, Bierner SM, Chouteau W, et al.  Needle electromyography predicts outcome after lumbar epidural steroid injection.  Muscle Nerve 2012;45(3):346-55.
  27. McCormick Z, Cushman D, Caldwell M, et al.  Does electrodiagnostic confirmation of radiculopathy predict pain reduction after transforaminal epidural steroid injection? A multicenter study. J Nat Sci 2015;1(8).
  28. Knutsson B. Comparative value of electromyographic, myelographic, and clinical-neurological examinations in the diagnosis of lumbar root compression syndromes. Acta Orthop Scand 1961; Suppl 49:1–135.
  29. Falck B, Nykvist F, Hurme M, et al.  Prognostic value of EMG in patients with lumbar disc herniation – a five year follow up.  Electromyogr Clin Neurophsyiol 1993;33(1)19-26.
  30. Tullberg T, Svanborg E, Isaccsson J, et al.  A preoperative and postoperative study of the accuracy and value of electrodiagnosis in patients with lumbosacral disc herniation.  Spine 1993;18(7):837-42.
  31. Alrawi MF, Khalil NM, Mitchell P, et al. The value of neurophysiological and imaging studies in predicting outcome in the surgical treatment of cervical radiculopathy. Eur Spine J 2007;16(4):495-500.
  32. Gertken JT, Patel AT, Boon AJ.  Electromyography and anticoagulation.  Physical Medicine & Rehabilitation 2013; 5(5S):S3-S7.
  33. Nagarajan E, Dyer N, Bailey E, et al.  Hematoma Risk after needle electromyography in patients using newer anticoagulants.  J Clin Neurophysiol 2019 Nov 8.
  34. Pease WS, Grove SL.  Electrical safety in electrodiagnostic medicine.  Physical Medicine & Rehabilitation 2013; 5(5S):S8-S13.
  35. Mackin GA, Horowitz SH, Leonard JA, et al. Guidelines for ethical behavior relating to clinical practice issues in electrodiagnostic medicine.  Muscle Nerve 2005; 31:400-405.
  36. Hoffman TC, Del Mar C. Patients’ expectations of the benefits and harms of treatments, screening, and tests: a systematic review. Published online December 22, 2014. JAMA Internal Medicine. doi:10.1001/jamaintern-med.2014.6016

Original Version of the Topic

Todd R. Lefkowitz, MD, Richard Kim, MD. Electrodiagnosis of Radiculopathies (Cervical, Thoracic, and Lumbar). 9/10/2015

Previous Revision(s) of the Topic

Todd R. Lefkowitz, MD. Electrodiagnosis of Radiculopathies (Cervical, Thoracic, and Lumbar). 7/24/2020

Author Disclosure

Todd R. Lefkowitz, MD
Nothing to Disclose