Electrical Safety and Iatrogenic Complications of Electrodiagnostic Studies

Electrical Issues

All electrical devices, including EMG machines, require current to operate. Current is delivered from an electrical cord plugged into a wall receptacle ( Fig. 43.1 ). A typical electrical receptacle in the United States contains three inputs: a black “hot” lead that carries 120 volts (V) of 60-Hz alternating current, a white “neutral” lead near 0 V, and a green ground lead that is used to dissipate leakage currents. When a circuit is created, current flows from the hot lead to the EMG machine and then returns via the neutral lead, based on the amount of resistance between the two leads as determined by Ohm’s law (see Chapter 42 ). Every wire, including power cords, has some small resistance; thus, a small voltage develops on the neutral lead, which equals the current flowing multiplied by the resistance in the power cord ( Fig. 43.2 ). The voltage increases with the length of the power cord and increases further if extension cords are added to the power cord. In addition, small voltage leaks are often present on the machine chassis, caused by stray capacitance and inductance from internal electronics ( Fig. 43.3 ). Thus, leakage currents can be transmitted onto the patient either from stray voltages on the machine chassis or on the neutral (reference) lead. As the ground electrode is close to true electrical neutral, the ground lead allows a pathway for stray current leaks to dissipate.

The risk of electrical injury depends on the amount of leakage current and whether the circuit passes through the heart. A very small current (e.g., 200 microamperes [μA]) applied directly to the heart can result in ventricular fibrillation and death. However, the normal, healthy individual typically is well protected by two important mechanisms. First, dry and intact skin provides a high resistance. Second, the large volume of soft tissue that surrounds the heart dilutes any current applied to the body (e.g., a current applied from arm to arm degrades to 1/1000 of the original signal when it reaches the heart, due to the dissipation from surrounding tissues).

The risk of electrical injury from leakage current increases in the following situations:

• Malfunctioning of the electrical equipment

• Multiple electrical devices attached to the patient

• Loss of the body’s normal protective mechanisms

The latter two (multiple electrical devices attached to the patient and loss of the body’s normal protective mechanisms) result in the “electrically sensitive” patient, a common situation in the intensive care unit.

To prevent the possibility of an electrical injury during EDX studies, it is essential for equipment to be regularly maintained, to always use a ground electrode, and to follow simple guidelines when using electrical devices attached to the patient ( Box 43.1 ). A wooden examining table is preferable to a metal table, as it does not conduct electricity ( Fig. 43.4 ). Machines should be turned on before attaching electrodes to the patient and turned off after disconnecting the patient, to minimize the risk of power surges. Equipment should be periodically inspected by a biomedical engineer to measure leakage current and verify proper grounding. In general, the maximum amount of acceptable leakage current is 100 μA or less, measured from chassis to ground, and 50 μA or less from any input lead to ground. Extension cords should be avoided to reduce the risk of voltages developing on the reference electrodes. Ground electrodes should always be used to avoid current flows from reaching the patient. The ground needs to be placed on the same limb as the active electrodes so that leakage currents cannot flow in a path through the heart ( Fig. 43.5A ).

The issue of an intact ground electrode and proper ground placement is most important when a patient is connected to other electrical devices. If the ground from the EMG machine is not functioning (i.e., ground fault), stray current from the EMG machine could flow to a ground electrode from a different electrical device. If the pathway included the heart and the amount of current was large enough, a cardiac arrhythmia could theoretically occur ( Fig. 43.5B ).

However, most modern medical devices, including EMG machines, are designed with electrical isolation. With electrical isolation, the current generated by the stimulator and potentials recorded by the electrodes are physically separated from the main EMG device, which is connected to a wall current and an earth ground. This limits the possibility that leakage current from the main machine can travel to the patient and return via some other earth ground attached to the patient. Electrical isolation is accomplished optically: the electric signals from the amplifier are converted to light by a light-emitting diode (LED). The brightness of the LED is proportional to the electric signal. A photodetector in the main device then picks up the light and converts it into an electric signal, which is then processed in the EMG machine. In this way, the electrical circuit in contact with the patient is physically separated from the electricity in the EMG machine. If leakage current from the EMG machine does reach a patient, the recording electrodes, including the ground electrode from the preamplifier, offer no physical path for the that current to flow through a patient to return to an earth ground.

