Chapter 50. Managing Adverse Outcomes during Regional Anesthesia

1. Safe regional anesthesia begins with a thorough knowledge of anatomy. In Labat's words, "anatomy is the foundation upon which the entire concept of regional anesthesia is built." Studying the anatomy of the major plexuses and peripheral nerves is critical for learning regional anesthesia and avoiding its complications.

2. Prior to performing regional anesthesia, it is imperative to thoroughly discuss the techniques, and their limitations, with the patient. Assessing which patients are most appropriate for performing these techniques on is important, as some are not suitable candidates (eg, those with major anatomic distortion or serious mental illness).

3. One of the most important principles for safe regional anesthesia is provision of a comfortable patient environment. If a patient suffers as a result of one's intervention, a basic principle of the practice of anesthesia has been violated.

4. Resuscitation equipment must be immediately available when performing regional anesthesia, and one must be prepared, at all times, to anesthetize and resuscitate the patient when necessary.

5. A skilled clinician must be willing to abort their technique in the face of excessive challenge. Dogged persistence is inadvisable. One must seek assistance when faced with difficulties and be prepared to change to an alternative route of anesthesia if persistent failure (more than 3 attempts or 20 minutes) occurs.

6. Do not perform regional anesthesia procedures in anesthetized adult patients unless the benefits outweigh the risks. If this principle is violated, the reasoning must be documented in the patient's file.

7. Always be accompanied by a skilled assistant when performing regional anesthesia.

8. The patient must be adequately monitored at all stages during regional anesthesia. Close monitoring can only be discontinued when the block has adequately resolved.

9. If neurologic injury is suspected after regional anesthesia, the cause should be determined quickly to minimize injury. Timely advice should be sought from appropriate consultants (neurologists, radiologists).

10. One must not assume that all patient injury is from regional anesthesia, as other possibilities exist. Do not hesitate to involve other disciplines in the quest to determine the cause of injury.

No matter how skillful an anesthesiologist may be, adverse perioperative events are inevitable in anesthesia practice. Adverse events have been associated with regional anesthesia since local anesthetics were first introduced by Koller in 18841 and will persist despite advances in technique, skill, and safety mechanisms. It is nearly impossible to address all of the described and potential complications of regional anesthesia; instead, we focus here on those most relevant to current practice.

Importance of Prevention

The time-honored statement that "an ounce of prevention is worth a pound of cure" is essential to remember2 when considering the management of adverse outcomes in regional anesthesia practice. It is most effective to prevent and minimize the risk of regional anesthesia complications. Neurologic injury is one of the most dreaded complications associated with all anesthesia techniques, including regional anesthesia, and it is important to realize that once a serious neurologic injury occurs, the chances of full recovery are unlikely.

Safe regional anesthesia begins with the first encounter with the patient. The first task at hand is to perform a thorough preoperative assessment of the patient. Good outcomes are obtained when a skilled anesthesiologist uses appropriate equipment and technique and sound judgment. Skilled intraoperative sedation and monitoring are vital to the practice of regional anesthesia. To this end, we advocate for specialized regional anesthesia areas in the operating room space where the appropriate equipment, monitoring, and assistance are readily available. Resuscitation drugs and equipment must always be immediately available in the event of a problem, and one must always be ready to convert to general anesthesia if required. Patients must be carefully observed during the postoperative period by appropriately trained postanesthetic care team when most of the serious complications become evident. Early identification and intervention are of the utmost importance in preventing neurologic injury.

Poor outcomes and serious complications are not prima facie evidence of negligence. However, the risk of litigation in contemporary anesthesia practice is closely associated with the severity of patient injury rather than the occurrence of negligence; this is particularly true of regional anesthesia injuries. The American Society of Anesthesiologists (ASA) Closed Claims study in the United States reported a high incidence of successful suits against anesthesiologists involved in regional anesthesia cases, even though the standard of care was met.3 The importance of effective communication and truthful disclosure with the patient and the patient's family cannot be overemphasized. Litigation is more frequent with the combination of an adverse event and a poor physician–patient relationship. A patient who feels that the physician has the patient's best interests at heart is less likely to pursue litigation than is a patient who does not respect or trust the physician; thus the anesthesiologist should pay particular attention to developing a good rapport with the patient in the limited time allotted. Table 50-1 summarizes basic recommendations for maintaining a standard of care in regional anesthesia practice. Maintaining a standard of care at all times does not guarantee against legal action, but it certainly will minimize the risk and is to be encouraged at all times.4

Patient Selection

Proper patient selection is a critical consideration for the safe and successful performance of regional anesthesia, as not all patients are suitable candidates for regional anesthesia (Fig. 50-1).5 Some patients are not psychologically suitable for regional anesthesia. A number of patients suffer from needle phobias and faint at the least provocation. In general, patients with severe mental illness are not suitable candidates for regional anesthesia unless it is combined with general anesthesia. Gross anatomic distortion may hinder the performance of regional anesthesia in some patients. Neuraxial techniques may be associated with hemodynamic disturbances such as bradycardia and hypotension; therefore, they are contraindicated in hemodynamically unstable patients and in those with fixed cardiac outputs. Regional anesthesia should be used cautiously in patients with preexisting neurologic disease, and if regional techniques are used in these patients, neurologic deficits must be clearly documented before the performance of regional anesthesia. Some patients are rigidly opposed to regional anesthesia, and it is important not to sell them on it. If it is clearly evident that a patient will benefit from regional anesthesia, it is reasonable to explain in detail the rationale behind the procedure; however, the decision to undergo regional anesthesia must be finally left to the discretion of the patient. It is very important to discuss the options of anesthesia with patients even if they are undergoing minor operative procedures. It is imperative to follow national and international guidelines pertaining to regional anesthesia practice (eg, ASA monitoring, American Society of Regional Anesthesia and Pain Medicine [ASRA] guidelines for anticoagulated patients, etc.), and if there is any deviation, it is important to specify the reasons and to document them. Table 50-2 summarizes the important factors involved in selecting suitable patients for regional anesthesia.

Consent

Informed consent and an explanation of the risks and alternatives to regional anesthesia must be provided to the patient (the anesthesiologist should never coerce a patient to accept or reject any anesthetic plan). Potentially serious complications associated with regional anesthesia should be disclosed to patients, including convulsions and the risk of cardiac toxicity from systemic injections of local anesthetics, spinal cord/nerve injury leading to paralysis or neurologic deficit, pneumothorax, hematoma, infection, cardiac arrest, and death. Interestingly, a recent study revealed that fewer than half of anesthetists disclose the risks of seizures, respiratory failure, and cardiac arrest before the administration of either neuraxial blocks or peripheral nerve blocks.6 Any common side effects specific to certain procedures such as the failure to achieve surgical anesthesia, patient awareness during conscious sedation, nausea, pruritus, headache, shivering, backache, dizziness, and urinary retention should also be discussed. However, anesthesiologists should bear in mind that even with informed consent, proper disclosure, effective communication, and an appropriate ethical approach, there is no guarantee that they will be legally protected.7

Use of Appropriate Equipment and Technique

As advances in regional anesthesia continue, techniques used must be continuously revised in light of any new, clinically relevant information and developments. For years, we have been percutaneously inserting needles toward neural targets and have relied solely on our knowledge of anatomy and techniques of paresthesia and the loss of resistance (LOR). The introduction of nerve stimulation was an important advance in regional anesthesia because it provided some objective evidence that the needle tip was close to the neural target. One of the most exciting recent advances has been the introduction of ultrasonography as a method to accurately place needles in close proximity to neural targets. Ultrasonography allows real-time visualization of anatomical structures and offers the potential to guide needle and catheter placement in regional anesthesia. This section highlights recent advances in nerve stimulation and ultrasonography that may play an important role in preventing complications during regional anesthesia practice.

Nerve Stimulation in Regional Anesthesia: Peripheral Nerve Blockade

Despite years of clinical use, the electrophysiologic effect of injectates on nerve stimulation has never been fully explained. The classic unanswered question concerning electrical stimulation is why is it that one may not be able to consistently stimulate a nerve with a current of less than 0.5 mA, even after eliciting a paresthesia in that nerve? Another phenomenon that is poorly understood is the "Raj test." The Raj test is when nerve stimulation is used to locate a nerve and a twitch is observed when the needle tip is close to the neural target. Ideally, the twitch is required to persist at a current of 0.5 mA. The clinician then injects a small volume of local anesthetic or normal saline through the needle. If the needle tip is in the correct location, the muscle twitch disappears immediately.8 Until very recently, the disappearance of the twitch was thought to be caused by physical displacement of the nerve by the injectate.9 We recently learned that this mechanism is best explained in electrical terms and is not entirely a result of the physical displacement of the nerve.10 In a porcine model, the injection of 0.9% sodium chloride solution (NaCl) abolished the motor response, and a subsequent injection of 5% dextrose reestablished a motor response during peripheral nerve stimulation.10 Simulation and in vitro experiments show that injections of 0.9% NaCl and 5% dextrose alter the electrical field in different ways.11 It was concluded that the injection of electrically conducting solutions (saline or local anesthetic) increases the conductive area surrounding the stimulating needle tip, leading to a decrease in the current density surrounding the target nerve (Fig. 50-2). The current density surrounding the needle tip is then no longer sufficient to stimulate the desired nerve.10 This observation suggests that effective nerve stimulation is sensitive to changes that occur at the needle–tissue interface, such as the angle of the needle or the injection of the local anesthetic. The net effect of these changes is to alter the current density at the tip of the needle or the path of the electric current, ultimately resulting in a change in the quality of the motor response.12 This phenomenon has also been reported in a clinical setting.13,14 In a clinical study, the mean current required to stimulate the supraclavicular, axillary, femoral, and sciatic nerves when using an insulated needle was significantly higher (up to 6 times) after the injection of normal saline.15 This effect has also been verified by others in the femoral and sciatic nerves.16 The clinical implications are as follows:

• One may use a nonconducting solution, such as 5% dextrose in water (D5W), rather than saline to open up the perineural space.10
• Reports of the use of nonconducting injectates (eg, D5W) in peripheral nerve block are promising and appear to provide stability when using electrical stimulation techniques.13,14,16
• The motor response resulting from electrical stimulation is augmented after an injection of D5W in a very high percentage of cases (96%). This allows one to increase the accuracy of placement of continuous catheters.14,16

Nerve Stimulation in Regional Anesthesia: Neuraxial Blockade

Epidural stimulation has been used to confirm and guide catheter placement in the epidural space. The epidural stimulation test confirms catheter placement through stimulation of the spinal nerve roots (not the spinal cord) with a low-amplitude electrical current conducted through normal saline via an electrically conducting catheter.17 Correct placement of the epidural catheter tip (1-2 cm from the nerve roots) is indicated by a motor response elicited with a current between 1 and 10 mA.17,18 There has been 1 case report of a current response indicating epidural placement with radiologic evidence of subdural placement.19 Any motor response observed with a significantly lower threshold current (<1 mA) suggests that the catheter is in the subarachnoid or subdural space or is in close proximity to a nerve root20-22; in these rare cases, a motor response is elicited with a significantly lower threshold current because the stimulating catheter may be very close (<1 cm) to the nerve roots or because it may be in direct contact with highly conductive cerebrospinal fluid (CSF) (Table 50-3).

Electrical stimulation has been applied to neural structures for neurophysiologic evaluation and pain control for many years23-26 and has proven to be safe. Although no known complication or patient discomfort has resulted from the epidural stimulation test, it has been recommended to keep the current below 15 mA and the stimulation time as brief (less than a few minutes) as possible.17,18,27,28 In particular, the current output must be carefully increased from zero and stopped once motor activity is visible to ensure that all motor responses, even those elicited with a low current (<1 mA), are detected. The nerve stimulator must allow a gradual increase in current output to at least 10 mA.

Table 50-3 compares features of the epidural "test dose" (lidocaine with 1:200 000 epinephrine) and the epidural stimulation test. Epidural stimulation is a tool for clinician use that may have a significant impact on 3 of the most significant complications associated with epidural anesthesia: systemic toxicity, accidental subarachnoid or subdural injections of local anesthetics, and neural damage.

Ultrasound Imaging in Regional Anesthesia

The application of ultrasound in regional anesthesia was first published in 1989 by Ting and Sivagnanratnam.29 Since then there has been an increasing number of reports in the world literature on this exciting application in regional anesthesia. With the widespread availability of highly portable ultrasound machines, ultrasound techniques have taken hold as a mainstay in the practice of regional anesthesia.

Ultrasound-Guided Peripheral Nerve Blockade

With the use of ultrasonography, one can observe neural targets, vascular structures, the advancing needle, and the actual spread of the local anesthetic solution after the injection of the local anesthetic in real time. Brachial plexus anesthesia is one of the most challenging techniques in regional anesthesia; therefore, ultrasound has potential to improve success rates with this technique. The application of ultrasound in regional anesthesia has renewed interest in the classic supraclavicular brachial plexus block (Fig. 50-3).30,31 Advances in this technology and, more importantly, familiarity with it (via appropriate training programs)32 continue to facilitate better imaging of nerve trunks, blood vessels, pleura, and the approaching needle.33,34 The safety advantages to ultrasound have been studied with great fervor.35 It seems that ultrasound convincingly can decrease the amount of local anesthetic used, it can decrease the incidence of unintentional vascular punctures, and it can significantly decrease the number of needle passes to perform a block.36-40 One can draw all sorts of conclusions regarding the positive effects that ultrasound-guided blocks will have on patient outcomes, but this has not yet been proven.41 Ultrasonographic technology is an undeniable advance in regional anesthesia; however, it does not completely eliminate difficulty in accurately identifying structures and observing the advancing needle in detail in all cases.

