Neurologic Complications of PNBs

Admir Hadzic, MD
Director of Regional Anesthesia,
St. Luke’s-Roosevelt Hospital Center, New York, NY
Associate professor of anesthesiology, College of Physicians and Surgeons, Columbia University, New York, NY

Alain Borgeat, M.D.
Professor and Chief of Staff
Department of Anesthesiology,
Orthopedic University Hospital Balgrist, Zurich, Switzerland

Steven Deschner, MD
Neurologist
Research Fellow
Department of Anesthesia, division of Regional Anesthesia
St. Luke's-Roosevelt Hospital Center, New York, NY

Contents
SECTION A

SECTION B - PERIPHERAL NERVE BLOCKS IN ANESTHETIZED PATIENTS

SECTION C - METHODS AND MEANS TO DECREASE THE RISK OF NEUROLOGIC COMPLICATIONS ASSOCIATED WITH NERVE BLOCKS

SECTION D - MANAGEMENT OF PATIENTS WITH NEUROLOGIC INJURY

INTRODUCTION

Although there are relatively few published reports of anesthesia-related nerve injury associated with the use of peripheral nerve blocks (PNBs), it is likely that the commonly cited incidence (0.4%) of severe injury is underestimated owing to underreporting.1-3 Most complications of peripheral nerve blocks were reported with upper extremity blocks. The less frequent clinical application of lower-extremity nerve blocks may be the main reason that there are even fewer reports of anesthesia-related nerve injury associated with lower-extremity PNBs as compared with upper-extremity PNBs.4 While neurologic complications after PNBs can be related to a variety of factors related to the block (e.g., needle trauma, intraneuronal injection, neuronal ischemia, and toxicity of local anesthetics), a search for other common causes should also include positional and surgical factors (eg, positioning, stretching, retractor injury, ischemia, and hematoma formation). In some instances, the neurologic injury may be a result of a combination of these factors.

In this chapter, mechanisms and consequences of acute neurologic injury related to the nerve block procedure are discussed and where appropriate, methods and techniques to reduce the risk of complications are suggested. Specific nerve injuries with upper and lower nerve block techniques, neuraxial anesthesia, and local anesthetic toxicity elsewhere in this volume.

FUNCTIONAL HISTOLOGY OF THE PERIPHERAL NERVES

The functional histology of the peripheral nerve is important to understand the mechanisms of peripheral nerve injury; the reader is refered to Chapters 3 and 4 for more in-depth discussion on this subject. Here we will only briefly review salient features of the organization of the peripheral nerves. A peripheral nerve is a complex structure consisting of fascicles held together by the epineurium—an enveloping, external connective sheath, Figure 1. Each fascicle contains many nerve fibers and capillary blood vessels embedded in a loose connective tissue, the endoneurium.5 The perineurium is a multilayered epithelial sheath that surrounds individual fascicles and consists of several layers of perineural cells. Therefore, in essence, a fascicle is a group of nerve fibers or a bundle of nerves surrounded by perineurium. Of note, fascicles can be organized in 1 of 3 common arrangements: monofascicular (single, large fascicle); oligofascicular (few fascicles of various sizes); and polyfascicular (many few fascicles of various sizes) 6.

Figure 1. Histology of the peripheral nerve. Shown are a large fascicle of the peripheral nerve with its axons, surrounded by perineurium, epineurium and nourishing blood vessels.

Nerve fibers can be myelinated or unmyelinated; sensory and motor nerves contain both in a ratio of 4:1, respectively. Unmyelinated fibers are composed of several axons, wrapped by a single Schwann cell. The axons of myelinated nerve fibers are enveloped individually by a single Schwann cell. A thin layer of collagen fibers, the endoneurium, surrounds the individually myelinated or groups of unmyelinated fibers.

Nerve fibers depend on a specific endoneurial environment for their function. Peripheral nerves are richly supplied by an extensive vascular network in which the endoneurial capillaries have endothelial “tight junctions,” a peripheral analogy to the “blood-brain barrier.” The neurovascular bed is regulated by the sympathetic nervous system, and its blood flow can be as high as 30-40 mL/100 g/minute.7 In addition to conducting nerve impulses, nerve fibers also maintain axonal transport of various functionally important substances, such as proteins, and precursors for receptors and transmitters. This process is highly dependent on oxidative metabolism. Any of these structures and functions can be deranged during a traumatic nerve injury, with the possible result of temporary or permanent impairment or loss of neural function.

The size and the number of the fascicles greatly in a peripheral nerve substantially vary from one peripheral nerve to another. In general, the larger the nerve, the greater the number and the size of the fascicles. Additionally, the larger the fascicle, the greater is the risk of intraneural injection as large fascicles can accommodate the tip of the needle.8 Of note, the fascicular bundles are not continuous throughout the peripheral nerve. They divide and anastomose with one another as frequently, as every few millimeters.9 However, the axons within a small set of adjacent bundles redistribute themselves so that the axons remain in approximately the same quadrant of the nerve for several centimeters. This arrangement is of practical concern to the surgeons trying to repair a severed nerve. If the cut is clean, it may be possible to suture individual fascicular bundles together. In such a scenario, there is a good probability that the distal segment of nerves synapsing with the muscles will be sutured to the central stump of motor axons and the same for sensory axons. In such cases, good functional recovery is possible. If a short segment of the nerve is missing, however, the fascicles in the various quadrants of the stump may no longer correspond with one another, good axial alignment may not be possible and functional recovery is greatly compromised or improbable.9 This arrangement of the peripheral nerve helps explain why intraneural injections result in disastrous consequences as opposed to clean needle nerve cuts which tend to heal much more readily.

CLINICAL PEARLS

  • The larger the nerve, the greater the number and the size of the fascicles. Additionally, the larger the fascicle, the greater is the risk of intraneural injection as large fascicles can accommodate the tip of the needle.

  • The delicate arrangement of the peripheral nerve offers an explanation as to why intraneural injections result in disastrous consequences as opposed to clean needle injuries which heal much more readily.

The connective tissue of a nerve is tough, compared to the nerve fibers themselves. The connective tissue of a nerve permits a certain amount of stretch without damage to the nerve fibers. The nerve fibers are somewhat “wavy,” and when they are stretched, the connective tissue around them is also stretched – giving it some protection.5 This feature, perhaps, plays a “safety” role in nerve blockade by allowing the nerves to be “pushed” rather than pierced by the advancing needle during nerve localization. For this reason, it is prudent to avoid stretching the nerves and nerve plexii during nerve blockade (such as in axillary brachial plexus block and some approaches to sciatic block).

Nerves receive blood from the adjacent blood vessels running along their course. These feeding branches to larger nerves are of macroscopic size and irregularly arranged, forming anastomoses to become longitudinally running vessel(s) that supply the nerve and give off subsidiary branches. Although the connective tissue sheath enveloping nerves serves to protect the nerves from stretching, it also believed that neuronal injury after nerve blockade may be due, at least partly, to the pressure or stretch within poorly stretchable connective sheaths and the consequent interference with the vascular supply to the nerve.

