What is the AANEM's position on the type of provider qualified to perform electromyography (EMG)?
If you have questions, please contact the policy department at firstname.lastname@example.org
Back to Top
Are physical therapists allowed to perform electromyography (EMG)?
The scope of practice for physical therapists is defined by state laws and the rules enacted by the Board of Physical Therapy in each individual state. However, in many instances, no specific provisions exist on whether or not physical therapists can perform or interpret needle EMG, leaving the issue open for interpretation by the Board of Physical Therapy. In some states, conflicting opinions of the Board of Physical Therapy and the Board of Medicine may co-exist on the issue of needle EMG. Since each board is autonomous in regulating its respective profession and cannot impose its rules or opinions on other health professions (for example, the Board of Medicine is unable to define scope of practice for physical therapists), such conflicting opinions would be considered equally valid, unless it could be demonstrated that one of them violates state laws.
It is the AANEM’s position that physical therapists are not qualified to perform or interpret EMG. You can refer to the AANEM’s Who is Qualified to Practice Electrodiagnostic Medicine Policy
for more information. This position is supported by the research of Dr. Timothy R. Dillingham, Chair of the Department of Physical Medicine and Rehabilitation at the University of Pennsylvania, School of Medicine. Dr. Dillingham’s research examined private payer claims data from 6381 electrodiagnostic encounters for persons with diabetes. Polyneuropathy diagnosis rates were highest for electrodiagnostic physicians—over 12%—compared to 2.1% by physical therapists. A later Dillingham study analyzing Medicare claims also demonstrated a significant difference between electrodiagnostic physicians and non-physician providers (including chiropractors, podiatrists, physical therapists, occupational therapists, nurse practitioners and physician assistants); the specialist physicians were 26% more likely to diagnose polyneuropathy than non-physicians. Significant disparities in recognizing a particular condition may suggest “that quality of electrodiagnostic care may be deficient when it is provided by those without requisite training and specialized knowledge.”
A review of several physical therapy training programs found little evidence that needle EMG is a part of the standard curriculum for physical therapists. There is, however, a program at the Rocky Mountain University that is providing a Doctor of Science (DSc) in clinical electrophysiology. The American Board of Physical Therapy Specialties (ABPTS) has developed an examination for select physical therapists that perform electrophysiology studies. The ABPTS requires that the applicant self-reports evidence of 2000 hours of direct patient care, 500 of which must have occurred in the previous 3 years. Not all of those hours must be dedicated exclusively to electrodiagnostic testing. However, the applicant is required to submit evidence of 500 complete electroneuromyography examinations during those 2000 hours.
Medicare provides coverage for EMGs performed by physical therapists if the physical therapist has completed certification in clinical electrophysiology through the ABPTS. Additionally, Medicare allows certified clinical electrophysiology physical therapists to personally supervise other physical therapists’ performance of EMG. The analysis of Medicare claims demonstrates a steady increase in a number of needle EMG claims filed by physical therapists in recent years. The AANEM maintains scope of practice information on each state and all AANEM members are encouraged to contact the organization with questions about the status of physical therapists in their state. AANEM members are also asked to submit examples of poor quality studies (in a manner that is compliant with HIPAA and any other regulations protecting patient’s privacy) for use in future advocacy efforts regarding this issue.
Back to Top
What is the difference between a national and local coverage decision?
While Medicare is a federal insurance program, local contractors administer the program in each jurisdiction. Jurisdictions are comprised of several states in a region. The local contractors in each jurisdiction are called Medicare Administrative Contractors (MACs), which were previously referred to as Local Carriers. Coverage decisions are made both by Medicare for national coverage issues and by MACs for local coverage decisions. MACs are required to follow national coverage decisions in adjudication of claims. An example of a national coverage decision is the policy that states physical therapists certified in clinical electrophysiology by the American Board of Physical Therapy Specialists will be paid by Medicare for performing and interpreting EMGs unless prohibited by state law. The MACs cannot make a policy that contradicts this policy. Local Coverage Determinations (LCDs) vary in each jurisdiction. An example is the electrodiagnostic LCD for Florida developed by First Coast Service Options. This LCD includes several elements that the AANEM believes support the performance of quality electrodiagnostic medicine studies. The LCD notes that both EMG and nerve conduction study are required for some diagnoses. The LCD states that consistent, excessive claims will be reviewed.