To minimize risks of regional anesthesia, potential exists to combine ultrasound and nerve stimulation techniques when performing regional anesthesia. Ultrasonography allows the clinician to see the advancing needle approaching what appears to be the target nerve or trunk. Nerve stimulation allows one to identify which nerve is being approached and, if indeed, what is being approached is a neural structure. Needle/catheter-tip visibility is enhanced using D5W as a preinjectate. Moreover, by using ultrasonography, one can observe the pattern of spread of D5W before committing to the injection of local anesthetic. Individually, ultrasonography (anatomic locating tool) and nerve stimulation techniques (physiologic response aid) have their limitations, but, when used in combination, these techniques may serve to compensate for each other's weaknesses and may facilitate optimal needle placement for peripheral nerve blocks, although this remains to be proven.35

Ultrasound Imaging for Neuraxial Blockade

Ultrasonography is useful for guiding peripheral nerve block placement in adult patients42,43; however, its application for guiding neuraxial blockade in adults and children remains limited, despite increasing interest.44

• Real-time ultrasound imaging of the lumbar spine is a simple procedure. Ultrasound aids the placement of lumbar epidural catheters and enhances the performance of combined spinal–epidural anesthesia.45,46
• Ultrasonography use improves the learning curve of obstetric lumbar epidural catheter placement for anesthesia trainees.47
• In patients with anticipated difficult epidural localization, this technology is helpful for estimating lumbar epidural depth; it also facilitates ease of placement.48,49
• Although ultrasound imaging has been used to guide lumbar epidural needle placement, it may be of limited value in the thoracic region, particularly in older children and adults, when visualization of the spinal cord and relevant structures is sought.50,51 Calcification of the posterior vertebral bodies in children older than 6 months of age prevents reliable imaging of the spinal cord.50

At the present time, ultrasonography guidance is helpful for viewing the lumbar region of selected patients, although its use for thoracic epidural placement is of value only in infants and small children, as their vertebrae are not fully ossified.

Monitoring Regional Anesthesia

We believe it is important to have an assistant observe and aid the patient at all times during the performance of regional anesthesia. As many as 15% of patients have a fear of needles and vasovagal episodes occur when performing regional anesthesia.52

• Standard electrocardiogram and pulse oximetry are essential monitors while performing regional anesthesia.
• Before performing the neural block, a baseline blood pressure reading should be obtained. Once the regional anesthesia procedure is complete, the monitors should remain attached. In conscious patients, end-tidal carbon dioxide monitoring is not used; however, there are special nasal prongs available for monitoring spontaneously ventilating patients.
• Evidence of regressing sensory and motor blockade and stable vital signs must be present to fulfill the criteria for discharge from the recovery area.
• Local anesthetic infusions are now routinely used in many medical centers. Patients receiving local anesthetic infusions should be visited regularly by a qualified physician postoperatively (ie, Acute Pain Service).

Record Keeping/Documentation

Accurate and meticulous recording of anesthesia information is essential for maintaining the quality of care in regional anesthesia. When determining the true nature and extent of complications, an accurately documented preoperative examination is important.

• Detailed documentation of patient consent and the clinical procedure is very important.
• Open and honest communication with the patient is essential for providing good quality patient care.

Physician–Patient Communication

Effective communication with each patient is essential for the prevention and early diagnosis of any potential complications. Patients undergoing regional anesthesia may report anxiety, and appropriate preoperative education for the patient can help mitigate this.53 Discussing the procedures, including their benefits and any significant risks involved, is the legal and professional responsibility of all anesthesiologists. It is equally important to maintain good rapport with the patient during the postoperative period.

• A telephone call to the patient on the first postoperative day is a reasonable and practical alternative to a visit.
• Specific common risks for certain blocks should be discussed with the patient before discharge. For example, patients undergoing supraclavicular blocks should be warned about the risk of pneumothorax and be informed about potential symptoms and what to do if they develop.
• Caution patients about the risk of burns (ie, from radiators) or the consequences of applying pressure to desensitized areas when sensory anesthesia continues after discharge.
• Warn patients about lying on paralyzed extremities for any length of time or letting them become dependent.
• Patients should receive written instructions and information about when to seek medical attention before discharge from the hospital.
• As per hospital policy in local practice areas, ambulatory patients should be discharged to the care of a family member or guardian. The consequences of being left unmonitored must be made clear to patients.

Allergic Reactions

Although allergies to local anesthetics are rare, a full array of allergic symptoms and signs ranging from mild skin irritation to full-blown anaphylaxis have been described. These signs and symptoms are almost always associated with amino ester preparations or preservatives (eg, methylparaben). Allergic reactions are more common after exposure to ester compounds than amides.54

It is quite difficult to obtain good statistical information on the problem of allergic reactions to local anesthetics, and most anesthesiologists are convinced that the risk of allergic reactions to lidocaine is infinitesimal. Lidocaine is frequently administered intravenously by cardiologists, emergency medicine specialists, and anesthesiologists, and reports of allergic reactions using it in this fashion are indeed rare.

The United Kingdom uses a tracking system for adverse reactions to all medications referred to as the ADROIT system (Adverse Drug Reporting on Line Tracking System). Nazir and Holdcroft55 accessed that database and studied adverse reactions to local anesthetics between the years 1967 and 2005 and recently published their findings. What was really surprising was that of 797 reports of adverse events to lidocaine, 331 of the reports involved lidocaine with no additives, and 96 of 331 of these were considered to be allergic reactions. Lidocaine in combination with methylprednisolone or prilocaine generated the most allergic reactions. There were 16 deaths linked with lidocaine alone, and at least 4 of them were deemed to be anaphylactic reactions. Allergic reactions were reported in 16% of the adverse events associated with bupivacaine. One can argue of course that most of these allergic reactions were not true allergies, but merely systemic effects mislabeled as "allergic" reactions.

Systemic Toxic Reactions

Nazir and Holdcroft's report provided further evidence of the minimal risk of local anesthetics.55 They report approximately 1 fatality per year in the United Kingdom by all users of local anesthetic drugs since 1967. Unfortunately, we do not have a denominator for the number of local anesthetics administered in the United Kingdom each year, but we can assume that it would be a large number. This study again proved that lidocaine caused fewer deaths than bupivacaine. There were 17 bupivacaine fatalities of 160 adverse events as compared with 20 of 797 for lidocaine. Fifteen of these fatalities had a cardiovascular origin. Ropivacaine was released for clinical use worldwide in 1996. Since that time there were 16 adverse events reported to the ADROIT tracking system, including 1 allergic reaction and no fatalities, though the editor is aware of ropivacaine-related deaths.

Systemic toxic reactions to local anesthetic drugs occur almost always as a result of unintentional intravascular injection and rarely follow the injection of an excessive quantity of local anesthetic into an appropriate site. The incidence of systemic toxicity has substantially decreased within the past 30 years. In 1969, Dawkins56 reported the incidence of seizures after local anesthetic injections to be 0.2% after epidural anesthesia. A recent study from France reported an incidence of seizures of 0.01%, which represents a 20-fold decline in a 40-year period.56,57 A higher occurrence of systemic reactions is after peripheral nerve blocks, especially brachial plexus and caudal blocks in adults.58 The maximum plasma concentration of local anesthetic (Cmax) resulting from an unintentional intravascular injection depends on a number of factors,59 including the total dose of local anesthetic injected, the speed and site of injection, and whether the injection is administered intravenously or intra-arterially. The lungs are an important repository for local anesthetic drugs; plasma concentrations of these drugs will be substantially higher if the lungs are bypassed (eg, an accidental intra-arterial injection in the head, face, or neck region).60 Plasma concentrations of local anesthetics are also influenced by the tension of carbon dioxide (CO2) and the pH. An elevated arterial CO2 tension increases cerebral blood flow, and an acidotic state increases intracellular ion trapping and the amount of free drug available. This combination of factors has a synergistic effect on the seizure threshold.61

Systemic toxic reactions occur much less frequently when local anesthetics are administered in peripheral sites. A number of factors influence the degree of absorption taking place from the periphery to the central circulation. The most important factor influencing absorption is the site of injection—absorption is more rapid in highly vascular tissues and less so in poorly perfused tissue. Rapid absorption also occurs from intrapleural injections, and very slow absorption occurs from the bladder and skin. Consequently, local anesthetic absorption increases from the highest to the lowest rates in the following anatomic sites: intercostal, epidural, brachial plexus, lower extremity, and subcutaneous tissue (Fig. 50-4).

The rate of absorption is reduced by the addition of epinephrine to local anesthetic drugs, but this also depends on the local anesthetic used. Furthermore, the addition of epinephrine itself may lead to other complications (see Complications of Peripheral Nerve Blocks later). As the plasma concentration of lidocaine increases, there is a typical progression of effects on the CNS and the cardiovascular system. This pattern of symptomatology is not typically seen with the more potent local anesthetics (Fig. 50-5).62 CNS excitation after epidural anesthesia almost always arises as a result of unintentional intravascular injections. Local anesthetics are amphiphilic molecules, having both lipophilic and hydrophilic properties; these drugs enter a variety of cellular compartments and have the potential to interact with a wide variety of molecules, including inotropic signaling pathways (sodium, potassium, and calcium ion channels), and also influence adrenergic and lysophosphatide signaling systems, cardiac bioenergetics, and mitochondrial dynamics.63 As plasma concentrations of local anesthetic increase, signs of local anesthetic toxicity increase in severity. Concurrent treatment with CNS depressant medications may modify the typical clinical signs of a toxic reaction and can mask some of the early warning signs (Table 50-4).

The cardiovascular system is more resistant to the toxic effects of local anesthetics than the CNS, especially after toxic doses of lidocaine. Local anesthetics affect both electrical and mechanical cardiac activity. Tachycardia and hypertension are early signs of cardiac toxicity, and with increasing doses patients develop bradycardia and hypotension; however, this pattern of symptomatology may not be seen when the patient receives a rapid intravascular injection and this pattern of symptoms and signs is not evident with potent local anesthetics.

Prevention

Early recognition of an intravascular injection is the key to prevention. Aspiration should be performed before any injection of local anesthetic. The aspiration should be repeated with any change in needle position.

The administration of epinephrine and isoproterenol with the local anesthetic can aid in ruling out an intravascular injection.64 Increased heart rate and systolic blood pressure in addition to T-wave changes are considered sensitive and specific end points in response to an intravascular injection of a test dose containing epinephrine. A single injection of 15 μg of epinephrine produces a heart rate increase of greater than 10 beats/min, a blood pressure increase greater than 15 mm Hg, and a decrease in T-wave amplitude of 25%.65 In the sedated patient, changes in heart rate may not be as reliable as T-wave and blood pressure changes.66 Older patients (>60 years of age) and those on β-blockers and anesthetized patients are less sensitive to β-adrenergic stimulation.

The accurate deposition of local anesthetic, in potentially smaller doses than traditionally used, holds promise to reduce the incidence of systemic toxicity. Ultrasound guidance may help to prevent intravascular needle placement and, more importantly, injection of local anesthetic.67 This has yet to be proven, although there is 1 report of a reduction in the number of seizures associated with ultrasound versus nerve stimulation guidance.68

Recently, an additional measure to warn of intravascular needle placement has been investigated, making use of simple observation of electrical impedance (as displayed on currently available nerve stimulators) during block placement.69 The failure to observe a rise in impedance upon injection of a test dose of D5W seems highly predictive of intravascular needle placement. The basis for this observation is that the injected fluid is quickly dispersed within the systemic circulation and does not increase the current density in the local vicinity as does a nonconducting solution (D5W). Although this "test" would not replace other precautionary measures, it may provide an additional warning sign of intravascular injection before any local anesthetic is injected.

Following the precautions outlined next can minimize the chances and impact of unintentional intravascular injection.

