REGIONAL ANAESTHESIA
Complications of regional anaesthesia
Learning objectives After reading this article you should: C be aware of the potential risks and complications of regional anaesthesia C understand the mechanisms resulting in severe complications C be able to explain the benefits and risks of regional anaesthetic techniques to patients, aiding them to give informed consent for the proposed procedure
Megan C Dale Matthew R Checketts
Abstract Complications of regional anaesthesia can be divided into those specific to central neuraxial blockade, those specific to peripheral nerve blockade, and those that pertain to both. Fortunately, severe complications e namely, spinal cord damage, vertebral cord haematoma and epidural abscess e are rare. Here we have given an overview of these complications, with reference to updated incidences available following the 3rd National Audit Project (NAP3) of the Royal College of Anaesthetists. A thorough knowledge of anatomy and pharmacology, and a meticulous, unhurried technique are essential to prevent such complications. When considering the use of a regional anaesthetic technique, the risks and benefits for that patient should be assessed on a case-by-case basis.
Authority (NHSLA) database. Eighty-nine percent of claims were related to CNB, with epidurals being the main culprit in 81% of these claims.2 Here, we hope to give a broad overview of the potential complications of regional anaesthesia, which include those specific to either central neuraxial or peripheral nerve blockade, or complications relevant to both (Table 1).
Central neuraxial blockade Post-dural puncture headache (PDPH) This is the most common complication following CNB, during which the dura are breached e whether intentionally or accidentally. The resultant cerebrospinal fluid leak causes sagging of intracranial structures, with traction and vasodilatation in the meninges. The onset of the headache is usually within 2e3 days following dural puncture, but can occur sooner or later. Patients will complain of a searing frontal or fronto-occipital headache, exacerbated by upright posture and relieved by lying down. Additional symptoms include neck stiffness, nausea and vomiting, hearing loss and tinnitus, vertigo, paraesthesia and visual disturbances. It is prudent to also consider the possibility of other diagnoses. These include viral, chemical or bacterial meningitis, intracranial haemorrhage, cerebral venous thrombosis, intracranial tumour, cerebral infarction and, in obstetric patients, preeclampsia.3 The risk of accidental dural puncture during an epidural procedure is about 1%. If a Tuohy needle breaches the dura, the incidence of headache can be as high as 70%.4 The incidence of headache following spinal anaesthesia is considerably less: 0.5e1%. This reduction in PDPH is due to advances in needle design; namely, smaller gauge needles (25e27 gauge) with pencilpoint tips that separate as opposed to cut the dural fibres. Cutting tip needles such as the Quincke are associated with a higher incidence of PDPH.
Keywords Central neuraxial block; complications; nerve damage; peripheral nerve block; regional anaesthesia Royal College of Anaesthetists CPD Matrix: 1F01, 2B04, 2G01 and 2G04.
Regional anaesthesia is by no means an advance of modern medicine. The Incans knew of the numbing effects of chewing coca leaves hundreds of years prior to cocaine being isolated from the coca plant. Cocaine was first used topically for eye surgery in 1884 and in 1897 the first subarachnoid block was performed by August Bier. Following the devastating Woolley and Roe case in 1947, detailing two men who became paraplegic following spinal anaesthesia, the place of spinal anaesthesia was uncertain despite its previously good safety record. Advances in local anaesthetic drugs and equipment have done much to improve the safety of regional anaesthesia. This, coupled with improved supervision and training in regional anaesthetic techniques, has helped to minimize complications. Reassuring data are available on major complications of central neuraxial blockade, following the Third National Audit Project (NAP3) of the Royal College of Anaesthetists. This audit estimated the risk of permanent nerve injury to be 4.2 per 100,000 and the risk of death or paraplegia 1.8 per 100,000 following central neuraxial blockade (CNB).1Szypula et al looked at the extent, patterns and cost associated with litigation claims related to regional anaesthesia in England. Regional anaesthesia itself was the largest clinical category within the NHS Litigation
Infective complications Epidural abscess is a rare but serious medical emergency, requiring prompt diagnosis and treatment to make full recovery a possibility.5 Delay in treatment can result in permanent neurological injury and even death. Factors that may predispose to the development of an epidural abscess following CNB include: immune-compromise, systemic corticosteroid drugs, diabetes and systemic sepsis.4 Technical difficulty with the block can result in localized haemorrhage, which may provide a focus of infection. An epidural catheter maintains the needle tract, providing a pathway for organisms. Another obvious risk factor
Megan C Dale FRCA is a Specialty Registrar at Ninewells Hospital, Dundee, UK. Conflicts of interest: none declared. Matthew R Checketts FRCA is a Consultant Anaesthetist at Ninewells Hospital, Dundee, UK. Conflicts of interest: none declared.
