Serious and permanent neurologic complications in the obstetric population are rare. Most neurologic complications following childbirth are intrinsic obstetric palsies. The most common intrinsic obstetric palsy is lateral femoral neuropathy. Palsies of the femoral, obturator, sciatic, common peroneal nerves and lumbosacral plexus have also been reported. Meticulous technique during neuraxial anaesthesia will decrease the risk of injury secondary to neuraxial procedures. Direct trauma to the spinal cord and spinal nerves may occur during neuraxial anaesthesia. Sterile technique should be employed during neuraxial procedures and precautions should be taken to ensure that the proper drug/concentration is being injected. Postpartum complaints should be addressed promptly. For infection and space-occupying lesions of the neuraxial canal, prompt diagnosis and treatment are essential to prevent permanent injury or death. Survey studies have demonstrated that women want to be told of the risks of neuraxial procedures, even when the incidence is rare.
Neurologic injury during childbirth has long been recognised. In the past half century, advances in labour analgesia, and the widespread introduction of neuraxial labour analgesia and anaesthesia for vaginal and caesarean delivery, have led to increased concern that postpartum neurologic injury is a complication of the anaesthetic procedure. Fortunately, if proper procedures and techniques are employed, permanent and serious complications as a result of neuraxial anaesthesia are rare.
Often the patient may first complain of neurologic symptoms to the primary care provider. Both the patient and the obstetric provider may assume that the injury is related to the neuraxial procedure. However, neurologic injury intrinsic to the childbirth process still accounts for the vast majority of neurologic injury after childbirth. Although quite distressing to the new mother, these injuries are usually transient. Potentially life- and limb-threatening central nervous system (CNS) complications do, however, occur after neuraxial anaesthesia, albeit rarely. Thus, it is important for the generalist, obstetrician and anaesthesiologist to have a basic understanding of the neurologic complications associated with neuraxial analgesia/anaesthesia and childbirth. Potentially catastrophic and reversible complications must be recognised and treated in a timely fashion to ensure the best possible outcome for the patient. This article reviews the current knowledge of intrinsic obstetric palsies, neuraxial anaesthesia-related nerve injury, evaluation of postpartum complaints suggestive of nerve injury and the medicolegal implications of postpartum nerve injury.
Intrinsic obstetric palsies
The vast majority of nerve injuries associated with childbirth can be attributed to the labour and delivery process itself. These are referred to as intrinsic obstetric palsies. The reported incidence of intrinsic obstetric palsy in modern-day obstetric practice varies from 0.6 to 92 per 10 000. The large differences in the reported incidence are likely secondary to study methodology. Studies in which there was individual patient follow-up report a higher incidence than retrospective or prospective survey studies.
Nerve injury during childbirth is presumably related to stretch or compression injury to the lumbosacral plexus or lower extremity peripheral nerves during childbirth. Compression of the nerve vascular supply is another possible mechanism. Review of anecdotal reports suggests that nulliparity, long labour, cephalopelvic disproportion, non-vertex foetal presentations and mid-forceps deliveries are associated with an increased risk of obstetric palsy. In a prospective study, nulliparity and prolonged second stage of labour were identified in a multivariate logistic regression analysis as factors associated with intrinsic obstetric palsy. Maternal and foetal body habitus, mode of delivery and neuraxial analgesia/anaesthesia were not associated with nerve injury. However, the study may have been underpowered to identify this association as the rate of neuraxial analgesia/anaesthesia was over 80% and the number of women with nerve injury without neuraxial anaesthesia was small. In contrast to these findings, Dar and colleagues reported a higher rate of nerve injury in women who received neuraxial analgesia than those who did not. However, follow-up for the control group differed from that of the neuraxial group and therefore conclusions drawn from these data may be not reliable.