• Incremental administration of the local anesthetic
• Frequent aspiration
• Observation of heart rate, systolic blood pressure, and T-wave changes
• Observation of the patient

Management of Local Anesthetic Toxicity

The initial treatment recommended for the management of patients with systemic toxicity is very similar to that used for any resuscitation. The SAVED mnemonic outlined in Fig. 50-6 can be used to guide the management of allergic reactions and systemic toxicity to local anesthetics.70

Because hypoxia, hypercapnia, and acidosis exacerbate all local anesthetic toxic reactions,71,72 control of the airway and ventilation is of paramount importance in the treatment of local anesthetic toxicity. Recent studies demonstrate improved hemodynamics and survival in animal models of bupivacaine toxicity with the administration of intravenous lipid emulsion.73,74 It is suggested that lipid emulsions function to remove local anesthetic molecules from binding sites that are responsible for cardiovascular depression.75 This has been verified clinically in one instance in which the serial serum concentrations of bupivacaine decreased faster than its half-life would predict after the administration of a lipid emulsion.76 It is likely by this same mechanism that lipid emulsions have been used to treat other toxidromes. Lipid emulsions may also act by reversing the local anesthetic inhibition on myocardial fatty acid oxidation.77 In this manner, they would restore myocardial adenosine triphosphate supply. Propofol, which is formulated in lipid, may reduce susceptibility to local anesthetic toxicity; however, its negative inotropic effects may mitigate against its use as an antidote in the face of cardiovascular collapse, and its use is not recommended for the treatment of local anesthetic systemic toxicity.78 Vasopressin is no longer recommended for treatment during resuscitation from local anesthetic toxicity.

Of course, studies to fully evaluate the use of lipid emulsions are difficult given the nature of its use. Regardless, there is a large and growing body of evidence to support its efficacy. It has been used in cases to treat local anesthetic-induced cardiac arrest due to bupivacaine, levobupivacaine, ropivacaine, and mepivacaine.79-81 It has even been reported to reverse the neurologic effects of local anesthetics.82 The use of lipid emulsions in pediatrics has not been extensively reported, but there are case reports of successful outcomes.83 Thus far, the safety profile of lipid emulsions seems to be quite good. There has been a case of an elevated amylase without clinical sequelae after its administration.84 The limits of safe dosing and systemic adverse effects of high-volume lipid emulsion administration are unknown at this time, and there is caution for administering high volumes in patients with abnormal pulmonary, renal, hepatic, and pancreatic function. The recommended dosing for Intralipid ® 20% (Fresenius Kabi, Uppsala Sweden) is to first administer a bolus of 1.5 mL × kg–1 IV over 1 minute and then start an infusion of 0.25 mL × kg–1 × min–1. Compressions should be maintained to circulate the lipid. Bolus doses can be administered every 3 to 5 minutes to a total of 3 mL × kg–1. Lipid emulsions should be administered (during resuscitation) to patients in cardiac arrest due to local anesthetic toxicity, and, preemptively, to those displaying overt neurologic toxicity.

The following are some recommendations for the management of significant local anesthetic systemic toxicity:

• Bronchospasm and generalized edema, sometimes associated with allergic reactions, may require use of bronchodilators, antihistamines, and corticosteroids.
• Ventilation and oxygen are required to correct acidosis and prevent hypoxia and hypercarbia; tracheal intubation should be individualized.
• Chest compressions and defibrillation may be required to restore organ perfusion and should be instituted based on patient hemodynamics.
• Lipid emulsion therapy can be lifesaving and should be administered early. Lipid emulsions should be available in locations where local anesthetics are used in potentially toxic doses.85
• Seizures may be managed with benzodiazepines (eg, midazolam 0.05-0.1 mg/kg), but alternatively with propofol (0.5-1.5 mg/kg) or barbiturates (thiopentone 1-2 mg/kg). Evidence exists that lipid emulsions can halt local anesthetic induced seizures, though this needs further evaluation.
• Profound hypotension can occur in both allergic reactions and systemic toxicity and usually responds well to vasopressors (eg, epinephrine) and plasma "expansion."
• Because reduced cardiac contractility is a core element in this condition, it is thought that the maintenance of coronary perfusion with the administration of epinephrine and norepinephrine improves outcome.86 There is now laboratory evidence to suggest that epinephrine may impair resuscitation and reduce the efficacy of lipid therapy.87
• Animal model evidence exists that epinephrine can interfere with lipid emulsion rescue.87 Current recommendations are to use epinephrine bolused in 10- to 100-mcg aliquots in the treatments of local anesthetic-induced hypotension.85
• However, malignant dysrhythmias, which particularly occur with bupivacaine systemic toxicity, should be controlled in a timely fashion because epinephrine can exacerbate these dysrhythmias.
• Therapeutic agents that are less arrhythmogenic have been investigated, including the use of vasopressin88,89 and phosphodiesterase inhibitors such as milrinone and amrinone, though their use still needs to be validated clinically.89,90
• In addition to lipid treatment, ventricular arrhythmias should be suppressed primarily with amiodarone 300 mg IV, with repeat administration of up to 150 mg 3 to 5 minutes later.
• Effective resuscitation in this setting is difficult, and atrioventricular pacing and cardiopulmonary bypass are additional options in refractory cases.91

Direct Needle Trauma to the Nerve

Auroy et al57 reported an incidence of serious nerve injury (permanent sensory and/or motor loss) after peripheral nerve blocks (PNBs) as 1.9 per 10 000 nerve block cases; in this study, all patients with serious injury experienced either pain on injection, paresthesia, or both during the performance of the block. A more recent analysis of nearly 7000 blocks reported a similar incidence of 0.04% for late neurologic deficits after PNBs.92 The incidence of minor neural injury after PNBs is in the range of 1% to 2%; most of these injuries are transient neurapraxias, which represent axonal disruption. Typically neurapraxia injuries are observed postoperatively when patients complain of persistent numbness in the distribution of a peripheral nerve. One cannot simply assume that all neurapraxia injuries are anesthesia related, as patient positioning, surgical trauma, and tourniquet application can all give rise to these symptoms. Numbness gradually regresses over a period of weeks and is rarely observed beyond 3 months, which is the amount of time required for axonal regeneration to occur.93 Regional anesthesia-related injuries may be the result of needle trauma, injection pressure, or the toxic effects of local anesthetics or additives. It is generally agreed that the location within the nerve which is most sensitive to injury is the intrafascicular compartment.

Prevention

Some suggest that most neural injuries are associated with either paresthesia or pain on injection. Needle damage or pressure generated during injection of local anesthetics account for most of these injuries.57,94 Needle insertion without injection is a routine technique used during microneurography and surgical repair procedures, which results in insignificant nerve damage. It is more likely that significant nerve damage is caused or worsened by both mechanical and chemical injury during intraneural injections of neurotoxic substances (ie, local anesthetics). High-pressure injections cause mechanical destruction of the neural fascicular architecture, pathophysiologic damage, and neural scarring.95 Chemically induced damage is also possible from high concentrations of local anesthetics, vasoconstrictors, preservatives, and other additives.

There is an ongoing debate among anesthesiologists about the safety of deliberately seeking paresthesia in regional anesthesia.96 There is also concern about performing regional anesthesia in comatose/anesthetized patients because of the inability to detect paresthesia. Currently, there is no substantial evidence that performing nerve blocks in awake patients is any safer than performing them in anesthetized patients, and attitudes are shifting toward the acceptance of performing blocks in anesthetized individuals as the safety of PNBs in general continues to increase.97

Ultrasound imaging has been used to examine the location of needle placement as positioned using paraesthesia98 and nerve stimulation as localization techniques.99,100 The resounding finding has been that intraneural injection of local anesthetic may occur with a greater frequency than previously thought, without inevitably leading to neurologic complications. Moreover, the generally held assumption that nerve stimulation thresholds of less than 0.3mA may not reliably prevent intraneural injection. Comparing intraneural and extraneural stimulation thresholds in ultrasound-guided supraclavicular block, Bigeleisen et al101 found that a stimulation current of 0.2 mA or less reliably predicts intraneural needle placement, but that a threshold current greater than 0.2 mA does not preclude intraneural placement. These results are based on the assumption that ultrasound can reliably visualize intraneural injection. Studies using pigs have shown that ultrasonographic nerve expansion during injection is consistent with intraneural injection as confirmed by histologic analysis.102,103 It is important to note that at this time it is wise to avoid intraneural needle placement, particularly because ultrasound imaging at this time cannot ensure avoidance of the intrafascicular compartment, which has been shown to be the principal location for nerve trauma via mechanical or chemical means.

Common sense dictates that small-gauge needles are less likely to damage nerves than larger-gauge ones. More than 30 years ago, Selander et al104 recommended using blunt needles when performing regional anesthesia, as it was thought that blunt needles were less likely to penetrate neural structures, and resultant intraneural injections would be less likely to occur; this recommendation was based on information derived from animal experiments. Selander's influence on this topic persists to this very day, even though subsequent studies showed that blunt needles, although far less likely to penetrate neural structures, are far more disruptive to neural tissue than sharp needles.105 Nevertheless, there are no clinical trials supporting any recommendations regarding what type of needle is best for regional anesthesia procedures. Most clinicians prefer small-gauge, short, blunt needles.

Injection pressures may influence the amount of damage inflicted on a nerve. One study suggested that persistent motor deficits were observed in animals injected with pressures of 1293 mm Hg or greater.106 Monitoring injection pressures is possible, although it is neither specific or sensitive enough to reliably predict neural injury.106 Clinicians should avoid rapid and high-pressure injections. The difficulty with monitoring injection pressures is that subjective "resistance" is often relied on, which has shown to be unreliable.107 One standardized technique that one of the authors (B.T.) routinely uses is the compressed air injection technique.108 Using Boyle's law, the pressure can be kept significantly below 1293 mm Hg (mean was 745 mm Hg) by compressing air, previously aspirated above the fluid, to 50% of its initial volume before and during injections.

Sterilizing agents, skin-cleansing substances, detergents, and certain preservatives (eg, metabisulfite) all cause neurotoxicity and should be carefully avoided when introduced into perineural spaces.

The neurotoxicity of a local anesthetic is related to its potency and its concentration.109 High-concentration local anesthetics such as 2% lidocaine and 0.75% bupivacaine should be avoided in peripheral nerve blocks. The addition of vasoconstrictors (eg, epinephrine) to local anesthetics may enhance the damage caused by an intraneural injection.110-112

Generally, neural damage resulting from peripheral nerve blocks is rare; consequently, it may be difficult to clearly demonstrate the safest equipment and techniques to be used. However, Table 50-5 summarizes several measures to prevent nerve injuries.

Management

When a neurologic injury is suspected postoperatively, a thorough history must be taken, and a complete physical examination must be performed and documented. Because most regional anesthesia procedures involve the percutaneous insertion of needles toward nerves, the burden often lies with the anesthesiologist to prove that damage was not caused as a result of improper technique and unsafe practice. Clinicians are obliged to maintain a very open mind when dealing with such challenging cases, as there can be a variety of causative factors (Table 50-6). The anesthesiologist should play a major role in determining the cause of the injury, as anesthesiologists have far more information concerning preoperative and intraoperative events than do most neurologists. Symptoms and signs of compression of the spinal cord must be dealt with urgently (within 6-8 h); otherwise, permanent paraplegia or quadriplegia may result. The anesthesiologist, neurologist, neurosurgeon, and radiologist must work as a team and strive to arrive at a diagnosis before serious permanent injury occurs. Diagnostic tools should be used judiciously to support the team of clinicians in arriving at a correct diagnosis; this is when we must rely on our neurology and radiology colleagues to guide us. Electrodiagnostic and imaging techniques can often take the guesswork out of many diagnostic dilemmas and allow for quick and precise diagnoses. Figure 50-7 summarizes the key steps involved in determining neurologic injury.

Diagnostic Tools for the Determination of Nerve Injury

The initial step in the diagnostic workup is to perform a thorough and well-documented history and physical examination, paying careful attention to the muscles innervated by the nerve(s) in question. Direct injury to the spinal cord, nerve roots, or peripheral nerves is best evaluated using imaging techniques, especially in the early stages of an injury. Electrophysiologic techniques are more useful in the later phases of an injury. The most common imaging modalities used are computerized tomography (CT) and magnetic resonance imaging (MRI). Electrodiagnostic techniques include evoked potentials, nerve conduction studies, and needle electrode examination of muscles (electromyography [EMG]). The use of these tools should complement the clinical examination, rather than replace an examination. Choosing the best tools/technologies for diagnosis should be a joint decision made with the neurologist, surgeon, and radiologist.

• CT is best suited for evaluating bony abnormalities.
• MRI is ideally suited for the examination of soft-tissue abnormalities, especially the spinal cord.
• For peripheral nerve, nerve plexus, and peripheral nerve complications, imaging is less likely to be useful for the demonstration of nerve injury.
• MRI may demonstrate the accumulation of blood and edema fluid, which can lead to compartment syndrome; MRI may also indicate neural compression caused by injury from the needle and local anesthetic injection.
• Nerve conduction studies test the function of large sensory and motor nerve fibers. Evaluating nerve conduction can reveal axonal loss or demyelination of the nerve; however, nerve conduction is less useful in timing lesions when the injury occurs.
• EMG is preferentially used for evaluating smaller motor units. EMG can be useful for the diagnosis of axonal injury and is also useful for quantitating the severity of the neurologic injury and for identifying the actual site of injury. EMG studies are typically recommended 2 to 3 weeks after an injury, although an early EMG studies are recommended 2 to 3 weeks after an injury.