ANAESTHESIA AND INTENSIVE CARE MEDICINE 14:4
142
Ó 2013 Published by Elsevier Ltd.
REGIONAL ANAESTHESIA
patients who receive a perioperative epidural was estimated as high as 1 in 19,500 in the NAP3 UK National Audit.5
The incidences of major complications of regional anaesthesia Complication
Spinal cord injury Traumatic injury of the spinal cord, nerve root or peripheral nerve may be caused by the block needle or by an indwelling catheter. These injuries need to be differentiated from neurological injury cause by cord compression, e.g. epidural abscess or haematoma. Early radiological imaging is essential to exclude the latter two. Establishing a causal link between procedural trauma and nerve damage is rarely straightforward. Confounding or contributing factors such as surgical positioning, the operation itself, the pathology under treatment and pre-existing conditions such as diabetes mellitus or spinal canal stenosis need to be considered.5 Nerve injury from needle or catheter trauma may or may not be associated with pain or paraesthesia during the procedure, but it is still prudent to perform CNBs on conscious or lightly sedated patients so that they may report these symptoms.5 There are three scenarios that could potentially increase the risk of injury to the spinal cord:6 Failure to identify the correct vertebral interspace. This is notoriously difficult to do when using surface anatomical landmarks such as Tuffier’s line. The spinal cord may terminate anywhere from T12 to L4. The combination of performing a subarachnoid block at a higher level than estimated, in a patient whose conus medullaris ends lower than usual may result in spinal cord puncture. The ligamentum flavum may not fuse in the midline, which can alter the customary ‘feel’ of performing a central neuraxial block, particularly an epidural block. The anterior-to-posterior diameter of the epidural space decreases from 5 mm in the lumbar region to 1e2 mm in the upper thoracic and cervical region, which can result in an accidental puncture of the dura and cord when performing a high epidural block.6 Permanent nerve injury is rare following CNB. NAP3 reported the incidence of permanent harm from nerve or spinal cord damage to be 1 per 100,000.
Incidence
Central neuraxial block C Post-dural puncture headache: Following spinal block Following epidural dural tap C Infective complications: Epidural abscess Meningitis C Vertebral canal haematoma C Spinal cord injury C Transient neurological symptoms
1 in 47,000b 1 in 200,000c 1 in 117,000 1 in 100,000 Up to 3 in 100d
Peripheral nerve block C Peripheral nerve injury
1.9 per 10,000
1 in 100 7 in 10a
Complications common to both central and peripheral nerve block Local anaesthetic toxicity Unknown C Total spinal block Unknown C Failed block 1 in 100 C
a
Incidence following accidental dural puncture with 16G Tuohy needle, in obstetric population.11 b Incidence for all adult epidurals, including perioperative, acute and chronic pain and obstetric epidurals. c Incidence for all adult perioperative and obstetric spinal anaesthesia, epidural and combined spinal/epidural anaesthesia. d Incidence with bupivacaine.
Table 1
is failure to adhere to a strict aseptic technique,4 which includes wearing a mask, cap, sterile gown and gloves, and adequate hand washing. The incidence of spontaneous epidural abscess occurring in the population is 0.2e1.2 per 10,000.5 Recent data from the NAP3 report the incidence of epidural abscess in patients following CNB as 2.1 per 100,000 (1 in 47,000), with permanent harm occurring in 1.3 per 100,000 (1 in 88,000) and paraplegia occurring in 0.42 per 100,000 (1 in 236,000).5 Meningitis is another complication of CNB, most commonly following spinal block. The same risk factors for epidural abscess apply to the risk of meningitis. This complication is rare (incidence within the UK is less than 1 per 200,0005) but potentially fatal if not promptly diagnosed and treated.