Risk of Electrical Injury

Central Lines and Electrical Wires

One of the more common ways a patient can become electrically sensitive is when the normal protective function of the skin is breached by intravenous lines and wires. This danger increases if the lines are actually in contact with or in close proximity to the heart, as occurs in central intravenous catheters ( Fig. 43.6 ). Most dangerous is the presence of an external wire near or in the heart, such as occurs with placement of a temporary external pacemaker or during the use of a guidewire while placing or changing a central line. Skin resistance typically is several million Ohms (MΩ). A central catheter traversing the skin reduces this resistance to 300,000 Ohms. Any fluid spill where a catheter enters the body decreases the resistance even further. If a catheter has an internal guidewire, the resistance drops to 70 Ohms (Ω). An external pacemaker wire essentially has no resistance. In situations where the resistance is so low, small leakage voltages may result in small leakage currents, known as microcurrents . Whereas microcurrents are completely harmless in a patient with intact skin, they are potentially very dangerous in an electrically sensitive patient (i.e., a patient with a central line, external pacemaker wires, etc.).

Thus EDX studies should never be performed on patients with external wires in place (i.e., external pacing wire, guidewires, etc.) because the conductive pathway to the heart is so vulnerable. However, studies can be performed on patients with central lines provided certain precautions are followed. Equipment must be maintained. Ground electrodes must always be used. If an upper extremity must be studied, in general, it is preferable and safer to study the upper extremity contralateral to the side with the central line. If that is not possible, one should refrain from proximal stimulation sites (i.e., axilla, Erb’s point, and root). Likewise, one should never proceed if there is a fluid spill where the central catheter enters the skin. It is important to note, however, that there is NO contraindication to performing routine NCSs on patients with peripheral IVs. Studies have been reported that specifically address this question and they find that NCSs are completely safe in patients with peripheral IVs, regardless of whether they are infusing saline or any other solution. Although NCSs in patients with central lines have not demonstrated any adverse reactions, the numbers that have been studied are relatively small. Thus it remains reasonable to follow the precautions provided here.

Implanted Pacemakers and Cardioverter-Defibrillators

Patients with implantable cardiac pacemakers and cardioverter-defibrillators are at much lower risk from stray current leaks than patients with central lines or external wires in place, because these devices are implanted under the skin, which leaves the normal protective mechanism of the skin intact. Implantable pacemakers and cardioverter-defibrillators both have electronic-sensing and electronic-delivery functions. Pacemakers are designed to treat bradycardia, as opposed to cardioverter-defibrillators, which are primarily used for tachyarrhythmias, especially ventricular fibrillation. In theory, stimulation delivered during NCSs might be mistaken as an abnormal cardiac rhythm. If the stimulator has a pulse duration greater than 0.5 ms and a stimulus rate greater than 1 Hz, a demand pacemaker might theoretically confuse such a stimulus with the ECG signal. There is only a single case report of an implantable pacemaker failure thought to be related to peripheral nerve stimulation. Other studies have shown no pacemaker inhibition or dysfunction with NCSs. Less is known about implantable automatic cardioverter-defibrillators (IACDs), which are now common. In theory, IACDs could be triggered by stimulation during NCSs, resulting in subsequent cardiac arrhythmias; however, there are no such reported cases. One study directly addressed the safety of NCSs, including stimulating Erb’s point, in patients with IACDs. Schoeck and colleagues studied 10 patients with pacemakers and 5 with IACDs. No electrical impulse was detected by either the atrial or ventricular amplifiers of the pacemakers or of the IACDs during median and peroneal NCSs. These studies included Erb’s point stimulation on the left side. The authors emphasized that all modern pacemakers and IACDs use bipolar leads wherein both leads (active and reference for sensing, and cathode and anode for stimulating) are imbedded in the cardiac wall. This is in contradistinction to the pacemakers used 25 years ago wherein a single wire lead was placed in the heart, and the metal body of the pacemaker in the chest served as the reference. In modern pacemakers and IACDs, the bipolar leads are very close together in the heart, and very far away from the surface, making any electrical contamination from NCSs extremely unlikely. Although the number of patients in this study was small, the results are reassuring that NCSs can be safely performed in patients with pacemakers and IACDs.

