Bringing the needle close to the nerve(s)
Insulated versus non insulated needles
Short versus long-bevel needles
Persistent Paresthesia, Clinical presentation
Pre-existing neurologic condition and regional anesthesia
Persistent paresthesia prevention and management
Peripheral Nerve Blocks
A successful peripheral nerve block results from injecting an adequate volume of an adequate concentration of local anesthetic in the proximity of the target nerve(s). Intraneural injection (especially intrafascicular) is harmful to the nerve and can lead to permanent damage. Therefore, a balance must be achieved between the need to get close to a nerve and safety.
There are many ways to ascertain the correct placement of a needle with respect to a nerve. A good knowledge of the anatomy makes things easier and safer. The methods are:
This anatomical method, practiced alone, has limited success, because it does not take into account anatomical variations, lacks depth perception and can not gauge proximity to a nerve with any degree of certainty. Therefore, the needle might end up too far from the nerve (failed block) or too close to it (intraneural).
For the longest time, Dr. Moore’s dictum “no paresthesia no anesthesia”, was the “law of the land” in regional anesthesia. Works by Selander and others, starting in the 1970s, have questioned the safety of this practice. Although, there is not enough evidence to believe that paresthesias lead to nerve damage, there seems to be enough circumstantial evidence to be cautious, especially if repeated paresthesias are elicited.
The nerve stimulator is connected to a needle, usually insulated, that delivers a current to its tip. The A alpha fibers (motor) are readily depolarized by the small currents used, but not the sensory fibers. As the needle approaches a mixed nerve, a painless muscle twitch is produced. The intensity of the response is inversely proportional to the needle tip-nerve distance (actually to the square root of it). A visible response at lower currents (less than 0.5 mA), suggests close proximity between the needle tip and the target nerve. There is a good amount of clinical evidence to suggest that a current of 0.5 mA or less, capable of eliciting a visible response, is a reliable indicator of critical proximity. However, evidence is lacking as to what exactly that distance is, and as to whether the distance is different for different nerves. In general it is thought that 1 mA of current will produce depolarization of a motor nerve at a distance of about 1 cm (10 mm).
Nowadays nerve stimulator techniques are widely practiced around the world. Since they do not necessarily rely on patient cooperation, they are sometimes used in unconscious or heavily sedated patients. We do not encourage this practice, as it can lead to complications that a conscious patient could perhaps help prevent (e.g., intraneural injection). With modern nerve stimulators the practitioner can adjust the pulse intensity (magnitude of the current) in mA; the pulse frequency (amount of pulses per second) in Hz (1 or 2) and the pulse width (duration of the pulse) in milliseconds (ms). The pulse duration most suitable for stimulating motor fibers in a mixed nerve is 0.1 ms (100 microsec).
Ultrasound could theoretically produce warming of tissues or gas formation. This technology is still expensive, and requires competency on interpretation of cross-section anatomy from “grainy” images. However, it has been rapidly progressing and it should be a matter of time before it becomes the method of choice.
Characteristics of ultrasound
The human ear can hear sounds between 20 and 20,000 Hz (cycles per second) or 20 KHz. Ultrasounds waves travel at a higher frequency than the highest frequency detectable by the human ear. Ultrasound waves used in medicine usually are in the 1 to 20 MHz range (1 MHz = 1 million Hz). High frequency waves are shorter and good for superficial structures. Ultrasound waves travel easily through fluids and soft tissue, but have problems traveling through bone and air. Ultrasound is better reflected at the transition between two different types of tissues like soft tissue-air, bone-air and soft tissue-bone.
The ultrasound is delivered from a small probe that contains a transducer. The transducer converts electrical signals into ultrasound waves. The transducer detects the reflected waves and converts them back into electrical signals, which are eventually the source of the image we see on the screen. Therefore, the transducer delivers ultrasound for part of the time and for part of the time it “listens” for the returned waves. Some of the wave sounds pass through tissues, while others get reflected back into the transducer. The distance is calculated as a function of the time it takes for the wave to return. Tissues with high density like bones, reflect most of the waves and produce a bright image. These tissues are known as hyperechoic. The hypoechoic structures are soft tissue structures with different degrees of echo.
The more perpendicular the probe is to the structure being searched, (e.g., nerve), the better the image obtained, because more bouncing sound waves can be detected by the transducer. This is also true when trying to visualize the needle. Changes as small as 10 degrees from the perpendicular, can distort the echogenicity of a nerve, by reducing the amount of waves returning to the transducer.
