INTRODUCTION — Neuraxial blockade is the most effective and most commonly used therapy for labor analgesia and cesarean delivery anesthesia. Neuraxial techniques (ie, spinal, epidural, combined spinal-epidural [CSE]) have a strong safety record, but both transient, mild neurologic complications and serious, life-threatening complications can occur.
This topic primarily focuses on the rare, serious, and/or life-threatening neurologic complications of neuraxial blockade in obstetric patients, with emphasis on prevention, early diagnosis, and prompt treatment. Post dural puncture headache is the most common complication of neuraxial procedures in obstetric patients, and is discussed separately. (See "Post dural puncture headache".)
Obstetric nerve palsies comprise the vast majority of neurologic injuries associated with labor and delivery, and are also discussed separately. (See "Neurologic disorders complicating pregnancy", section on 'Postpartum compression neuropathies'.)
INCIDENCE OF NEUROLOGIC INJURIES — Serious neurologic complications of neuraxial anesthesia in obstetric patients are rare. The true incidence of most such complications is unknown, but it appears that they occur less often in obstetric patients than in general surgical patients. As examples, pregnant patients tend to be healthier than older adult and/or immunocompromised patients who are at greatest risk of infectious complications, such as epidural abscess. The incidence of spinal epidural hematoma (SEH) after neuraxial anesthesia procedures is far lower in obstetric than in nonobstetric patients (see 'Spinal epidural hematoma' below).
However, obstetric patients may be predisposed to other complications (eg, epidural vein cannulation, misidentification of spinal level) as a result of engorgement of the epidural veins and distortion of anatomical landmarks in late pregnancy. Estimates of the incidence of individual neurologic complications are discussed in the relevant sections below.
Sources of data on incidence of injuries — Accurate estimates of the incidence of severe complications are limited by the rarity of these events and the scarcity of prospective studies. Much of the available information comes from case reports, case series, and limited institutional cohorts, and from large, older surveys that may not reflect currently accepted practices [1-4]. Several more recent surveys have been conducted in mixed populations or in sick, older adult patients who are at higher risk for neurologic complications of neuraxial blockade; the generalizability of these findings to the obstetric population has been questioned [5-7]. Further, these and other surveys may not reflect divergent practices, skills, and training of anesthesia providers in different geographic regions, and are often underpowered to detect rare events. Reporting bias, misdiagnosis, and late diagnosis (ie, after patient discharge) are also likely to contribute to the paucity of accurate data. In some cases, spontaneous neurologic complications (eg, epidural abscess, SEH) may be attributed to neuraxial blockade due to the temporal relationship between onset of symptoms and the procedure itself.
Four sources of relevant data are the Serious Complication Repository (SCORE) project, the Anesthesia Closed Claims Project, and the Third National Audit Project (NAP3) and National Obstetric Anaesthetic Database (NOAD) from the United Kingdom. Findings from these sources are discussed in relevant sections below.
●Findings from a multicenter database, the SCORE project , are presented where applicable. The SCORE project collected data on more than 257,000 obstetric anesthetics from 30 participating (largely academic) institutions over nearly a five year period (2004 to 2009). Although this study too was underpowered to capture significant numbers of rare events, the results may be used to guide discussions of informed consent.
●When applicable, data from the Anesthesia Closed Claims Project are presented to identify the number of maternal injury claims paid on behalf of anesthesia providers from 2000 to 2011 . This database captures claims paid by professional liability insurance companies that insure an estimated 36 percent of practicing anesthesiologists from a variety of practice settings throughout the United States. These data are limited by lack of a uniform definition of standard of care among experts who participate in the claim settlement and by the absence of a denominator population for evaluating the incidence of complications.
●NAP3 was a prospective study of the major complications of neuraxial anesthesia in the United Kingdom from 2006 to 2008 . It included over 700,000 neuraxial anesthetics, 45 percent of which were performed in obstetric patients or were obstetric procedures. Only 12 major complications were reported in obstetric patients during NAP3. However, risk factors and features of neurologic complications in nonobstetric patients may be relevant for obstetric patients as well.
●The Obstetric Anaesthetist Association in the United Kingdom collects some data on complications through its NOAD survey. Data from over 3,000,000 obstetric anesthetics collected over a five year period (2009 to 2014) were published in 2020 . There were 28 cases of permanent neurologic injury reported after neuraxial analgesia/anesthesia over a three year period, for a rate of 5 per 100,000, a rate that is significantly higher than reported in other studies. However, no details of the injury, etiology, or anesthetic management were provided. In addition, since the survey is voluntary and doesn't capture all anesthetics, the results cannot be used to determine a true incidence of events.