GO TO TOP

MECHANISMS OF PERIPHERAL NERVE INJURY

The etiology of peripheral nerve injury related to the use of PNBs falls in 1 of 4 categories, Table 1. Laceration results when the nerve is cut partially or completely, such as by a scalpel or a large-gauge cutting needle. Stretch injuries to the nerves may result when nerves or plexuses are stretched in a nonphysiologic or exaggerated physiologic position, such as during shoulder manipulation under an interscalene block. Pressure, as a mechanism of nerve injury, is relatively common. Typical example of this mechanism is chronic compression of the nerves by neighboring structures, such as fibrous bands, scar tissue, or abnormal muscles where they pass through fibro-osseous spaces if the space is too small, such as the carpal tunnel. Such chronic compression syndromes are called entrapment neuropathies. Examples of pressure injuries applicable to PNBs include external pressure over a period of hours (e.g., a “saturday night palsy” resulting from pressure of a chair back on the radial nerve of the insensate arm). The pressure may be repeated and have a cumulative effect (e.g., an ulnar neuropathy resulting from habitually leaning on the elbow). Such a scenario is conceivable, for instance, in a patient who positions the anesthetized arm (e.g., long-acting or continuous brachial plexus block) in a nonphysiologic position for a few hours. Another example of pressure-related nerve injury is prolonged use of a high-pressure tourniquet. Finally, an intraneural injection may lead to sustained high intraneural pressure, which exceeds capillary occlusion pressure, and leads to nerve ischemia 10. Vascular nerve damage after nerve blocks can occur when there is acute occlusion of the arteries from which the vasa nervora are derived or from a hemorrhage within a nerve sheath. With injection injuries, the nerve may be directly impaled and the drug injected directly into the nerve, or the drug may be injected into adjacent tissues, causing an acute inflammatory reaction or chronic fibrosis, both indirectly involving the nerve. Chemical nerve injury is the result of tissue toxicity of injected solutions (e.g., local anesthetic toxicity, neurolysis with alcohol or phenol, etc.)

Table 1. Mechanism of peripheral nerve injury related to peripheral nerve blocks

Mechanical-acute
Laceration
Stretch
Intraneural injection
Vascular
Acute ischemia
Hemorrhage
Pressure
Extraneural
Intraneural
Compartment syndrome
Chemical
Injection of neurotoxic solutions

Clinical Classification of Acute Nerve Injuries

Classification of acute nerve injuries is useful when considering the physical and functional state of damaged nerves. In his classification, Seddon 11 introduced the terms neurapraxia, axonotmesis, and neurotmesis (Table 2); Sunderland 12 subsequently proposed a 5-grade classification system.

Neuropraxia refers to nerve dysfunction lasting several hours to 6 months after a blunt injury to the nerve. In neuropraxia, the nerve axons and connective tissue structures remain intact. The nerve dysfunction probably results from several factors, of which focal demyelination is the most important abnormality. Intraneural hemorrhage, changes in the vasa nervora, disruption of the blood-nerve barrier and axon membranes, and electrolyte disturbances all may add to the impairment of nerve function. Because the nerve dysfunction is rarely complete, clinical deficits are partial and recovery usually occurs within a few weeks, although some neurapraxic lesions (with minimal or no axonal degeneration) may take several months to recover.

Axonotmesis consists of physical interruption of the axons but within intact Schwann cell tubes and intact connective tissue structures of the nerve (ie, the endoneurium, perineurium, and epineurium). Sunderland subdivided this group, depending on which of the 3 structures were involved, Table 2. With axonotmesis, the nerve sheath remains intact, enabling regenerating nerve fibers to find their way into the distal segment. Consequently, efficient axonal regeneration can eventually take place.

Neurotmesis refers to a complete interruption of the entire nerve including the axons and all connective tissue structures (epineurium included). Clinically, there is total nerve dysfunction. With both axonotmesis and neurotmesis, axonal disruption leads to wallerian degeneration, from which recovery occurs through the slow process of axonal regeneration. However, with neurotmesis, the 2 nerve ends may be completely separated, and the regenerating axons may not be able to find the distal stump. For these reasons, effective recovery does not occur unless the severed ends are sutured or joined by a nerve graft. With closed injuries the only way to distinguish clearly between axonotmesis and neurotmesis is surgical exploration and intraoperative inspection of the nerve.

Table 2. Classification of nerve injuries*

Seddon

Sunderland

Structural and functional processes

Neurapraxia

1

Myelin damage, conduction slowing, and blocking

Axonotmesis

2

Loss of axonal continuity, endoneurium intact, no conduction

Neurotmesis

3

Loss of axonal and endoneurial continuity, perineurium intact, no conduction

4

Loss of axonal, endoneurial, and perineurial continuity; epineurium intact; no conduction

5

Entire nerve trunk separated; no conduction

Based on data from Seddon11, Sunderland12, and Lundborg13.
It should be noted that most acute nerve injuries are mixed lesions.11 Different fascicles and nerve fibers typically sustain different degrees of injury, which may make it difficult to assess the type of injury and predict outcome even by electrophysiologic means. Recovery from a mixed lesion is characteristically biphasic; it is relatively rapid for fibers with neurapraxic damage, but much slower for axons that have been totally interrupted and have undergone wallerian degeneration.

Mechanical nerve injury

Intraneural Injection

As opposed to a relatively clear injury caused by a sharp needle cut, intraneural injection has the potential to create structural damage to the fascicle(s) that is more extensive and less likely to heal, Figure 2. Indeed, the devastating sequelae of sensory and motor loss after injection of various agents into peripheral nerves has been well documented.14 Nearly all experimental studies on this subject have demonstrated that the site of injection is critical in determining the degree and nature of injury. More specifically, to induce neurologic injury, the injectate must be injected intrafascicularly; extrafascicular injections of the same substance typically do not cause nerve injury.15 Thus, the main factor leading to a substantial peripheral nerve damage associated with injection techniques is injection of local anesthetic into a fascicle. This causes mechanical destruction of the fascicular architecture and sets into motion a cascade of pathophysiologic changes including inflammation, cellular infiltration, axonal degeneration, and others, all possibly leading to nerve scaring.
Figure 2. An example of mechanical injection injury to the peripheral nerve. Shown is a large fascicle, with a needle track, syrinx created by hydrostatic pressure of the injectate, as well as the needle track into the fascicle. Perineurium is seen bulging off the surface of the fascicle.


Histologic features of injury after intraneural injection are rather nonspecific and range from simple mechanical disruption and delamination, to fragmentation of the myelin sheath and marked cellular infiltration, Figure 3. Using a variety of animal models of nerve injury, a vast array of cellular changes following peripheral nerve trauma have been documented.15 The extent of actual neurologic damage after an intrafascicular injection can range from neuropraxia with minimal structural damage to neurotmesis with severe axonal and myelin degeneration, depending upon the needle-nerve relationship, agent injected and dose of the drug used.16,17,18,19 In general, subperineural changes tend to be more prominent, compared with the central area of the fascicle.20 Additionally injury to primary sensory neurons which is not detectable hsitologically, causes a shift in membrane channel expression, sensitivity to algogenic substances, neuropeptide production, and intracellular signal transduction, both at the injury site and in the cell body in the dorsal root ganglion. All of this leads to increased excitability and the occurrence of acute or chronic pain often experienced by patients with neurologic injury. It should be noted that intraneural injection and its resultant mechanical injury are merely the inciting mechanisms; a host of additional changes occur involving inflammatory reactions, chemical neuritis, intraneural hemorrhage, all of which eventually leading to nerve scaring and chronic neuropathic pain.
Figure 3. Fascicular injury after an intraneural injection. Shown is loss axonal degeneration, extravasation of erythrocytes and inflammatory cell infiltration.


Prevention of Intraneural Injection

Pain on injection

Little is known about how to avoid an intraneuronal injection. Pain with injection has long been thought of as the cardinal sign of intraneuronal injection; consequently, it is commonly suggested that blocks be avoided in heavily premedicated or anesthetized patients. However, numerous case reports have suggested that pain may not be reliable as a sole warning sign of impending nerve injury, and it may present in only a minority of cases.21-25 Fanelli and colleagues have reported unintended paresthesia in 14% of patients in their study; however, univariate analysis of potential risk factors for postoperative neurologic dysfunction failed to demonstrate paresthesia as a risk factor.3 In addition, the sensory nature of the pain-paresthesia can be difficult to interpret in clinical practice.26 For instance, a certain degree of discomfort on injection (“pressure paresthesia”) is considered normal and affirmative of impending successful blockade because it is thought that this symptoms indicates that injection of local anesthetic has been made in the vicinity of the targeted nerve.26 In clinical practice however, it can be difficult to discern when pain-paresthesia on injection is “normal” and when it is the ominous sign of an intraneural injection.27 Moreover, it is unclear how pain or paresthesia on injection, even when present, can be used clinically to prevent development of neurologic injury. For instance, in a prospective study on neurologic complications of regional anesthesia by Auroy and colleagues,2 neurologic injuries after paresthesia ensued, although the participating anesthesiologists stopped the injection when pain on injection was reported by the patients.

