To the Editor:—
We read with great interest the article by Urmey and Stanton. 1It would appear, however, that the validity of the results must be viewed with some caution and the authors’ conclusions, in particular, call for several comments.
First, use of fresh batteries for a nerve stimulator is not sufficient to prove its adequacy. A peripheral nerve stimulator in use for several months might not perform as well as a new one. 2For use of this importance, the stimulator output should have been tested with a calibrated oscilloscope. A stimulator that underestimates current output can mislead the operator. Moreover, a stimulator with clearly marked polarity would have been preferable. If polarity is inversed inadvertently, more current is needed.
Nerves are not fixed in the surrounding tissue. Despite the needle presumably being near the nerve and immobilized when paresthesia are reported, imperceptible patient movement or breathing can dislodge the needle. There is no doubt that for patient comfort and security, as well as ethical reasons, the solution was injected after applying the stimulator only long enough to achieve paresthesia and not while maintaining paresthesia. To date, how close a needle must be to a sensory fascicle to produce mechanical paresthesia has not been established. 3Is the needle situated inside the nerve in contact with the sensory fascicle of the reported paresthesia, just touching the nerve, or somewhere in the vicinity of the nerve? Shoulder paresthesia can be provoked by deep palpation of the interscalene groove in thin patients.
Why varied responses are observed after mechanical or electrical stimulation of the same root or trunk can be explained by fine details of anatomy. With the interscalene technique, the plexus is located at the level of the trunk and/or anterior branch of the spinal nerve. As a result, the response elicited is clearly metameric. The upper trunk and C5-C6 contain sensory fascicles that distribute to the upper lateral brachial cutaneous nerve (axillary nerve; C5), the lower lateral brachial cutaneous nerve (radial nerve; C5-C6), the lateral antebrachial cutaneous nerve (musculocutaneous nerve; C6), and the palmar cutaneous branch of the median nerve (C6). Consequently, paresthesia extending from the shoulder to the hand is not surprising with the interscalene approach. Similarly, with a nerve stimulator, the contractions commonly mediated by the cephalad roots (C5-C6) or upper trunk are those of the supraspinatus, infraspinatus (C4-C6), deltoid (C4-C6), biceps brachii (C5-C6), brachioradialis (C5-C6), extensor carpi radialis longus (C6-C7), pronator teres (C6-C7), and flexor carpi radialis (C6-C8) muscles. A distal paresthesia or motor response does not necessarily imply that the needle has been inserted too deeply, contrary to what had been suggested elsewhere. 4Lastly, satisfactory results require taking care not to regard as positive the shoulder responses that are due to stimulation of a collateral branch either directly or with the shaft of an uninsulated needle.
In addition, electrophysiologic aspects of nerve stimulation and its clinical application are important to consider. Nerves coursing from the plexus are mixed, consisting of different nerve fascicles, including groups of axons that determine a common and specific motor or sensory innervation. With the minimum current able to produce a stimulus, large A α motor fibers are stimulated, while small A δ and C fibers are not. In such cases, a motor response is obtained without pain or patient discomfort. The shoulder is innervated by 30% of the fibers of the brachial plexus, 28.4% for motor innervation, and only 1.6% for sensory innervation. 5The assumption by Urmey and Stanton that each paresthesia constituted evidence of contact with a sensory fascicle alone was probably unfounded. Sensory and motor fascicles do not appear to be very distant from each other in the brachial plexus. Eliciting a motor response rather than a sensory response should occur frequently.
However, when pure sensory fibers were stimulated, confirmation of the needle tip being in contact with the nerve is obtained by eliciting paresthesia with each pulsation. 2For successful peripheral nerve blockade, electrical paresthesia at 0.4–0.6 mA (100–200 μs) have been used as endpoint without evidence of mechanical paresthesia or nerve damage. 6,7If contact persisted with a sensory component of the nerve root, how was it possible for none of the patients to report electrical paresthesia to the two observers in spite of the maximum attained current of 1.0 mA, when they clearly indicated having the corresponding mechanical paresthesia several seconds earlier? As a very high stimulus is needed once the tip is some distance away from the nerve according to Coulomb's law, the only credible explanation is that, in 70% of the patients studied, the needle was sufficiently remote from the nerve after the initial paresthesia.
On the contrary, a very low current was sufficient to produce the response in the other 30% of cases. An uninsulated needle would have required more current to stimulate the nerve. Achievement of a response with such minimal intensity indicates that the needle remained in very close contact or was touching the nerve. 8Urmey and Stanton are to be congratulated because they provided good evidence that the needle has to be quite close to the nerve when paresthesia is elicited. A motor response at 0.1–0.2 mA for 200 μs was obtained in several patients after the paresthesia disappeared. When responses are observed for currents this low, injection should be avoided, notably in unconscious subjects. 9When performing nerve stimulation with an insulated needle, the proper endpoint is the minimal stimulating current. The needle is then released and if the response persists, the needle-nerve relationship should remain unchanged at low current (0.2–0.5 mA for 100 μs). The operator should verify that the motor response disappears by applying decreased current once again before injection. The response should be abolished instantaneously with a painless and easily injected 1 ml-test dose.
Paresthesia occurs and serves as a warning with all techniques including use of nerve stimulators. Paresthesia is usually difficult to elicit with a short-bevelled, insulated needle. 3The observed 30% increase in ability to obtain nerve stimulation with a long-bevelled needle is most likely due to a tendency of short-bevelled needles to press or push the nerve away. 10Long-bevelled needles are sharper, may penetrate the nerve easier, and potentially increase the risk of postoperative dysesthesia. In contrast, despite frequent unintentional paresthesia during block placement, the withdrawal and redirection of stimulating short-bevelled needles is not associated with an increased incidence of neurologic complication, even when using a multiple injection technique. 11Urmey and Stanton would probably have consistently observed a motor response before mechanical or electrical paresthesia had they chosen to begin with an adequate procedure of nerve stimulation at a recommended higher current.
Currently, no compelling evidence exists to endorse a single technique or needle as superior with respect to success rate or incidence of complications. There are no prospective randomized controlled studies that compare the relative risks of regional anesthesia performed on anesthetized or conscious patients. Nevertheless isolated case reports, 12and medicolegal reports provide a background of suspicion, which is difficult to refute with a lack of data either for or against the practice. The incidence of dramatic complications appears to be higher following interscalene approaches to the brachial plexus. 12In the absence of firm data to the contrary, the weight of clinical practice suggests that the majority of peripheral nerve blocks, with the possible exception of the interscalene approach to the brachial plexus, can be performed under sedation. The anesthesiologist should carefully consider whether the benefits of regional anesthesia performed on an anesthetized patient are greater than the risk of a catastrophic outcome. It is important that authors not draw conclusions erroneously based on slight imperfections in a study design, especially conclusions leading to recommendations or contraindications with significant medicolegal connotations.