Anesthetics, local

Children

Neonates and infants absorb local anesthetics more rapidly after topical application to the airways, and peak plasma concentrations can be reached within 1 minute of application. In the first few months of life they have a larger volume of distribution, reduced hepatic clearance, and lower concentrations of albumin and alpha1-acid glycoprotein [ ].

Respiratory

Laryngospasm is a serious event that results in partial or complete upper airway obstruction. It can occur after airway anesthesia by local anesthetic spray.

A 54-year-old patient scheduled for flexible fiberoptic bronchoscopy, following a lung transplantation 18 months before, had intravenous induction with propofol [ ]. The vocal cords and vocal folds adducted immediately after a rapid injection of 2 ml of 2% lidocaine via a bronchoscope injection port. Ventilation ceased and the end-tidal carbon dioxide concentration fell to zero. There was spontaneous recovery after 40 seconds and ventilation resumed.

The authors postulated that direct application of a drug on the vocal cords has the potential to induce laryngospasm, although this has never been described before in clinical practice.

There has been a report of total airway obstruction after topical anesthesia of the larynx before fiberoptic intubation [ ].

A 69-year-old man with a neck cancer had inspiratory and expiratory stridor, and it was decided to perform an awake fiberoptic intubation. Four minutes after topicalization of the larynx using 7.5 ml of lidocaine 2% he had total airway obstruction. An attempt at cricothyroidotomy failed, but oral fiberoptic intubation finally succeeded. The vocal cords were abducted and not swollen.

The authors speculated that either laryngospasm or depression of laryngeal muscle tone due to the use of lidocaine could have caused sudden obstruction.

Nervous system

Seizures can occur after airway anesthesia.

• A 70-year-old man was given lidocaine 1200 mg to anesthetize the airway before bronchoscopy and 5 minutes later had a tonic–clonic seizure lasting 2 minutes before self termination [ ]. There were no long-term harm sequelae. The lidocaine concentration 30 minutes later was 33 μmol/l and may have been as high as 40 μmol/l during the procedure.

Lidocaine is potentially toxic at concentrations over 30 μmol/l. The authors stressed that local anesthetics should be used sparingly in airway anesthesia.

Transient cerebellar ataxia a few minutes after the topical use of lidocaine on mucosal surfaces has been described in two patients, a 58-year-old man who had lidocaine 10% spray for bronchoscopy and a 66-year-old woman who received lidocaine 2% orally for transesophageal echocardiography [ ]. In neither case was another cause of cerebellar ataxia identified. The second patient had previously had a similar reaction to lidocaine.

Sensory systems

Permanent anosmia after topical nasal anesthesia with lidocaine 4% has been described.

• A 62-year-old man had fiberoptic endoscopy with lidocaine 4% spray and 10 minutes later complained of anosmia [ ]. Computed tomography ruled out tumor, infection, and obstruction.

The authors postulated, in the absence of other obvious causes, that lidocaine had caused mitochondrial dysfunction, with activation of apoptotic pathways. They concluded that endoscopic topical local anesthesia should be done with the subject sitting and the head upwards to reduce contact of the anesthetic with the olfactory cleft.

Cardiovascular

Cardiovascular complications can arise from unintended stellate ganglion block [ ].

• A 67-year-old man had an axillary plexus block for a right palmar fasciectomy with mepivacaine 850 mg and adrenaline 225 micrograms. Twenty minutes later he became agitated and confused and an electrocardiogram showed fast atrial fibrillation. Rapid systemic absorption of the combination of high-dose mepivacaine and adrenaline in a patient who was also taking amiodarone, sotalol, captopril, and amiloride for pre-existing cardiac disease was felt to be responsible [ ].

Pulmonary embolism has been attributed to brachial plexus block.

• A 43-year-old man with end-stage renal disease became acutely hypoxic after an interscalene brachial plexus block with 35 ml of 1.5% mepivacaine for primary placement of an arteriovenous fistula in the left arm [ ]. He had been undergoing hemodialysis for 1 month using subclavian and internal jugular vascular catheters for temporary access. Immediately after an apparently straightforward block, his oxygen saturation fell from 99 to 85%, he complained of chest pain and shortness of breath, and he developed hemoptysis. A CT scan suggested acute pulmonary embolism.

The authors proposed that manipulations and vasodilatation related to the interscalene block may have facilitated the dislodgement of a pre-existing thrombus in the arm.

• A 34-year-old man undergoing acromioplasty of the right shoulder had a sudden cardiac arrest after an interscalene brachial plexus block with a mixture of ropivacaine 150 mg and lidocaine 360 mg [ ]. After successful resuscitation, severe hypotension persisted, necessitating the use of an adrenaline infusion. The patient developed pulmonary edema and was mechanically ventilated for 22 hours. He eventually made a good recovery.

A similar report with the use of a combination of lidocaine and levobupivacaine has been published [ ]. Tachycardia was the only cardiovascular symptom, while seizures were easily treatable. Both reports are in line with the improved cardiovascular safety reported with enantiomer-specific local anesthetics as discussed below.

Transient vascular insufficiency has been reported after axillary brachial plexus block [ ].

In a 3-year-old child, an axillary plexus block using 7 ml of bupivacaine 0.5% and 3 ml of lidocaine 2% with adrenaline 1:200 000 was established to allow re-implantation of an amputated thumb. After the injection, the hand became pale and no pulses were palpable; 15 minutes later the color and pulses returned.

The author noted that this is a rare event, with only one previous published report, and proposed that several mechanisms may have been causative: intra-arterial injection of adrenaline or local anesthetic, mechanical obstruction from subintimal injection into the arterial wall, severe vasospasm, and a pressure effect on the axillary sheath.

Respiratory

Large volumes (30–40 ml) of local anesthetics for interscalene block cause hemidiaphragmatic paresis in nearly all patients. An interscalene brachial plexus block in 11 volunteers using 10 ml of either 0.25% bupivacaine or 0.5% bupivacaine, both with adrenaline 1:200 000, resulted in significant impairment of lung function (forced vital capacity fell by 75% and FEV1 by 78%) and in hemidiaphragmatic excursion in those given 0.5% bupivacaine, but not 0.25% bupivacaine [ ]. The authors suggested that 10 ml of 0.25% bupivacaine provides adequate anesthesia, with only occasional interference with respiratory function.

However, reducing the volume of local anesthetic (1.5% mepivacaine) from 40 to 20 ml, and applying proximal digital pressure, did not reduce the incidence or intensity of diaphragmatic paralysis during interscalene block in 20 patients, in whom arterial oxygen saturation fell significantly [ ].

• A 55-year-old man with newly diagnosed non-small-cell lung cancer developed difficulty in breathing, cyanosis, agitation, and confusion, 10 minutes after interscalene supplementation of an axillary nerve block with only 3 ml of 2% mepivacaine with adrenaline [ ]. He was anesthetized, intubated, and ventilated. Surgery proceeded and postoperative radiographic examination of the lungs showed ipsilateral elevation of the diaphragm with reduced respiratory excursion. Phrenic nerve block after the interscalene injection was the postulated cause of the deterioration in respiratory function. He was successfully extubated at the end of the procedure.

Pneumothorax has occasionally been observed [ ]. The axillary technique is recommended to prevent this complication [ ].

Phrenic nerve palsy, resulting in paralysis of the ipsilateral hemidiaphragm, can rarely cause severe respiratory compromise, depending on pre-existing lung dysfunction. In unpremedicated patients who underwent supraclavicular brachial plexus block for upper limb surgery, blocks were performed using a peripheral nerve stimulator and 0.5 ml/kg of bupivacaine 0.375% [ ]. Spirometric and ultrasonographic assessments of diaphragmatic function were made at intervals. Of 30 patients, 15 had complete paralysis of the hemidiaphragm, 5 had reduced diaphragmatic movement, and 10 had no change. Those with complete paralysis all had significant reductions in pulmonary function and those with reduced or normal movement had minimal changes. Only one of the patients had respiratory symptoms and the oxygen saturation remained unchanged. This may not be the case, however, in patients with significant pre-existing respiratory disease or in obese people; the authors therefore suggested caution in choosing this approach as a safer alternative to general anesthesia in such individuals.