Informed consent — For purposes of informed consent, we inform otherwise healthy pregnant patients that the risk of severe permanent and/or life-threatening neurologic complications is unknown, but likely extremely low. Specifically, we explain that infection (ie, meningitis, epidural abscess) is one of the most common causes of serious neurologic sequelae of neuraxial blockade, but still rare (see 'Incidence of meningitis' below and 'Incidence of spinal epidural abscess after neuraxial procedures' below). In contrast, epidural hematoma is among the least frequently encountered complications of neuraxial blockade in obstetric patients, with an estimated incidence of 1/200,000 to 1/250,000, or less [5,8]. (See "Adverse effects of neuraxial analgesia and anesthesia for obstetrics", section on 'Spinal epidural hematoma'.)
We explain that we take several precautions to perform epidural, combined spinal-epidural (CSE), and spinal procedures below the level of the spinal cord, and that permanent, disabling nerve or spinal cord injuries are extremely rare. Nonetheless, we encourage patients to notify us if they experience a radiating, burning sensation (ie, paresthesia) during needle or catheter insertion or severe pain upon injection of neuraxial medication, which may alert us to proximity to the spinal cord or a spinal nerve root. We explain that the risk of chemical injury, such as arachnoiditis, is unknown, but likely very low. (See 'Arachnoiditis' below and 'Spinal cord and nerve root injury' below.)
EVALUATION OF POSTPARTUM NEUROLOGIC COMPLICATIONS — A neurology consultant should be involved in the evaluation of serious neurologic complications that are of unclear or concerning etiology and/or that persist despite treatment, or persist beyond the expected timeframe. Neurologists can help determine which diagnostic aids are appropriate (eg, nerve conduction studies, electromyography, imaging studies), and the degree of urgency of evaluation. Diagnostic delay in some cases (eg, intracranial or spinal epidural hematoma [SEH], neuraxial infection) can result in permanent or life-threatening injuries.
Most postpartum nerve injuries have obstetric etiologies. However, neuraxial anesthesia techniques are often suspected as the cause. Thus, it is important for anesthesia clinicians to understand the most common obstetric nerve palsies, and to perform a careful history and physical examination to evaluate complaints. In general, peripheral nerve injuries result from obstetric trauma; injuries in a dermatomal distribution suggest a more central injury and are more likely to be attributable to neuraxial blockade. Obstetric nerve injuries are discussed separately. (See "Neurologic disorders complicating pregnancy", section on 'Postpartum compression neuropathies'.)
Incidence of meningitis — Meningitis is a rare but potentially life-threatening complication of neuraxial blockade in obstetric patients. Meningitis can occur after dural puncture during a spinal or combined spinal-epidural (CSE) technique, or after an unintentional dural puncture during an epidural technique.
●Bacterial meningitis – The reported incidence of iatrogenic bacterial meningitis in obstetric patients varies considerably in the literature. In a retrospective survey of serious complications after neuraxial procedures in Sweden between 1990 and 1999, no cases of infectious meningitis were identified from among an estimated 55,000 neuraxial procedures (mostly spinal techniques) for cesarean delivery and 200,000 epidural techniques for labor pain . A 2008 review of surveys of spinal and CSE techniques in obstetric patients reported an incidence of 1/39,000 . No cases of meningitis were identified in the Serious Complication Repository (SCORE) project , in which four cases reported as meningitis/epidural abscess by survey respondents were confirmed to have been epidural abscess after review by the study authors .
A 2017 analysis of the Anesthesia Closed Claims Project database identified three cases of meningitis in obstetric patients (including one case potentially related to a blood patch) among the 19 headache claims .
●Chemical meningitis – Chemical meningitis is a rare complication of neuraxial techniques that may result from contamination of drug solutions or equipment with detergents, preservatives, or foreign substances, although several case reports attributed to chemical meningitis may actually have been infectious in origin . Chemical meningitis is a vanishingly rare complication of neuraxial anesthesia in current clinical practice using disposable equipment.
Risk factors and ports of entry — It has been widely accepted that dural puncture is a requirement for iatrogenic meningitis but lack of a dural puncture should not eliminate meningitis (with potentially fatal consequences) from the differential, as cases of meningitis have been reported without known dural puncture .
Poor adherence to sterile technique, including failure to wear a mask, is the most common procedural risk factor for meningitis. However, multiple attempts at placement (which may predispose to accidental dural puncture during an epidural technique and to vascular trauma), bacteremia [16,17], and labor itself may also place parturients at risk. The most common route of iatrogenic infection related to poor aseptic technique is direct inoculation into the cerebrospinal fluid (CSF) of aerosolized droplets originating from the upper airway of medical personnel . In case reports in which the source of infection has been determined, DNA fragments of the causative pathogen have been matched with the medical providers' nasopharyngeal swabs . Other routes of direct contamination include contamination from incompletely sterilized skin, contaminated equipment, and contaminated operator hands.