The easiest way to identify a peripheral nerve is on a transverse scan, also called “short axis view”. The needle, on the other hand, can be advanced with the “out-of-plane” approach, crossing the ultrasound beam perpendicularly. The needle becomes practically invisible as its cross section is one more of the thousands of dots that form the ultrasound image. With the “in-plane” approach the needle is advanced parallel to the probe. Depending on its depth and angle of insertion the whole needle can be visualized. With either approach the needle is aimed to the nerve surroundings.
Scanning superficial structures, like the brachial plexus, requires high frequency probes (10-15 MHz) that provide good resolution, but limited penetration (3-4 cm). For deeper structures like the brachial plexus in the infraclavicular region or sciatic nerve in the buttocks, lower frequencies (4-7 MHz) are needed. Deep scanning of intra abdominal organs requires frequencies of 3-5 MHz.
Insulated needles (Teflon-coated) are the needles most commonly used in conjunction with a nerve stimulator in the United States and Europe. They are also used commonly for ultrasound-guided peripheral blocks. The current applied to this needle concentrates at its tip, making the localization of nerves more accurate. Several brands of these needles exist in the market and they come ready with a connection that only fits the negative electrode. Connecting the negative electrode to the exploring needle lowers the amount of current necessary to depolarize a nerve.
Non-insulated needles transmit the current preferentially to the tip, but also along the shaft of the needle making the localization of nerves less accurate. Insulated needles are more expensive than non-insulated needles.
Standard needles have a tip angle of around 14 degrees and are known as “sharp’ needles. It is frequently recommended to perform regional block with short-bevel needles with an angle of 30 to 45 degrees. This recommendation comes from studies by Selander et al who demonstrated more neural damage in isolated sciatic nerves when sharp needles were used. The damage with sharp needles was also more extensive when the orientation of the sharp bevel was perpendicular to the fibers. With short bevel needles, the damage was less frequent as the fibers were pushed away by the advancing needle.
This concept has been challenged by Rice et al. In fact it may be more difficult to penetrate a nerve fascicle with a short-bevel needle than with a sharp needle, but should it occur, the lesions are more severe.
Persistent paresthesias can occur after regional anesthesia, although severe neurologic injury is extremely rare. Neal estimates the incidence of persistent neuropathy after regional anesthesia to be less than 0.4%.
A large survey by Auroy et al in France in 1997, involving 71,053 neuraxial blocks and 21,278 peripheral nerve blocks, a low incidence (0.03%) of nerve complications after regional anesthesia was confirmed. The survey showed that neurological deficits were extremely low, but relatively more frequent after spinal (70%) than epidural (18%) or peripheral nerve block (12%). In two third of the cases that developed neuropathy after spinal, and 100% of the cases after epidural, a paresthesia was elicited by the needle or during injection. Among the neurological deficits that developed after non-traumatic spinals, 75% of them were in association with the use of 5% hyperbaric lidocaine.
Cheney et al in 1999 reviewed the American Society of Anesthesiologist closed-claims database and found that out of 4,183 claims, 670 (16%) were considered “anesthesia-related nerve injury. Injury to the ulnar nerve represented 28% of the total, and in 85% of the cases it was associated to general anesthesia. Other nerve injuries were brachial plexus in 20%, lumbosacral trunk in 16% and spinal cord 13% and they were more related to regional anesthesia. In 31% of the brachial plexus injuries the patient had experienced a paresthesia with the needle or after injection. They concluded that prevention strategies are difficult because the mechanism for nerve injury, especially of the ulnar nerve, is not apparent.
Lee et al in 2004 conducted a new review of the Closed Claims Data for the 1980 to 1999 period focusing in regional anesthesia. A total of 1,005 regional anesthesia-related claims were reviewed. These claims were 37% obstetric related and 63% non-obstetric. All regional anesthesia, obstetric claims were related to neuraxial anesthesia/analgesia. In 21% of the non-obstetric claims peripheral nerve blocks were involved. The most common block was axillary block (44%). Upper extremity blocks were more involved in claims than lower extremity blocks. Nerve injury temporary or permanent was claimed in 59% of the peripheral nerve injury claims. Death or brain damage was usually the result of cardiac arrest associated with neuraxial block. Pneumothorax accounted for 10% of the claims and “emotional distress” was claimed in 2% of the cases. Eye blocks accounted for 5% of the claims.