Phrenic nerve paresis is an extremely common adverse effect of interscalene block, with rates of up to 100%. In contrast to this, a recent study has shown a rate of only 13% (8/60 patients) [ ]. In this comparative study of high-volume/low-concentration versus low-volume/high-concentration local anesthetic for interscalene block, all patients received the same total dose of lidocaine 300 mg, bupivacaine 75 mg and adrenaline 150 micrograms, in either 60 or 30 ml. There were no complications in the high-volume/low-concentration group, compared with a 27% rate of phrenic nerve paresis and a 66% rate of bradycardia and hypotension in the low-volume/high-concentration group.

Persistent hemidiaphragmatic paralysis after interscalene brachial plexus block has been described again, this time with a mixture of bupivacaine 0.5% and lidocaine 1% [ ]. The patient developed long-term dyspnea with significantly impaired pulmonary function.

Two cases of respiratory compromise after infraclavicular brachial plexus blockade have been described [ ].

• An 84-year-old woman weighing 74 kg had a past history of hypertension, emphysema, and ischemic heart disease. She had an infraclavicular brachial plexus block with 40 ml (400 mg) of prilocaine 1% and 10 ml (75 mg) of ropivacaine 0.75%, and 20 minutes later developed difficulty in breathing and became desaturated. She had received midazolam 2 mg before the block.

• A 47-year-old woman with a history of hypertension, gastric reflux, and obesity was premedicated with oxazepam 10 mg and had an infraclavicular brachial plexus block with the same doses of ropivacaine and prilocaine as in the first case; 10 minutes later she developed dyspnea and became desaturated.

Each patient’s symptoms settled with supplementary oxygen, and surgery proceeded uneventfully. In both instances a chest X-ray showed a raised hemidiaphragm on the side of the block, but pneumothorax was excluded. The respiratory compromise was probably caused by paresis of the ipsilateral diaphragm due to blockade of the phrenic nerve, which is likely to occur after an infraclavicular plexus block but is well tolerated in most patients. Dyspnea in these two patients may have resulted from several factors. Both had been lightly sedated with benzodiazepines (although both were alert and cooperative, so this probably had a minimal contribution). The first had emphysema, which may have been an important factor; in such patients diaphragmatic function is important for sufficient gas exchange and a 50% loss of function can result in significant impairment. The second woman was obese, and obesity is associated with a reduction in functional residual capacity and respiratory function, so she may have had reduced respiratory reserve. The authors suggested that in patients with reduced pulmonary reserve, infraclavicular brachial plexus blockade should be avoided and an axillary approach considered. In addition they speculated that a smaller volume of local anesthetic may reduce the risk of phrenic nerve blockade.

Ear, nose, throat

Vocal cord paralysis can occur when local anesthesia is used after previous damage.

• A 71-year-old patient with unrecognized pre-existing left vocal cord paralysis developed severe stridor and airway compromise after right-sided subclavian plexus block [ ]. The paralysis was the consequence of partial glossectomy and neck dissection 18 months earlier for squamous cell carcinoma of the tongue.

The authors recommended evaluation of vocal cord function before brachial plexus block in patients with previous surgery or radiotherapy to the neck.

Nervous system

Neurological injury after peripheral blockade has an incidence of less than 1%. However, it has been suggested that for axillary nerve blocks, neurological damage is more likely if paresthesia is the endpoint for location of the nerve sheath, in contrast to the transarterial method. This is probably due to the increased likelihood of direct damage from a needle, intraneural injection of local anesthetic, or toxicity of the local anesthetic to the nerve [ ]. However, published results on this issue remain contradictory [ ].

Ropivacaine is less toxic than bupivacaine. However, there have been reports of brachial plexus blockade after ropivacaine, associated with unusual symptoms of nervous system toxicity; none of the patients recalled the events and there were no subsequent sequelae [ ].

• A 46-year-old man received an axillary nerve block using 40 ml of 0.5% ropivacaine with 1:200 000 adrenaline and 45 seconds later developed a sinus tachycardia and started screaming, appearing terrified. He struck out violently with all limbs and sat upright, attempting to leave the bed. The pulse oximeter reading (SpO2) fell to 90% and his symptoms were interpreted as a seizure and treated successfully with 100% oxygen, sodium thiopental, and intubation.

• A 60-year-old woman received an interscalene block using 30 ml of 0.5% ropivacaine with 1:200 000 adrenaline. Immediately after the injection, she sat up and began screaming in a loud high-pitched voice, appearing terrified and enraged. She then attempted to get off the stretcher in an uncoordinated manner and became unresponsive to verbal commands. She had a sinus tachycardia and hypotension. Treatment with 100% oxygen and propofol was effective.

• A 76-year-old woman received an interscalene block using 20 ml of 0.75% ropivacaine with 1:400 000 adrenaline. At the end of the injection, she sat up and appeared extremely terrified; she screamed twice, fell back on the stretcher, and began moving the unblocked arm and both legs in clonic movements, remaining unresponsive to verbal command. She had a sinus tachycardia and hypertension (205/70 mmHg). The seizure abated with thiopental.

The authors suggested that these signs of anxiety, vocalization, and agitation may have been due to the administration of ropivacaine formulated exclusively as the S(–) enantiomer, which has a spectrum of nervous system and cardiovascular toxicity different from the racemic mixture.

Reverse arterial flow can cause nervous system toxicity, even during peripheral regional blocks with only small volumes of local anesthetic [ ].

• A 47-year-old woman received an axillary brachial plexus block with 3 ml of 1% lidocaine after negative aspiration. She became dysphoric 30 seconds later, with muscle twitching in the face and distal arms, became unresponsive, and required ventilation.

During a study of 104 adults to compare the efficacy and safety of 40 ml of 0.75% ropivacaine (300 mg) and 40 ml of 0.5% bupivacaine (200 mg) for axillary plexus block, significantly more patients reported postoperative dizziness in the ropivacaine group (5 versus 0) [ ]. However, this occurred 4–5 hours after the injection in two patients and the day after in the other three, and was therefore unlikely to have been due to high serum concentrations. One patient developed dizziness, dysarthria, and unconsciousness, with convulsions shortly after an injection of ropivacaine, indicating an intravenous injection.

In some cases adjuvants should be considered as well as the local anesthetic after a toxic reaction [ ].

• A 52-year-old woman received an axillary plexus block with 20 ml of 1% ropivacaine, clonidine 70 micrograms, and 15 ml of 1% mepivacaine with 1:400 000 adrenaline. Generalized tonic-clonic seizure activity developed, even though careful incremental aspiration was performed. She was still comatose 90 minutes later, but this was reversed by intravenous naloxone.

The authors suggested that clonidine could have been responsible for the maintenance of her unconscious state.

Axillary blockade using high-dose mepivacaine with adrenaline was performed in 50 patients, each of whom received 850 mg of mepivacaine; two patients had symptoms of toxicity associated with this combination (euphoria, dizziness, and tinnitus) 13 and 15 minutes after the procedure with doses of 14.1 and 16.4 mg/kg of mepivacaine respectively [ ]. One patient who received 10.9 mg/kg developed hypertension and atrial fibrillation, became agitated, and lost consciousness 12 minutes after the block was performed, and required beta-blockade and midazolam before waking up 15 minutes later. Another received 6.5 mg/kg, became light-headed, agitated, and hypertensive, and reported whole body numbness 18 minutes later, with resolution of symptoms after 10 minutes with beta-blockade. The author thought that adrenaline had probably been responsible for the reaction in the first patient. As high-dose mepivacaine did not greatly improve the quality of the block and can obviously produce serious systemic reactions, it would be prudent to limit the dose to under 10 mg/kg.

Horner’s syndrome is a well-recognized complication of interscalene brachial plexus block, stellate ganglion block, and occasionally epidural blockade. It occurs when the local anesthetic reaches the cervical sympathetic trunk and is usually transient. However, persistent Horner’s syndrome is a rare complication, and may represent traumatic interruption of the cervical sympathetic chain. Cases of prolonged Horner’s syndrome related to prevertebral hematoma formation at the site of continuous interscalene blockade have been described [ ].