Theoretically, neuraxial labour analgesia may indirectly contribute to obstetric palsy by several mechanisms. Neuraxial analgesia is associated with a longer second stage of labour. Women with neuraxial analgesia-induced sensory blockade may not recognise symptoms of impending nerve injury and thus may fail to take action (shift their position) in order to relieve the nerve compression. The presence of motor blockade may contribute to an inability to easily move and reposition oneself. Women in pain tend to change positions spontaneously and frequently, while women with neuraxial analgesia often remain in one position in bed, either by choice or hospital policy. Finally, women with neuraxial analgesia often push during the second stage of labour in a semi-Fowler or lithotomy position with the thighs hyperflexed on the abdomen. This position may stretch nerves as they course from the pelvis to the lower extremities. In one study, the thigh-hyperflexion pushing position was associated with a greater incidence of nerve injury than the other positions.
Symptoms of intrinsic obstetric palsies improve or resolve in almost all patients; the median duration of symptoms is 6–8 weeks. Based on the duration of symptoms, it is likely that these nerve injuries are secondary to minor degrees of axon loss or focal demyelination.
Specific obstetric palsies
Intrinsic obstetric palsies include lesions of the lumbosacral trunk and lower extremity peripheral nerves ( Table 1 ). The most commonly injured nerve is the lateral femoral cutaneous nerve. Other plexuses and nerves include the lumbosacral plexus, and the femoral, obturator, sciatic and common perineal nerves. In two prospective studies, the incidence of obstetric palsies associated with a motor deficit was 24% and 37%.
| Nerve | Nerve roots | Sensory deficit | Motor deficit |
|---|---|---|---|
| Lateral femoral cutaneous nerve | L2–L3 | Anterolateral thigh | None |
| Femoral nerve a | L2–L4 (posterior division) | Anterior thigh, medial leg and foot | Hip flexion, knee extension, patellar reflex |
| Obturator nerve | L2–L4 (anterior division) | Medial thigh and knee | Thigh adduction |
| Lumbosacral plexus | L1–S4 | Lateral leg and dorsum of foot b | Foot dorsiflexion and eversion (footdrop) b |
| Sciatic nerve | L4–S4 | Buttocks, posterior thigh, leg and foot | Knee flexion a |
| Common peroneal nerve | Anterolateral leg and dorsum of foot | Foot dorsiflexion and eversion (footdrop) | |
| Posterior tibial | Sole of foot | Foot plantar flexion and inversion |
a The femoral nerve extends into the leg as the saphenous nerve.
b The peroneal component of the lumbosacral plexus is injured most often. Sensory and motor deficits may be more extensive if other components of the plexus are injured.
The lateral femoral cutaneous nerve (L2, L3) is a purely sensory nerve innervating the anterolateral thigh ( Fig. 1 ). Lateral femoral cutaneous nerve palsy is also known as meralgia paresthetica. The nerve emerges from the lateral border of the psoas muscle medial to the anterior superior iliac spine, and then passes under or through the inguinal ligament into the thigh. It supplies sensory innervation to the lateral thigh ( Fig. 2 ). Risk factors for meralgia paresthetica include increased abdominal pressure, pregnancy, obesity, diabetes mellitus, external pressure (by a belt or tight clothing) or prolonged hip flexion. Exacerbation of lumbar lordosis (e.g., during pregnancy) may lead to compression of the nerve by the posterior fascicle of the inguinal ligament when the nerve bisects the ligament, or may lead to stretch injury secondary to fixation of the nerve at the ligament. Indeed, meralgia paresthetica sometimes presents during pregnancy, typically after 30 weeks’ gestation. The nerve may be injured during caesarean delivery by cutting a wide incision, or stretching or pressure by a retractor.
More than 30% of the intrinsic obstetric palsies identified in a prospective study were palsies to the lateral femoral cutaneous nerve. Although injury is usually unilateral, bilateral injury has been described. Symptoms include numbness and paresthesias to the anterolateral thigh and are usually self-limiting, although the numb sensation can be quite uncomfortable and occasionally painful.
At the turn of the 20th century, postpartum femoral neuropathy was quite common. Due to modern obstetric management, the incidence has decreased markedly in the past half century, although it is still the second most common intrinsic obstetric palsy. The nerve is composed of the posterior divisions of ventral rami of the L2, L3 and L4 spinal roots. It courses through the psoas muscle and emerges from its lateral border to descend into the thigh under the inguinal ligament in the groove between the psoas and iliac muscles, providing sensory innervation to the anterior thigh and motor innervation to the iliopsoas and the quadriceps muscles of the thigh. The nerve extends into the leg as the saphenous nerve, providing sensory innervation to the medial leg and foot.