Transient neurological radiculopathy (TNR) This is a symptom complex that may occur after spinal anaesthesia. Typically, there is complete recovery from the spinal block, followed by dysaesthesia in the back and buttocks, radiating to the thighs. Symptoms can last up to 72 hours. It is unclear whether the pain is related to local anaesthetic toxicity or if it is musculoskeletal or myofascial in origin. This complication has occurred with different local anaesthetics, of differing baricity. The incidence following hyperbaric lidocaine 5% (no longer licenced for subarachnoid anaesthesia) is up to 37%, compared with 3% with bupivacaine.4 Another possible aetiological factor is patient position: lithotomy position can stretch the lumbosacral nerve roots. This compromises perfusion, placing nerves at increased risk of injury.
Vertebral canal haematoma Vertebral canal haematoma is also a medical emergency: delay in diagnosis and treatments beyond 8 hours will result in paraplegia. Patients at risk include those with disordered coagulation, patients on antiplatelet or anticoagulant drugs, and patients in whom the procedure proved technically difficult, resulting in multiple attempts. Most case reports are associated with the use of epidural catheters, and often the only sign is inappropriate motor weakness, or progressive weakness and sensory disturbance. The incidence of vertebral canal haematoma is 0.85 per 100,0005 for all CNBs, with permanent harm occurring in 1 in 140,000 patients.5 However, the incidence of permanent harm in
ANAESTHESIA AND INTENSIVE CARE MEDICINE 14:4
Peripheral nerve blockade Peripheral nerve injury The incidence of transient neuropathy after peripheral nerve block may be as high as 10e15%7 but the actual incidence of
143
Ó 2013 Published by Elsevier Ltd.
REGIONAL ANAESTHESIA
permanent injury is unknown. This is due to a limited number of large-scale prospective studies looking at neurological complications following peripheral nerve blockade. A study by Auroy8 quotes an incidence of 1.9 per 10,000 for serious nerve injury following peripheral nerve blockade. Patients with underlying peripheral neuropathy (e.g. diabetics) are more susceptible to perioperative nerve injury. Mechanisms of peripheral nerve injury can be classified as follows: Direct needle trauma Pain and paraesthesia on injection is not always a reliable indicator of intraneural needle placement. Toxic injury Caused by intraneural local anaesthetic injection. Exposure of peripheral nerves to high concentrations of local anaesthetics, particularly if the injection is intrafascicular, can result in nerve damage, although this remains controversial. Compressive injury May be caused by a haematoma or abscess at the site of injection, or by prolonged tourniquet use. Early diagnosis and treatment is important to prevent permanent injury. Stretch injury Usually a result of excessive traction or improper positioning. Ischaemic injury Rare, and can result from a number of factors. Vascular injury can result in impaired blood supply to a nerve or nerve plexus. Similarly, the combination of hypotension, abnormal positioning in a patient with atherosclerosis, patients with pre-existing neuropathy and controversially, the addition of vasoconstrictors to the local anaesthetic solution, can contribute to peripheral nerve ischaemia. Blunt trauma, toxic and ischaemic injuries all cause axonal loss or damage and carry a poor prognosis. Compressive and stretch injuries result in a neuropraxia. These lesions have a better prognosis, as the axon is still intact. Fortunately, the great majority of injuries are transient; 95% will resolve in 4e6 weeks, and more than 99% will resolve within one year7 (Box 1).