Finally, the LCD highlights that services provide by neurologists and physiatrists are covered; all others must demonstrate competency through training in an applicable residency or fellowship or very extensive CME. This LCD is only applicable to the state within First Coast’s jurisdiction, which includes Florida, Puerto Rico and U.S. Virgin Islands. Sometimes MACs have similar LCD language, but often they are different. You can find the MAC that covers your jurisdiction and each MAC’s coverage decisions at aanem.org.
Back to Top
I have a patient with a deep brain stimulator.
Deep brain stimulators (DBS) are increasingly prevalent in patients with Parkinson’s disease, dystonia, and other disorders. The DBS devices consist of either (1) a single stimulator implanted on either side of the pectoralis muscle which is capable of stimulating the subthalamic nucleus bilaterally through two separate leads or (2) two stimulating devices, one placed on each side of the chest, stimulating only the ipsilateral side. The DBS leads typically traverse subcutaneously from the subclavicular area to the lateral-posterior neck and then over the occiput to penetrate the skull at variable sites in the parietal area.
Electromagnetic interference from medical and household devices may cause DBS devices to switch ON or OFF. Also, some patients may “experience a momentary increase in their perceived stimulation” described as uncomfortable (Medtronic Physician and Hospital Staff Manual; Soletra® & Kinetra® devices). More importantly, nerve conduction studies pose a theoretical risk of introducing electrical current through the leads, which could be transmitted directly to the brain. The typical stimulation intensity of DBS devices range from 12–50 ìA, which is far below current employed in routine nerve conduction study (personal communication, Medtronics). The course of the DBS leads through the supraclavicular and occipital areas may pose additional risks to Erb’s point and cervical root stimulation. As there currently are no studies assessing the safety of performance of nerve conduction study in patients with DBS devices, the physician should evaluate the risks and benefits of electrodiagnostic testing in each case.
Back to Top
I have a patient with a pacemaker.
Cardiac pacemakers and implanted cardiac defibrillators (ICDs) are increasingly used in clinical practice, and no evidence exists indicating that performing routine electrodiagnostic studies on patients with these devices poses a safety hazard. However, there are theoretical concerns that electrical impulses of nerve conduction studies could be erroneously sensed by devices and result in unintended inhibition or triggering of output or reprogramming of the device54. In general, the closer the stimulation site is to the pacemaker and pacing leads, the greater the chance for inducing a voltage of sufficient amplitude to inhibit the pacemaker. Despite such concerns, no immediate or delayed adverse effects have been reported with routine nerve conduction study.
A single study of 10 patients with pacemakers with bipolar sensing configurations and 5 patients with ICDs found no evidence of cardiac device sensing or malfunctioning with routine nerve conduction study utilizing surface stimulation (including left Erb’s point stimulation in 9 patients).54 The authors did not evaluate ICDs placed in the lower abdomen or pacemakers with unipolar sensing mode. Although percutaneous nerve stimulation may be performed in patients with implanted cardiac pacemakers with little risk, 39 complete inhibition of a unipolar pacemaker in conjunction with an interscalene nerve stimulator (utilized for regional anesthesia) was reported.25 A stimulator, therefore, should be used only with extreme caution if it is necessary to stimulate the brachial plexus ipsilateral to a pacemaker or ICD implantation site, particularly if it is unknown if the sensing mechanism is unipolar or bipolar. Caution is advised when performing intramuscular percutaneous stimulation, especially in the upper extremities, as the safety of its performance has not been established. Special care should be given to proper grounding of the patient. Routine consultation with the patient’s cardiologist is not required. In patients with external cardiac pacemakers, the conductive lead, inserted into the heart (usually transvenously) and connected to the external cardiac pacemaker, presents a serious potential hazard of electric injury to the heart.2 Nerve conduction studies are not recommended in any patient with an external conductive lead terminating in or near the heart. The nature of recurrent and frequent electrical impulses that may occur with repetitive stimulation or eliciting somatosensory evoked potentials (SSEP) poses a special circumstance. Nerve stimulation in the lower extremities or in distal upper extremities would be unlikely to have untoward effects upon pacemakers or ICDs. Repetitive stimulation for assessing integrity of the neuromuscular junction typically necessitates study of proximal and/or cranial nerve-innervated muscles, which may place the stimulating electrode closer to the cardiac device.