The femoral nerve does not descend in the true pelvis, making compression injury by the foetus unlikely. Postpartum femoral neuropathy often involves weakness of hip flexion, suggesting the injury is proximal to the inguinal ligament. Thigh flexion, external rotation and abduction (a common position for pushing during the second stage of labour) may compress the nerve or tether the nerve at the inguinal ligament, leading to a stretch lesion to the intrapelvic portion of the nerve. Variant slips of the psoas and iliacus muscles may split the femoral nerve in the pelvis, placing tension on the nerve and leading to nerve entrapment. Compression of the nerve at the knee may result in an isolated saphenous neuropathy. Patients complain of numbness of the anterior thigh (and sometimes the medial leg), and weakness of thigh flexion and knee extension. On examination, the patellar reflex is decreased.
The obturator nerve is composed of the anterior divisions of the ventral rami of the L2, L3 and L4 spinal roots. It descends through the psoas muscle and emerges near the brim of the pelvis over the sacroiliac joint and traverses the lateral wall of the lesser pelvis before entering the obturator foramen and dividing into two branches and descending to the thigh. The obturator nerve supplies sensory innervation to the medial thigh and posterior knee and motor innervation to the adductor muscles of the thigh.
Because the obturator nerve descends in the true pelvis, the most likely mechanism of injury during childbirth is compression of the nerve between the pelvis and foetal head or forceps applied to the foetal head. The lithotomy position exaggerates angulation of the nerve as it exits the obturator foramen. Haematomas as a complication of pudendal nerve block have led to entrapment neuropathy of the obturator nerve. Women with postpartum obturator neuropathy complain of numbness of the medial thigh and an abnormal gait secondary to weakness of thigh adduction. This weakness may be masked by compensation from muscles innervated by the femoral and sciatic nerves.
The lumbosacral plexus is composed of the lumbar plexus (ventral rami of the L1 through L4 spinal nerves) and the sacral plexus (ventral rami of L5 through S4 spinal nerves). Components of the lumbosacral plexus cross the pelvic brim, sacral ala and sacral iliac joint, or originate in the true pelvis. The lumbosacral plexus may be injured by compression of the plexus between the bony pelvis and foetal presenting part, or it may be compressed during instrumental vaginal delivery. Pelvic and birth features which predispose to plexus injury during childbirth include macrosomia, malpresentation, a straight sacrum, a flat and wide posterior pelvis, posterior displacement of the transverse diameter of the inlet, wide sacroiliac notches and prominent ischial spines. The portion of the plexus most often injured is L4–L5. As it crosses anterior to the sacral ala and sacral iliac joint, the nerve can be compressed by the descending foetal head. Because of its relationship to the bony pelvis, the common peroneal nerve component of the lumbosacral plexus is injured more often than the tibial component.
Injury to the lumbar plexus results in variable symptoms depending on the degree and site of injury to the plexus. Patients often present with foot drop and numbness of the lateral lower leg and dorsum of the foot because the common peroneal component of the plexus is more often injured. The diagnosis of the precise site of injury (e.g., plexus, sciatic or the common peroneal nerve) may require electrodiagnostic studies. Katirji and colleagues reported six women of short stature with postpartum foot drop. Electrodiagnosis testing localised the demyelinating lesions to the lumbosacral plexus. Injury to the plexus may involve dysfunction of the anal sphincter, whereas more peripheral nerve injury will not. Table 2 lists signs and symptoms of various lesions that cause foot drop.