Intensity of the stimulating current

The optimal current intensity resulting in accurate localization of a nerve has been a topic of controversy.28-32 For instance, stimulation at currents higher than 0.5 mA may result in block failure because the needle tip is distant from the nerve, whereas stimulation at currents lower than 0.2 mA theoretically may pose a risk of intraneuronal injection.33 Other authors suggest that a motor response with a current intensity between 1.0 and 0.5 mA is sufficient for accurate placement of the block needle,28 while some advise using a current of much lower intensity (0.5 to 0.1 mA).29,31 Others simply suggest stimulating with currents less than 0.75 mA, 32,34 or progressively reducing the current to as low a level as possible while still maintaining a motor response.30

CLINICAL PEARLS

  • Most authors suggest that nerve stimulation with current intenisity 0.2-0.5 mA (0.1 msec) indicates intimate needle-nerve placement

  • Stimulation with current intensity 0.2mA my be associated with intraneural needle placement

  • Motor response to nerve stimulation may be absent even when the needle is inserted intraneurally.

Most recently published reports on nerve blocks have suggested obtaining nerve stimulation with currents of 0.2-0.5 mA (100 msec) before injecting local anesthetics, believing that motor response with current intensities lower than 0.2 mA may be associated with intraneural needle placement. However logical these beliefs might sound, there are no published clinical reports substantiating these concerns.

Consequently, in current clinical practice, development of nerve localization and injection monitoring techniques to reliably prevent intraneural injection remains elusive.22 Nerve stimulators are very useful for nerve localization; however, the needle-nerve relationship cannot be adequately precisely and reliably ascertained as early literature suggests.28 Response to nerve stimulation with a commonly used current intensity (1 mA) may be absent even when the needle makes physical contact with or is inserted into a nerve, Figure 4.35-37 Occurrence of nerve injuries despite using nerve stimulation to localize nerves further suggests that nerve stimulators can at best provide only a rough approximation of the needle-nerve relationship.1 One fundamental problem with the nerve stimulation is that the current flows in all directions following the path of the least resistance and not necessarily only towards the nerve. Miniscule changes at the needle tip-tissue interface can make a substantial difference on the preferential flow of current away from the nerve. This may results in cessation of the motor response even when needle is in intimate relationship with the nerve or even intraneurally. The current interest for ultrasound-assisted nerve localization holds promise for facilitating nerve localization and administration of nerve blocks; however, the image resolution of this technology is insufficient to visualize nerve fascicles and prevent intrafascicular injection.

Figure 4. Intensity of the electrical current required to obtain a motor response in a sciatic nerve block model in pigs. As the distance of the needle to the nerve decreases from 0.1 mm to the intraneural location of the needle, stimulation can be obtained with a current of progressively lesser intensity (minimum 0.08 mA (0.1 msec) with needle intraneurally). However, in 5 (25%) of the attempts to stimulate with needle intraneurally motor response could not be obtained with currents of 0.5mA-1.7mA.*

* Kapur E, et al. 2006 Unpublished data.


Resistance to injection

Assessing resistance to injection is a common practice, similar to loss of resistance to injection of air or saline using a “syringe feel” during administration of epidural, paravertebral, or lumbar plexus blocks. Similarly, assessing tissue resistance and injection compliance is another means of estimating the anatomic location of the needle tip during the practice of PNBs. For this, clinicians use a "syringe feel" to estimate what may be an abnormal resistance to nerve block injection and thus, reduce the risk of intraneural injection. 10,31,38 However, this practice has significant inherent limitations.20 For instance, the resistance to injection is greater with smaller needles, introducing additional confusion as to what constitutes “normal” or “abnormal” resistance. Secondly, as opposed to “loss of resistance” in an epidural injection, there is no baseline pressure information or a change in tissue compliance during nerve block injection. In other words, with nerve block injection there is no change in pressure that can be relied upon. For instance, in a study by Claudio and colleagues, all anesthesiologists detected a change in pressure of as little as 0.5 psi during a simulated nerve block injection.20 However, when gauging the absolute pressure, clinicians substantially varied (by as much as 40 psi) in their perception of what constituted an abnormal resistance to injection. Finally, no information has been available on what constitutes “normal” and “abnormal” injection pressure during nerve block performance. For these reasons, subjective estimation of resistance to injection is at least as inaccurate as perhaps estimating blood pressure by palpating radial artery pulse; objective means of assessing resistance to injection should be far superior in standardizing injection force and pressure.

CLINICAL PEARLS

  • Normal injections into epineurium of the nerves should not result in significant resistance to injection

  • When injection proves to be difficult (injection pressures >20 psi using clinically relevant injection speed), the injection is best aborted and the needle flashed to assure patency before trying to re-inject

  • Manual assessment of the resistance to injection using a hand-feel method is highly subjective and depends on the speed of injection, needle size, and ability of the person injecting to consistently discern the "normal" from "abnormal" resistance.

To explain the mechanisms responsible for development of neuraxial anesthesia after an interscalene block, 39,40 Selander and coworkers, injected solutions of local anesthetic into rabbit sciatic nerves and traced the spread of the anesthetic along the nerve sheet.41 They postulated that an intraneural injection results in significant intraneural spread of local anesthetic. In their model, these investigators incidentally noticed that intraneural injections often resulted in higher pressures (up to 9 psi) than those required for perineural injections (< 4 psi). Injection into a nerve fascicle resulted in rupture of the perineurium and histologic evidence of disruption of the fascicular anatomy. This study, however, used a small-animal model, microinjections (10-200 mcl), miniature needles, clinically irrelevant injection rates (100-300 mcl/min), and did not study neurologic consequences after intraneural injections. It is perhaps for these reasons that their foretelling results on the possible association of injection pressure with intrafascicular injection did not change the clinical practice.

More recent studies, however, have used clinically more-applicable injection speeds and volumes of local anesthetic in a canine model of nerve injury.4 The results of these studies suggest that intrafascicular injection is associated with high-injection pressures (> 20 psi) and carry a risk of neurologic injury, Figures 5 and 6.4 Only intraneural injections resulting in pressures greater than 20 psi have been associated with clinically detectable neurologic deficits (Figure 7) as well as histologic evidence of injury to nerve fascicles.

Figure 5. Injection pressures recorded during perineural injection of 2% lidocaine in a sciatic nerve block model in pigs. Using an injection speed of 15 ml/min and 25 gauge insulated nerve block needle injections pressures were at or bellow 20 psi in all but one injection.*

 * Kapur E, et al. 2006 Unpublished data.
 

Figure 6. Injection pressures recorded during intraneural injection of 2% lidocaine in a sciatic nerve block model in pigs. Using an injection speed of 15 ml/min and 25 gauge insulated nerve block needle injections pressures were at significantly above 20 psi in all but two injections.*

* Kapur E, et al. 2006 Unpublished data.
 

Figure 7. Twenty four hours after perineural or intraneural application of 2% lidocaine in a sciatic nerve block model in pigs, the intraneural group continues to exhibit signs of paresis in the sciatic nerve distribution.