Nonetheless, as there are no data to determine the safety of performing these procedures in patients with pacemakers or ICDs, proximal upper extremity and cranial nerve stimulation sites should be avoided for repetitive and SSEP stimulation. Needle EMG recording does not introduce electrical current into the body and, therefore, poses no risk of interference with implanted cardiac devices.
Back to Top
I have a pregnant patient.
No known contraindications exist from performing needle EMG and nerve conduction studies on pregnant patients. In addition, no complications from these procedures have been reported in the literature. Evoked response testing, likewise, has not been reported to cause any problems when performed during pregnancy.
Back to Top
My patient has a prosthetic joint.
Prosthetic joints may become infected postoperatively due to hematogenous spread of bacteria. Bacteria may enter the circulatory system through infections involving the surgical site or other noncontiguous tissues or following procedures producing bacteremia including dental procedures and gastrointestinal studies. 4,26,5,61,43 The risk for prosthetic joint infection declines rapidly during the first few postoperative months and continues to decline during the first 2 postoperative years.11 There are no published reports of complications related to needle EMG in patients with prosthetic joints. Based upon current published literature, it is the opinion of the AANEM that there is no contraindication to needle EMG in patients with prosthetic joints when sterile single patient use or properly autoclaved needle electrodes are utilized and infected spaces are not traversed by the needle electrode.
Back to Top
What are the electromyography (EMG) risks to patient in critical care?
ELECTRICALLY SENSITIVE PATIENTS
EDX Studies in the Critical Care Unit
The critically ill patient is at particular risk for electrical injury because certain protective factors may not be operative.2 Two important defenses against electric injury are frequently lost in these patients. First, the high skin resistance provided by dry, intact skin is often breached by intravenous and intra-arterial catheters with leakage and spills around the catheter site. With lowered resistance, current applied in these areas will be conducted more efficiently to the rest of the body, including the heart.42,35,32,24
The second important protection against electric injury is the large volume of soft tissue which surrounds the heart (i.e., the trunk) and dilutes any electric current applied to the body, protecting the heart from direct electric current application. In the critically ill patient, intracardiac catheters are now commonplace. Such catheters bypass this large electric sink and provide small, otherwise harmless currents, potentially lethal, direct access to the immediate vicinity of the heart(microshock).2 Most manufacturers make intracardiac devices electrically isolated so that they will not conduct electricity. The same attention, however, must be given to these catheters as to other percutaneous catheters; for example, stimulation in the immediate vicinity of the catheter should be avoided and should never be done in the presence of fluid spills or leakage.