• A 48-year-old obese woman had a 22G interscalene catheter inserted under local anesthesia via a short-bevel stimulating needle. Anesthesia was achieved using 0.6% ropivacaine 40 ml followed by an infusion of ropivacaine 0.2% for effective analgesia. On day 3, she reported blurred vision and a painful neck swelling. She had developed a hematoma around the catheter insertion site (confirmed by ultrasound) and had an ipsilateral Horner’s syndrome including myosis, ptosis, enophthalmos, ipsilateral anhidrosis, and conjunctival hyperemia.

• An interscalene catheter was inserted in an awake 20-year-old woman for analgesia after shoulder surgery. Analgesia was achieved with ropivacaine 0.2% as a 30 ml bolus followed by an infusion of the same solution. One day later she had visual disturbance and neck swelling due to a hematoma between the prevertebral and scalene muscles.

Neither patient was taking NSAIDs, aspirin, or anticoagulants. Catheters were removed immediately on diagnosis of hematoma formation. There was no neurological or sympathetic fiber damage to the upper limb in either patient, as tested by electroneuromyography and sympathetic skin response. Remission in both cases occurred within 1 year. There has been one previous report of prolonged Horner’s syndrome in the absence of any obvious technical complication [ ]. Further studies into the use of interscalene catheters are needed to assess their propensity to cause this rare complication.

Late Horner’s syndrome has been reported during postoperative continuous infraclavicular brachial plexus analgesia [ ].

• A 33-year-old healthy woman (55 kg, 170 cm) was given analgesia via a continuous infraclavicular brachial plexus catheter after osteosynthesis of a complex fracture of the elbow. On the second postoperative night, about 50 hours after catheter placement, she developed nasal congestion. She had classical Horner’s syndrome and the perineural infusion was stopped (568 ml of 0.2% ropivacaine in all). The Horner ’s syndrome vanished about 2 hours later.

The authors assumed that the delayed presentation had been due either to accumulation of solution during the infusion or to atypical proximal migration of the solution to the supraclavicular paravertebral area.

In another rather unusual case, a patient developed motor and sensory blockade in all four limbs, other than the ipsilateral arm, after an axillary brachial plexus block [ ]. Epidural blockade was suspected after ruling out nervous system events such as a stroke.

In 60 patients receiving patient-controlled interscalene analgesia with either ropivacaine 0.2% or bupivacaine 0.15%, there was a significant reduction in hand motor function and an increased incidence of paresthesia in the bupivacaine group, with no difference in pain scores [ ]. This finding contrasts with that in a comparison of epidural bupivacaine or ropivacaine, in which there was no difference in motor function between the two groups [ ].

Inadvertent injection into the subarachnoid space, occasionally causing cerebral or neurological problems, is a life-threatening complication of brachial plexus anesthesia. It can also cause postdural puncture headache [ ].

Interscalene block can cause paralysis of the arm.

• A 33-year-old woman received combined regional and general anesthesia for a shoulder repair [ ]. Preoperatively an interscalene catheter was placed uneventfully. The next day, she had almost complete paralysis of the arm with hypesthesia of dermatomes C5–7. The symptoms persisted and 4.5 months later, during surgical exploration of the brachial plexus, electrical stimulation of the three trunks was possible and there were electrophysiological signs of recovery. Despite extensive neurophysiological tests a clear cause could not be established and there was no improvement at 2 years.

In 133 patients who had interscalene block for elective shoulder surgery (mepivacaine 1.5% alone or combined with bupivacaine 0.5–0.75%, all via paresthesia technique), successful surgical anesthesia was achieved in 98% [ ] There was one major perioperative complication, a seizure within 5 minutes of the block. Two patients developed transient postoperative apraxias and 37 (28%) had neck pain and bruising.

Nerve damage after regional anesthesia has many causes, including stretching, compression, ischemia, surgical trauma, and local anesthetic toxicity. Puncture by the block needle and intraneural injection of local anesthetic are thought to be major risk factors that lead to nerve injury. However, the results of a prospective study in 26 patients having ultrasound guided axillary plexus block have challenged these widely held beliefs [ ]. There was ultrasound evidence that 22 of the patients had puncture of at least one nerve, and 21 had intraneural injection into at least one nerve. Despite this, there was no evidence of damage to these nerves, either subjectively or on sensory and motor testing up to 6 months postoperatively. An alternative explanation for the nerve was injection into a fascicle rather than into the stroma of the nerve (i.e. piercing the perineurium causes injury while breaching the epineurium does not appear to) [ ]. Sharp needles, high injection pressures, and dense fascicles within the nerve all increase the risk.

Sensory systems

• An intolerable metallic taste appeared and disappeared in a 48-year-old woman within hours of infusion of bupivacaine via an axillary catheter, and its severity changed with the rate of infusion [ ]. The mechanism was postulated to be through sodium channels or taste bud disturbances.

Psychological, psychiatric

• A 59-year-old woman, grade ASA I, had psychiatric effects associated with local anesthetic toxicity after receiving bupivacaine 50 mg and mepivacaine 75 mg for an axillary plexus block. She complained of dizziness and a “near death experience” [ ].

Hematologic

Methemoglobinemia has been reported in a woman who received a combination of local anesthetics [ ].

• A 60-year-old woman with medical problems including severe coronary vascular disease and anemia, taking multiple medications, including isosorbide dinitrate, received axillary plexus blockade with bupivacaine 150 mg + 10 ml of 1% lidocaine injected into the operative field; 90 minutes later her SpO2 fell to 85–89% on oxygen 10 l/minute. She became drowsy, disoriented, and tachypneic, and an arterial blood gas showed a metabolic acidosis and a methemoglobin concentration of 6.4%. Her mental status improved 10 minutes after methylthioninium chloride and sodium bicarbonate; her SpO2 rose to 96% on air, her methemoglobin concentration fell to 1.6%, and her acidosis partly resolved.

The authors assumed that displacement of lidocaine from protein binding by bupivacaine, in combination with metabolic acidosis and treatment with nitrates, had caused methemoglobinemia.

Susceptibility factors

Mepivacaine toxicity has been studied in 10 patients with end-stage chronic renal insufficiency undergoing vascular access surgery [ ]. These patients represent a high-risk group for general anesthesia, as they often have concomitant coronary artery disease, hypertension, and diabetes. Brachial plexus block is often used: as well as avoiding systemic effects, it enhances regional blood flow. However, high doses of local anesthetic are required, and this block carries one of the highest rates of seizures. In this study, following axillary block with mepivacaine 650 mg, plasma concentrations were greater than the threshold of 6 micrograms/ml, above which signs of nervous system toxicity reportedly occur. The authors suggested that the absence of nervous system signs may have been due to slow systemic absorption of the local anesthetic. Peak concentrations occurred after 60–90 minutes, but were still high at 150 minutes, raising the question of more prolonged monitoring after these blocks.

Cardiovascular

There has been a report of T wave changes on the electrocardiogram during caudal administration of local anesthetics [ ].

• A 4.2 kg 2-month-old baby was given a caudal injection under general anesthesia for an inguinal hernia repair. A mixture of 1% lidocaine 2 ml and 0.25% bupivacaine 2 ml was injected. Every 1 ml was preceded by an aspiration test and followed by observation for electrocardiographic changes for 20 seconds. On administration of the third 1 ml dose, there was a significant increase in T wave amplitude. The aspiration test was repeated and was positive for blood. The caudal injection was stopped and the electrocardiogram returned to normal after 35 seconds. The baby remained cardiovascularly stable with no postoperative sequelae.

Previous reports have suggested that an increase in T wave amplitude could result from inadvertent intravascular administration of adrenaline-containing local anesthetics. This is the first case report of local anesthetics alone causing significant T wave changes.

During administration of a test dose of bupivacaine and adrenaline to exclude intravascular injection, the recommended signs to look for are increased amplitude of the electrocardiographic T wave by 25% or an increased heart rate by more than 10/minute. In three children there were no changes in the electrocardiogram or heart rate, despite intravascular administration [ ]. The authors offered a range of possible explanations, including too low a dose of adrenaline, too short a time taken to assess the effect of the test dose, or lack of atropine pretreatment. They also recommended repeated aspiration during injection, as in all three cases blood could be aspirated with this approach.