The only effective way to manage neurologic complications is to prevent any mishaps from occurring in the first place, as there is limited chance of recovery once the damage has occurred. Neurologic consultation and testing should be considered if there are any persistent symptoms or signs after a procedure. If symptoms are mild and do not interfere with the patient's daily activities, reassurance can be offered after evaluating the extent and severity of the patient's symptoms. It is of prime importance to continue to follow patients suffering from nerve injury after discharge from the hospital; it is also necessary to instruct patients to seek medical attention if their symptoms worsen or do not improve. Most residual dysthesias or hypesthesias resolve in 4 to 6 weeks, and the majority are resolved (>99%) within 1 year.113,114

Needle Trauma to Surrounding Structures

Tissues within the vicinity of the PNB may be unintentionally injured. Vascular injury can also occur during PNB because many peripheral nerves travel in parallel with vascular structures. Other injuries may be caused by direct needle trauma, including pneumothorax and direct spinal cord injury.

Spinal Cord Injury

Permanent spinal cord injury after brachial plexus block is the most severe complication resulting from PNB. There are a number of reported spinal injuries associated with peripheral nerve blocks: an interscalene block performed with an 8-cm needle resulted in a permanent neural deficit at the C8-T1 level.115 In another instance, an anesthetized patient suffered permanent spinal cord injury after an interscalene block.116 A patient has suffered from Brown-Séquard syndrome after an attempted interscalene block using a spinal needle.117 Other peripheral blocks may also increase the risk of spinal cord injury. Permanent spinal cord injuries may occur after paravertebral blocks because the needles used for paravertebral blocks are inserted in close proximity to the spinal cord.

Prevention

It is important to note that most of the serious nerve injury cases reported involve deviations from the recommended anesthetic practice standards. Currently, it is recommended to perform these blocks in awake patients to detect paresthesia or pain on injection. Ultrasonography may help to reduce the risk of spinal injury as needle advancement can be observed in real time. It is important to note here that the use of ultrasound visualization should not preclude other precautionary means during performance of PNB. When performing ultrasound-guided techniques in close proximity to the spine, using a technique that directs the needle toward to spine should still be avoided. For instance, when aligning the needle to a sagittally positioned probe during lumbar plexus blockade, it is important to avoid out-of-plane technique from lateral to medial, which would direct the needle toward the spinal column.

The best option for avoiding the risk of nerve injury is to select a block insertion site remote from the spinal cord. Axillary blocks appear to be safer than supraclavicular blocks; however, supraclavicular approaches to the brachial plexus are required in shoulder surgery.

Small-gauge needles and short needles are strongly recommended when performing brachial blocks in the supraclavicular region. Longer-than-usual needles were associated with many of the spinal cord injuries associated with brachial plexus blocks.

Management

There is no specific treatment for primary needle damage to the spinal cord. Nevertheless, the initial step in the management of spinal cord injury is the recognition and identification of neural dysfunction. Acute, potentially reversible causes of spinal cord injury, such as nerve compression from hematomas, must be identified and dealt with early (within 6-8 h); otherwise permanent paraplegia or quadriplegia may result. The anesthesiologist, neurologist, and radiologist must work as a team to arrive at a correct diagnosis. Appropriate electrodiagnostic and imaging techniques must be used to make a quick and precise diagnosis.

Pneumothorax

Any regional technique requiring needle insertion toward the lung involves the risk of pneumothorax. Pneumothorax has been an unwelcome complication of supraclavicular techniques since Kulenkampff118 first described the classic supraclavicular approach in 1911. The incidence of pneumothorax is difficult to determine and varies depending on the approach. Brand and Papper119 reported an incidence of 6.1% in a large teaching hospital using the classic Kulenkampff technique. The risk of pneumothorax has deterred many anesthesiologists from using the supraclavicular approach and is the most likely reason that axillary and infraclavicular approaches are popular. The risk of pneumothorax is much lower following the interscalene approach to the brachial plexus compared with the classic supraclavicular approach.120 Ward121 reported a 3% incidence of symptomatic pneumothorax following the interscalene technique. The risk of pneumothorax is reduced following the vertical technique mainly because the needle is not directed toward the lung.122 More recently, however, case series of supraclavicular blocks performed with ultrasound guidance have reported the absence of any single clinically significant pneumothorax.20,123 Although these studies are not large enough to detect the true incidence, the practice of supraclavicular blocks is being revisited.

Prevention

Anesthetic techniques requiring the insertion of a needle directed toward the lung in the supraclavicular region all carry the risk of pneumothorax. Extra care should be exercised in tall, thin patients, as they appear to be at greater risk of pneumothorax. For all patients, a right-sided pneumothorax occurs more frequently because the cupola of the lung is higher on the right side. Patients should be warned in advance of this risk, and ambulatory patients should be given careful instructions on how to proceed should symptoms develop.

Supraclavicular approaches should be avoided in patients with severe impairment of pulmonary function. Blocks should never be performed bilaterally. Intuitively, complications can be avoided or reduced if clinicians are able to visualize the advancing needle approaching the target nerve or trunk. Ultrasonographic technology facilitates this goal in real time. However, this technique requires significant training and practical experience.

Management

Because upper-extremity operations are carried out on ambulatory patients who are discharged within a few hours of surgery, it is important that patients be warned about the risk of pneumothorax before leaving the hospital. Patients who develop chest pain, dyspnea, or cyanosis after discharge should be instructed to go to the nearest emergency center. Symptoms and signs may not develop for hours, and patients may not become symptomatic until a 20% pneumothorax is present. Evacuation of the pneumothorax is usually required when the degree of lung collapse is 25% or greater.

Ultrasound Detection of Pneumothorax

Ultrasound imaging can rule out or confirm pneumothorax both effectively and efficiently and is better than bedside chest radiographs.124 This said, acquiring the prerequisite imaging technique to master interpretation of both normal and pathologic lung signs requires time and effort. Ultrasonography to detect pneumothorax should be performed at the anteroinferior aspect of the thorax in a supine patient; a large majority of cases involve the anterior zone, and this area is involved in most if not all life-threatening cases.

Neural Toxicity

Local anesthetics are considered harmless substances when injected perineurally in appropriate concentrations and quantities. High concentrations of local anesthetics can permanently damage neural tissue in some instances.125 Preservatives in local anesthetic drugs may also damage nerves and other surrounding tissues. In the United States during the 1970s, it was noted that a change in the constitution of a sodium metabisulfite, a preservative found in chloroprocaine, resulted in several cases of cauda equina syndrome.126 The addition of ethylenediaminetetraacetic acid (EDTA) to chloroprocaine is associated with severe back pain in some patients after epidural anesthesia.127,128 Studies show that 5% hyperbaric lidocaine for spinal anesthesia is linked to the syndrome transient neurologic symptoms (TNS).129

Myotoxicity

Myotoxicity is a recognized complication of intramuscular injections of local anesthetics.130 Local anesthetics are proposed to cause a pathologic efflux of Ca2+ from the sarcoplasmic reticulum, resulting in contracture, cell destruction, and necrosis. After this occurrence, the regeneration of fibrils occurs within a few weeks. Among the local anesthetics tested, bupivacaine caused the most damage, and procaine caused the least.131 Injury was noted to be worse with repeated injections and when epinephrine was used.132-134 In clinical practice, myotoxicity is largely unnoticed except in ophthalmic regional anesthesia. Diplopia has been reported after retrobulbar blocks; however, this symptom is short-lived in most cases, and permanent damage is rare. Ropivacaine was found to be less myotoxic than bupivacaine in an animal model.135 The full implications of the effects of local anesthetics on muscle have not yet been evaluated.

Phrenic Nerve Paralysis

The incidence of hemidiaphragmatic paresis and decreased respiratory function after supraclavicular blocks in 8 healthy volunteers was noted; the overall incidence of paresis was 50% after the administration of 30 mL of lidocaine 1.5% with epinephrine, and none of the volunteers reported respiratory symptoms.152 However, anecdotal reports exist concerning patients who are devoid of respiratory disease and who later became symptomatic after an interscalene block.121,137-139

Temporary phrenic nerve paralysis after interscalene brachial plexus block is expected in up to 100% of cases. Permanent phrenic nerve palsy has been observed after interscalene block,140 but is extremely rare.141

Phrenic nerve paresis is common after supraclavicular blocks, regardless of the technique used, yet patients do not usually become symptomatic.142,143 In another study, the effects of ipsilateral hemidiaphragmatic paralysis on respiratory function after continuous interscalene blocking showed that all patients had a 27% reduction in forced vital capacity, a reduced forced expiratory volume of 26%, and a decreased peak expiratory flow rate.144

Thus the use of supraclavicular techniques in certain groups of patients must be considered carefully. Supraclavicular techniques may need to be avoided in patients with advanced pulmonary disease. Bilateral supraclavicular techniques are absolutely contraindicated.

Detection of Hemidiaphragmatic Paresis Using Ultrasonography

By observations of diaphragmatic excursion, ultrasound is a practical, sensitive, and low-risk method for detecting ipsilateral hemidiaphragmatic paresis after PNB.142,145 The primary diagnosis is paradoxical cephalad motion as compared with normal active caudad motion on inspiration.

Horner Syndrome

Horner syndrome (ipsilateral, miosis, ptosis, enophthalmos, loss of sweating) is frequently observed after supraclavicular approaches to the brachial plexus, although its incidence may be lower when ultrasound is used to guide the supraclavicular approach,30,146 and patients and other caregivers should be informed of this temporary distortion to avoid diagnostic confusion.

Hoarseness

Hoarseness may occur if the local anesthetic spreads to the recurrent laryngeal nerve. Specific management is not required, as the symptoms will abate as the anesthetic wears off. Persistent hoarseness should urge the clinician to consider an alternative cause.

Prevention/Management of Localized Trauma/Toxicity

The key to preventing complications associated with PNB is to increase the accuracy of needle placement and to use the minimum volume and concentration of local anesthetic required to produce a successful and high-quality block. Ultrasonography has revolutionized regional anesthesia, as sonography makes it possible to directly visualize the needle tip and local anesthetic spread in real time while ensuring the avoidance of critical structures. Perhaps the greatest advantage of ultrasound guidance is the ability to reposition the needle after injecting a test dose of local anesthetic or alternative (D5W) and failing to observe adequate or appropriate spread.146 Although the evidence base at this time is limited for improved safety with ultrasound-guided brachial plexus blocks,147 and although there have been reports of nerve injury after and vascular punctures during ultrasound-guided PNB,41,42 our increasing experience as well as others' has shown promise in this area. Like any other technological advance, receiving proper training and gaining experience is critical with real-time ultrasonography. This is particularly critical for obtaining and maintaining needle control and visibility. As with other adverse effects, careful diagnosis of brachial plexus complications and the provision of supportive measures are essential for proper clinical management.

Direct Needle Trauma

As a needle or catheter is advanced into the epidural space, direct trauma to the spinal cord, conus medullaris, and spinal nerve roots can occur. Sensory loss and, less commonly, motor deficits occur as a result of spinal cord trauma. Some patients recover completely; however, the injury persists in the many of patients. The incidence of spinal cord trauma is very low, and much of the data available comes from retrospective sources. In a prospective multicenter study, Auroy et al57 found 5 cases of radiculopathy after 30413 epidurals. In each of these patients, pain or paresthesia was noted during needle insertion and drug administration, and the radiculopathy was observed in the distribution of the associated paresthesia. One of the most disturbing complications of neuraxial blockade is neurologic injury. Three well-known syndromes are associated with damage to the spinal cord, roots, and coverings: cauda equina syndrome, adhesive arachnoiditis, and anterior spinal artery syndrome. These syndromes are addressed in other chapters.

Prevention

To avoid nerve trauma, a studied technique and accurate anatomic knowledge is a prerequisite.149 Although epidural placement in the anesthetized child is considered safe, similar placement in the adult population remains controversial. Recent case reports highlight the potential for neurologic trauma when performing epidural anesthesia in the anesthetized patient.150-152 The use of the stimulating epidural catheters allows pediatric anesthesiologists to place lumbar or thoracic epidurals from the caudal space, minimizing the risk of needle-mediated nerve injury.28 When performing an epidural in an awake, cooperative adult, needle advancement should be halted if the patient complains of pain. In most adults, the spinal cord terminates at the lower portion of the body of L1; however, there are considerable variations among individuals. The ability of the clinician to correctly identify lumbar spinous interspaces has been questioned by Broadbent et al153 using magnetic resonance imaging. In this study, only 29% of the interspaces were correctly identified, whereas 51% of the time clinicians were at a higher vertebral level than anticipated; furthermore, the spinal cord terminated below L1 in 19% of subjects. One explanation for inaccuracy in determining the level of needle placement in the lumbar region is that the Tuffier line is not a fixed landmark, nor is the conus medullaris.154-156 Fortunately the variation in these landmarks is such that there is almost always a safety margin of 2 to 4 vertebrae between the Tuffier line and the conus medullaris. With advancing age, however, this safety margin narrows. Furthermore, with the increasing incidence of obesity in our population, there is a tendency to estimate the Tuffier line to be at a higher level. Therefore, one should be particularly attuned to anatomic variation when performing spinal anesthesia in elderly obese patients. There are very few reports of spinal cord damage caused by errant needle placement in the literature, the reason being that many of these cases enter the court system and are subsequently not published, at least in the medical literature. Reynolds157 published some information about this problem about 10 years ago. She accumulated 7 case reports (6 of which were obstetric) of spinal cord trauma after spinal anesthesia and indicated that there were several more cases pending in the United Kingdom. It is very important to take the time to identify the landmarks in every patient and if in doubt seek verification of the landmarks using ultrasonography, performed by an experienced spinal sonographer, or just a plain x-ray.