The current evidence suggests that neurologic injury does not always develop after an intraneural injection. 42 In fact, injection after an intraneural needle placement is more likely to result in deposition of the local anesthetic between and not into the fascicles.4 Intraneural, but extrafascicular (interfascicular) injection probably occurs more commonly than thought in clinical practice.37 Such an injection results in a block of unusually fast onset and long duration rather than in a neurologic injury. This is because an intraneural but extrafascicular injection leads to intimate exposure of nerve fascicles to high concentration and doses of local anesthetics. However, permanent neurologic injury does not develop since the local anesthetic is deposited outside the fascicles and the blocks slowly resolve after the injection without evidence of histologic derangement.

Needle design and direct needle trauma

Needle tip design and risk of neurologic injury have been matters of considerable debate for more than 3 decades. Nearly 30 years ago, Selander and colleagues,43 suggested that the risk of perforating a nerve fascicle was significantly lower when a short-bevel (e.g., 45°) needle was used as opposed to a long-bevel (12°-15°) needle.43 The results of their work is largely responsible for the currently prevalent trend of using short-bevel needles (i.e., angles 30°- 45°) for the majority of major peripheral nerve conduction blocks. However, the more recent work of Rice and McMahon44 suggested that when placed intraneurally, short-beveled needles cause more mechanical damage than the long-beveled needles.44 In their experiment in a rat model, deliberate penetration of the largest fascicle of the sciatic nerve with 12°- to 27°-beveled needles, with short-beveled needles resulted in the greatest degree of neural trauma. Their work suggests that sharp needles produce clean, more-likely-to-heal cuts, whereas blunt needles produced noncongruent cuts and more extensive damage on the microscopic images. In addition, the cuts produced by the sharper needles were more likely to recover faster and more completely than were the irregular, more traumatic injuries caused by the blunter, short-beveled needles.44 Although the data on needle design and nerve injury have not been clinically substantiated, the theoretical advantage of short-beveled needles in reducing the risk of nerve penetration has influenced both practitioners and needle manufacturers. Consequently, whenever practical, most clinicians today prefer to use short-beveled needles for major conduction blocks of the peripheral nerves and plexuses. Sharp bevel, small-gauge needles however, continue to be used routinely for many nerve block procedures, such as axillary transarterial brachial plexus block, wrist and ankle blocks, cutaneous nerve block, and others.

Regardless of the considerations related to the needle design and risk of nerve injury, the actual clinical significance of isolated, direct needle trauma remains unclear. For instance, it is possible that both paresthesia and nerve stimulation techniques of nerve localization may often result in unrecognized intraneural needle placement, yet the risk of neurologic injury remains relatively low. Similarly, during femoral arterial cannulation (arterial line insertion), it is likely that the needle is often inadvertently inserted into the femoral nerve, yet injuries to the femoral nerve are rare, and when they occur, they are usually attributed to hematoma formation rather than needle injury.45 It is possible that a needle-related trauma without accompanying intraneural injection results in injury of a relatively minor magnitude, which readily heals and may go clinically undetected. In contrast, needle trauma coupled with injection of local anesthetic into the nerve fascicles carry a risk of much more severe injury.20

Toxicity of injected solution

Nerves can be injured by direct contact with a needle, injection of a drug into or around the nerve, pressure from a hematoma, or scarring around the nerve 8,46-48. Experimental studies have shown that the degree of nerve damage following an injection depends on the exact site of the injection and the type and quantity of the drug used 49. The most severe damage is produced by intrafascicular injections, although extrafascicular (subepineurial) injections of some particularly noxious drugs can also produce nerve damage.17,18 Benzylpenicillin, diazepam, and paraldehyde are the most damaging; however, a number of other medications such as, antibiotics, analgesics, sedatives, and antiemetic medications are also capable of damaging peripheral nerves when injected experimentally or accidentally49.

Local anesthetics produce a variety of cytotoxic effects in cell cultures, including inhibition of cell growth, motility, and survival, as well as morphologic changes. The extent of these effects is proportionate to the length of time the cells are exposed to the local anesthetic solution and occur using local anesthetic at normal clinical concentrations. Within normal ranges, the cytotoxic changes are greater as concentrations increase. In the clinical setting, the exact site of local anesthetic deposition plays a critical role in determining the pathogenic potential.50 After applying local anesthetics outside a fascicle, the regulatory function of the perineural and endothelial blood-nerve barrier is only minimally compromised. High concentrations of extrafascicular anesthetics may produce axonal injury independent of edema formation and elevated endoneural fluid pressure.51 As with the effects of local anesthetics in cell cultures, the duration of exposure and concentration of local anesthetic determine the degree and incidence of local-anesthetic–induced residual paralysis. Neurotoxicity of local anesthetics will be dealt with in greater detail elsewhere in this text.

Neurologic complications following regional anesthesia may also be caused by the direct effects of local anesthetics on the nervous tissue. Toxicity has been reported primarily with the intrathecal use of local anesthetics, but with the increasing popularity of peripheral nerve block anesthesia, reports are surfacing about the direct toxic effects of local anesthetics on peripheral nerves.2 Several theories regarding the mechanism of injury have been suggested. Prolonged exposure, high doses, high concentrations, body positioning and the specific agent used may cause transient or permanent neurologic injury by a number of intracellular mechanisms. Once the neurologic injury has occurred, it has been suggested that additives such as epinephrine, or the existence of a pre-existing neurologic condition may predispose the patient to neurotoxic effects of local anesthetics (the “double-crush” concept).

Experimental models of neurotoxicity of local anesthetics have included application of local anesthetic to the sciatic nerve in animals, desheathed nerve preparations, and dorsal root ganglion cells in culture using concentration of local anesthetic comparable to those used clinically.52 These studies have revealed considerable information about the mechanism of injury. Sakura and colleagues discovered that the mechanism did not involve voltage dependent sodium channels. They substituted tetrodotoxin for lidocaine and found that tetrodotoxin blocked these channels as effectively as lidocaine without producing the toxicity associated with lidocaine.53 Johnson and colleagues discovered that cell toxicity may be related to mitochondrial degradation. Local anesthetics caused the mitochondria to depolarize and stop producing ATP. With the loss of ATP, energy-dependent mechanisms are compromised, leading to the accumulation of calcium intracellularly and activation of enzymes that cause cell degredation.54 He found that this was unrelated to hypoxia, because lidocaine actually reduced oxygen demand.52 Cell death or apoptosis was related to the concentration and/or the length of exposure. 1% lidocaine at an exposure for more than 90 minutes was required to kill 50% of the cells. Exposures of less than an hour were completely reversible, but exposure to lidocaine at 5% concentration caused immediate cell death or necrosis.52

In addition to electrolyte imbalance (leading to cell death), the loss of ATP has been found to cause failure of axonal transport compromising the ability of the neuron to transport materials synthesized in the perikaryon to the axon terminal.55 Fast axonal transport moves neurotransmitters from the cell body to the nerve terminal. Lidocaine has been shown to produce a reversible blockade of rapid axonal transport. Recovery is dependent on the concentration and the exposure time of the local anesthetic on the nerve tissue. High concentrations and/or prolonged exposure has been postulated to cause prolonged or permanent nerve injury.56 Furthermore, the loss of ATP leads to the failure of the sequestration of neurotransmitters within the cells, leading to an increase in the extracellular concentration of glutamate. Excessive glutamate in the extracellular space through NMDA receptors can exacerbate the elevation of calcium within the cells ultimately leading to further cell degredation.55,57 This effect is noted only in the central neuraxis where glutamate is found.