Evaluation of acute nerve injury7 Focused history from patient Detailed examination: Localize problem to specific nerves, nerve plexus, nerve roots or spinal cord
C C
If no deficit or deficit minimal and improving: Observe and reassure patient
C
If neurological deficit present: Referral to neurologist and senior anaesthetist with experience in nerve injury Additional investigations: Baseline electromyography (EMG). Repeat at 14e21 days to quantify the magnitude of acute nerve injury Appropriate imaging, e.g. magnetic resonance imaging
C
Management: Usually conservative If no clinical recovery or evidence of re-innervation by EMG by 2e5 months, consider referral to peripheral nerve surgeon
C
Box 1
anaesthetic. Initially, patients will experience effects on the central nervous system, progressing to effects on the cardiovascular system, including cardiovascular collapse and death. Vasovagal reactions This is a phenomenon described mostly in young patients; fit males in particular. The reported incidences are between 13 and 28%.3 Regional anaesthesia such as spinal anaesthesia and interscalene blocks, performed in an awake, sitting, fasted patient reduces ventricular filling. This, in combination with anxiety and systemically absorbed exogenous adrenaline results in vigorous ventricular contraction. If the ventricular walls touch, massive vagal outflow results, causing profound bradycardia or even asystole (BezoldeJarisch reflex).3
Complications common to both central and peripheral nerve blockade
Respiratory complications There are three main mechanisms by which respiratory complications can occur: Hemidiaphragmatic paresis: Hemidiaphragmatic paresis occurs in nearly 100% of patients receiving interscalene blocks, and in 50% of patients receiving supraclavicular blocks.3 Improved precision with ultrasound, enabling smaller volumes of local anaesthetic, may decrease this risk or the extent of the block. This complication is more problematic in patients who already have a degree of respiratory compromise (e.g. chronic obstructive airways disease). Pneumothorax: This is a potential complication with supraclavicular and paravertebral blocks, as a result of the close proximity of the pleura to the nerves to be blocked. Ultrasound guidance may afford a degree of safety. High spinal block: This can occur after in deliberate or inadvertent placement of subarachnoid local anaesthetic. Pregnant women are particularly
Local anaesthetic toxicity This includes local and systemic effects of local anaesthetics. The mechanism of local anaesthetic toxicity is unclear and remains debatable. Local anaesthetics injected perineurally can compromise the bloodenerve barrier to a small degree, resulting in the hypertonic endoneurial fluid becoming hypotonic, with the accumulation of oedema and resultant increased intrafascicular pressures.9 High concentrations of local anaesthetics can also produce axonal injury not related to oedema. Injections of local anaesthetic intrafascicularly are clearly neurotoxic, especially with ester local anaesthetics. The extent of these effects is proportional to the duration of exposure of nerves to local anaesthetic and the concentration of local anaesthetic. Thus, the perineurium has an important role in protecting the nerve fascicles from direct local anaesthetic toxicity.9 Systemic toxicity usually results from inadvertent intravascular injection or systemic absorption of toxic doses of local
ANAESTHESIA AND INTENSIVE CARE MEDICINE 14:4
144
Ó 2013 Published by Elsevier Ltd.
REGIONAL ANAESTHESIA
during surgery or labour, highlighting the need to improve intraoperative management of a failed regional technique.
susceptible, because of compression of the thecal sac by engorged epidural veins. The high spinal block compromises intercostal muscle function with subjective dyspnoea, poor cough and sometimes even transient somnolence.3 Hypotension is also common because of sympatholysis. Symptomatic treatment, supplemental oxygen and reassurance are appropriate management in most cases. Resuscitation and airway equipment should always be to hand in case further intervention is required.