Although the use of ultrasound has been used to assess determination of the intervertebral level through palpation,153 the learning curve for having accuracy with determining interspaces is shallow, with only 27% of anesthesiologists achieving competency within 20 trials (2 min each).158 Paresthesia associated with spinal cord injury can occur at the time of needle placement, yet it may also occur during the injection of the solution or as a secondary consequence of irritation, edema, or hematoma.159,160 Pain is more commonly associated with extra-axial lesions affecting the nerve roots or blood vessels that are innervated by pain-mediating sensory neurons.161 In contrast, because there are no pain receptors within the spinal cord (or the brain), intra-axial trauma may be painless161; this allows percutaneous cervical cordotomy to be performed in awake patients.162,163 During this procedure, the cervical cord is typically punctured multiple times with a 22-gauge needle electrode, yet the patient generally describes neither pain nor paresthesia.164 In addition, pain after dural puncture is rare in clinical practice. Thus anesthesiologists should be reminded that they should not simply assume that paresthesia will always be reported as the needle encroaches on the spinal cord.165,166 However, one might expect a motor response if a needle encroaches on a motor tract during attempts at epidural anesthesia. Electrical stimulation during epidural needle advancement may provide an additional warning sign.147,148

Ischemic injuries are among the rarest complications reported after regional anesthesia procedures; however, when such injuries occur, several factors play a role, including hypotension, abnormal positioning, vascular disease, diabetes mellitus, and the clamping of major vessels.94,169 In short, the addition of epinephrine to local anesthetic solutions has both advantages and disadvantages.170 In an animal study in which epinephrine and phenylephrine were administered, a significant reduction in dural blood flow but no reduction in spinal cord blood flow occured.171

Management

The management of postoperative neurologic sequelae requires the cooperation of the anesthesiologist, surgeon, and neurologist. Advice may also be required from the radiologist and neurosurgeon. Although it is easy to blame an adverse neurologic outcome on the presence of an epidural, it should be borne in mind that other factors can lead to demonstrable nerve injury, including undiagnosed preexisting neurologic disorders, ligation of nutrient spinal cord vessels during abdominal or thoracic surgery, injury to the femoral nerve during pelvic surgery, injury to the lateral cutaneous nerve of the thigh during retraction close to the inguinal ligament, and pressure on the fibular head leading to neurapraxia of the lateral popliteal nerve. If an adverse outcome occurs, the lesion should be localized by taking the patient's history and by performing a thorough neurologic examination. Some factors enabling differential diagnosis are as follows:

• Bilateral symptoms associated with pain should alert one to the possibility of neuraxial pathology.
• Injury at the nerve roots affects both posterior and anterior rami.
• Preservation of sensation over the paraspinous muscles suggests a more distal injury.
• Investigations should include blood cultures and coagulation studies.
• Immediate MRI is the standard for evaluating neuraxial lesions.
• EMG can be used to determine the site of injury and the degree of axonal loss, although it may take up to 3 weeks for changes to appear on the electromyogram. It may be useful to perform this immediately upon recognition of neural dysfunction to establish the possibility of a preexisting lesion.

Hematoma

Epidural hematoma after neuraxial anesthesia is a rare event. Bleeding from an epidural vein may occur on needle or catheter insertion, but is usually self-limiting. Neurologic symptoms and signs caused by an epidural hematoma are atypical in the presence of normal coagulation. The true incidence is unknown, but is estimated to occur in fewer than 1 in 150000 cases of neuraxial anesthesia.172 Vandermeulen et al173 reviewed 61 case reports between 1906 and 1994 and found that two-thirds of the cases had a hemostatic abnormality. Early diagnosis and intervention are essential to preventing any long-term adverse outcomes.

Prevention

In recent years, new anticoagulant and antiplatelet drugs have been introduced and have given rise to new challenges in the management of the anticoagulated patient undergoing neuraxial blockade. The American Society of Regional Anesthesia has released guidelines in response to this evolving shift in medical practice172; it is important to follow these guidelines to minimize the risk of hematoma.

Management

Similar to the peripheral nerve injury, the evaluation of neurologic injury after regional anesthesia is a vital part of adverse outcome management. Several issues should be considered:

• Back pain with lower-limb weakness and sensory deficit should alert the clinician to the presence of a central compressing lesion.
• Bowel and bladder incontinence can be an associated finding.
• Painless evolution of this complication has been reported, and early warning signs may be masked by the administration of local anesthetic via an epidural catheter and the presence of a urinary catheter.
• If MRI confirms the diagnosis, then rapid surgical intervention within 6 to 8 hours is recommended.
• Epidural catheters containing metal elements should be avoided while undergoing MRI, as it will generate artificial interference and inaccurate diagnoses and poses other additional risks.174 Such catheters should be removed if it is safe to do so; if catheter removal is unsafe, a CT scan should be considered instead of an MRI.

Infection

Epidural abscess formation, although rare, is a serious, potentially devastating complication. Kane's175 retrospective review of 50000 epidurals found no case of abscess formation, whereas Moen et al176 reported 12 cases of abscess formation from an estimated 250000 patients after epidural insertion. Nine of these patients had risk factors predisposing them to infection. Six of the patients received their epidural for analgesia after trauma, and of these, 5 were thoracic epidurals for chest trauma. The authors speculated that the overrepresentation of thoracic trauma patients might in part be a result of a lesser hygienic standard being observed, where placement likely occurred outside the "cleaner" environment of the operating suite.

Although the immunocompromised patient may carry a greater risk of developing infective complications with epidural use, extensive experience using epidural analgesia and anesthesia with patients who have HIV has countered early fears surrounding regional anesthesia in this population. Regional anesthesia is particularly beneficial for the HIV carrier population, as it eliminates delayed metabolism of systemic opioids caused by protease inhibitors.177 Patients with AIDS often have neurologic manifestations of their disease, and the prevalence of peripheral neuropathy increases as the disease progresses.178 Attention should be given to assessing preoperative neurologic status, as this allows the clinician to correctly attribute post-block neurologic sequelae to the true underlying cause.

Epidural abscess presentation can be variable, but the cardinal symptoms and signs involve back pain with localized tenderness and fever that often develop days after the puncture. Leukocytosis would be expected in the presence of an epidural-induced abscess and may occur several days or months after needle and catheter insertion. After the formation of an epidural abscess, the patient can develop progressive weakness and may develop paraplegia if untreated. Meningitis may develop if the patient has endured a lumbar puncture in this setting. The most common pathogen involved in abscess formation is Staphylococcus aureus; it should act as a standard for guiding antibiotic treatment until definitive culture results are available. As with an epidural hematoma, prompt surgical consultation is warranted for potential abscess drainage.

In pediatric patients, there is some concern regarding catheter infection with prolonged use of caudally placed catheters because of the proximity of the sacral hiatus to the anal region. Although studies have not found clinical evidence of greater infection rates with the caudal approach to catheter placement, increased bacterial colonization has been reported with this technique. Staphylococcus epidermidis was the predominant microorganism colonized on the skin and catheters of lumbar and caudal epidurals, and gram-negative bacteria were also found on tips of caudal catheters.179 Although the overall infection rate associated with caudal epidural catheters appears to be quite low, tunneling caudal catheters or simply fixing the catheter with occlusive dressing in an immediate cephalad direction has been recommended to reduce the risk of contamination by stool and urine.28,180

Prevention

Epidural abscesses can occur spontaneously (a reported incidence of 0.2 to 2 per 10 000 hospital admissions per year),181 and lumbar puncture has been safely performed in potentially bacteremic patients.182 Following is a list of guidelines for the prevention of epidural abscess formation:

• Sound aseptic technique, monitoring of the infection site, antibiotic prophylaxis, and bacterial filter use all contribute to a lower incidence of epidural space infections.
• Although both lidocaine and bupivacaine are bactericidal in high concentration, this property is likely not clinically significant at the concentrations used in practice.183
• The performance of neuraxial block should be avoided where local infection exists at the needle entry site.

Management

Following is a list of guidelines for the management of epidural abscess formation:

• Daily catheter site inspection is essential for the early identification of epidural abscess formation.
• Prompt removal of the catheter is essential when erythema and local discharge are present.
• Carefully assess any symptoms or signs of back pain.
• If any neural dysfunction occurs, a diagnosis must be immediately made in order to evaluate infective causes.
• Once a diagnosis of epidural abscess is made, a combination of medical (antibiotic) and surgical (incision and drainage) treatment may be needed.

Total Spinal Anesthesia

Total spinal anesthesia occurs when an excessive dose of local anesthetic is injected into the subarachnoid space; this is usually the result of an unintentional injection of a dose of local anesthetic intended for the epidural space. A high spinal may be seen when a small epidural dose or a large spinal dose of local anesthetic enters the subarachnoid space. Obstetric patients are particularly vulnerable because the engorged epidural venous plexus reduces spinal CSF volume and predisposes this population to cephalad local anesthetic spread. Total spinal anesthesia is rarely seen in nonobstetric cases, as observed by Dawkins, who reported an incidence of 0.2% of total spinal anesthesia in 48000 patients undergoing epidural anesthesia.56

Prevention

To prevent total spinal anesthesia from occurring, it is essential to use standard technique when aspirating, and an epidural test dose should be used. The subsequent use of small, incremental doses of local anesthetics may reduce the risk of this complication. The use of electrical stimulation is a useful and reliable real-time technique that can be used as an adjuvant tool for ruling out intrathecal placement before the administration of a potentially large test dose (Table 50-3).

Management

Total spinal anesthesia is a true medical emergency, as patients become profoundly hypotensive, apneic, and unconscious with remarkable pupillary dilation. Resuscitation with endotracheal intubation, mechanical ventilation, and vasopressor therapy is frequently required, and recovery may take between 30 minutes and 6 hours, depending on the agent used and the type of dose administered. Cerebrospinal fluid lavage via an epidural catheter has been used to successfully treat a total spinal in a 14-year-old child; in this case, the patient recovered within 30 minutes.184

With a high spinal, the patient may complain of numbness in the hands or may have difficulty breathing; if this occurs, the situation can usually be managed with reassurance, careful use of sedation, and treatment of hypotension. Respiratory function should be closely monitored with pulse oximetry, and measurements of adequate airflow should be made.185 The potency of sedative agents is increased in the presence of a high spinal,186-188 and one should be prepared to intervene in the event of significant respiratory compromise.

Subdural Injections of Local Anesthetic Drugs

The subdural space is a potential space between the dura and the arachnoid that extends from the level of the second sacral vertebra up to the floor of the third ventricle; the subdural space differs from the epidural space in that it is both extra- and intracranial. This space envelops the cranial and spinal nerves for a short distance and is widest in the cervical area. The incidence of subdural injections of local anesthetic drugs is reported to range from 0.1% to 0.8%,189 and this occurs more frequently after epidural injections.190 However, subdural injection may be an explanation for the occasional failed spinal anesthesia when pencil-point needles with side apertures are used. The design of a pencil-point needle with side apertures makes it possible for the opening to exist partially in both the subarachnoid and the subdural spaces.191 The diagnosis of a subdural catheter placement is best achieved using an injection of radiopaque dye. A typical radiologic pattern is pathognomic of subdural catheter placement (Fig. 50-9).

Prevention

Extra care should be exercised in patients who have had previous back surgery or a dural puncture at the same or adjoining interspace, as subdural injections are more likely to occur in these patients. Clinically, the subdural injection of local anesthetic drugs should be suspected when motor or sensory changes do not follow the expected pattern. Subdural injections result in a very slow onset of motor and sensory anesthesia and extensive and/or patchy sensory blocking.22 Patients may also complain of respiratory difficulties and may appear obtunded. The degree of cardiovascular depression may vary but hypotension is usually not severe; however, rapid onset of cardiovascular depression with concurrent loss of consciousness can result within 2 minutes, and cardiorespiratory arrest has been reported in the obstetric setting.192 Based on cases reported, the epidural stimulation test appears to be a potential diagnostic test providing information about the location of the needle or catheter in the subdural space.21,22

Management

The treatment of subdural injections of local anesthetics is predominantly supportive. Patients sometimes require intubation, ventilation, and sedation and usually recover within 6 hours of the injection.