| L5 nerve root | Lumbar plexus | Sciatic nerve | Peroneal nerve | |
|---|---|---|---|---|
| Motor | Weakness of paraspinous muscles | Weakness of gluteal muscles and anal sphincter | ||
| Ankle inversion | Weak | Weak | Normal or weak | Normal a |
| Plantar flexion | Normal | Normal | Normal or weak | Normal |
| Toe flexion | Weak | Weak | Normal or weak | Normal |
| Sensory loss | Poorly demarcated, predominately big toe | Well demarcated to L5 dermatome | Lower 2/3 of lateral leg, and dorsum of foot | Lower 2/3 of lateral leg and dorsum of foot |
| Ankle jerk | Normal b | Normal b | Normal or weak | Normal |
| Pain | Common, radicular | Common, may be radicular | Can be severe | Rare |
a Attempt at inversion should be made with the foot dorsiflexed passively to 90°.
Although less common than postpartum lateral femoral and femoral neuropathies, sciatic nerve injury has also been reported. The sciatic nerve (L5, S1, S2, S3 and S4) arises from the lumbosacral plexus and emerges from the pelvis below the piriformis muscle to enter the thigh between the ischial tuberosity and the greater trochanter. The nerve splits in the upper part of the popliteal space to become the common peroneal and the tibial nerve. The sciatic nerve and its extensions supply sensory innervation to the posterior thigh, and most of the leg and foot (except the medial aspects supplied by the femoral nerve). It supplies motor innervation to the hamstring muscles of the thigh, and the muscles of the leg and foot.
Historically, the most common cause of iatrogenic sciatic nerve injury was intramuscular injections and for this reason the gluteal muscle should be discouraged as a site for injections. Sacroiliac joint dysfunction is a common complaint of pregnancy, and the relaxed joint may cause sciatic nerve symptoms as the nerve courses anterior to the joint. Stretch injury to the nerve has been described in women placed in the lithotomy position using ‘candy cane’ stirrups (resulting in hip flexion with minimal knee flexion). Postaci and colleagues reported sciatic nerve palsy after caesarean delivery and attributed it to compression of the nerve by a wedge placed under the patient’s right buttock to laterally displace the uterus. The presentation of sciatic nerve injury is variable, depending on the specific nerve fibres that have been injured. Foot drop is common; the differential diagnosis includes lumbosacral plexopathy and peroneal nerve injury.
The common peroneal nerve arises from the lateral side of the sciatic nerve, having separated from the tibial nerve component of the sciatic nerve high in the popliteal fossa. It then winds around the fibular neck before separating into the deep and superficial branches. The superficial peroneal nerve is primarily sensory to the lateral leg and dorsum of the foot while the deep peroneal nerve primarily provides motor innervation to muscles of the ankle and foot responsible for foot dorsiflexion and eversion.
Compression of the common peroneal nerve at the head of fibula is the primary mechanism of nerve injury. During childbirth, compression can be caused by inappropriate positioning of the legs in stirrups. Postpartum peroneal nerve palsy has been attributed to prolonged external compression by the mother’s hands applying pressure to the lateral knees as she pulled her thighs up during second-stage pushing. In addition, common peroneal nerve palsy has also been attributed to prolonged maternal squatting during childbirth.
Anaesthesia-related neurologic deficits
Over the past several decades, neuraxial analgesia and anaesthesia have consistently been shown to be safe procedures, although not free from risk. Of major concern to the patient and the anaesthesiologist is the risk of permanent neurological injury after a neuraxial technique. The consequences of permanent neurological injury may be disastrous in an otherwise young and healthy obstetric patient.
Neurologic complications directly attributable to neuraxial anaesthesia/analgesia during childbirth are rare; therefore, the real incidence is difficult, if not impossible to determine. Auroy and colleagues performed a prospective audit of regional anaesthetics in France over a 10-month period in 1998–1099. Over 35 000 obstetric neuraxial procedures were reported. Serious events included three cases of respiratory failure, two seizures after epidural anaesthesia (due to local anaesthetic toxicity) and two ‘peripheral neuropathies’. The reason that peripheral neuropathies were considered complications of neuraxial anaesthesia is unclear. Moen and colleagues reported a retrospective survey in Sweden of obstetric and non-obstetric procedures. Eight complications were identified in 205 000 neuraxial procedures, including spinal haematoma, epidural abscess, trauma spinal cord injury (from a needle), intracranial subdural haematoma, permanent abducens nerve palsy and Horner’s syndrome with facial pain. The Third National Audit Project of the Royal College of Anaesthetists (United Kingdom) prospectively estimated that the incidence of permanent injury after central neuraxial block in all patients ‘pessimistically’ at 4.2 per 100 000 (95% confidence interval (CI): 2.9–6.1) and ‘optimistically’ at 2.0 per 100 000 (95% CI: 1.1–3.3). The corresponding estimates for obstetric procedures ( N = 320 425) were lower at 1.2 and 0.3 per 100 000, respectively.