Local anesthetics have been shown to cause membrane solubilization at high concentrations. At clinical concentrations, they can form micelles that may act as detergents to disrupt the cell membrane, although this has not been proven in nerve cell membranes.58-61 Oda and colleagues demonstrated that 5% lidocaine and 0.5% dibucaine were minimum concentrations causing irreversible neurologic damage. No neurologic damage was seen with 2% lidocaine or 0.2% dibucaine.62

Neurotoxicity varies with the local anesthetic solution. In histopathologic, electrophysiologic, and neuronal cell models, lidocaine and tetracaine have been shown to have a greater potential for neurotoxicity than bupivacaine.63 Additives, i.e. epinephrine, can increase the toxicity of both lidocaine and bupivacaine.64 A preexisting neurologic condition, i.e. peripheral neuropathy, injury, surgery, may predispose the patient to nerve injury from toxicity at clinical doses (i.e. double crush concept).65

In summary, local anesthetics have potentially cytotoxic effects. The mechanisms appear to involve disruption of mitochondrial function, electrolyte imbalance leading to detrimental intracellular calcium accumulation, loss of axonal transport and release of glutamate. The toxicity and ultimate damage to nerve tissue is related to concentration of the agent, site of action, time of exposure and the specific local anesthetic agent used. Most studies have demonstrated a greater effect on the intrathecal use compared to epidural or peripheral nerve exposure. This may be reflect the typically higher baricity, more concentrated dose of local anesthetic bathing the spinal cord for a prolonged period of time as compared to a large volume, less concentrated solution typically used in epidural and peripheral nerve anesthesia.

Neuronal ischemia

Lack of blood flow to the primary afferent neuron results in metabolic stress. The earliest response of the peripheral sensory neuron to ischemia is depolarization and generation of spontaneous activity, symptomatically perceived as paresthesias. This is followed by blockade of slow-conducting myelinated fibers and eventually all neurons, possibly through accumulation of excess intracellular calcium, which accounts for the loss of sensation with initiation of limb ischemia. Nerve function returns within 6 hours if ischemic times are less than 2 hours. Ischemic periods of up to 6 hours may not produce permanent structural changes in nerves. However, detailed pathological examination after ischemia initially shows minimal changes, but with 3 hours or more of reperfusion, there develops edema and fiber degeneration that last for 1 to 2 weeks, followed by a phase of regeneration lasting 6 weeks. In addition to neuronal damage, oxidative injury associated with ischemia and reperfusion also affects the Schwann cells, initiating apoptosis.

The perineurium is a tough and resistant tissue layer. An injection into this compartment or a fascicle can cause a prolonged increase in endoneurial pressure, exceeding the capillary perfusion pressure. This pressure, in turn, can result in endoneural ischemia.10,41 The addition of vasoconstricting agents theoretically can enhance ischemia because of the resultant vasoconstriction and reduction in blood flow. The addition of epinephrine has been shown in vitro to decrease the blood supply to intact nerves in the rabbit.66 However, in patients undergoing lower-extremity surgery, addition of epinephrine to the local anesthetic solution used in combined femoral and sciatic nerve blocks has not been shown to be a risk factor for developing postblock nerve dysfunction.3

Tourniquet Neuropathy

Tourniquet-induced neuropathy is well documented in the orthopedic literature and ranges from mild neuropraxia to permanent neurologic injury.67-70 The incidence of tourniquet paralysis has been reported as 1 in 8000 operations.71 A prospective study of lower-extremity nerve blockade suggests that higher tourniquet inflation pressure (> 400 mm Hg) was associated with an increased risk of transient nerve injury.3 Current recommendations for appropriate use of the tourniquet include the maintenance of a pressure of no more than 150 mm Hg greater than the systolic blood pressure and deflation of the tourniquet every 90 to 120 minutes.70 Even with these recommendations, posttourniquet-application neuropraxia may occur, particularly in the setting of preexisting neuropathy.72,73

Compressive Hematoma

Little data exists regarding the safety of PNB in patients treated with anticoagulants. Compressive hematoma formation leading to neuropathy has been associated with needle misadventures when performing lower extremity PNB, particulary with concomitant treatment with anticoagulants.74,75 However, as opposed to spinal or epidural hematoma, peripheral neuropathy from this etiology typically resolves completely.72,76-78 Regardless, these reports emphasize the important differences in the risk-benefit ratio of PNBs compared with neuraxial blocks in patients receiving anticoagulant therapy.

GO TO TOP

 

SECTION B - Peripheral Nerve Blocks In Anesthetized Patients

Regional anesthesia-associated nerve injury is a significant source of concern for patients, surgeons, and anesthesiologists alike. In addition nerve injury is a potential medico-legal liability for anesthesiologists.32 Peripheral nerve blocks (PNBs), in particular, are of significant concern because the typical technique involves placing the needle tip in the immediate vicinity of the nerve or plexus. Consequently, any postoperative neurological impairment is automatically, and often unjustly, attributed to the PNB procedure.

Few issues in regional anesthesia have been the subject of as intense controversy as to whether peripheral nerve blocks (PNBs) carry a higher risk of neurological complications when performed in anesthetized patients versus awake patients. Opinions vary from heavy premedication being essential to the success of regional anesthesia63 to its being equated with negligence. Unfortunately, no large-scale controlled studies of the safety of PNBs in awake versus anesthetized patients exists, nor are such studies likely to be available in the future. In the absence of randomized, controlled studies, experts are left to draw conclusions and make logical recommendations solely on their interpretation of the few available case reports, and anecdotal experiences. However, any such recommendations regarding the use of sedation or general anesthesia in patients receiving PNBs could have significant medicolegal repercussions. Therefore, the purpose of this article is to review the available literature that supports or refutes the practice of administering nerve blocks in anesthetized patients, discuss the current controversies, and provide insight into the future of the subspecialty with regard to this issue. Specific concerns regarding performance of PNBs in anesthetized patients are presented and addressed below. Although the scientific value of case reports are often discounted, they may actually be more informative than large epidemiological studies because they include a detailed account of peri-block events, which is often lacking in epidemiologic studies.

GO TO TOP


SYMPTOMS OF INTRANEURAL INJECTIONS

The premise behind the common recommendation that PNB should only be performed in awake patients is that an awake patient can provide information that will prevent intraneural injection and therefore avoid neurological injury. This is because it is believed that intraneural injections are excruciatingly painful, and an awake protesting patient is the best available monitor. However, there are three significant problems with this logic.

(i) The first is that the literature does not support the widespread notion that relying on an a fully awake patient's report of pain on injection is reliable method to prevent nerve injury. In fact, most neurologic complications reported in the literature have not been associated with pain on injection.1-3,20,23-25,27,79-83 For instance, of 49 cases of nerve injury found in the literature search (TABLE 3), 48 patients (98%) were awake. Of these, 42 cases included specific information about the patient’s response to the injection; only 4 (10%) patients reported pain on injection. Some reports specifically state that the patient did not have pain, whereas in most others the authors commented that the block performance was uneventful. Interestingly, the pain on administration of local anesthetic (LA) into nerve tissue may be absent even in the central neuraxial area. For instance Kao et al. reported a case of neural injury that was clearly related to spinal cord trauma from a thoracic catheter that had been inserted while the patient was anesthetized; the patient did not have pain during administration of LA postoperatively.84 More recently, Tripathi et al reported a case report of paraplegia after an intracordal injection during attempted steroid injection85, whereas Tsui and Armstrong reported a case of direct spinal cord injury after epidural injection86. Both complications occurred in awake patients who did not report pain on needle placement and consequent injection. These report clearly indicate that reliance on pain as a symptom of injection into neurologic tissue is unreliable.

(ii) The second problem refers to the value of the pain (if present) as monitor in preventing nerve injury, because in cases in which the pain does accompany an intraneural injection it may already be too late to prevent neurologic injury. For example, in the ASA closed-claims study, Cheney and coworkers indicate that on those occasions where pain did occur during injection, the anesthesiologist stopped the injection, however, the patients still went on to develop nerve injury.87 Similarly, studies utilizing animal models of intraneural injection suggest that nerve fascicles become injured at the very onset of the injection and with injection of very small volume of local anesthetic (as little as 0.5-1 ml).79,88 Thereafter, as injection continues, the fascicle ruptures and the injectate simply leaks out through the ruptured perineurium into the epineurial sheath. At this stage however, the damage to the fascicle(s) may already have been done.