If NCSs are performed in patients with implantable pacemakers or IACDs, several simple procedures are recommended to be followed to preserve safety ( Box 43.2 ). Stimulation should not be performed near the actual implanted device. There should always be a minimum of 6 inches between the implanted device and the stimulator. Just as with NCSs performed in a patient with a central line, it is preferable to use the contralateral arm if possible. High stimulus intensities should be avoided and stimulus pulse duration should be 0.2 ms or less so that the stimulation is not misinterpreted as a QRS complex. Stimulation rates should be no greater than 1 Hz so as to prevent the theoretical risk that the stimulation is misinterpreted as a cardiac rhythm. Thus, the typical repetitive stimulation done during neuromuscular junction testing is best avoided.

Box 43.2

Guidelines for Pacemakers and Implantable Cardioverter-Defibrillators

With permission from Al-Shekhlee A, Shapiro BE, Preston DC. Iatrogenic complications and risks of nerve conduction studies and needle electromyography. Muscle Nerve . 2003;27:517–526, with permission.

• Do not perform studies on patients with external pacer wires.

• Ensure that all ground electrodes are functional.

• Limit all electrodes, including the ground, to the extremity of interest, and keep all electrodes as far away from the heart as possible, without crossing cardiac devices or their wires.

• Do not stimulate near the device (allow a minimum of 6 inches) and avoid ipsilateral proximal stimulation sites (i.e., axilla, Erb’s point, root stimulation).

• Use a stimulus duration of 0.2 ms or shorter and a stimulus rate of 1 Hz or slower. Thus, the typical repetitive stimulation done during neuromuscular junction testing is best avoided.

• Consult a cardiologist regarding performing studies in patients with an implantable automatic cardioverter-defibrillator.

• Laboratory emergency drugs should be available, including crash carts.

Pneumothorax

Pneumothorax is the most potentially serious iatrogenic complication of needle EMG. At any time during or just after the EMG examination, unexpected chest pain, shortness of breath, or cyanosis in a patient should alert the electromyographer to the possibility of a pneumothorax. If such symptoms develop, a prompt chest x-ray film is indicated to confirm the diagnosis, followed by urgent consultation with a thoracic surgeon as to whether a chest tube or observation is required. Use of neuromuscular ultrasound to aid in EMG needle placement in muscles at high risk for pneumothorax is discussed in detail in Chapter 40 . Although rare, this complication has been most often reported when sampling the following high-risk muscles ( Fig. 43.7 ):

• Diaphragm. Needle EMG of the diaphragm is sometimes used to help determine whether respiratory insufficiency has a neuromuscular basis. However, because the pleural fold is in close proximity to the diaphragm, a relatively small error in needle position may increase the risk of inadvertent pleural puncture and possible pneumothorax. The decision to sample the diaphragm depends on the experience of the electromyographer along with weighing the risk of pneumothorax to the potential benefit in that particular patient. Because patients for whom this study is ordered often have respiratory problems that prompt the study to be ordered, they may be the least able to handle an additional respiratory complication. In Chapter 13 , we purposely did not include the diaphragm as a routine muscle to study during needle EMG. In our opinion, the risk-to-benefit ratio of sampling this muscle by using surface landmarks is too high to justify its use as a routine muscle to be sampled. However, if ultrasound guidance is employed, safety of needle EMG of the diaphragm is assured. Ultrasound can be used either directly or indirectly to safely perform diaphragmatic needle EMG (see Chapter 40 for full discussion).