Nervous system

Inadvertent dural puncture is a recognized complication in up to 1% of caudal anesthetics. It can be due to excessive needle insertion or sacral abnormalities. Potentially serious consequences, such as total spinal anesthesia, can result [ ].

Neuromuscular

In a controlled, randomized study in 60 children undergoing subumbilical surgery three different concentrations of levobupivacaine were used for caudal anesthesia [ ]. The caudal block was performed with levobupivacaine 0.125%, 0.2%, or 0.25% (total volume 1 ml/kg). The 0.125% solution was associated with significantly less early motor blockade but also a significantly shorter duration of postoperative analgesia.

Hematologic

In eight episodes of toxic methemoglobinemia in seven premature infants after the combination of caudal anesthesia (prilocaine 5.4–6.7 mg/kg) and Emla cream (prilocaine 12.5 mg) for herniotomy, the highest methemoglobin concentration 5.5 hours after anesthesia was 31% [ ]. All the infants were symptomatic, with mottled skin, pallor, cyanosis, and poor peripheral perfusion. The most severe symptoms occurred at 3–8 hours and disappeared within 10–20 hours. The authors stressed the importance of recognizing the poor tolerance of premature infants to methemoglobinemia and that whereas topical prilocaine is relatively safe, caudal administration is not.

Susceptibility factors

Nervous system

Deep cervical plexus block can cause ipsilateral phrenic nerve palsy. A patient with pre-existing respiratory disease and a contralateral raised hemidiaphragm developed hypoxia and respiratory distress when given 20 ml of plain bupivacaine 0.375% by this route for carotid endarterectomy [ ]. Local anesthetic spread resulted in presumed stellate ganglion block, which caused nasal congestion and aggravated the respiratory distress. The symptoms resolved without intubation, but the authors advised against deep cervical plexus block in patients with diaphragmatic motion abnormalities or chronic respiratory disease.

Nerve palsies can occur during deep cervical plexus anesthesia.

• A woman complained of being unable to clear secretions effectively from her throat, had a paroxysm of coughing, and developed a large neck hematoma requiring surgical re-exploration [ ].

• A 71-year-old man complained of difficulty in breathing and was desaturated on pulse oximetry for 5 minutes after cervical plexus blockade [ ]. He required tracheal intubation, was ventilated for 110 minutes, and was then successfully extubated. It was thought that the most likely diagnosis was cardiorespiratory failure exacerbated by phrenic nerve blockade.

• A 67-year-old man developed transient hemiparesis and facial nerve palsy before becoming unconscious and apneic 10 minutes after a right cervical plexus block [ ]. His trachea was intubated without the need for anesthetic drugs and he was ventilated. Hypotension was treated with intravenous ephedrine. He woke up, started breathing, and was extubated 75 minutes later. The authors postulated brainstem anesthesia following accidental injection of local anesthetic into a dural cuff as a cause of loss of consciousness.

Hemidiaphragmatic paralysis can occur with cervical plexus anesthesia and can be particularly risky in cases of pre-existing airways obstruction [ ].

Infiltration of even small doses of a local anesthetic in the region of the carotid artery is likely to cause nervous system toxicity if injected intra-arterially [ ].

• A 76-year-old man had already received a deep and superficial cervical plexus block for an awake carotid endarterectomy. One hour later, during manipulation of the carotid artery discomfort was treated with infiltration of 1 ml of 0.5% lidocaine in that region. Immediately he became unresponsive, with generalized tonic-clonic seizure activity of the face and arms. He was given 100% oxygen and within 30 seconds the seizure terminated spontaneously with no sequelae.

This demonstrates the requirement for constant vigilance in a patient undergoing awake carotid endarterectomy.

Cardiovascular

Acute hypertension leading to myocardial infarction and pulmonary edema has been described after the use of mepivacaine with levonordefrin [ ].

In 54 patients with coronary artery disease undergoing dental extraction under local anesthesia randomized to two groups with and without adrenaline 1:100 000 [ ]. Three had ST segment depression after administration of adrenaline and two others had increased CK-MB activity. Surprisingly, the authors concluded that dental extraction performed under local anesthesia with 1:100 000 adrenaline does not imply additional ischemic risks.

Nervous system

An unexplained case of permanent neurological deficit, consisting of left facial palsy, right sensorineural hearing loss, gait ataxia, and hemisensory loss in the body and face, has been described after inferior alveolar nerve block [ ].

Facial paralysis is occasionally reported and is not necessarily due to poor technique; in one case vascular spasm seemed to provide an explanation [ ].

• An 8-year-old girl received prilocaine for a dental procedure performed under 70% oxygen/30% nitrous oxide [ ]. The dose of 288 mg was 2.7 times higher than the recommended safe dose of 6 mg/kg. Toward the end of the procedure, she became unconscious and had a convulsion.

Two reviews have highlighted the fact that the degree and incidence of neurological damage after dental anesthesia is probably underestimated. Some drugs, such as articaine and prilocaine, seem to cause a higher incidence of paresthesia than others [ , ].

In seven subjects articaine with adrenaline caused distortion of lingual nerve function with effects on vowel pronunciation and therefore the potential to impair speech [ ].

• A 49-year-old man developed uvular deviation as a result of palatal muscle paralysis following intraoral mandibular block of the inferior alveolar nerve with 1.8 ml of 2% lidocaine with adrenaline 1 in 100 000 [ ]. A few minutes after injection he had swallowing difficulties and a foreign body sensation in his throat. There was paralysis of the velum palatinum, with deviation of the uvula towards the non-paralysed side opposite the point of anesthetic infiltration. This resolved after the anesthetic had worn off.

The authors suggest that a high inferior alveolar nerve block can easily affect the mandibular nerve if the anesthetic solution diffuses to the internal trunk of the third trigeminal branch and the supply to the tensor veli palatini.

Paresthesia associated with the use of local anesthetics as part of dental care is infrequent, although its incidence has increased over the last 30 years.

Prolonged dysesthesia has been reported in seven cases of inferior alveolar nerve block injection, all associated with articaine [ ]. The author recommended a widespread survey of the relation between prolonged dysesthesia and particular local anesthetic choices to clarify this apparent adverse effect.

In fact, such a review was published in 2003 [ ]. The use of articaine, and to a lesser extent prilocaine, for lingual and inferior alveolar nerve blocks is associated with a higher incidence of paresthesia in these nerves compared with lidocaine, bupivacaine, or mepivacaine. This raises doubts about the suitability of articaine and prilocaine for local anesthesia in dentistry. The incidence of paresthesia associated with the use of these agents should be considered when selecting a local anesthetic for anesthesia of the mandible and associated structures.

Nerve injury after mandibular nerve block has been previously reported. Articaine, and to a lesser extent prilocaine, have been implicated in an increased incidence of permanent paresthesia after mandibular nerve anesthesia, and lingual nerve injury is most common and most incapacitating. These conclusions have been supported by a recent case series of 54 nerve injuries in 52 patients, in which standardized assessment of neurosensory function showed that toxicity was most likely the central causative factor [ ]. Over half of these cases were associated with the use of articaine. The authors concluded that articaine produced a more than 20-fold increase in the incidence of injection injury after mandibular nerve block. Recent reviewers have recommend avoiding articaine and prilocaine for mandibular and lingual nerve block, although they have concluded that it may be the high concentration rather the drug itself that is responsible for nerve damage [ , ].

Inferior alveolar blockade is commonly used in dental practice. However, a rare case of upper lip blanching with temporary lateral rectus nerve palsy, leading to diplopia, has been reported [ ].

Sensory systems

Adverse ocular effects, such as ptosis, are on record [ ]. Transient dizziness, diplopia, and partial blindness have been reported after the entry of lidocaine with adrenaline into the ophthalmic artery following mandibular block [ ]. A similar case after posterior alveolar block resulted in dizziness and diplopia for 3 hours when the patient stood up, possibly due to the entry of local anesthetic into the ophthalmic artery [ ].