Two common sources of current which might affect the hospitalized patient are leakage current from attached electric equipment and applied current from stimulators that are a part of electrodiagnostic machines. Stimulator complications can be avoided by refraining from stimulating next to areas with percutaneous catheters and especially avoiding areas where there is fluid leakage. “Leakage current” is current that leaks to the instrument chassis and then can be delivered to the connected patient if improper grounding conditions exist. The maximum current allowed to leak from the case or from patient connections is 10μA.59,9,8,7,6,37 The electrodiagnostic physician is responsible for ensuring the machine in use meets these minimum specifications. Providing proper patient grounding is necessary to protect patients from electric injury. The third ground wire is required on all electric equipment for patient use because it provides a harmless route for any chassis leakage current. The current flows directly to the ground, instead of to the patient. Testing of the third ground wire integrity and outlet grounds should be performed at regular intervals.55 Special safety considerations arise when patients are connected to multiple machines. Defects in outlet grounds or ground faults may occur in individual outlets. Thus, if a person is connected to equipment supplied from different outlets, one with a functional ground and the other nonfunctional, leakage current may flow from the machine connected to the nonfunctional outlet round, through the patient into the functional outlet ground wire of lower voltage. Thus, it is recommended that the patient be disconnected from all nonessential electric equipment. The remaining equipment should be plugged into the same outlet or, at least, outlets in the same vicinity which are likely to share a common ground. When using electrodiagnostic equipment, it is recommended that the ground be placed between the stimulator and the recording electrodes, as well as keeping ground and needle electrodes in close proximity. This practice helps ensure that any leakage current or applied current will return to ground and not spread to the rest of the body. If proper attention is given to equipment leakage current, grounding, and location and type of percutaneous catheters, electrodiagnostic testing of the electrically sensitive patient can be performed without risk.
Electrodiagnostic laboratories should have in place a power outage and surge rotection policy. In an effort to assist laboratories in meeting this standard, the AANEM has developed model policies: Model Power Outage Protection Policy
Back to Top
What are the risks associated with performing an electromyography (EMG) on a patient taking blood thinners?
Bleeding and hematoma are potential risks of needle EMG in patients with or without disorders of hemostasis. There is limited data regarding the incidence of clinically significant bleeding complications from needle EMG and any additional risk in patients who are receiving antiplatelet or anticoagulant therapy or who suffer from thrombocytopenia or clotting factor deficiencies. There are many different blood thinning medications now available, in addition to some herbal remedies that have anticoagulant properties (see Appendix). The degree of blood thinning cannot be monitored with some of the newer medications such as rivaroxaban but there may be a ceiling effect related to limited solubility with increasing doses.
Despite the inherent risk of needle EMG in patients with and without increased bleeding tendencies, since the technique was first developed in the 1960s, there have been only two case reports of compartment syndrome occurring after needle EMG, and in neither case was the patient taking blood thinning medication.[1, 2] There have only been four reports of symptomatic hemorrhage following needle EMG in patients taking blood thinning medication. However two patients suffered trauma between the time of EMG and diagnosis of the hemorrhage; thus it is not clear that the EMG was the cause.[3, 4] The third case was an 81-year-old woman with thalassemia, on warfarin (INR 2.5) who developed painful swelling at the procedure site 2 days after EMG and was found to have calf hematoma and a posterior tibial pseudoaneurysm which was managed conservatively. The fourth case was published in 2013, by Crisan et al., and reported a large gluteal hematoma diagnosed 3 days after needle EMG due to a 4 g/dl drop in hemoglobin. CT scan revealed extensive intramuscular hematoma in the piriformis, gluteus medius/minimus, tensor fascia lata and vastus lateralis (muscles examined during the EMG included gluteus medius and vastus lateralis). Angiogram showed active bleeding from a distal branch of the left superior gluteal artery. The patient also developed spontaneous hematoma in the unexamined posterior thigh muscles and had ecchymosis in one arm and in the abdomen but coagulopathy workup was negative. The patient was taking aspirin, subcutaneous heparin and oral methylprednisolone at the time of EMG. In a survey of 47 electrodiagnostic laboratories with ACGME-approved fellowships, 3 laboratories reported a single instance of serious bleeding complications (requiring intervention) occurring in anticoagulated patients and one laboratory reported two instances of serious bleeding complications, in the history of their lab recollection.