Regional anesthesia could result in nerve damage directly from a needle or catheter or be the result of ischemia. Ischemia could be the potential result of vasoconstrictor use or by an intraneural injection that produces a raise of the intraneural pressure leading to nerve ischemia. Local anesthetic toxicity could play a role in cauda equine and transient neurological symptoms. Another mechanism of nerve injury could be hematoma and infection leading to scar formation.
A preexisting neurological injury should always be documented. It is important to realize that nerve damage can occur perioperatively for a reason other than regional anesthesia. Nerves can be injured during surgery by direct trauma, use of retractors and tourniquets and by improper positioning. Nerves can also be damaged postoperatively by a tight cast or splint, wound hematoma or surgical edema.
Epinephrine containing local anesthetic solutions may theoretically produce nerve ischemia by vasoconstriction of the epineural and peri-neural blood vessels. Patients at increased risk would be those with previous impaired microcirculation (e.g., diabetics). There is no evidence at this time to suggest a detrimental effect of epinephrine in regional anesthesia, as used in clinical practice. We use epinephrine 1:400,000 extensively, in all kind of patients, and we appreciate its role in helping to diagnose an inadvertent intravascular injection (please see discussion on epinephrine in local anesthetic chapter).
The symptoms can appear within 24 h after the injury, but sometimes they do not present until days or weeks after the offending procedure took place. The degree of symptoms is usually related to the severity of the injury. The symptoms can be mild, like tingling and numbness that usually disappear within weeks to more rarely severe cases of neuropathic pain and motor involvement that can last months and even years.
A pre existing neurologic condition per se is not a contraindication to regional anesthesia. However a careful preoperative assessment must be made, and any neurological deficit must be documented in the patient’s chart. A thorough discussion with the patient and the surgeon is always important.
Certain progressive neurologic conditions like multiple sclerosis, acute poliomyelitis, amiotrophic lateral sclerosis, Guillian Barre syndrome are relative contraindications to regional anesthesia, because the development of new symptoms postoperatively may be confounded with a nerve block’s complication. In these cases the risks and benefits must be carefully evaluated before proceeding with regional anesthesia.
There are other stable neurologic conditions like a preexisting peripheral neuropathy, inactive lumbosacral radiculopathy and neurologic sequelae of stroke that can be adequately managed with regional anesthesia, provided that all preexisting neurological deficits are well documented in the chart.
In order to minimize the risk of neurologic injury after regional anesthesia several factors are important, including patient selection and type of surgeon. A meticulous nerve block technique, avoiding direct trauma to the nerve and an appropriate selection of local anesthetic concentration, is also important. The role of vasoconstrictors, especially low dose (1:400,000), on clinical development of neural ischemia, has not been elucidated.
When a neuropathy develops in the postoperative period, a prompt evaluation is necessary and a multidisciplinary approach, with participation of neurology, radiology, and surgery, is recommended. A detailed history must be obtained including the timing and nature of symptoms. A physical exam should look for any signs or hematoma or infection. A neurological exam by a neurologist is also crucial.
Although electrophysiological studies remain normal for 14 to 21 days after the injury, ordering them early will help to establish a baseline and rule out any preexisting condition. These tests have limitations, as they only assess large motor and sensory fibers and not small unmyelinated fibers. They usually include nerve conduction velocity studies and electromyography and sometimes may include evoked potentials.
They assess functional integrity of sensory nerves by measuring amplitude and velocity of peripheral nerve conduction. Injuries involving fascicular damage primarily show a decrease in the amplitude of the action potential, a sign that the impulses are being transmitted by a reduced amount of fibers. Conduction velocity in these cases may be minimally affected. When the lesion is demyelinating, like the ones seen after tourniquet compression, nerve conduction velocity is greatly affected while the amplitude remains normal.
It records electrical activity in the muscle helping to locate the denervated muscles in reference to the level at which the nerve damage has occurred. Within 2-3 weeks post injury, spontaneous activity can be recorded from the muscle, in the form of sharp waves and muscle fibrillation. After 3 months the pattern may change, as nerve regeneration by “sprouting” takes place. In permanent injuries, electromyography remains abnormal.