(iii) Third, pain is notoriously difficult to assess in terms of quality and intensity. Therefore, distinguishing between the discomfort that is commonly seen during local anesthetic injection (which is considered normal) and that of intraneural injection can be difficult. As an example, in the study by Borgeat et al, 21% of patients undergoing interscalene block reported transient, burning pain, yet none of these patients went on to develop neurologic complication.

Table 3. Reports of peripheral nerve injury with major conduction blocks

GO TO TOP


NORMAL VERSUS ABNORMAL DISCOMFORT/PAIN ON INJECTION

Injection of LA in the close proximity to the nerves is often associated with discomfort on injection.81,93 This is thought to result from “spraying” of the LA in the vicinity of the components of the brachial plexus. Based on his studies of brachial plexus anesthesia, Winnie coined the term “pressure paresthesia” to describe the discomfort patients feel during local anesthetic injection and implied that this was a desirable sign of impending successful blockade.81 In actual clinical practice however, the variability of patients’ pain thresholds, their ability to verbalize a sensation pain during a procedure, and an anesthesiologist’s subjective interpretation of any such response make it very difficult to recommend where a “line” could be drawn between normal and abnormal pain or paresthesia on injection. In fact, a number of published case reports demonstrate that patients’ complaints of pain during PNB may not be helpful in preventing the development of the neurological complication. For instance, Barutell et al published a report where a patient communicated discomfort on injection which was perceived as “normal pressure paresthesia” by the anesthesia team and when the injection was carried out; the patient went on to develop permanent neurologic damage.26 Similarly, in the report by Kaufman et al,26 all 7 patients had discomfort at some point during block injection; however this information could not be used to prevent the development of permanent neurologic injury which occurred in all patients. This however may be an example of case-report bias. In other words, it is possible that patients’ reports of pain may have prevented injuries but such events are unlikely to get reported.

GO TO TOP

 

DOGMAS ON COMPLICATIONS OF REGIONAL ANESTHESIA

Blanket statements and dogmas are common in the field of medicine and regional anesthesia in particular. Much too often, a wide range of recommendations based on a single observation are inappropriately extrapolated. As an example, Walton et al. reported the occurrence of brachial plexus palsy after total shoulder arthroplasty under an otherwise uneventful interscalene block.27 However, the authors went on to suggest that, “to minimize the risk of brachial plexus injury with interscalene block” PNBs should not be performed in anesthetized patients, and if paresthesia of “unusual severity” occurs, the injection should be immediately stopped. Ironically, their patient was neither anesthetized prior to the block injection, nor did he have paresthesia or pain on injection!

Benumof reported 4 cases of severe neurological injury that resulted in cervical paraplegia in patients receiving interscalene brachial plexus blocks under GA.91 The discussions that followed however, often recommended that sedation and GA be abandoned to decrease the risk of nerve injury. Such recommendations are based on this case report are inappropriate because none of the patients in this report suffered a peripheral nerve injury. Rather, these patients received an intracordal injection, a complication entirely avoidable with restriction of the needle insertion depth and/or use of more lateral approach to interscalene block.4,92,94

Those who base their criticism of "heavy" premedication or GA prior to performing peripheral nerve blocks on the cases reported by Benumof95 forget that interscalene block is a superficial procedure, devoid of significant discomfort where excessive sedation and analgesia are usually unnecessary except in children who otherwise would not hold still during the procedure. In contrast, many other PNB procedures involve deeper placement of the needle and several attempts at nerve localization that result in significant patient discomfort.94,96,97 Therefore, generalized recommendations to avoid premedication during PNBs carry the risk of limiting the use of PNBs because of an inevitable decrease in patient acceptance. In our practice at St. Luke's-Roosevelt Hospital Center in New York, patients are sent an anesthesia satisfaction survey after their discharge home. Analyses of more than 5,000 of patient responses clearly indicate that the most satisfied patients are those with no recollection of any anesthetic procedure being performed on them. Therefore, if PNBs are to be accepted by patients, adequate premedication is necessary to avoid patient discomfort during needle placement.

GO TO TOP


LITERATURE REVIEW

No study has compared the risk of neurologic complications in awake versus anesthetized patients, and it is unlikely that such studies will ever be done. A review of published reports of injury after PNBs indicates that significant neurologic injury after PNBs in awake patients occurs at a rate of 0.2%-0.4%.97 Most of these reports included brachial plexus blocks only, probably because these techniques are used more frequently than lower extremity nerve blocks.20 In a recent similar prospective evaluation by Bogdanov and Loveland, none of 548 patients who received an interscalene brachial plexus block after induction of GA developed permanent or long-term neurologic complications.81 Similarly, in a report presented at the 2005 Annual ASRA Spring Meeting, Tsai et al presented the data on 226 PNBs of both upper and lower extremities, all performed in heavily sedated or anesthetized patients, none of whom developed neurologic complications. Bogdanov et al. used a modified classical approach to interscalene block proposed to reduce complications whereas Tsai et al. used objective assessment of injection pressures to reduce the risk of intraneural injection during PNBs of both upper and lower extremity, Figure 8.98 While the relatively small number of patients in these reports do not allow accurate comparison, these studies at least indicate that the risk of complications of PNBs after GA may not be substantially more common than complications reported in other similarly powered studies in awake patients.99 In fact, performance of PNBs in heavily premedicated patients or after induction of GA is undoubtedly a common practice, and a routine in the pediatric anesthesia practice. A recent informal poll conducted during the ASRA 2005 session on complications of PNBs indicated that approximately half of the present attendees performed blocks in heavily sedated or anesthetized patients.

Figure 8. Injection of local anesthetic in lateral approach to popliteal block with in-line monitoring of the injection pressure to avoid pressures >20 psi which may be associated with intraneural injection.

GO TO TOP


RESCUE BLOCKS, MULTIPLE INJECTION TECHNIQUES, AND “REVERSE” AXIS BLOCKS

Several PNB techniques that are equivalent of a PNBs in anesthetized patients are accepted as sound practice. These include “rescue” blocks, multiple injection techniques and “reverse” axis blocks; they are all similar to the practice of PNBs in anesthetized patients because the PNB is performed in a partially or fully anesthetized limb.

“Rescue blocks”

Missed nerve blocks may occur from 3% to30% of the time and usually involve only one or two of the terminal nerves.20 Several well established and universally accepted techniques of PNBs, such as repetition of the block, peripheral injection of the nerve and others are often used to rescue failed blocks despite the risk that the needle may be inserted or injection be made into an anesthetized nerve.81 Common examples of rescue blocks after failed axillary or interscalene brachial plexus blocks include elbow and wrist blocks.100,101

“Multiple injection techniques”

Multiple injection techniques for both upper and lower limb blockade have been introduced relatively recently in clinical practice with the suggestion that they decrease onset time, increase the success rate, and decreases the required dose of LA.102-109 The withdrawal and redirection of the needle to elicit multiple motor responses however, carry a greater risk of direct needle trauma and intraneural injection into already anesthetized nerves. Regardless, this technique has been uniformly accepted by most experts in the field.

“Double blocks” and “repeat” blocks after failed blocks

Regional anesthesia for elbow surgery has traditionally been a challenge. The most commonly used brachial plexus blocks--the classical approach to interscalene block and the axillary block--are not ideal because they either do not result in reliable block of the ulnar nerve or do not result in adequate analgesia for the tourniquet, respectively. For that reason, in patients undergoing elbow surgery successful regional anesthesia requires the concomitant use of two separate approaches, an interscalene and an axillary approach.110 Obviously, performance of brachial plexus block at either level precludes the sensory or motor feedback information during performance of the block at the second level.