Ophthalmological complications after intraoral anesthesia occurred in 14 cases over 15 years [ ]. The most common symptom was diplopia. Three patients developed Horner’s syndrome, with ptosis, enophthalmos, and miosis on the same side as the anesthesia. Three patients developed mydriasis and ptosis. There was complete resolution in all patients. The authors postulated that direct diffusion of anesthetic solution from the pterygomaxillary fossa through the sphenomaxillary cavity to the orbit had caused the ophthalmological effects.

• A 45-year-old man developed temporary monocular blindness, ophthalmoplegia, ptosis, and mydriasis immediately after a mandibular block injection [ ]. Unidentified intra-arterial injection into the maxillary artery, with backflow of the local anesthetic solution to the middle meningeal artery was the postulated cause.

• A 73-year-old man with a history of infective endocarditis was admitted for multiple dental extractions and received prilocaine 144 mg after aspiration [ ]. Within 2 minutes he reported that he could not see in his left eye. Fundoscopy showed diffusely obstructed retinal vessels, with multiple segmented clear fluid emboli and an incomplete cherry-red spot. There was no evidence of choroidal abscess or central nervous system signs of recent thromboembolism. Anterior chamber ocular paracentesis with ocular massage was attempted without improvement. Five days later his visual acuity remained at light perception only. Two months later his vision was unchanged.

The authors noted that this is a rare event and proposed causative mechanisms: intra-arterial injection causing retrograde flow in an abnormal anatomy or injection through vascular abnormalities from previous trauma or inflammation. It was difficult to implicate endocarditis, in the absence of calcific or platelet fibrin emboli. They concluded that delivery of local anesthetic must be done with aspiration before and care during injection. This will possibly prevent intravascular injection.

Transient cranial nerve palsies have been described after dental nerve block. The exact mechanisms of local anesthetic spread have not been clearly defined, although intra-arterial injection and local spread have been discussed. Two cases of transient blurred vision with loss of accommodation in one eye after routine inferior alveolar nerve block have been reported [ ]. These were caused by ipsilateral ciliary nerve blockade, possibly due to inadvertent intravascular injection with spread of local anesthetic retrogradely to the orbit.

Another case of transient diplopia due to accidental sixth cranial nerve blockade with ipsilateral lateral rectus paresis followed maxillary injection of articaine with adrenaline for dental extraction [ ]. In all cases resolution was complete within minutes to hours.

Immunologic

True allergic reactions to amide local anesthetics are extremely rare. Anaphylaxis after local lidocaine administration has been reported [ ].

• A 4-year-old child, previously healthy, received an intrapulpal injection of 0.5 ml of lidocaine 2% with 1:100 000 adrenaline for a dental procedure; 15 minutes later he became severely cyanotic and short of breath, and had a respiratory arrest and sinus bradycardia. Cardiopulmonary resuscitation was started immediately, followed by rapid blood volume expansion and adrenaline administration. After 24 hours his vital; signs stabilized and he recovered completely.

Allergic reactions to lidocaine in dental cartridges and reusable vials can occur because of preservatives such as parabens. However, in this case the preservative was sodium sulfite, which has not been reported to cause anaphylactic reactions. The cause of the anaphylaxis was not determined in this case, as the parents refused tests.

Additives

Additives in local anesthetic solutions can cause allergic reactions [ ].

• A 34-year-old man developed swelling and redness of the face after receiving lidocaine as Lignospan® for dental treatment. Patch testing showed allergic contact dermatitis due to the preservative disodium ethylenediamine tetra-acetic acid (EDTA).

Comparative studies

After thoracotomy, 106 patients received a thoracic epidural infusion of either 0.1% or 0.2% bupivacaine, both with fentanyl 10 micrograms/ml, compared with epidural fentanyl alone; there was no difference in the number of episodes of postoperative hypotension (systolic pressure below 90 mmHg) or in the number of interventions for postoperative hypotension, but intraoperative vasopressors were used significantly more in the bupivacaine groups [ ]. In addition, two patients given 0.2% bupivacaine reported slight weakness of both hands and another Horner’s syndrome and weakness of the right hand. There was a similar incidence of nausea and pruritus in all the groups; however, the incidence of respiratory depression with fentanyl was high (4.2%).

Random allocation of 150 women in labor to either an intermittent epidural bolus, a continuous epidural infusion, or patient-controlled epidural analgesia with 0.125% bupivacaine and sufentanil 0.5 micrograms/ml resulted in significantly more frequent motor blockade with continuous infusion compared with intermittent boluses (22 versus 4%), with similar frequencies of pruritus, hypotension, and high sensory level in each group [ ].

In 52 patients who received either epidural bupivacaine (0.10–0.28 mg/kg/hour) or lidocaine (0.44–0.98 mg/kg/hour), both with epidural morphine, there were no significant differences in the times to mobilize, motor function (as measured by the Bromage grade), and the incidence of hypotension [ ]. Most of the patients had no motor blockade, and the Bromage grade did not help predict which of them could be mobilized.

In 90 parturients who received epidural analgesia during labor with bolus administration of either 10 ml of 0.125% bupivacaine or 0.125% ropivacaine, each with sufentanil 7.5 micrograms, there were comparable onset times and duration of analgesia in the two groups, but patients given ropivacaine had significantly less motor blockade after the third and subsequent epidural injections compared with those given bupivacaine: 93% of those given ropivacaine had no motor impairment compared with 66% of those given bupivacaine [ ]. There were no differences in hemodynamic effects and pruritus.

An epidural infusion of 0.2% ropivacaine plus sufentanil has been compared with 0.175% bupivacaine plus sufentanil in 86 patients postoperatively after major gastrointestinal surgery; there was no statistically significant difference in the incidence of adverse effects (respiratory depression, sedation, nausea, vomiting, pruritus, and motor blockade), but those given ropivacaine mobilized more quickly [ ].

In 60 women who underwent elective cesarean section under epidural anesthesia, 0.5% levobupivacaine or 0.5% bupivacaine (30 ml) were equally efficacious in terms of anesthesia [ ]. The incidence and severity of motor blockade, hypotension, changes in QT interval, nausea, and vomiting were not significantly different, and neither were the neonatal Apgar scores.

Drug combinations are often used in epidural anesthesia to enhance the analgesic effect and minimize adverse effects. Continuous epidural analgesia (0.125% bupivacaine 12.5 mg/hour and morphine 0.25 mg/hour) has been compared with patient-controlled analgesia (morphine) in 60 patients after major abdominal surgery. Analgesia was superior in the epidural group, satisfaction and sedation scores were similar in both groups, whilst episodes of moderate nocturnal postoperative hypoxemia (SaO2 85–90%) were more frequent in the epidural group [ ].

The addition of opioids to local anesthetic to improve the efficacy of epidural analgesia for cesarean section has been advocated [ , ]. A test dose of lidocaine 60 mg was given to 24 patients undergoing elective cesarean section, followed by either bupivacaine 45 mg or bupivacaine 45 mg plus fentanyl 50 micrograms [ ]. Sensory blockade to T6 was achieved in both groups, but pain scores were significantly lower in the fentanyl group. Rescue fentanyl on uterine exteriorization was required in 40% of the control group, but in none in the fentanyl group. There were no significant differences in adverse effects, specifically pruritus, hypotension, nausea and vomiting, maternal respiratory depression, and Apgar scores.

Analgesia after major surgery has been evaluated in a prospective study in 2696 patients, who received either epidural or intravenous analgesia for postoperative pain relief [ ]. Epidural analgesia consisted of bupivacaine 0.25% with morphine 0.05 mg/ml and was used in 1670 patients. Intravenous analgesia with morphine 1 mg/ml was used in 1026 patients. The patients with epidural analgesia had better pain relief both at rest and during mobilization compared with intravenous analgesia. However, orthostatic dysregulation in 6%, pruritus in 4.4%, and technical problems in 6.2% were more frequent with epidural analgesia. In comparison, intravenous morphine analgesia had a higher frequency of opioid related adverse effects, such as sedation/hallucinations/nightmares/confusion in 2.5% and respiratory depression in 1.2%. This study used background infusion plus patient-controlled analgesia in both groups, which might have affected the adverse effects in the intravenous group; perhaps another choice of epidural solution would have caused less hypotension.