The aetiology of injury directly attributable to neuraxial anaesthesia/analgesia includes direct trauma by needles or catheters, injection of unintended (toxic) substances, injection into unintended places, central neuraxial infection or haematoma, transient neurologic syndrome and vascular catastrophes.
Direct trauma
Direct trauma to the spinal cord, conus medullaris and spinal nerve roots can be caused by needles or catheters. The spinal cord terminates at the L1 vertebral body in most adults; however, it terminates below L2 in a small percentage of patients ( Fig. 3 ).
Trauma to the spinal cord or conus medullaris may result because the anaesthesiologist inaccurately estimates the vertebral level of dural puncture. Indeed, several studies have shown that anaesthesiologists are poor at identifying specific lumbar epidural interspaces. In an magnetic resonance imaging (MRI) study, experienced anaesthesiologists correctly identified the vertebral interspace 29% of the time using traditional landmarks. Sixty-eight percent of the time the actual interspace was one to four levels higher than that estimated by the anaesthesiologist. Several changes to body habitus during pregnancy may contribute to misestimation of the vertebral interspace. Tuffier’s line (the imaginary line drawn between the iliac crests) may be more cephalad in pregnant women than in non-pregnant women because of difficulty in flexing the lumbar spine. In addition, in pregnancy, the hips are generally wider than the shoulders, and in the lateral position, a line dropped perpendicular to the horizontal from the upper iliac crest will cross the spine in a more cephalad position than when the hips are narrow.
Direct injection into the spinal cord was identified in two of 89 nerve injuries in obstetric patients in the American Society of Anesthesiologists Closed Claims Project from 1990 to 2003 and in three patients in the retrospective Swedish registry study. Reynolds reported six cases of damage to the conus medullaris after spinal or combined spinal–epidural anaesthesia for obstetric procedures. All patients reported pain during needle insertion. A spinal cord syrinx was identified on MRI that corresponded with the distribution of the pain during the procedure and the postpartum neurologic deficit. These findings reinforce several aspects of spinal anaesthetic procedures that are important to avoiding direct trauma to the spinal cord, including (1) choosing a lumber puncture site below L3, (2) halting needle advancement immediately if the patient perceives any pain or paresthesia and (3) injecting anaesthetic solution only if the pain or paresthesia completely resolves.
Infection
Spinal epidural abscess and meningitis are rare, but potentially catastrophic complications of neuraxial anaesthesia. In an analysis of the American Society of Anesthesiologists Closed Claims Project database from 1980 to 1999, meningitis and abscess accounted for 46% of neuraxial injuries associated with neuraxial procedures in obstetric patients. Both meningitis and epidural–spinal abscess seem to occur less commonly in obstetric patients than in general surgical patients. In a retrospective survey study in Sweden, 29 cases of meningitis were identified after 1 260 000 spinal anesthetics for non-obstetric surgical patients, and zero cases were identified among the 55 000 women who received spinal anaesthesia for caesarean delivery. Similarly, 12 epidural abscesses were report among 450 000 surgical patients, whereas only one abscess was identified among 200 000 labour epidural procedures. The incidence of meningitis in obstetric patients estimated from 10 summated survey studies was 1 in 39 000 neuraxial procedures and the incidence of epidural abscess was estimated at 1 in 302 757. In a 2008 review of the literature, Reynolds identified 16 reports of epidural abscess in obstetric patients. All cases involved an epidural catheter; three followed a combined spinal–epidural technique and none followed single-shot spinal anaesthesia.