Use of crude compression devices to control surgical bleeding from the extremities, can be traced back, according to Bailey, to ancient Rome. The term “tourniquet” was apparently first used by Petit in France in 1718, to describe a mechanical screw-like contraption that he introduced to provide surgical hemostasis. Lister in 1864, was the first surgeon, who used the tourniquet to produce a bloodless surgical field. Modern tourniquet devices have a microprocessor, use an air pump and are able to accurately and safely maintain the desired pressure. A fail-safe mechanism protects from pressure ever exceeding 500 mmHg.
Tourniquet time: Recommended tourniquet time varies, but the most commonly accepted limit is 2 hours. This recommendation is based on a work by Wilgis, published in 1971 in which he demonstrated more acidosis after 2 hours of use. Surgeons should be made aware of 2-hour tourniquet time and tourniquet should be deflated at that time, unless the surgeon is at a crucial time of the operation and requests more time. This communication with the surgical team needs to be documented in the chart.
Despite the widely accepted 2 hour limit, Klenerman, as cited by Bailey, has shown minimal muscle damage under electron microscopy, with tourniquet times not exceeding 3 hours.
Some people advocate deflating the tourniquet at 1.5h for 5-15 minutes and then re inflate it for an additional 1.5 h.
Tourniquet inflation pressure: It is believed that inflation pressure is even more important a factor than time, in influencing injury. It is recommended to use the minimum inflation pressure that accomplishes ischemia. In general 100 mmHg above the systolic pressure is a common setting. Roekel and Thurston in 1985 showed that 200 mm Hg for the upper extremity and 250 mm Hg for the lower extremity were adequate parameters. Adding layers of padding is important. Wrinkles in the padding should be avoided, since they may become pressure points.
Tourniquet associated problems: The exsanguination with an Esmarch bandage prior to tourniquet inflation causes an increase in preload, which can be significant in cases of bilateral lower extremity tourniquets. The elimination circulation in part of one extremity also leads to an increase in afterload. This may cause problems in patients with cardiad problems and decreased cardiac output. Exsanguination of lower extremities has also been associated with pulmonary embolism and cardiovascular collapse.
Some patients may develop post-tourniquet nerve palsy, affecting more frequently larger motor fibers than sensory fibers. These lesions are usually reversible. The magnitude of the compression and the time of it dictate the severity of the injury.
Patients also can develop “post-tourniquet syndrome”, a clinical picture characterized by interstitial edema, arm weakness and numbness secondary to cell injury and alteration or permeability. It usually resolves within a week.
When the tourniquet is deflated, blood pressure drops (sudden drop in preload and afterload) and heart rate increases as blood rushes into an ischemic, vasodilated bed (reactive hyperemia).
Carbon dioxide and potassium levels increase and so does lactic acid leading to acidosis. These effects peak at about 3 minutes post deflation. There is also a decreased in patient’s temperature.
Tourniquet pain: It is commonly observed despite signs of otherwise good anesthesia of the extremity. Unpremedicated volunteers refer intolerable pain by 30 minutes. Signs of tourniquet pain, manifested as a gradual rise in blood pressure, are also observed under neuraxial blocks and general anesthesia. Patients report this pain under the tourniquet and distal to it.
Controversy exists as to how this pain is transmitted. De Jong and Cullen in 1963 proposed that tourniquet pain was transmitted by small non-myelinated sympathetic fibers. However tourniquet pain can arise even when high thoracic levels of anesthesia are present.
It seems that tourniquet pain is transmitted, as other painful sensations, by A-delta myelinated fibers and C unmyelinated fibers. Tourniquet pain is usually described as burning, cramping or heaviness. The burning and aching sensations, characteristics of ischemia, are believed to be conducted by unmyelinated fibers (MacIver and Tanelian, 1992), while the sharp pain, usually a small component of tourniquet pain, is transmitted by A-delta fibers. MacIver and Tanelian proposed that C fiber activation by ischemia-induced alterations are responsible for tourniquet pain. They studied in an in-vitro model the effects of ischemic alterations (i.e., hypoxia, hypoglycemia, lactic acid, and decreased ph), on A-delta and C pain fibers. They showed that hypoxia and hypoglycemia induced under ischemia, increased C fiber tonic action potential activity, but did not affect A-delta fibers. Increased lactate and decreased pH did not alter the discharge frequency of C fibers in this model. The activation of C fibers by ischemia products seems crucial in tourniquet pain. Whether these C fibers eventually enter the spinal cord at a level above the somatic nerve block is debatable.