Systematic reviews

An attempt has been made to identify the ideal epidural test dose, which by its adverse effects would allow detection of a misplaced (intravascular instead of epidural) catheter or needle, in order to avoid more serious adverse effects later [ ] The following tests had a sensitivity and positive predictive value of at least 80%.

• 1.

  Non-pregnant adults: adrenaline 10 or 15 micrograms increased systolic blood pressure (SBP) by at least 15 mmHg or either increased SBP by at least 15 mmHg or increased heart rate by at least 10/minute.

• 2.

  Children: adrenaline 0.5 micrograms/kg increased SBP by at least 15 mmHg.

• 3.

  Pregnant women: fentanyl 100 micrograms caused sedation, drowsiness, or dizziness within 5 minutes.

The author stated that the systematic use of adrenaline in non-pregnant adults and children to detect intravascular needle/catheter position is reasonable, considering the absence of serious adverse effects. In contrast, routine use in pregnant women could not be justified, because of the low positive predictive value and potential adverse effects; here an analysis of the benefit to harm balance of each scenario for the parturient is necessary, and its use in testing epidurals for cesarean section appears to be more appropriate than for labor analgesia. There were too few data to determine the best method of detecting intrathecal or subdural catheter misplacement.

Cardiovascular

Hypotension is a frequent adverse effect of epidural anesthesia. In a comparison of the effects of bupivacaine and ropivacaine in 60 women undergoing cesarean section, 90% had a fall in blood pressure to below 90 mmHg, or by more than 30% of baseline [ ].

Abrupt onset of arterial hypotension is also a complication of cervical epidural anesthesia, particularly in elderly patients [ ]. However, supplementation with adrenaline in this high-risk group is no longer defensible; it is better to be cautious with dosage and to monitor the patient closely.

• Severe hypotension during a lumbar epidural anesthetic in a 61-year-old woman taking amitriptyline was refractory to high doses of ephedrine and other indirect alpha-adrenergic agents [ ]. It eventually responded to one dose of noradrenaline 200 micrograms, illustrating the importance of the choice of vasopressor for treating hypotension in the presence of chronic tricyclic antidepressant use.

• A 27-year-old woman developed significant myocardial depression and pulmonary edema after administration of 5 ml of bupivacaine 0.5% via an epidural catheter [ ]. The bupivacaine followed a test dose of 3 ml lidocaine 2%.

Although initial aspiration on the epidural catheter was negative, the most likely explanation must be inadvertent intravascular administration of lidocaine and bupivacaine.

Hypotension during epidural anesthesia can be due to functional hypovolemia. It is usually treated with intravenous fluids and/or vasopressors. In order to validate the changes in intravascular volumes after thoracic epidural anesthesia over a longer time, a study was undertaken in 12 healthy volunteers, who were randomized to receive either colloidal fluid (hydroxyethyl starch 7 ml/kg) or a vasopressor (ephedrine 0.2 mg/kg) 90 minutes after the administration of 10 ml of bupivacaine 0.5% through a thoracic epidural inserted at T7–10 [ ]. Thoracic epidural anesthesia in itself did not lead to any changes in blood volume, despite a fall in blood pressure. The authors concluded that fluid administration leads to dilution and recommended that hydroxyethyl starch may be preferred to ephedrine in patients with cardiopulmonary disease, in order to avoid perioperative fluid overload.

In a randomized double-blind study of the cardiovascular effects and neonatal outcome of epidural blockade in healthy parturients scheduled for elective cesarean section, the patients were allocated to either epidural ropivacaine 0.75% or bupivacaine 0.5% [ ]. The two agents produced equally satisfactory blockade, but ropivacaine 0.75% produced a more pronounced reduction in maternal heart rate. However, this had no effect on neonatal outcome. The authors concluded that both bupivacaine 0.5% and ropivacaine 0.75% could be recommended for epidural anesthesia in elective cesarean section.

Unusually, a mother and her child died after repeated administration of a local anesthetic for cesarean section; pulmonary edema was believed to have been the cause [ ].

Intracardiac conduction disturbances should not be considered as absolute contraindications to epidural anesthesia: there were only nine cases of sinus bradycardia, easily reversed with atropine sulfate, in 66 patients [ ]. However, rare cases of complete heart block and complete left bundle branch block have occurred [ ].

Unexpected cardiopulmonary arrest can result from accidental dural puncture during epidural blockade [ ].

• Asystolic cardiac arrest has been described in a 55-year-old man who underwent partial hepatectomy under combined general and epidural anesthesia [ ]. During postoperative recovery he developed asystole followed by ventricular fibrillation. Resuscitation was unsuccessful.

The authors concluded that in the absence of any other abnormality the arrest had been the result of an autonomic imbalance due to spreading sympathetic block, although other postoperative causes of death should not be discarded.

Prolongation of the QT interval can predispose to dysrhythmias with local anesthetics.

• Intraoperative cardiac arrest occurred in a 9-year-old child with Pfeiffer syndrome (craniosynostosis, mild syndactyly of hands and feet, and dysmorphic facial features) undergoing reversal of a colostomy [ ]. All previous anesthetics had been uneventful. The child received an epidural catheter at the L3/4 interspace. A test dose of 2 ml of lidocaine 1% with adrenaline 1: 200 000 was administered and aspiration for spinal fluid was negative. One minute after the first dose of bupivacaine 0.25% 3 ml with adrenaline 1: 200 000 he developed cardiac dysrhythmias and 3 minutes later, and before surgical incision, ventricular fibrillation. After chest compression, 100% oxygen, adrenaline, and sodium bicarbonate, sinus rhythm returned. Blood was aspirated from the epidural catheter. Postoperative investigation showed a long QT syndrome.

Prolongation of the QT interval predisposes to ventricular dysrhythmias and can be triggered by adrenaline. In this case the authors concluded that accidental intravascular injection of bupivacaine and adrenaline may have triggered the dysrhythmia.

Brugada syndrome (right bundle branch block and raised ST segments), can cause sudden cardiac death, potentially hastened by class I antidysrhythmic drugs. Intravenous sodium channel blockers such as local anesthetics can unmask Brugada syndrome.

• A 77-year old man with no previous symptoms of ischemic heart disease underwent elective gastrectomy for carcinoma of the stomach [ ]. Preoperative electrocardiography showed partial right bundle branch block. An epidural catheter was inserted at interspace T9/10 before induction. Aspiration of the catheter was negative for blood and cerebrospinal fluid. Bupivacaine 0.25% 10 ml was given in 2 ml increments, and an infusion of 0.125% bupivacaine and fentanyl 2.5 μg/ml was begun at 8 ml/hour. The operation was uneventful. Three epidural bolus doses were given postoperatively over 11 hours, consisting of 0.125% bupivacaine with fentanyl 2.5 μg/ml, 8 ml, 5 ml, and 5 ml. After the last dose, his systolic blood pressure fell to 80 mmHg. An electrocardiogram showed right bundle branch block with new convex-curved ST segment elevation in V1-V3. Acute myocardial infarction was ruled out and a diagnosis of Brugada syndrome was made. Bupivacaine was withdrawn after a total infusion time of 17 hours (total dose of bupivacaine 443 mg). The patient made a complete and uneventful recovery.

As Class Ib drugs such as lidocaine do not induce the characteristic electrocardiographic changes, the authors suggested that bupivacaine causes greater inhibition of the rapid phase of depolarization in Purkinje fibers and ventricular muscle, and remains bound to sodium channels for longer than lidocaine.

Bradycardic arrest developed rapidly secondary to a reflex vagal response following abdominal wall traction in a patient with a thoracic epidural [ ].