The most common causative organism of epidural abscess is Staphylococcus aureus . The source of S. aureus may be the direct contamination from skin or other sources, or haematogenous spread (e.g., from the vagina). In two non-obstetric cases the causative strain was isolated from a contaminated bottle of lidocaine and the nares of the anaesthesiologist. The risk of infection increases with the duration of epidural catheterisation.
Onset of symptoms of epidural abscess is usually 4–10 days after epidural catheterisation. The most common symptoms are severe backache and localised tenderness. Other signs and symptoms include fever, neck stiffness and headache, and elevated white blood cell count and erythrocyte sedimentation rate. The catheter entry site may be inflamed and there may be a fluid leak. Late-onset symptoms include radicular pain, lower limb and sacral numbness, loss of muscle stretch reflexes and bladder dysfunction. Early diagnosis and treatment are essential to a good outcome. Gadolinium MRI is the gold standard for diagnosis, followed by surgical drainage and appropriate antibiotic therapy.
Meningitis is a rare, but potentially deadly complication of neuraxial anaesthesia. It occurs more commonly after neuraxial procedures involving dural puncture and more commonly in women following labour and vaginal delivery than following scheduled caesarean delivery. A 2008 review identified 38 published reports of meningitis in obstetric patients. There are no reports of postpartum meningitis in the absence of neuraxial anaesthesia/analgesia.
Postdural puncture meningitis is most commonly caused by viridans type streptococci ( S. salivarius, sanguis, uberis ). By contrast, community-acquired bacterial meningitis is commonly caused by S. pneumoniae , N. meningitis and, less frequently, H. influenza . Viridans streptococci colonise the upper respiratory tract, the female genital tract and the gastrointestinal tract. They are not normally virulent, but they thrive in watery media such as cerebrospinal fluid (CSF). Normally the blood–brain barrier protects the central nervous system against bacterial invasion. Dural puncture may be associated with vascular trauma, and in this manner, the blood–brain barrier may be breeched. Viridans type streptococci do not grow well in conventional culture media, and as such, some reports of aseptic meningitis may have, in fact, been misdiagnosed bacterial meningitis.
Reports of iatrogenic meningitis tend to occur in clusters, suggesting that contamination occurs because of a breakdown in sterile technique. Contamination by organisms residing in the nasopharynx of medical personnel during dural puncture has been identified as a source of infection in the anaesthesiology, radiology and neurology literature, suggesting that droplet contamination plays a role in transmission.
The onset is typically 12 h to several days after delivery. Symptoms of meningitis include fever, headache, photophobia, nausea, vomiting and neck stiffness. Confusion, drowsiness and a positive Kernig’s sign (inability to straighten the knee when the hip is flexed) may also be present. Diagnostic lumbar puncture reveals an elevated CSF protein level and white blood cell count, and a CSF glucose level that is lower than the plasma glucose concentration. Because viridans streptococci are the most likely pathogens, bacteria should be cultured in broth. Treatment with the appropriate antibiotics should not await culture results. Vancomycin and a third-generation cephalosporin have been recommended as first-line treatment.
Several national organisations have published guidelines for infection control precautions for anaesthetic procedures. Traditionally, povidone–iodine has been used for skin disinfection prior to initiation of neuraxial analgesia. Chlorhexidine in alcohol is a superior disinfectant and has been recommended by the American Society of Regional Anesthesia and Pain Medicine (ASRA) for skin disinfection before neuraxial anaesthesia. Reports of transmission of airway organisms from the provider to the patient suggest that wearing a clean, good-quality mask is an important part of sterile precautions for neuraxial procedures. The ASRA recommends that the anaesthesia provider wear sterile gloves, mask and cap. Guidelines published by the Association of Anaesthetists of Great Britain and Ireland (AAGBI) recommend full-barrier precautions for neuraxial procedures, which includes sterile gloves and gown, a cap, mask and the use of a large sterile drape. Handwashing, removal of watches and jewelry and appropriate catheter dressing are also important components of sterile technique.