There have been several retrospective studies examining the risk of paraspinal hematoma following needle EMG. An often quoted study by Caress et al. was triggered after an asymptomatic but quite large hematoma was noted in the lumbar paraspinals of a young woman after needle EMG. The authors then performed an uncontrolled retrospective review of 17 further cases and found 4 other small, asymptomatic hematomas on MRI, but none of these hematomas were diagnosed on the original MRI report. Since then Gertken et al. have published a large case series of 370 patients who underwent paraspinal EMG within the 7 days preceding spine MRI (a total of 431 paraspinal areas were examined with both EMG and MRI). There were no hematomas detected by 2 radiologists who independently reviewed the images. London et al. published a smaller controlled blinded study, comparing paraspinal hematoma rates in patients with and without EMG preceding the MRI. No hematomas were detected in the 29 patients who underwent EMG prior to MRI (many of whom were taking aspirin and/or NSAIDs), and 2 hematomas were found on MRI in control patients who had not undergone EMG prior to the MRI. There were some equivocal findings labeled "possible hematoma" but these were distributed equally between the 2 groups.
There have been two prospective studies using ultrasound to visualize muscles post EMG to evaluate for hematoma formation. In the first study by Lynch et al. three very small, subclinical hematomas were seen in 209 muscles (the tibialis anterior). There was no statistical difference in hematoma rates in subjects taking anti-platelet or anticoagulant therapy compared to control subjects. In the second study by Boon et al. seven "high risk" muscles were examined with ultrasound after needle EMG (the 3 segments of paraspinal muscles, iliopsoas, flexor pollicis longus, posterior tibialis and flexor digitorum longus). Out of 323 muscles examined, only 2 showed evidence of a small, asymptomatic hematoma, one in a patient taking clopidogrel (posterior tibialis muscle) and the other in a patient taking warfarin with INR of 2.3 (flexor pollicis longus). Gertken et al. has since published a review of the literature to date, in which the total number of muscles imaged post EMG is 1037, including 488 controls, 222 patients taking anticoagulants, 328 taking aspirin or clopidogrel, 35 taking NSAIDs and 3 taking herbal remedies that could affect clotting. There were 10 asymptomatic hematomas found in this group, giving an overall rate of 0.96% risk of hematoma formation post EMG and in the specific subgroups: controls 1.02%, antiplatelet agents 0.61%, anticoagulants 1.35%.
In summary, there is a growing body of literature to support the safety of needle EMG, in patients with and without increased bleeding risk. In patients taking anticoagulation medication, the thrombotic risk of discontinuing anticoagulation prior to EMG outweighs the risk of the needle examination while on anticoagulation.[14,15] Nonetheless, needle EMG is an invasive procedure, and each case should be considered individually with regard to the potential benefits of the study relative to the risks of intramuscular hemorrhage or other bleeding. In such situations, needle EMG should be performed with added caution and it is prudent to first examine small, superficial muscles which are easily compressible to watch for bleeding problems. Prolonged pressure over the needle site will usually produce hemostasis. Some practitioners utilize vapocoolant spray to improve hemostasis, although there are no studies assessing the utility of this technique. Likewise, no data indicate that various needle parameters (e.g. gauge, monopolar vs. concentric, etc.) present different risks for bleeding complications.
Farrell CM, Rubin DI, Haidukewych GJ. Acute compartment syndrome of the leg following diagnostic electromyography. Muscle Nerve 2003;27(3):374-377.
Vaienti L, Vourtsis S, Urzola V. Compartment syndrome of the forearm following an electromyographic assessment. J Hand Surg [Br] 2005;30(6):656-657.
Baba Y, Hentschel K, Freeman WD, Broderick DF, Wszolek ZK. Large paraspinal and iliopsoas muscle hematomas. Arch Neurol 2005;62(8):1306.
Butler ML, Dewan RW. Subcutaneous hemorrhage in a patient receiving anticoagulant therapy: an unusual EMG complication. Archives of physical medicine and rehabilitation 1984;65(11):733-734.
Crisan E, Patil V, Chawla J. Gluteal hematoma after a needle electromyography examination of the gluteal medius muscle. Muscle & Nerve 2013;Suppl.
Rosioreanu A, Dickson A, Lypen S, Katz DS. Pseudoaneurysm of the calf after electromyography: sonographic and CT angiographic diagnosis. AJR Am J Roentgenol 2005;185(1):282-283.