Systemic and Local Toxicity

Unintentional intravascular catheter placement can go unrecognized and may lead to local anesthetic toxicity. There has been a dramatic decline in the incidence of systemic toxic reactions to local anesthetics after epidural anesthesia within the past 30 years. Dawkins56 reported a 0.2% incidence of toxicity in a retrospective analysis of 48292 cases of epidural anesthesia in 1969; this series included thoracic, lumbar, and sacral epidurals. More recently, Brown et al193 reported a 0.01% incidence of toxicity after lumbar epidural anesthesia in a retrospective study of 16 870 cases and a 0.69% incidence of toxicity after caudal epidural anesthesia in a series of 1295 cases. This 20-fold reduction in toxic reactions after epidural anesthesia during the past 30 years is explained, in part, by significant changes in regional anesthesia practice and by the influence of regional anesthesia societies in North America, Europe, Asia, Australia, and New Zealand. In the early 1980s, several deaths were reported in the United States after accidental intravascular injections of bupivacaine while performing epidural anesthesia. These deaths occurred as a result of cardiac toxicity that had not been previously reported with bupivacaine.194 Several deaths were also reported in the United Kingdom when bupivacaine was used for intravenous regional anesthesia.195 These tragedies led to a practice change in regional anesthesia. Single-injection epidural techniques commonly practiced 30 years ago have been replaced by continuous techniques involving injections of small incremental doses of local anesthetics; subsequently, "test dosing" has become a standard when using local anesthetics in regional anesthesia. Bupivacaine was implicated in 50% of toxic reactions reported by Auroy et al57 and in a large percentage of cases in Brown et al's study.193

Prevention/Management

To avoid potential local toxicity, local anesthetic free of preservatives in an appropriate concentration should be considered for use in the neuraxial space.

For epidural catheter anesthesia, using soft-tipped catheters (eg, metal-reinforced catheter) may reduce the introduction of the catheter into the vessel.196 The most important aspects of prevention were discussed in the local anesthetic section earlier. In summary, aspiration, test doses, and incremental dosing are vital to preventing local and systemic toxicity.

The epidural stimulation test has the potential to detect intravascular catheter placement and should not be overlooked.20

Postdural Puncture Headache

Postdural puncture headache (PDPH) is a widely discussed and published topic in regional anesthesia; it is also one of the most common complications of epidural and spinal anesthesia. Advances in needle design and gauge, as well as a better understanding of the physiologic mechanism of PDPH, have dramatically reduced the incidence of PDPH associated with spinal anesthesia, even in the obstetric population.197 However, the incidence of PDPH after epidural anesthesia in obstetric patients remains unchanged, with headaches ranging from 0 to 2.6% of cases.197

Prevention

Prevention of PDPH relies on the education of clinicians regarding the factors influencing the incidence of PDPH; such information is based on previous clinical case reports and studies. There is a strong link between onset of headache and needle gauge, age, sex, pregnancy, bevel design, and bevel orientation.

The dura consists of a mixture of elastic collagen and elastin fibers contained in a viscous intercellular ground substance196; it is primarily a longitudinally oriented structure, and its greatest tensile strength and stiffness exists in the longitudinal orientation. When a needle penetrates the dura, the size of the defect will be dependent on the number of elastin fibers cut, as well as by the tendency of those cut fibers to recoil in opposing directions, creating a crescent-shaped defect. As the gauge of the needle increases, more elastic fibers are cut. Fink197 examined the dura of elderly cadavers and found less viscoelastic material and more fibrous connective tissue. Young patients are at greatest risk of PDPH, as their greater dural elasticity maintains a patent defect compared with the less elastic dura of the elderly.

Norris et al200 demonstrated the importance of bevel orientation in relation to the incidence of PDPH after penetration of the dura with an epidural needle. Lybecker et al201 suggested that bevel orientation may be even more important than needle gauge and was unable to show any difference in PDPH when using 22- and 25-gauge needles, provided that the bevel was vertically oriented. Ready et al202 suggested that the incidence of PDPH is reduced when the needle is placed in an oblique direction. The arachnoid is closely adherent to the dura, and when a needle is advanced perpendicularly, the holes made by the bevel in the dura and arachnoid regions are directly in line with one another. When a needle is directed obliquely, the dural puncture does not line up with that in the arachnoid layer, thus obstructing CSF leakage.

Needle design has been implicated as a factor in the development of PDPH. Blunt-pointed needles (eg, Sprotte, Whitacre) are linked to a reduced incidence of PDPH as opposed to sharp cutting-point needles (eg, Quincke). Blunt needles rather than sharp needles are the tool of choice, particularly in patients with a high risk of developing PDPH (eg, adolescent and young adult patients). The most effective way to treat PDPH is to prevent this problem in the first place. The principal factor responsible for the development of PDPH is the size of the dural perforation. Thus smaller, blunt needles should be used for spinal anesthesia. On the other hand, the most commonly used epidural needle used is the 16- or 17-gauge Tuohy needle for continuous epidural anesthesia.

Treatment

After unintentional dural puncture with a Tuohy needle during epidural catheter placement, some authors suggest that the epidural catheter should be advanced through the puncture hole in an effort to reduce the incidence of PDPH. Intrathecal placement of the epidural catheters after accidental dural puncture in the obstetric setting is common practice in some centers.203 It is thought that the presence of the epidural catheter generates an inflammatory response, leading to early closure of the dural defect.204 Because the epidural catheter is in the intrathecal space in this circumstance, extreme caution should be exercised to treat this catheter as a spinal catheter to avoid possible neurologic complications and infection.197 Conservative measures, including bedrest and oral hydration, remain popular therapies for PDPH, despite no evidence to support them. Bedrest may postpone the occurrence of the headache, yet it does not prevent the onset.205 Obstetric patients should be encouraged to mobilize soon after delivery, so that PDPH, if present, can be diagnosed and treated while yet in the hospital.

Mild headaches can be treated with intravenous fluids, caffeine, and theophylline; methylxanthines may block cerebral adenosine receptors, leading to cerebral vasoconstriction, although this proposed mechanism is controversial.206 Camann et al207 have demonstrated the efficacy of caffeine in 40 postpartum patients. A single oral dose of caffeine is safe, less expensive than intravenous caffeine, and may offer temporary relief. The evidence for both the therapeutic and prophylactic use of caffeine has been called into question after a systematic review of the literature.206 Caffeine is a potent CNS stimulant and should be avoided in women who have pregnancy-induced hypertension, as it may lower the seizure threshold.208 When considering the usage of caffeine as treatment for mild headache, it is worth noting that the cerebral vasoconstrictive properties of caffeine are transient, and the headache may return after 48 hours.

Sumatriptan is a serotonin type 1-d receptor agonist and has been used for cluster headaches and migraine and has been suggested as a treatment of PDPH.209 Currently, it is not routinely recommended for use for the treatment of PDPH.210 Further analysis of frovatriptan is required before it can be considered a useful prophylactic agent.211

Cosyntropin, the synthetic form of adrenocorticotropic hormone, has been used to treat PDPH; this pharmaceutical is thought to work by stimulating CSF production and β-endorphin output.212

Epidural Blood Patch

The epidural blood patch (EBP) was introduced by Gormley in 1960 and is known to be the most effective treatment for PDPH. Gormley213 observed that patients who bled during myelography had a lower incidence of PDPH, and when he himself subsequently developed PDPH, he requested an injection of autologous blood into his epidural space with the positive result of the alleviation of his headache. In 1970, the report of DiGiovanni and Dunbar214 of the successful use of epidural blood patching in 41 of 45 patients led to its popularization. This form of treatment is indicated when conservative measures have failed and the headache is severe or is likely to extend the hospital stay. The success rate for a first epidural blood patch is 85%, rising to 98% after a second patch.

DiGiovanni et al215 suggested that an epidural blood patch acts as a gelatinous tamponade, and when injected, the blood generates sufficient pressure to lift the brain; this author has suggested that blood acts as a sealant, plugging the hole created by the needle, thus preventing further CSF leakage. Magnetic resonance images displaying the lumbar region after blood patching show a mass effect that compresses the thecal sac and conus. The blood spreads 3 to 5 spinal segments from the injection site and spreads mostly in the cephalad direction. This finding was confirmed by Szeinfeld et al216 using tagged red cells demonstrating extension of the blood 6 segments in the cephalad direction and 3 in the caudad direction after the injection of an average of 14.8 mL of autologous blood. The mass effect persists beyond 3 hours, and clot resolution occurs in 7 hours.217 Symptoms are frequently relieved within minutes of the procedure, and this response supports the counterpressure theory.

When performing an EBP, care should be taken to maintain a sterile field, and the epidural space should be identified in the usual manner. An assistant draws 15 to 20 mL of autologous blood aseptically. The administration of blood should be done at a rate of 1 mL/3 s. The end point of injection occurs when the patient complains of back, neck, or buttock pain as classically described by Szeinfeld et al.216 Much less blood is required for blood patches in the midthoracic region than in the lumbar region, usually in the order of 5 to 10 mL.

To ensure adequate healing, the patient should remain recumbent for 1 to 2 hours after a blood patch and may resume ambulation thereafter; the patient should refrain from any strenuous activity for several days.

Complications from EBPs are rare, but can be serious. Transient bradycardia, lumbovertebral syndrome, and facial palsy have all been reported.218-221 One case of cauda equina syndrome has been reported in a patient who was subjected to 6 blood patches; the patient made a full recovery after evacuation.222

EBP has been successfully performed in children. A caudal blood patch has been reported by Kowbel in a 4-year-old child who developed a subarachnoid cutaneous fistula after repeated lumbar punctures for chemotherapy.223 In this situation, the epidural blood patch was performed by passing an epidural catheter via the caudal canal and by injecting 8 mL of blood. A lumbar EBP has been performed in a 7-year-old child.224

Blood patches have been safely performed in HIV-positive patients. HIV crosses the blood–brain barrier and infects the CNS early in the clinical course. EBP is unlikely to introduce HIV into the CNS.203

Prophylactic Epidural Blood Patching

Prophylactic blood patches are controversial and have supporters and detractors.225,226 The effectiveness of using EBP as a prophylactic depends on the proximity of the catheter tip to the dural tear. Although blood patching is a relatively safe procedure, there are some risks associated with its use, and patients do not always get a headache after dural puncture, even with a large-gauge needle. Aldrete and Brown227 describe a case of intrathecal hematoma and arachnoiditis after prophylactic blood patching through a catheter. Meta-analysis of studies considering the prophylactic use of EBPs concluded that there are not enough trial participants to reliably draw conclusions.228

Alternatives to the Epidural Blood Patch

• Epidural saline treatment has been used for PDPH, but it is significantly less effective than EBP.228 Successful use of prolonged saline infusion has been reported in patients with failed EBP.229,230
• Fibrin glue, a pooled plasma product, has been used to treat CSF leak in cancer patients229 and in PDPH cases after spinal anesthesia where 2 EBPs had failed.230
• Dextran-40 has also been used to treat PDPH as it undergoes delayed absorption from the epidural space because of its high viscosity and molecular weight.231

Failure of Spinal/Epidural Anesthesia

Failure of neuraxial blockade is more common with epidural rather than spinal anesthesia. Thus this section focuses on discussing failed epidural anesthesia. Anesthesiologists recognize entry into the subarachnoid space by the tactile sensation produced and the visual element of CSF. On the other hand, anesthesiologists generally recognize entry into the epidural space by the tactile sensation produced when using the LOR technique and the ease of epidural catheter insertion. Thus entry into the epidural space is purely tactile in many cases, and the end point of entry is subject to more misinterpretation than that of spinal anesthesia. In epidural anesthesia, false LOR may occur, and quite often the only proof that the needle is correctly positioned is that a successful block occurs. False losses of resistance are more frequently encountered in obese patients in whom anatomy may be ill-defined. The less fibrous tissue planes in neonates limit tactile feedback.234 Moreover, Sharrock235 suggested that false loss of resistance may also occur in the elderly who have a high incidence of cyst formation within the interspinous ligaments.

Prevention

An important distinction that should be made during epidural anesthesia is that of complete failure versus a partial blockade/failure. The inability to pass a catheter into the epidural space frequently indicates that the needle is not in the epidural space. Catheters may become occluded with blood, or the catheter may kink, take a unilateral course, break, or become knotted, all of which can contribute to the complete failure of epidural anesthesia. The presence of a midline epidural band has been suggested236 and may explain why difficulty may be encountered when threading the catheter through the Tuohy needle. When epidural local anesthetic dosing for anesthesia approaches the maximum safe limit without noticeable analgesia, a failed epidural must be considered and should prompt the clinician to pursue an alternative course of anesthesia.

Careful matching of the dermatomal level of the catheter tip to that of the surgical site will yield greater success of epidural blockade. The epidural stimulation test has been used to verify accurate epidural tip placement,28 ensuring that the dermatomes involved in the surgical procedure are selectively blocked.

Management

Effectively managing partially working and/or failed epidurals is very important in patient satisfaction and safety, particular in obstetric patients. The partially working epidural is commonly encountered when undergoing anesthesia for cesarean sections; reported failure rates are in the order of 2% to 13.1%.237 A poorly functioning epidural or partially working spinal should be identified early before the decision to proceed to cesarean section is made. In an emergency situation, the anesthesiologist has a number of options available to rescue the situation; such options include converting to general anesthesia, supplemental epidural or caudal injections, and local infiltration anesthesia. Intraoperative discomfort and visceral pain may occur in up to 50% of cesarean patients.238 A block to the T4 level is considered optimal in most cesarean patients; however, debate exists concerning the best modality with which to test the upper level of the block. Loss of pinprick and cold sensation are popular testing options, but may have poor predictive value.239,240 The loss of touch is considered by some to best equate with surgical anesthesia.241

Surgical factors increasing the likelihood of intraoperative discomfort include exteriorization of the uterus and round ligament stretching, both of which exceed the analgesia provided during an apparently adequately dense nerve block. Subdiaphragmatic blood or amniotic fluid may cause back, chest, or shoulder discomfort.