Spinal/epidural haematoma
In a 2000 review of the English language literature, Loo and colleagues identified no reports of spinal/epidural haematoma after spinal anaesthesia, and seven reports after epidural anaesthesia in obstetric patients. Three of the seven diagnoses were made clinically, and no imaging studies confirmed the diagnosis; three of the cases had risk factors for bleeding. Indeed, most epidural haematomas in non-pregnant patients occur in elderly patients with arterial disease, or in the setting of vascular trauma in the presence of coagulopathy. Pregnant patients are hypercoagulable and this may confer some degree of protection. A ‘bloody tap’ in a healthy parturient does not increase the risk for spinal/epidural haematoma. Spontaneous epidural haematomas during pregnancy have also been described.
Unfortunately, is it not possible to determine laboratory coagulation parameters for ‘safe’ administration of neuraxial anaesthesia. A survey of obstetric anaesthesiologists found that most feel safe initiating neuraxial anaesthesia if the platelet count is above 80 × 10 l −1 , but would not initiate a procedure if the platelet count is less than 50 × 10 9 l −1 .
Pulmonary embolism has been identified as a major cause of maternal mortality. It is increasingly common for obstetric patients to be anticoagulated with unfractionated or low-molecular-weight heparin in both the ante- and postpartum period. National bodies have promulgated guidelines for the initiation of neuraxial anaesthesia in the presence of anticoagulation, including the ASRA and the Royal College of Obstetricians and Gynaecologist.
The benefits of neuraxial anaesthesia/analgesia must be assessed against the risk of spinal/epidural haematoma in patients with risk factors for bleeding. The risk of haematoma appears to be greater with continuous epidural anaesthesia compared to single-shot spinal anaesthesia. In addition, symptoms of spinal/epidural haematoma may be masked by continuous neuraxial neuroblockade. Patients should be queried about a history of abnormal bleeding before initiation of a neuraxial procedure and physical examination may reveal evidence of abnormal coagulation. Spinal/epidural haematoma has been associated with epidural catheter removal ; therefore, coagulation status should be near normal before removal of the catheter. Lower-extremity neurologic checks may be indicated in patients at risk for spinal/epidural haematoma.
Signs and symptoms of a spinal/epidural haematoma include acute onset of back and radicular leg pain, lower extremity weakness and numbness and bladder and bowel dysfunction. These complaints should generate prompt evaluation, including neurosurgical evaluation and MRI if indicated. Adverse neurologic outcome is directly related to the duration of symptoms. For best recovery of neurologic function the haematoma should be decompressed within 6–8 h of the onset of symptoms.
Dural puncture-related nerve injury
It is not within the scope of this article to discuss post-dural puncture headache; however, several complications of dural puncture can cause focal neurologic deficits. Isolated cranial nerve palsies, most commonly cranial nerves VI (abducens nerve) and VIII (vestibulocochlear nerve) can present after dural puncture, before or after post-dural puncture headache. Patients may present with diplopia or complaints of hearing loss. The mechanism is stretching of the nerves due to loss of CSF and intracranial hypotension. Cranial nerve palsies usually, but not always, resolve. The palsies do not consistently resolve following an epidural blood patch, even though the headache usually does.
Cortical vein and venous sinus thrombosis are rare complications of pregnancy; both have been associated with dural puncture and dural puncture headache. Patients commonly present with headache and may have raised intracranial pressure, focal sensory or motor deficits, seizure or coma.
Cranial subdural haematomas have been described after obstetric epidural anaesthesia complicated by inadvertent dural puncture. The mechanism is thought to be CSF leak leading to sagging of the brain and traction and rupture of the bridging cerebral veins. Diagnosis of both venous thrombosis and cranial subdural haematoma may be difficult, as the initial headache may be attributed to a post-dural puncture (‘spinal’) headache. Persistent headache after treatment of dural puncture headache with an epidural blood patch, particularly if accompanied by altered mental status, seizures, or focal neurologic deficits, should prompt imaging studies to rule out a cranial subdural haematoma.