GO TO TOP

AWAKE PATIENTS WILL HAVE SIGNS OF CNS TOXICITY AS A MONITOR BEFORE CVS TOXICITY ENSUES

Practice of PNBs involves administering large volumes and doses of LAs. A typical clinical presentation of LA toxicity is in an awake or sedated patient during or immediately after injection of LA, followed by sudden onset of confusion, seizure, arrhythmias or cardiac arrest.111,112 It has been suggested that heavy sedation or GA increases the risk of severe systemic toxicity of LAs because it diminishes the patient’s ability to report early signs and symptoms of rising LA serum levels. However, there are no reports of LA toxicity in adult patients under GA; essentially all reports of toxicity were in awake patients.23,91,113,114 For example, Edde and Deutsch reported an occurrence of cardiac arrest after interscalene brachial plexus block in an awake patient who had no symptoms of toxicity until the entire dose (20 mL of 0.5% bupivacaine) was administered.115,116 Conversely, others may argue that premedication offers protection because of its anticonvulsive effects. Moreover, since the critical steps in treating patients with severe toxicity is establishment of patent airway, hyperventilation, administration of oxygen, and hemodynamic support, one can argue that anesthetized and ventilated patients who develop systemic toxicity may actually be better off because they already have a secured airway, they are receiving a high concentration of oxygen and they are typically in, an environment that is ideally suited for aggressive resuscitation.

Bernards et al. reported that pigs given an intravenous infusion of bupivacaine failed to demonstrate signs of CNS toxicity prior to cardiovascular collapse if they had been pre-medicated with benzodiazepines.23 Indeed, in clinical studies of local anesthetic toxicity, volunteers clearly report an escalating series of symptoms as plasma concentration of local anesthetic rises during continuous, slow intravenous infusion. However, the occurrence of symptoms of local anesthetic toxicity depends on the rate at which drug is injected. The difficulty in extrapolating this into clinical practice is that local anesthetic for the purpose of neuronal blockade is given as a relatively rapid bolus, rather than a slow, escalating continuous infusion as is the case in animal models.

For these reason, any suggestion that GA predisposes to a greater risk of severe systemic toxicity of LA is purely theoretical, as there are no data to firmly support this belief. Finally, LA toxicity is of potentially greater concern in the pediatric patient due to lower dose requirements and need for stringent adherence to mg/kg dosage rather than a volume dosage as is the common practice in adult anesthesia. Regardless, the practice of regional anesthesia in anesthetized patients is universally accepted in this patient population.

GO TO TOP


NERVE BLOCKS IN ANESTHETIZED CHILDREN VERSUS ADULTS

As opposed to PNBs in adults, performing blocks in anesthetized pediatric patients is a universally accepted practice. This is by necessity, because pediatric patients are unlikely to be cooperative during needle insertion, nerve stimulation and manipulation necessary to accomplish PNB. In addition, most pediatric patients require concomitant administration of GA to allow for immobility during the surgical procedure. However, from the standpoint of risk of complications, this divergence in consensus on PNBs after GA between adults and pediatric patients does not make much sense. In other words, with perhaps the exception of infants, there are no sufficient anatomical, or neuro-physiological differences to justify this divergence in recommendations. While one can argue that complications of PNBs in children are rare, PNBs are not routinely used in the pediatric population and there are simply no series comparable to those in adults to allow one to draw a clear conclusion regarding the risk of nerve injury in children. However, a large prospective study performed in France in children demonstrated a small risk of complications with peripheral nerve blocks.117

MONITORING POSSIBILITIES DURING PNBs

Since this discussion focuses on the impact of heavy sedation or GA on the risk of neurologic complications, it is important to discuss the monitoring that is available to reduce the risk of nerve injury during PNBs. In general, there are two phases amenable to monitoring during placement of PNBs. These include needle placement guidance (percutaneous stimulation, ultrasound) and avoidance of intraneural injection (report of pain on injection by the patient (when present), nerve stimulation, and assessment of resistance to injection).118 With regards to intraneural injection, neither percutaneous stimulation or ultrasound guidance are helpful. Percutaneous stimulation may be helpful with approximating the needle insertion site but it is useless in estimating the needle-nerve relationship. Ultrasound on the other hand, offers real-time needle guidance below the skin level. However, in addition to the required skill, expense and inconvenience of the equipment, the image resolution is simply insufficient to rule out intraneural needle placement.119 Nerve stimulator-assisted PNBs entered the practice of regional anesthesia with the promise of decreasing the risk of neurologic complications associated with paresthesia technique.120 However, it soon became apparent that nerve stimulators could not prevent neurologic injury.37 More recently, it has been suggested that in many circumstances, the motor response to nerve stimulation may be absent at the point at which the needle makes contact with the nerve,121 and that intraneural needle placement is possible despite the use of nerve stimulators, Figure 4.1

In summary, few issues in the practice of regional anesthesia have evoked such strong and divergent opinion among clinicians as performance of PNBs in anesthetized or deeply sedated patients. This is because administration of PNBs has traditionally been based on individual preferences, clinical impressions, and other subjective criteria, rather then on established practice standards. Avoidance of deep sedation and GA is often suggested to decrease the risk of peripheral nerve injury with PNBs, however, there is no evidence in the literature to suggest that either practice is safer with regards to the risk of nerve injury. Regardless, the serious nature of complications resulting from inadvertent injections into the spinal cord suggests that regional anesthesia techniques close to the centroneuraxis should be practiced with extreme caution and insight into the appropriate depth of the needle insertion whether the patient is awake, sedated or anesthetized. It is the opinion of these authors that proper training, and use of proper techniques and nerve block/monitoring equipment are more likely to decrease the risk of complication than are unfounded blanket statements as to the advisability of performing peripheral nerve blocks in anesthetized patients. This is because such statements are unfounded by the relevant literature and can have a potentially negative impact on the practice of regional anesthesia. Many patients and PNB techniques require appropriate sedation for block performance and patient acceptance, however, clinicians may be reluctant to use them due to the medico-legal concerns engendered by admonitions against performing blocks in sedated or anesthetized patients. Future efforts must be directed toward developing more objective and exacting nerve localization and injection monitoring techniques to more reliably detect and prevent intraneural injection. The results of these efforts will inevitably be of far greater importance to the future of PNBs and their role in practice of modern anesthesiology than over-reaching, unsubstantiated opinions and statements.

GO TO TOP

SECTION C - METHODS AND MEANS TO DECREASE THE RISK PF NEUROLOGIC COMPLICATIONS ASSOCIATED WITH NERVE BLOCKS

The published data suggests that neurologic complications of PNBs are relatively rare. However, their the severity of consequences and lack of prevention strategies continue to present a source of significant concern for both clinicians and patients. The main inciting mechanism of neurologic injury with PNBs appears to be an intrafascicular or intraneural injection. It is fortunate that peripheral nerves possess an inherent natural protection; intraneural injections often do not result in intrafascicular needle placement and therefore, do not necessarily lead to nerve injury. It is commonly suggested that the use of short-beveled needles and avoidance of excessive sedation and general anesthesia should be employed to decrease the risk of nerve injury. However, these commonly voiced recommendations have been recently challenged. In addition, avoidance of adequate premedication may have a significant negative impact by decreasing the patient’s acceptance and satisfaction with PNBs. The relatively low incidence rate of complications with PNBs, coupled with the lack of objective documentation and means to more precisely monitor administration of nerve blocks make retrospective analyses of cases of nerve injury largely speculative with regard to the actual mechanism of nerve injury in clinical practice.

The following recommendations are suggested to decrease the risk of complications with PNBs:

  • Aseptic technique
    Most nerve block techniques are merely percutaneous injections. However, infections are known to occur and can result in significant disability. Since this complication is almost entirely preventable, every effort should be made to adhere to strictly aseptic technique.