Intervention by the clinician should involve direct communication with the patient. Pharmacologic management may be necessary depending on the level of distress. Intravenous ketamine in 10- to 20-mg increments and small doses of fentanyl or benzodiazepines are considered safe, although some advise waiting to administer these medications until the umbilical cord is clamped.237 Nitrous oxide has been used in the treatment of patients with breakthrough pain, but this treatment is controversial in obstetric anesthetics.242 If rescue efforts fail, general anesthesia should be considered, paying special attention to preoxygenation and potential airway difficulties.

For postoperative epidural analgesia, it is also important to confirm the working condition of epidural analgesia (sensory test, epidural stimulation test, low pain scores).

Hypotension

Hypotension is a common physiologic change associated with neuraxial blockade. Its presence predicts block success, but as a side effect, if left untreated or poorly managed, hypotension can lead to serious morbidity or death. Hypotension results from preganglionic sympathetic blockade that leads to a reduction in systemic vascular resistance (SVR) and cardiac output if the venous return is not maintained. SVR decreases as a result of a reduction in sympathetic tone, the extent of which is related to the number of spinal segments blocked. Cardiac output is altered by changes in heart rate and stroke volume. The reduction in stroke volume is a result of a fall in preload and contractility, which is load dependant. If the block involves the cardiac sympathetic nerve supply, bradycardia and reduced contractility can be expected.

High thoracic epidural anesthesia has the potential to block cardiac afferent and efferent fibers originating at the first to fifth thoracic levels. Interest has evolved concerning the potential positive effects of cardiac sympathetic blockade in patients with coronary artery disease: dilation of coronary vessels, reduced heart rate, and decreased myocardial oxygen demand.243 Clinically, improvement in cardiac function when using high thoracic epidural anesthesia is likely due to improved diastolic function of the left ventricle.242 Although measures of cardiac function can be improved with a high thoracic epidural, there does not appear to be a decrease in mortality or myocardial infarction incidence after coronary artery bypass grafting.244

Prevention

The effect of prophylactic administration of intravenous fluid, ephedrine, and methoxamine on cardiovascular responses to both epidural and combined epidural and general isoflurane anesthesia in 45 adult patients undergoing knee arthroplasty has been examined.245 In Wright and Fee's study,246 systolic blood pressure was significantly greater after ephedrine administration than after fluid preloading or methoxamine administration. An increase in plasma volume triggered by epidural-induced hypotension has been observed as a result of fluid movement from the interstitial to the intravascular space.244 A larger percentage of fluid administered is retained by hypotensive than normotensive patients,248 resulting in hemodilution. Holte et al's study249 showed that it was not the epidural that leads to changes in blood volume, but rather the infusion of fluid that effects blood volume. Hydroxyethyl starch and ephedrine have similar hemodynamic effects; ephedrine may be the preferred option for patients when excess fluid administration is undesirable. Fluid administration before induction of spinal and epidural analgesia usually can reduce the risk of hypotension. Patient position is crucial in preventing low cardiac output states in patients undergoing epidural anesthesia. If severe hypotension occurs during the course of epidural anesthesia, the most likely cause is inadequate venous return as a result of blood loss, an unfavorable patient position, or surgical obstruction.

Management

The first step in the management of hypotension is making sure that there is no interference with venous return. Place the patient in 5 degrees of Trendelenburg (and in the case of a pregnant patient, a leftward roll) to improve return. Subsequent treatment with fluid or pressor administration should be used to restore the systemic blood pressure to acceptable levels.

Significant changes in blood pressure are uncommon in pediatric patients after the proper administration of epidural analgesia. A high sympathetic single-shot caudal block to T6 caused no significant changes in heart rate, cardiac index, or blood pressure in children.250,251 Even when thoracic epidural blockade is combined with general anesthesia, cardiovascular stability is usually maintained in otherwise healthy pediatric patients.

Respiratory Complications

Several studies have examined high thoracic epidurals in both healthy people and in those with chronic obstructive airway disease. Peak expiratory flows, forced vital capacity, forced expiratory volume in 1 second, and maximum expiratory pressures are reduced252,253 in those suffering from this disorder. Kochi et al254 investigated the effect of high thoracic epidural anesthesia on the hypercapnic ventilatory response and ventilation pattern; duration of inspiration, rib cage excursion, and its contribution to tidal volume decreased significantly, whereas mean inspiratory flow rate and minute ventilation increased. Furthermore, end-tidal Pco2 and the tidal excursion of the abdomen remained unchanged, whereas hypercapnic ventilatory response decreased significantly. Lumbar and high thoracic region–induced epidurals do not interfere with the ventilatory response to hypoxemia.255 Gruber et al256 demonstrated the safety of thoracic epidural anesthesia with bupivacaine 0.25% in patients with severe chronic obstructive pulmonary disease. The potential for phrenic (C3 to C5) palsy is low with an epidural block, except during unintentional blockade after an interscalene brachial plexus block.257 Cervical epidural anesthesia has been used for upper limb, parathyroid, and carotid operations.258-260

Bonnet et al258 reported respiratory difficulties in 3 of 394 patient undergoing carotid endarterectomy using 15 mL of 0.5% bupivacaine or 0.37% to 0.40% bupivacaine plus fentanyl (50-100 μg). Many case series, although smaller in number than this study published by Bonnet, have not reported respiratory difficulties to be a significant problem.259,260

Capdevila et al261 reported that both 0.25% and 0.375% cervical epidural bupivacaine impaired diaphragmatic excursion, tidal volume, forced vital capacity, and hand grip strength in patients having postoperative hand rehabilitation and did not recommend the technique for this purpose.

The incidence of respiratory depression is closely associated with the use of neuraxial opioids. The rate of incidence of respiratory depression requiring intervention after conventional opioid dosing is approximately 1%.262

Prevention

To prevent respiratory complications, do the following:

• Avoid the use of high doses of opioids.
• Limit opioid dosages, especially in the intrathecal space.
• Avoid the concomitant use of parenteral opioids or sedatives.
• Avoid or limit doses in the patient with advanced age (>60 years of age), sleep apnea, and other coexisting diseases.
• Use hydrophilic drugs (eg, morphine) with caution.

Management

To manage respiratory complications, do the following:

• Treat mild respiratory depression with oxygen.
• If an infusion is used, then reduce the rate.
• Depending on the severity of respiratory complications, consider ventilatory support, the administration of narcotic antagonists, and the discontinuation of the opioid infusion.

Nausea and Pruritus

Nausea and pruritus are common side effects seen with the administration of neuraxial opioids. The reported incidence of postoperative nausea and vomiting (PONV) with opioid administration is 30% to 65%263-265 and for pruritus is 80%.266 Pruritus is thought to be multifactorial in nature and is speculated to operate via an "itch center" in the CNS via medullary dorsal horn activation and antagonism of inhibitory transmitters.267 Pruritus is a dose-dependant phenomenon; its onset involves possible mediators including C fibers in the skin, serotonin (5-HT3) receptors, prostaglandins, and micro-opioid receptors.268 The obstetric population seems to be at greater risk for developing pruritus.

PONV is a complex, multifactorial problem:

• Epidural administration of local anesthetics alone carries a low risk of causing the occurrence of PONV.269
• Factors such as surgery, age, and sex influence the reported incidence of PONV associated with epidural anesthesia.270
• Within 5 to 15 minutes of epidural administration, peak plasma opioid concentrations can reach levels similar to those seen after an intramuscular injection.271
• In patients receiving epidural morphine, there have been no differences in PONV onset or duration when different doses up to 5 mg were administered,272 whereas higher doses have been shown to lead to both an increase and a decrease in reported PONV.270
• Epidural fentanyl and meperidine do not appear to influence PONV in the same way that morphine does; as reported, fewer PONV cases have been documented with the use of fentanyl and meperidine after orthopedic surgery when compared with morphine.273

Prevention/Management

To prevent and/or manage PONV and pruritus, do the following:

• Reduce the dose administration and avoid neuraxial opioid administration. Doing so is effective in reducing the incidence of nausea and pruritus.
• Use antihistamines, opioid antagonists (naloxone and nalbuphine), propofol, nonsteroidal anti-inflammatory drugs (NSAIDs), and 5-HT3 receptor antagonists as both preventative and therapeutic measures.
• Dexamethasone has been shown to be a superior antiemetic for PONV-associated epidural morphine as compared with metoclopramide274 and 5-HT3 receptor antagonists.275
• Investigation into acupressure point P6 for the prevention of PONV has revealed inconsistent findings. There may be a reduction in nausea, but there is currently no evidence that it decreases postoperative vomiting.276

Postoperative Urinary Retention

Postoperative urinary retention (POUR) is common after major surgery and occurs in

• 20% to 68% of patients after abdominoperineal resection.
• 16% to 80% of patients after radical hysterectomy.
• 20% to 25% of patients after anterior resection.
• 10% to 20% of patients after proctocolectomy.277

POUR is a multifactorial condition and involves factors such as age, pain, bladder outlet obstruction, detrusor-inhibiting medication, pelvic autonomic nerve damage, and the inhibition of sympathetic reflexes.278 A single episode of bladder overdistension can result in significant POUR morbidity. Overfilling of the bladder can stretch and damage the detrusor muscle, leading to atony of the bladder wall, so that recovery of micturition may not occur when the bladder is emptied. On the other hand, the excessive use of an indwelling catheter can lead to urinary tract infection, urethral stricture, prolonged hospital stay, or death.279,280 Epidural use for postoperative pain management is usually reserved for patients undergoing major surgery, where urinary catheter placement may be performed for reasons other than anticipated postoperative urinary retention. Stenseth et al281 found an incidence of 42% for POUR in 1085 uncatheterized patients having epidural morphine for a variety of major operations. Epidural morphine relaxes the detrusor muscle with a corresponding increase in the maximal bladder capacity, whereas intramuscular and intravenous morphine have no effect on detrusor contraction. This is further supported by the fact that detrusor changes occur 15 to 30 minutes after epidural morphine administration and are reversed by intravenous naloxone, suggesting that spinal opioid receptors have an important role.278

Prevention/Management

Other epidural opioids such as fentanyl, meperidine, and methadone may also contribute to POUR, but contribute to a lesser degree than that observed with morphine.273,283 Diagnosis of POUR is best made with a portable ultrasound bladder scanner.278 There is a suggestion that use of the bladder scanner ought to be routine in the postanesthetic care unit for patients at high risk of retention.284,285 A threshold of 600 mL has been suggested as a diagnostic threshold.286 In addition to bladder catheterization, treatment options for opioid-mediated POUR may include intravenous naloxone administration.287 Nalbuphine is an opioid-mixed agonist–antagonist and has been used to restore detrusor function without reversing the analgesic effects of epidural morphine.288 Short-term (24 hours) urinary catheterization for major surgery involving morphine epidural analgesia may help prevent the morbidities associated with both POUR and longer-term epidural catheterization.289

Backache

Backache is a common complaint after epidural anesthesia, and its incidence ranges between 2% and 30% of patients.139,290 The causal relationship between epidural anesthesia and backache has been suggested by some studies291,292 and refuted by others.293,294 The etiology of backache is multifactorial in nature. Drug use, abnormal posture, muscle relaxation, and, in obstetric cases, exaggerated lumbar lordosis and the process of undergoing labor have been implicated as causes.205

In a retrospective study, MacArthur et al291 looked at 11 701 patients. Of the 1634 women who reported backache, 1132 (69%) had experienced it for more than 1 year. In this study, a significant association was found between backache and epidural anesthesia (relative risk: 1.8); 903 of 4766 women (18.9%) who had had epidural anesthesia reported this symptom, compared with 731 of the 6935 women (10.5%) who had not undergone epidural anesthesia. However, prospective data refute the findings of Macarthur, as noted by Breen,293 who interviewed 1185 women and found that of the 1042 (88%) for whom follow-up data were available, the incidence of postpartum back pain in those who received epidural anesthesia was equivalent to that of those who did not (44% vs 45%). Multiple logistic regressions revealed that postpartum back pain was associated with a history of back pain, younger age, and greater weight. Russell240 demonstrated that new-onset postpartum back pain was not associated with regional anesthesia. Among the women in Russell's study who received either 0.125% bupivacaine or 0.0625% bupivacaine, the incidence of new long-term back pain was 7.6% when compared with controls, and there was no difference found between the groups. Women who are seeking analgesia for labor should be reassured that back pain after epidural analgesia is minimal and is usually limited to the early postpartum period.

In 1987, 2-chloroprocaine was marketed by AstraZeneca in a new formulation (Nesacaine-MPF) involving disodium EDTA as a chelating agent for epidural and caudal use. Reports linking this new formulation with backache emerged128,294 and gave way to a possible explanation that the EDTA could cause hypocalcemic tetany of the paraspinous muscles. The drug now comes preservative-free and is prepared in dark-glass bottles to prevent light-induced disintegration. Despite the elimination of EDTA from this medication, backache continues to be reported with its use.296

Prevention/Management

Backache after epidural placement should not be ignored, as it can be a cardinal symptom of a space-occupying lesion within the spinal canal. Complications such as an epidural hematoma and abscess, although rare, can have catastrophic outcomes if unrecognized and untreated.