Chemical injury
Chemical injury of the neuraxis may lead to cauda equina syndrome or arachnoiditis. Spinal nerves in the epidural space are quite tolerant of the injection of potentially neurotoxic substances, whereas nerves in the subarachnoid space are much more susceptible to toxic injury. Nerves in the epidural space are covered by the dural cuff (i.e., dura, arachnoid and pia mater), whereas nerves in the spinal space are covered by pia mater. In addition, the sacral roots in particular are poorly myelinated. Blood flow in the epidural space rapidly removes injected toxins. There are many reports of accidental injections of substances into the epidural space (e.g., thiopental, potassium chloride and antibiotics) without permanent sequelae. By contrast, the accidental injection of toxic substances into the subarachnoid space has resulted in permanent nerve damage (e.g., cauda equina syndrome and arachnoiditis).
Many substances are neurotoxic, including drug preservatives and high doses of local anaesthetics. Spinal microcatheters were removed from the market in the United States after reports of cauda equina syndrome following failed spinal anaesthesia. Presumably, catheters were positioned in the subarachnoid space such that maldistribution of local anaesthetic occurred, resulting in failed spinal anaesthesia and a high concentration of drug surrounding individual nerve roots.
Transient neurologic syndrome
Transient neurologic syndrome (TNS) is characterised by transient pain in the lower back, buttock or lower extremities, which occurs approximately 12 h after the resolution of uncomplicated spinal anaesthesia. The syndrome is more commonly seen after spinal anaesthesia with lidocaine rather than bupivacaine. Typically, the pain is described as aching or cramping, and worse at night. It may improve with ambulation and treatment with non-steroidal anti-inflammatory drugs, and typically lasts several days. The neurologic examination is normal and permanent sequelae have not been reported. The aetiology of the syndrome is unclear; indeed, it is not known whether this syndrome represents actual neurologic injury. The incidence of TNS following spinal lidocaine appears to be lower in obstetric patients (0–7%) than in the general surgery population (10–30%).
Vascular injury
Ischaemic injury to the spinal cord is rare following neuraxial anaesthesia in obstetric patients. Blood supply to the spinal cord is via a single anterior, and bilateral posterior spinal arteries arising from the Circle of Willis. The major supply to the anterior lumbar spinal cord is the artery of Adamkiewicz. This unilateral artery usually arises from the lower thoracic or upper lumbar portion of the aorta between T9 and L2. The area of the cord supplied by the artery of Adamkiewicz is particularly susceptible to hypoperfusion. In 15% of individuals a secondary blood supply to the spinal cord ascends from the internal iliac arteries and may play a major role in spinal cord perfusion. Theoretically, the ascending artery that arises from the internal iliac artery in some patients may be at risk of compression from the foetal head or instrumentation during assisted vaginal delivery. Anterior spinal artery syndrome is characterised by a predominantly motor deficit, with or without loss of pain and temperature sensation, bowel and bladder incontinence, and sparing of vibration and joint sensations (which are transmitted in the posterior columns). Loo and colleagues summarised three case reports of anterior spinal artery syndrome in obstetric patients.
Spinal canal arteriovenous malformations may increase the risk of vascular injury during neuraxial anaesthesia. Dilated serpiginous epidural veins, supplied by small arterial feeder arteries, may extend over many segments of the spinal canal. The resulting haemangioma may increase pressure on epidural veins and reduce spinal cord blood flow and the risk of spinal haematoma or ischaemic damage and compression is increased. Pregnancy and epidural analgesia have been reported to precipitate paraplegia in previously asymptomatic patients.
Other neurologic injury
Several reports have described cases of non-specific sacral numbness after neuraxial (spinal and epidural) anaesthesia for caesarean delivery. There was no associated pain, dysesthesias or motor component and the numbness resolved within 8 weeks. The aetiology of this symptom is unclear.
Posterior reversible encephalopathy syndrome (PRES) is a CNS disorder associated with acute hypertension, renal insufficiency or immunosuppressive therapy. It has been reported in pregnancy in both normotensive and hypertensive women. Clinical signs and symptoms include headache, visual changes, seizures and mental status changes.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