• Serratus anterior. The serratus anterior muscle lies between the scapula and the chest wall and inserts laterally on the ribs. An inadvertent puncture through the muscle between the ribs may allow the needle to enter the pleural space. To reduce the possibility of pneumothorax, the muscle can be sampled with the electromyographer’s fingers placed in two adjacent inter-rib spaces while the needle is inserted into the muscle directly over the rib.

• Supraspinatus. The supraspinatus muscle lies within the supraspinous fossa of the scapula. The middle of the fossa may be very shallow in some individuals. Thus, if the muscle is sampled too deeply at this point, the needle may puncture the pleura ( Fig. 43.8 ). Complica tions can be prevented by either avoiding the muscle altogether or sampling it more medially in the supraspinous fossa. This is performed by first palpating the acromion, the spine of the scapula, and the vertebral border of the scapula. The needle is then inserted just above the spine of the scapula at a point three-quarters of the distance from the acromion to the vertebral border of the scapula. Often, the supraspinatus can be avoided by sampling the infraspinatus instead. The infraspinatus muscle and infraspinous fossa are much larger than the supraspinatus and supraspinous fossa above. When screening for a suprascapular neuropathy, the infraspinatus muscle is the preferred muscle to study. Only if the infraspinatus muscle is abnormal is it then necessary to sample the supraspinatus to differentiate a lesion at the spinoglenoid notch from one at the suprascapular notch or above (see Chapter 34 ).

  

• Rhomboids. The rhomboids are infrequently sampled. However, they are useful to study in two situations: (1) to differentiate a C5 from a C6 radiculopathy (the rhomboids are derived from the C4– C5 roots) and (2) to differentiate an upper trunk brachial plexopathy from a more proximal radiculopathy (the rhomboids are innervated by the dorsal scapular nerve, which arises directly from the ventral rami of the roots proximal to the brachial plexus). Because the rhomboids originate on the dorsal spine and insert onto the medial border of the scapula, a needle placed too deeply may pass through the rhomboids and thoracic paraspinal muscles, resulting in a pleural puncture.

• Cervical and thoracic paraspinal muscles. The cervical paraspinal muscles are commonly sampled in the evaluation of cervical radiculopathy. Thoracic paraspinal muscles are one of the key sites to study in the evaluation of suspected motor neuron disease. These muscles can be safely studied, provided the needle placement is neither too lateral nor too deep. Considering the proximity of the thoracic paraspinal muscles to the lungs in the thorax, it is not unexpected that pneumothorax is a potential complication of thoracic paraspinal muscle sampling ( Fig. 43.9 ). However, pneumothorax can also occur during EMG examination of the lower cervical paraspinal muscles or when an EMG needle is used for cervical nerve root stimulation. Some patients, especially those who are thin with longer necks, may have lung tissue that reaches above the clavicle ( Fig. 43.10 ). In one study of 23 patients, 22% had lung tissue above the level of the clavicle. The average distance between skin and lung in these individuals was 3.3 cm, a distance clearly within the reach of a conventional 37- or 50-mm EMG needle. This complication is easily prevented by ensuring that the needle remains close to the midline, within the bulk of the paraspinal muscles.

  
  

In a study by Kassardjian and colleagues at the Mayo Clinic, seven cases of symptomatic pneumothorax were seen over 18 years in the evaluation of 64,490 patients. The most common muscle sampled associated with pneumothorax was the serratus anterior (0.445%), followed by the diaphragm (0.149%). In these rare cases, the time period when patients became symptomatic ranged from during the performance of the EMG itself to up to 1 day later.