• A 53-year-old man with a long smoking history received a T7/8 thoracic epidural before induction for subtotal gastrectomy. He was given atropine 0.5 mg and midazolam 2 mg intramuscularly 30 minutes before surgery. An epidural test dose of 2% lidocaine 3 ml was given uneventfully, and after induction 0.375% ropivacaine 10 ml. The heart rate fell rapidly 1 minute after the abdominal self-retaining retractor was positioned and bradycardic arrest occurred. Prompt removal of the surgical stimulus, cardiac compression, atropine 0.5 mg and adrenaline 1 mg resulted in the return of normal hemodynamics.

Sympatholysis from neuraxial blocks results in loss of compensatory mechanisms, thereby potentiating the severity of vagal effects. Awareness and prompt management by removing the vagal stimuli and resuscitation including anticholinergic and/or sympathomimetic drugs should lead to excellent recovery.

Respiratory

Respiratory depression was noted in 0.24% of patients in a Chinese series of 10 978 epidural blocks [ ]. Direct paralysis of respiration probably plays an important role. Respiratory depression with adverse cardiovascular effects after miscalculated dose requirements or a misplaced catheter has also been described [ ].

In 15 patients receiving lidocaine 300 mg plus adrenaline by cervical epidural injection, the upper cervical nerve roots C3, 4, and 5 were anesthetized. None of the patients had pre-existing pulmonary disease. Only one had symptoms of impaired pulmonary function at 20 minutes after epidural, and complained of dyspnea, with a reduction in maximum inspiratory pressure, FEV1, FVC, and SpO2. Four patients had a bradycardia requiring atropine, eight complained of nausea, and one developed hypotension requiring ephedrine. At 20 minutes after the epidural, all the patients had a maximum reduction in FEV1 and FVC, ranging from 12 to 16% of preanesthetic measurements. The authors felt that as the maximum inspiratory pressure was virtually unchanged, this suggested that the motor function of the phrenic nerve was mostly intact, despite analgesia of the C3, 4, and 5 dermatomes [ ].

Hiccups that last longer than 48 hours are referred to as persistent hiccups, and those lasting more than 2 months are considered intractable. Persistent or intractable hiccups can lead to fatigue, sleep disturbances, dehydration, and even wound dehiscence in the perioperative period.

• A 65-year-old man received a series of three epidural injections, each with 11 ml of a mixture of 0.08% bupivacaine and triamcinolone 80 mg, in an anesthesia pain clinic for evaluation and treatment of lumbar spinal stenosis [ ]. After the first two injections he developed leg weakness, which resolved after about 4 hours. After the third injection he developed mild urinary retention, which resolved without consequence 6 hours later. All three injections were associated with hiccups after about 1 hour and persisting for 5–7 days. He received two further epidural injections of a glucocorticoid in isotonic saline and did not develop hiccups. All the procedures were 8 weeks apart. A year later, after an epidural injection for a total knee replacement he developed hiccups, which resolved 9 days later.

There are many causes of hiccups. They are most commonly gastrointestinal in origin, such as gastric distention or gastro-esophageal reflux disease. Metabolic derangements and drugs are also frequently implicated. Two cases of hiccups after thoracic epidural injections of glucocorticoids have previously been reported, but in this case a glucocorticoid injection without bupivacaine did not lead to hiccups.

During pregnancy and labor there are important respiratory changes. In a prospective study to clarify whether minor motor blockade brought on by lumbar epidural anesthesia in laboring women further compromises respiratory function, 60 parturients received lumbar epidural anesthesia at L2–4 [ ]. After a test dose of 3 ml of lidocaine 2% and then a total dose of 10–15 ml of bupivacaine 0.125%, followed by a bolus of fentanyl 50 micrograms, a continuous infusion of 10 ml/hour of bupivacaine 0.125% with fentanyl 0.0001% was started, when sensory blockade at T10 was reached. Most of the patients (87%) had significant improvements in respiratory function, suggesting benefits of epidural analgesia in parturients.

Nervous system

Three case reports have illustrated the neurological consequences of epidural anesthesia in predisposed patients.

• A 51-year-old man, ASA grade II, with non-insulin-dependent diabetes, underwent radical prostatectomy and enterocystoplasty under general anesthesia, before which a lumbar epidural catheter was inserted at L3–4 but was not used during surgery [ ]. In the recovery room, a test dose of 3 ml of lidocaine 1% with adrenaline 1:200 000 was administered, followed by a bolus dose of 10 ml of ropivacaine 0.75%, which resulted in a block that reached T10. One hour later an infusion of ropivacaine 0.2% was started at 5 ml/hour. Ten hours later he complained of pain, and the pump rate was increased to 10 ml/hour. The treatment was continued for 72 hours without any more dosage adjustments. Eight hours after the end of the epidural treatment he described a burning sensation and pain in the back, spreading to the legs and feet. These symptoms increased with movement, but there were no motor abnormalities. The symptoms persisted, and electromyography showed a sensory polyneuropathy in all four limbs. Eight weeks after the operation he described diminished pain and paresthesia.

The author suggested that local anesthesia in patients with pre-existing diabetic polyneuropathy may result in additional ischemic insult and intraneural edema. Patients with diabetes may therefore be at higher risk of local anesthetic toxicity.

• A 64-year-old man, with a history of multiple spinal operations, chronic low-back pain, and a transient cauda equina syndrome after the most recent operation, was given a left L2 transforaminal epidural injection, unsuccessfully [ ]. A further attempt at L1 was successful, and 5 ml of bupivacaine 0.125% and 40 mg of triamcinolone was injected. Two minutes later his legs became paralysed. An MRI scan showed signal changes consistent with acute spinal infarction. Four years later he showed no improvement.

Direct injury to the vascular supply of the spinal cord may have been one explanation for this adverse outcome; other reasons included vasospasm caused by either bupivacaine or the glucocorticoid, end-capillary occlusion by glucocorticoid particles, or needle-related factors.

• A 27-year-old primipara was admitted to hospital at week 36 because of a 10-day history of progressive weakness and numbness in all limbs [ ]. Guillain–Barré syndrome was diagnosed and she was given large doses of intravenous immunoglobulin. Her neurological symptoms improved after 5 days. Five weeks later she spontaneously delivered under epidural analgesia (L2–3), after a test dose of 3 ml of lidocaine 2% with 1:200 000 adrenaline and then 25 mg of ropivacaine 0.2% with 16 micrograms of sufentanil over 3 hours. At this stage she had increased sensory and motor block. Twelve hours postpartum she was unable to walk and had augmented symptoms from the arms, together with facial weakness. She was given large doses of intravenous immunoglobulin. Four months later her status had improved but she still depended on a walker.

The authors speculated that the worsening of the patient’s symptoms could have been due to local anesthetic toxicity, since local anesthetics can cause morphological changes in neurons in vitro, impairing their growth.

Trigeminal nerve palsy has been reported on a few occasions after lumbar epidural anesthesia. Horner’s syndrome is also a rare complication of epidural blockade, but it is more common in obstetric patients (0.4–5%). There has been a report that subdural placement of an epidural catheter caused both of these complications [ ]. The subdural space is a potential space containing a small volume of serous fluid between the dura mater and arachnoid mater, and in this case the authors confirmed the subdural position of the catheter by repeat injection and epidurography.

Peripheral paresthesia, in 1.13% of patients in a Chinese series [ ] and 0.16% of patients in a Japanese study of 15 884 epidurals [ ], is the most frequent neurological deficit attributed to spinal and epidural analgesia.

High spinal block has previously been reported as a rare complication of epidural anesthesia.

• A 31-year-old woman in labor had an epidural catheter sited at L3/4 [ ]. A test dose of 0.25% bupivacaine 10 ml was followed 90 minutes later by another 10 ml. After a further 90 minutes she required cesarean section, had a block to T7, and was topped up with 0.75% ropivacaine 10 ml. Within minutes she developed arm weakness, and over the next 15 minutes developed further ascending block requiring intubation. Three hours later the block had regressed to T8 and she had no further complications.

The cause was thought to be subdural injection, although other mechanisms could not be excluded; for example the catheter could have been partly intrathecal and the ultimate distribution of the dose could have been related to the speed of injection or catheter migration before the final dose was given.

High neuraxial block occurred after a large volume of local anesthetic was given to expand the epidural space after accidental dural puncture [ ].