Intravenous regional anesthesia (IVRA) is one of the oldest techniques used in anesthesia practice today; Bier297 first described IVRA in 1908. The technique was not widely practiced for the first 50 years of its inception because it was not practical to perform a venous cut down on patients undergoing relatively minor extremity procedures. Holmes298 made a very simple adjustment to the technique in 1963 and demonstrated that Bier's technique could be performed using a percutaneous intravenous approach. IVRA is now widely used for minor upper- and lower-extremity procedures all over the world,298 as it is relatively easy to perform and has a very high success rate. There are a number of recognizable risks associated with this procedure, the most serious being compartment syndrome and local anesthetic toxicity. A primary limitation of the technique is the duration of tolerability of the tourniquet, which, without opioid supplementation, is about 45 minutes. The addition of ketamine in very low doses (0.1 mg/kg) greatly extends the time that patients can tolerate the tourniquet. Clonidine and Ketoralac have also been recommended for this purpose. Fortunately systemic toxicity and compartment syndrome are rare complications when IVRA is properly administered.

Compartment Syndrome

Compartment syndrome has been reported after IVRA of the upper and lower extremities.299,300 In 1 reported case, hypertonic saline was mistakenly used as a diluent for the local anesthetic. Long-bone fractures of the forearm or leg increase the risk of compartment syndrome, and IVRA should not be used in this circumstance. Severe ischemia of the upper extremity has been reported in at least 1 case after IVRA in an otherwise healthy young female, where the etiology was unclear.301 Table 50-7 lists some of the possible causes of compartment syndrome.

Venous thrombosis is a recognized complication of tourniquet application. There are anecdotal reports of subclavian steal syndrome after sudden LOR in the upper extremity, leading to transient cortical blindness.302

Local Anesthetic Toxicity

The risk of local anesthetic toxicity is quite low after IVRA. Auroy et al57 reported an incidence of 2.7 seizures per 10000 cases after IVRA in their study. Deaths have been reported when increased amounts of toxic cardiac drugs (eg, bupivacaine) have been used for IVRA.195 The main cause of this complication is faulty tourniquet technique.

Inadequate exsanguination before the inflation of the tourniquet allows the operator to exceed the tourniquet inflation pressure during the injection, thereby allowing local anesthetic solution to escape into the circulation. Interosseous escape of the local anesthetic can occur during injection. Accidental or premature deflation of the tourniquet (within 20 minutes) allows the local anesthetic to enter the circulation in toxic concentrations. When an excessive dose of local anesthetic is injected, toxicity may occur on release of the tourniquet, even when following appropriate recommendations.

Prevention

Intravenous regional anesthesia is easy to perform, yet one must pay particular attention to the details surrounding the procedure. Preventing complications begins with appropriate patient selection and surgical indication. Good intravenous access is important, as is proper exsanguination of the limb. Table 50-8 lists points to consider for proper patient and surgical selection.

Proper techniques for effective and safe IVRA that are essential to prevent complications are as follows:

• Place the tourniquet above the elbow (tourniquet application is less reliable in the distal portion of the extremity).
• Thorough exsanguinations should take place before injection of the local anesthetic. Appropriate doses of preservative-free local anesthetic should be administered.
• Lidocaine free of preservatives is one of the most frequently used local anesthetics for IVRA; the recommended dose is 3 mg/kg.
• Other drugs, such as prilocaine, have been used because of their favorable pharmacokinetic profile; however, some of these drugs, such as prilocaine, are no longer available in many countries.
• A preservative-free form of chloroprocaine was recently introduced in Europe and has many potential benefits, especially with regard to toxicity.303
• Ropivacaine has also been studied as a potential local anesthetic for IVRA and may offer better tourniquet tolerance and better postoperative analgesia compared with lidocaine.304
• Bupivacaine is contraindicated in IVRA.

Management

The management of local anesthetic toxicity is the same as that discussed in the previous section. The clinician must have dedicated intravenous access to inject other medications if required. Proper equipment and personnel must be available for emergency cardiopulmonary support (airway, breathing, circulation).

Like other types of nerve blocks, there are some serious risks associated with the use of ophthalmic regional anesthesia. Common complications include hemorrhage, brainstem anesthesia, and myotoxicity. These complications vary with the mode of regional opthalmic anesthesia. The clinician should be aware of other less frequently occurring complications of ophthalmic regional anesthesia that include globe ischemia, perforation of the globe, optic nerve and facial nerve damage, and elicitation of the oculocardiac reflex.

General Prevention

To prevent complication as a result of ophthalmic regional anesthesia, do the following:

• Inform the patient of the procedures and its likely outcomes during anesthesia. For instance, many patients experience visual alterations with sub-Tenon's anesthesia, which can be terrifying.
• Provide sedation to those patients who require it. Routine use of sedation is discouraged.305
• Consider only selected patients who are taking anticoagulant medication with a current international normalized ratio of less than 2 as a potential candidate for ophthalmic regional blocks.306
• Carefully weigh the benefits and risks of performing ophthalmic regional anesthesia in patients who have discontinued their anticoagulant medication.
• Consider alternative methods of applying ophthalmic anesthesia if there is risk of thrombotic complications after discontinuation of anticoagulant medication.307,308
• Let patients on antiplatelet therapy continue their medications if medically indicated.309
• Postpone surgery in severely hypertensive patients.
• Consider using small-gauge disposable needles (25 gauge), less than 31 mm in length.310,311
• Consider the site of anesthetic injection. Vascular structures are larger in the apex of the orbit, and also the upper nasal area is particularly vascular. Areas with increased vascular architecture should be avoided to prevent complications during the induction of ophthalmic regional anesthesia.

General Management

As with other nerve blocks, once damage has occurred as a result of ophthalmic regional anesthesia, it is difficult to reverse. If complications do occur, supportive measures are recommended for patient care.

Retrobulbar Hemorrhage

The reported risk of retrobulbar hemorrhage varies substantially in anesthetic literature.312 In one of the largest reported series, Hamilton reported an incidence of 0.44% hemorrhages in 12 000 ophthalmic regional anesthesia cases.313 The severity of retrobulbar hemorrhage varies depending on the origin of the bleeding. Arterial bleeding is the most dangerous complication of retrobulbar injections because tamponade can occur, which leads to ischemia of the globe. In this situation, lateral canthotomy may be required to relieve the pressure. The site of injection is also important to consider in avoiding hemorrhage.

Subconjunctival Hemorrhage

The risk of subconjunctival hemorrhage in sub-Tenon's anesthesia has been reported as high as 100%, and as low as 19%.316,317 Its incidence is influenced by the type of cannula used (anterior cannulae causing more hemorrhage than others) and patient factors such as age as well as steroid, antiplatelet agent, and NSAID use. This complication can be prevented with careful dissection as well as the use of epinephrine locally.318 Some advocate for the use of bipolar cautery, but this remains controversial.305 Fortunately, this complication tends to be self-limiting, and ocular compression is often enough for treatment.

Brainstem Anesthesia

When local anesthetic spreads directly into the brain from the orbit, brainstem anesthesia occurs. The incidence of brainstem anesthesia is 1 case for every 350 to 1500 ophthalmic cases,310 and symptoms may appear within 2 minutes of the injection. Maximum effects usually occur within 20 minutes, and recovery occurs in 2 to 3 hours. Symptoms and signs can vary greatly and include those listed in Table 50-9.

Globe Perforation

Blindly inserting a needle into the orbit is associated with the risk of globe perforation. The site of injection and the axial length of the globe must be carefully considered before needle insertion to prevent damaging the globe. In 1 reported series, there were no globe perforations in 2000 cases of ophthalmic regional anesthesia,319 whereas another study reported an incidence of 1 case of globe perforation per 12 000 ophthalmic cases.313

Prevention

The risk of globe perforation can be reduced by presurgically assessing the axial length of the patient's eye. Patients susceptible to perforation of the globe include those with elongated globes (>26 mm), which occur in myopic patients; in those with retinal detachment; and in those who require refractive surgery. Myopic patients with staphyloma are particularly vulnerable to globe perforation.320

The clinician should attempt to visualize in the "mind's eye" the equator of the globe and to avoid repositioning the needle until it is located past the equator. All needles should be directed tangentially with the bevel facing the globe. Pain or resistance to needle advancement is a warning sign of perforation of the sclera. Some experts suggest aiming the needle midway between the inferior and lateral rectus muscles to allow a clear point of entry to the intraconal space. The inferior rectus muscle should be carefully avoided to prevent diplopia.

Management

The key to successful management of globe perforation is early diagnosis and treatment. The patient may report paresthesia at the time of needle insertion. Funduscopic examination by an ophthalmologist may confirm the diagnosis of globe perforation. Depending on the severity of damage, globe perforation can be managed by laser photocoagulation therapy, cryotherapy, and other prompt surgical procedures. Because the appropriate management of globe perforation is complex, careful consultation should take place with the ophthalmologist.

Myotoxicity

Myotoxic effects of local anesthetic drugs were discussed in this chapter's section on Toxic Effects of Local Anesthetics on the Nerves and Surrounding Structures. Typically, diplopia and ptosis can occur for up to 48 hours when using long-acting local anesthetics. However, direct injection of these drugs into the highly sensitive eye muscles can permanently damage them. The inferior rectus muscle appears to be particularly vulnerable to injury.139,321

Miscellaneous Complications Resulting from Ophthalmic Regional Anesthesia

Globe ischemia, optic nerve, and facial nerve damage and oculocardiac reflex are less frequent complications of ophthalmic regional anesthesia.

This chapter provides a comprehensive review of the management of complications after regional anesthesia.

• Although the overall incidence of complications after brachial plexus anesthesia is low, the proportion of complications after supraclavicular methods is higher than that reported after axillary blocks (ie, the incidence of seizure activity after interscalene or subclavian perivascular methods is at least 6 times that of the axillary approach).
• Many of the complications of supraclavicular methods have an impact on pulmonary function; consequently, supraclavicular methods should be avoided in patients with significant pulmonary dysfunction. In a sense, there should always be a clear-cut indication for selecting supraclavicular methods because of the risk of pulmonary complications.
• Local anesthetic drugs must be injected slowly and incrementally, and patients must be observed carefully for signs of local anesthetic toxicity. The addition of epinephrine to local anesthetics is a useful marker for detecting accidental intravascular injections.
• Persistent pain on injection of local anesthetics is a potential sign of intraneural injection.
• Large-gauge and long needles should be avoided, and brachial blocks should be avoided in comatose adult patients.

Finally, brachial plexus anesthesia is not yet an exact science. Ultrasound-guided needle insertion is renewing interest in regional anesthesia as a result of this advance. We can expect success rates close to 100% in the near future. However, complications will still occur.

Primary factors for consideration with respect to adverse outcomes during regional anesthesia are as follows:

• Patients must be selected carefully.
• Patients must provide informed consent.
• The safest approach must be selected.
• The appropriate dose of local anesthetic should be given.
• Patient discomfort must be minimized.

Drug delivery via neuraxial block is an effective method for providing analgesia for lower-extremity, abdominal, and thoracic procedures, as well as for managing labor pain. In the perioperative setting, epidural analgesia has numerous benefits in addition to providing effective pain relief. These benefits include earlier ambulation, rapid weaning from mechanical ventilation, reduced time spent in a catabolic state, and lowered circulating stress hormone levels.316 Although serious complications are uncommon, patients should be informed about common side effects, such as urinary retention and pruritus, and should be counseled concerning the risks of major neurologic complication such as paralysis. Newer technologies such as sonography and the stimulating epidural catheter may increase the safety and ease of catheter placement. For the individual patient, the risks and benefits of epidural analgesia should be carefully considered, and prevention of patient injury is the highest priority.

• Atlee JL. Complications in Anesthesia. 2nd ed. Philadelphia, PA: Elsevier/Saunders; 2007.
• Horlocker TT, Wedel DJ, Benzon H, et al. Regional anesthesia in the anticoagulated patient: defining the risks (the second ASRA Consensus Conference on Neuraxial Anesthesia and Anticoagulation). Reg Anesth Pain Med. 2003;28:172-197.
• Neal JM, Bernards CM, Butterworth JFt, et al. ASRA practice advisory on local anesthetic systemic toxicity. Reg Anesth Pain Med. 2010;35:152-161.
• Neal JM, Bernards CM, Hadzic A, et al. ASRA Practice advisory on neurologic complications in regional anesthesia and pain medicine. Reg Anesth Pain Med. 2008;33:404-415.
• Neal JM, Brull R, Chan VW, et al. The ASRA evidence-based medicine assessment of ultrasound-guided regional anesthesia and pain medicine: executive summary. Reg Anesth Pain Med. 2010;35:S1-S9.
• Tsui BC, Finucane B. Epidural stimulator catheter. Tech Reg Anesth Pain Manage. 2002;9:150-154.