 

  • Short bevel insulated needles
    Insulated needles are now widely available and result in much more precise needle placement. The short bevel design helps prevent nerve penetration.

 

  • Needles of appropriate length for each block technique
    Excessively long needles should not be used for nerve blockade. For instance, never use needles longer than 50 mm for an interscalene block. In addition to the safety reasons, needles of appropriate length are also advanced with far greater precision than excessively long needles.

 

  • Surface localization
    In patients with difficult anatomy, surface localization of superficially seated nerves or plexuses can help reduce the number of needle passes. 

 

  • Needle advancement
    During needle localization, advance and withdraw the needle slowly. Keep in mind that nerve stimulators deliver current of very short duration once (1 Hz) or twice (2 Hz) a second and that no current is delivered between the pulses. Thus, fast insertions and withdrawal of the needle passages may result in failure to stimulate the nerve because the needle may pass near by, or even through, the nerve between the stimuli without eliciting nerve stimulation.

 

  • Fractionated injections
    Inject smaller doses and volumes of local anesthetics (3-5 mL) with intermittent aspiration to avoid inadvertent intravascular injection. Always observe the patient during the injection of local anesthetic because negative aspiration of blood is not always present with an intravenous injection. This approach may allow detection of the signs of local anesthetic toxicity before the entire dose is injected.

 

  • Accuracy of the nerve stimulator
    Always make sure that the nerve stimulator is operational, delivering the specified current, and that the leads are properly connected to the patient and the needle.

 

  • Avoidance of forceful, fast injections
    Forceful, fast injections are more likely to result in channeling of local anesthetic to the unwanted tissue layers, lymphatic vessels, or small veins that may have been cut during needle advancement. Such injections may result in massive channeling of the local anesthetic in the systemic circulation, with consequent risk of severe CNS and cardiac toxicity. Finally, forceful, fast injections under excessive pressure are more likely to result in an unrecognized intraneuronal injection. Limit the injection speed to 15-20 mL/minute.

 

  • Avoidance of injection against abnormal resistance
    Intraneuronal needle placement may result in high resistance (pressure) to injection due to the compact nature of the neuronal tissue and its connective tissue sheaths. Always use the same syringe and needle size to develop a “feel” during the injection. As a rule, when injection of the first 1 mL of local anesthetic proves difficult, the needle should be slightly withdrawn and the injection attempted again. If the resistance persists, the needle should be completely withdrawn and flushed before repeating the insertion. Ultimately, objective injection pressure monitoring will likely become a standard monitor to standardize nerve block injections, reduce the risk of intraneural injection and for medico-legal documentation, Figure 8.

 

  • Abort injection if pain is reported by the patient
    Severe pain or discomfort on injection may signify intraneuronal placement of the needle and should be avoided. When this occurs, the injection should be abandoned, although the chances are that a damage may have already been caused at the time when the pain occurs. Lancinating, "shooting" pain on injection should not be confused with a mild “paresthesia-like” report by the patient when the needle is placed in the immediate vicinity to the nerve or plexus. In this case, local anesthetic can be injected slowly, provided that the resistance to injection is normal (<20 psi). Resuming the injection after waiting to see whether the pain will go away should not be done under any circumstances.

 

  • Chose your local anesthetic solution wisely
    Always choose a shorter acting (and less toxic) local anesthetic for short procedures where long-lasting postoperative analgesia is not required. Local anesthetic toxicity is the most common complication with neuronal blockade; The risk of sever toxicity is substantially lower with chloroprocaine or lidocaine than with bupivacaine.

 

  • Blocks in anesthetized patients
    In the absence of reliable monitoring, blocks in anesthetized patients should still not be a common practice. When it is necessary to place blocks in anesthetized patients, this should be done by practitioners with experience and pro-cons documented in the chart. The introduction of ultrasound-guided nerve blocks and injection pressure monitoring will likely change the practice and allow more routine performance of nerve blocks in anesthetized patients.

 

  • Repeating blocks after a failed block
    Repeating a block after a failed block should be avoided whenever possible. When indicated, it should be done only by those with substantial experience in the planned technique. Avoidance of abnormal resistance to injection or objective injection pressure monitoring is of utmost importance here as the clinician can no longer count on pain on injection as a sign of intraneural injection.

GO TO TOP

PART D - MANAGEMENT OF PATIENTS WITH NEUROLOGIC INJURY

PERIPHERAL NERVE INJURY AFTER PERIPHERAL NERVE BLOCKS: DIAGNOSIS, PROGNOSIS, AND TREATMENT

Nerve injury is recognized as a potential complication of peripheral nerve block anesthesia, but fortunately, severe or disabling injuries are rare. When they do occur, they can be a frightening complication for the patient, the surgeon, and the anesthesiologist. Fortunately, most symptoms of nerve injury usually resolve in four to six weeks in over 95% of patients, and in 99% of the patients by one year.35 Early intervention may help prevent the long-term sequelae that can occur with unrecognized, and therefore improperly treated nerve injuries. With the resurgence of interest and utilization of peripheral nerve block anesthesia, it is important to develop a coinciding plan for management of postoperative nerve injuries. The plan should include recognition, diagnosis, and treatment of the nerve injury. A working knowledge of the long-term prognosis, understanding the importance of appropriate neurology consultations and familiarity with available diagnostic tests and treatment options is essential for the proper management of patients with neurologic injury.

With any peripheral nerve block, inclusive documentation of the block is of utmost importance for diagnostic, therapeutic and medicolegal purposes. Documentation that includes the nerve(s) stimulated, the minimum current used, the number of attempts, the appearance of pain or paresthesia during the procedure and measures taken, resistance to injection and injection pressure if pressure monitoring is used, type/dose of local anesthetic agent, and patient condition during the block is essential in understanding the mechanism of the injury. This documentation can help the anesthesiologist and/or consulting specialist determine the possibly etiology of the injury, any associated conditions, and guide treatment modalities.

GO TO TOP


Mechanisms of Injury/Symptoms:

Following peripheral nerve anesthesia, if sensory and/or motor function remains depressed beyond the expected duration of action of the local anesthetic, potential causes for the neurologic deficits should be investigated. Neurologic deficits may be related to vascular injuries, compression injuries, local anesthetic action, or traumatic nerve injury with fascicular disruption. When faced with a neurologic deficit, especially in a patient in whom a peripheral nerve block has been placed, it is important to remember that there are many causes of nerve injury unrelated to the performance of the regional anesthetic.
Other factors, i.e. tourniquet use, improper positioning, postoperative swelling, surgical trauma, and pre-existing neurologic deficits may be contributing or causative influences. Proper and objective documentation of the nerve block procedures can go a long way in deciphering whether the injury was caused by a nerve block, surgery or other factors.

a. Vascular Injuries
When symptoms occur early in the postoperative period (i.e. within minutes to hours), the anesthesiologist should suspect a vascular or compression-type injury due to disruption of the blood supply from an insult to the vascular structures, from hematoma at the surgical or block site, or from deep venous thrombosis. This is particularly true should a neurologic deficit appear after apparent resolution of the block.

b. Compression Injuries

Compression neuropathies usually present as a focal mononeuropathy at sites where nerves pass through tissue tunnels, i.e. median through carpal tunnel, spinal nerve through vertebral foramina. Acute compression injuries can develop secondary to limb tourniquet paralysis, retractors used for surgical procedures, from expanding hematomas, or from improper intraoperative positioning, i.e. stretching of the cords of the brachial plexus during sternal retraction, or extending/pronating the forearm causing ulnar nerve compression.37,122 A compression injury can also occur with peripheral nerve block anesthesia from increased endoneural fluid pressure. This can occur with injection into a tight tissue compartment or when high injection pressures are used during nerve block administration.77 Metabolic diseases such as diabetes mellitus may make nerves more susceptible to compression injuries.123

c. Local Anesthetic Action

It is important to remember that neurologic defici