• A lumbar epidural was attempted in a 50-year-old woman, ASA grade 1, listed for transabdominal hysterectomy. The first attempt at L4/5 resulted in inadvertent dural puncture and another attempt was made at L3/4. After being satisfied that the needle was in the correct location, with a negative aspiration test and no response after a test dose of 0.25% lidocaine 3 ml with adrenaline 0.5 micrograms/ml, 13 ml of 0.5% plain bupivacaine was given via the needle to expand the epidural space to facilitate insertion of the catheter. Within 7 minutes, the sensory block had spread to T4 and hypotension and respiratory distress subsequently ensued. Intravenous fluids and mephentermine 6 mg were given, general anesthesia was induced, and surgery was performed. The patient was hemodynamically stable with sensory block at T10 at the end of the operation.

The authors suggested that preceding dural puncture is a contraindication to expanding the epidural space before catheter placement.

In a retrospective review of 139 patients with pre-existing nervous system disorders there was no evidence of new or worsening neurological deficits after neuraxial block [ ] Analysis of patients’ records enabled the characteristics of the disorders to be determined and review of the daily progress notes and follow-up visits identified new or progressive postoperative neurological problems. Previous poliomyelitis, multiple sclerosis, and traumatic spinal cord injury accounted for about 90% of this study population, and 74% of patients described active symptoms. There were technical complications in 15, most commonly paresthesia and traumatic needle placement. There was satisfactory block in 98% and no new neurological deficits in any of the subjects. The authors suggested that the risks of exacerbating pre-existing nervous system disorders with a neuraxial block may have been overestimated, and that neuraxial block should not be considered to be absolutely contraindicated in this population. The limitations of this study included potential selection bias, short (6–8 week) postoperative follow-up, and difficulty in detecting minor or subclinical complications.

Another patient experienced a high block via unintentional subdural injection of ropivacaine [ ].

• Epidural analgesia was planned for a 25-year-old pregnant woman. The first attempt resulted in dural puncture and another was made one level higher. After a negative aspiration test, 0.2% ropivacaine 12 ml was given in 3 ml increments over 15 minutes. She developed bilateral facial paresthesia towards the tip of her nose, but no other problems. The baby was delivered vaginally without event, and the patient’s numbness disappeared after 6 hours.

The authors suggested that the use of a dilute solution had spared any ventilatory or nervous system symptoms, despite evidence of subdural spread reaching the brainstem.

Bilateral foot drop occurred in a 9-year-old boy after removal of his epidural catheter, which had been present for 2 days [ ]. He initially underwent a 4-hour urological procedure with a combined general anesthetic and epidural technique, and then had 2 days of epidural analgesia. Electromyography showed focal demyelination of both peroneal nerves at the fibular head, supporting the authors’ belief that the deficit had resulted from direct pressure over a period of hours while local anesthesia persisted. This again highlights the need for careful positioning and pressure area care as long as effects of local anesthetics persist.

Lumbar extradural analgesia with bupivacaine increases intracranial pressure in some patients, apparently those who already have some reduced intracranial compliance, and who may be at risk [ ]. A sudden increase in intracranial pressure, due to an increased volume in the caudal space, can precipitate respiratory arrest because of direct midbrain stimulation.

• A watershed cerebral infarct with subsequent full recovery occurred in a 70-year-old man 8 hours after a hypotensive event following an incremental bolus of 1% lidocaine 10 ml via an established epidural catheter [ ].

A cause-and-effect relation cannot be established in such cases.

Epidural anesthesia can mask a neurological deficit, such as nerve compression of the femoral nerve and lateral femoral cutaneous nerve of the thigh from the lithotomy position [ ].

• Neurological effects after accidental intravenous injection of a large dose of levobupivacaine (142 mg) have been described during epidural anesthesia [ ].

• A 77-year-old woman had epidural anesthesia, following negative aspiration, with a 3 ml test dose of 0.75% levobupivacaine with 1:200 000 adrenaline and then incremental doses up to a total of 17 ml of 0.75% levobupivacaine. During the final 5 ml of injection, she became disoriented and drowsy, with slurred speech, immediately followed by excitation with shouting and writhing about. She was given thiopental for seizure prophylaxis with high-flow oxygen, and the excitatory signs abated. The catheter was withdrawn 1 cm and blood was freely aspirated. The serum levobupivacaine concentration 14 minutes later was 2.7 micrograms/ml.

Sensory systems

Hearing loss after epidural block has been reported [ ].

• A 30-year–old woman with a body mass index of 54 received an epidural catheter at the L3/4 interspace during labor. The procedure was uneventful. A test dose of 3 ml lidocaine 2% was administered. The first top-up dose consisted of 10 ml plain bupivacaine 0.25%. The sensory level was T10 bilaterally after 15 minutes. With the first top-up and every subsequent top-up dose (bupivacaine 0.1% with fentanyl 2 micrograms/ml) she complained of bilateral hearing loss, disappearing spontaneously after 30-60 seconds. After 10 hours a cesarean section was performed. Anesthesia was achieved with two injections of 10 ml of bupivacaine 0.5%. Transient deafness occurred with each top-up dose. The postoperative period was uneventful.

Transient hearing loss after epidural block occurs because the perilymph in the inner ear is in continuity with the cerebrospinal fluid and any pressure wave in the epidural space is conducted to affect the inner ear [ ].

Metabolism

A small reduction in glucose concentrations, rarely leading to hypoglycemic coma, can occur [ ]. This effect is in keeping with the finding that the catabolic stress response to surgery may be suppressed by epidural analgesia [ ]. However, in one study, thoracic epidural administration produced a degree of hyperglycemia [ ].

• Symptomatic hypoglycemia occurred in a healthy 30-year-old primigravida after a second 5 ml bolus of 0.25% bupivacaine administered epidurally during labor [ ]. She developed an altered mental state, which responded rapidly to 50 ml of 50% dextrose administered intravenously.

Urinary tract

Epidural anesthesia increases the risk of urinary retention [ ].

Skin

Delayed-type hypersensitivity to epidural ropivacaine has been described.

• A 74-year-old man with postherpetic neuralgia and no history of drug allergies developed a purpuric rash and widespread blotchy erythema on his legs, trunks, and arms following continuous epidural blockade with ropivacaine 0.2% without preservatives (up to 96 ml/day) [ ]. He had normal white cell and platelet counts and a slight eosinophilia (640 × 106/l). The epidural infusion and other drugs (amitriptyline, alprazolam, and laxoprofen) were withheld and the eruptions completely resolved within 7 days. Intradermal ropivacaine 0.2% produced erythema (maximum size 23 mm × 13 mm) at 8–72 hours. Histology showed perivascular infiltrates of lymphocytes and eosinophils in the dermis. Patch testing with amitriptyline, alprazolam, and loxoprofen induced no eruptions, and neither did restarting the drugs.

This report led to a correspondence questioning the duration of the infusion and also possible cumulative toxicity of ropivacaine [ ].

Musculoskeletal

Occasionally orthopedic patients have developed compartment syndrome postoperatively during epidural infusions of bupivacaine/fentanyl mixtures. However, although “aggressive analgesia” was blamed for the resulting disasters, there seems to have been a remarkable lack of adequate pressure area care, correct positioning, and regular review of both patients and splints [ ].

Reproductive system

The effect of continuous epidural bupivacaine 0.075% by infusion for analgesia during labor on uterine artery resistance has been evaluated in 20 pregnant women [ ]. There was evidence of increased uterine blood vessel resistance with Doppler derived velocimetry at 1, 2, and 4 hours. The authors concluded that the clinical significance of this was yet to be determined. Previous studies of the transient effects of bupivacaine 0.25% on uterine blood flow have been inconclusive.

Infection risk

Contamination of catheters, with subsequent clinical infection, is a potential hazard of epidural analgesia. But not every suspected infection is what it seems; aseptic meningitis has been described after an intradural injection of bupivacaine with methylprednisolone acetate [ ].

Death

Inadvertent intravenous administration, due to the accidental placement of an epidural catheter in a vein, is a high-risk complication; deaths have been reported [ ].