Gruis KL, Little AA, Zebarah VA, Albers JW. Survey of electrodiagnostic laboratories regarding hemorrhagic complications from needle electromyography. Muscle Nerve 2006;34(3):356-358.
Caress JB, Rutkove SB, Carlin M, Khoshbin S, Preston DC. Paraspinal muscle hematoma after electromyography. Neurology 1996;47(1):269-272.
Gertken JT, Hunt CH, Chinea NI, Morris JM, Sorenson EJ, Boon AJ. Risk of hematoma following needle electromyography of the paraspinal muscles. Muscle & Nerve 2011;44(3):439-440.
London Z, Quint DJ, Haig AJ, Yamakawa KS. The risk of hematoma following extensive electromyography of the lumbar paraspinal muscles. Muscle & Nerve 2012;46(1):26-30.
Lynch SL, Boon AJ, Smith J, Harper CM, Jr., Tanaka EM. Complications of needle electromyography: hematoma risk and correlation with anticoagulation and antiplatelet therapy. Muscle & Nerve 2008;38(4):1225-1230.
Boon AJ, Gertken JT, Watson JC, Laughlin RS, Strommen JA, Mauermann ML, Sorenson EJ. Hematoma risk after needle electromyography. Muscle & Nerve 2012;45(1):9-12.
Gertken JT, Patel AT, Boon AJ. Electromyography and anticoagulation. Pm R 2013;5(5 Suppl):S3-7.
Petersen, P., et al., Placebo-controlled, randomised trial of warfarin and aspirin for prevention of thromboembolic complications in chronic atrial fibrillation. The Copenhagen AFASAK study. Lancet, 1989. 1(8631): p. 175-9.
Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation. Analysis of pooled data from five randomized controlled trials. Arch Intern Med 1994;154:1449-1457.
Appendix: Examples of commonly prescribed medications that affect coagulation
Warfarin (Coumadin): monitored with PT/INR
Dabigatran (Pradaxa): no universal monitoring currently
Rivaroxaban (Xarelto): can be monitored with PT, aPTT, or anti-Xa activity
Apixaban (Eliquis): monitoring is not usually required
Intravenous and subcutaneous anticoagulants
Heparin: monitored with aPTT
Dalteparin (Fragmin): monitoring is not usually required
Enoxaparin (Lovenox): monitoring is not usually required
Fondaparinux (Arixtra): can be monitored with anti-factor Xa assay
Ardeparin (Normiflo): not currently available in the USA
Danaparoid (Orgaran): can monitor factor Xa assay
Lepirudin (Refludin): can monitor with aPTT ratio
Urokinase (Abbokinase): not routinely monitored
Alteplase (Activase): not routinely monitored
Cyclooxygenase inhibitors (antiplatelet agents)
Aspirin: monitoring is not usually required
Triflusal: monitoring is not usually required
Adenosine reuptake inhibitors (antiplatelet agents)
Dipyridamole (Aggrenox): monitoring is not usually required
ADP receptor inhibitors (antiplatelet agents)
Clopidogrel (Plavix): monitoring is not usually required
Ticagroler (Brilinta): monitoring is not usually required
Ticlopidine (Ticlid): monitoring is not usually required
Plasugrel (Effient): monitoring is not usually required
Phosphodiesterase inhibitors (antiplatelet agents)
Cilostazol (Pletal): monitoring is not usually required
Glycoprotein IIB/IIIA inhibitors
Abciximab (Reopro): monitoring is not usually required
Eftibatide (Integrilin): monitoring is not usually required
Tirofiban (Aggrastat): monitoring is not usually required
Terutroban: monitoring is not usually required
NSAIDs (such as naproxen, ibuprofen, meloxicam): no monitoring
Herbal supplements (including St John's wart, Saw palmetto herb, Ginkgo biloba, garlic, ginseng, Hsien ho Tsao): no monitoring
Abbreviations: PT: prothrombin time; INR: international normalized ratio; aPTT: activated partial thromboplastin time; NSAID: nonsteroidal anti-inflammatory
Back to Top