Conclusion Severe complications associated with regional anaesthesia are rare, but litigation claims still occur even with low severity harm, e.g. failed spinal block in obstetrics. Factors associated with litigation include epidurals, nerve injury, inadequate anaesthesia, obstetrics and ophthalmic blocks.2 The benefits of superior analgesia in the postoperative period, as well as suppression of the stress response to surgery, usually outweigh the risks, but the pros and cons of regional anaesthesia should be considered on a case-by-case basis. To ensure safe practice, a thorough knowledge of anatomy and pharmacology is essential. Patients should give appropriate and informed consent, and the risks discussed should be documented. A meticulous, unhurried technique is also important. Safe practice extends into the postoperative period as well. Ward visits by the anaesthetist or acute pain team, thorough communication with nursing and medical staff, as well as education in potential complications will help to identify warning signs early, so that appropriate investigations and management can be instituted. A
Total spinal block This is the extreme of a high spinal: a total spinal block implies anaesthesia of the brainstem, which will result in rapidly ensuing loss of consciousness, respiratory arrest and profound hypotension. This can occur during an intentional subarachnoid block with excessive spread of local anaesthetic drugs, for example a scenario of a failed epidural top-up, with conversion to spinal anaesthesia. It is conceivable that the thecal sac can be compressed by the epidural contents, or epidural local anaesthetic inadvertently leaks into the subarachnoid space.3 A total spinal block may also occur if a large volume of local anaesthetic meant for an epidural or plexus block, is accidentally placed subarachnoid. This is a known complication of lumbar plexus, interscalene and paravertebral/lumbar plexus blocks. Regarding interscalene blocks, epidural and subarachnoid needle placement have been reported using the lateral or Winnie approach as well as the posterior approach. The vertebral canal can be located as little as 3 cm to the skin. Nerve roots may have a cuff of dura for several centimetres from the neuraxis, making direct intrathecal injection of local anaesthetic a possibility. Similarly, it is also possible for an epidural catheter to migrate into the subarachnoid space.
REFERENCES 1 Cook TM, Counsell D, Wildsmith JAW. Major complications of central neuraxial block: report on the Third National Audit Project of the Royal College of Anaesthetists. Br J Anaesth 2009; 102: 179e90. 2 Szypula K, Ashpole KJ, Bogod D, et al. Litigation related to regional anaesthesia: an analysis of claims against the NHS in England 1995e 2007. Anaesthesia 2010; 65: 443e52. 3 Picard J, Meek T. Complications of regional anaesthesia. Anaesthesia 2010; 65: 105e15. 4 Faccenda KA, Finucane BT. Complications of regional anaesthesia. Incidence and prevention. Drug Saf 2001; 24: 413e42. 5 The 3rd National audit Project of the Royal College of anaesthetists. Major complications of central neuraxial block in the United Kingdom. Report and Findings. Also available at: http://www.rcoa.ac.uk/docs/ nap3_web-large.pdf; January 2009. 6 Neal JM. Anatomy and pathophysiology of spinal cord injury associated with regional anesthesia and pain medicine. Reg Anesth Pain Med 2008; 33: 423e34. 7 Sorenson EJ. Neurological injuries associated with regional anesthesia. Reg Anesth Pain Med 2008; 33: 442e8. 8 Auroy Y, Narchi P, Messiah A, et al. Serious complications related to regional anesthesia. Anesthesiology 1997; 87: 479e86. 9 Hogan QH. Pathophysiology of peripheral nerve injury during regional anesthesia. Reg Anesth Pain Med 2008; 33: 435e41. 10 Fettes PDW, Jansson J-R, Wildsmith JAW. Failed spinal anaesthesia: mechanism, management, and prevention. Br J Anaesth 2009; 102: 739e48. 11 Turnbull DK, Shepherd DB. Post-dural puncture headache: pathogenesis, prevention and treatment. Br J Anaesth 2003; 91: 718e29.
Failure of block A review by Fettes et al quoted the failure rate of spinal blocks to be about 1%.10 Reasons for a failed central neuraxial block are multi-factorial. Not only is this frustrating for the anaesthetist but, when managed badly, a failed block can be distressing for the patient. Similarly a peripheral nerve block may fail if the injected local anaesthetic does not reach the target nerves. This can happen in the following scenarios: If a nerve stimulator is used to identify nerves, the wrong motor response can be misinterpreted for the correct one.3 It is possible to stimulate a nerve through a fascial plane, but if the needle does not pass through that plane, spread of the local anaesthetic will be impeded and not reach the target nerve. Failure to identify anatomy correctly if ultrasound guidance is used. Ultrasound guidance may improve the successful rate of peripheral regional anaesthesia, but this does require sound knowledge of sonoanatomy and clinical experience. In the claims analysis by Szypula et al,2 31% of obstetric anaesthetic regional anaesthesia claims were related to pain
ANAESTHESIA AND INTENSIVE CARE MEDICINE 14:4
145
Ó 2013 Published by Elsevier Ltd.