Bacteremia can cause meningitis whether or not a neuraxial procedure is performed. The causative organism may be introduced into the CSF by dural puncture after a traumatic vascular injury during a neuraxial procedure. Meningitis can also result from meningeal invasion by pathogenic organisms after hematogenous spread, unrelated to neuraxial procedures. (See "Pathogenesis and pathophysiology of bacterial meningitis", section on 'Pathogenesis'.)
Risk factors for bacteremia during labor include underlying vaginal infections and manual removal of the placenta. Group B Streptococcus, which may be present in the parturient's vagina and bloodstream, has been implicated in some cases of iatrogenic meningitis after uncomplicated epidural techniques [20,21].
Most reported cases of meningitis in obstetric patients have occurred in patients who had neuraxial blockade for labor, rather than for elective cesarean delivery . Patients who labor may be at increased risk because labor can precipitate a mild bacteremia, labor and delivery suites are less sterile than operating rooms, and antibiotics are not routinely administered prior to vaginal delivery.
Clinical presentation of meningitis — Iatrogenic bacterial meningitis usually presents within 6 to 36 (often within the first 24) hours of performance of a neuraxial technique, often in a fulminant manner . Headache, usually severe, is a common initial presentation, and must be distinguished from post dural puncture headache, preeclampsia, posterior reversible encephalopathy syndrome, and other concerning sources of headache in the peripartum period. Neck pain and rigidity, fever, mental status changes (eg, disorientation, confusion, lethargy), photophobia, and nausea and vomiting are typical. Kernig's sign (inability to straighten the leg below the knee when the hip is flexed) and Brudzinski's sign (involuntary knee and hip flexion when the neck is flexed) should also prompt suspicion for meningitis. If untreated, seizures, cerebral edema, coma, and maternal death may ensue. (See "Clinical features and diagnosis of acute bacterial meningitis in adults".)
If meningitis is suspected, a neurology consultation should be obtained for further diagnosis, which is discussed separately. (See "Clinical features and diagnosis of acute bacterial meningitis in adults".)
Pathogens — The causative organisms in iatrogenic (ie, post dural puncture) infectious meningitis differ from those commonly seen in community-acquired meningitis (ie, Neisseria meningitidis, Streptococcus pneumoniae, or Haemophilus influenzae). Strains of Viridans group streptococci (Streptococcus salivarius, Streptococcus sanguinis, Streptococcus uberis, Streptococcus vestibularis, Streptococcus mitis, etc) are the dominant causative organisms in iatrogenic meningitis. These organisms are commensals found in the mouth, upper airway, and vagina, and are normally of low virulence but thrive in watery media, such as the CSF . In a review of 176 cases of post dural puncture meningitis between 1952 and 2005, Viridans group streptococci were isolated in 76 percent of cases in which a pathogen was identified, with S. salivarius implicated most often .
Other, less common causative organisms of iatrogenic meningitis include Pseudomonas aeruginosa and Staphylococcus aureus, as well as several other organisms that have been identified only in individual case reports. In many cases, no organism is identified . (See "Lumbar puncture: Technique, indications, contraindications, and complications in adults", section on 'Infection'.)
Treatment of meningitis — Outcomes are favorable if iatrogenic bacterial meningitis is diagnosed promptly and treated properly. Treatment and prognosis of meningitis are discussed separately. (See "Initial therapy and prognosis of bacterial meningitis in adults".)
Prevention of infectious complications — Adherence to strict aseptic technique is the most effective way to minimize the risk of bacterial meningitis and other infectious complications of neuraxial techniques. We adhere to the American Society of Anesthesiologists (ASA) and the American Society of Regional Anesthesia and Pain Medicine (ASRA) updated recommendations for the prevention of infectious complications of neuraxial blockade , as described separately. (See "Spinal anesthesia: Technique", section on 'Aseptic technique'.)
In addition to basic aseptic technique, we follow these procedures to minimize the risk of neuraxial infection:
●Prepare the patient's skin with a single-use ChloraPrep applicator (not supplied in commercially prepared epidural and spinal kits) as per packet instructions. (See "Lumbar puncture: Technique, indications, contraindications, and complications in adults", section on 'Aseptic technique'.)
●Administer preprocedure antibiotic therapy in patients with known or suspected systemic bacterial infection, or consider an alternative to neuraxial blockade on a case-by-case basis.
●Replace catheters after unwitnessed catheter disconnects. If a catheter disconnect is witnessed, we may replace the catheter, remove the catheter without replacement (ie, if delivery is imminent), clean the exposed tip with alcohol or ChloraPrep, or cut the catheter with sterile scissors below the disconnect site, depending on the clinical scenario.
●Remove epidural catheter as soon as clinically appropriate, and without delay in the presence of any signs of infection (eg, erythema at the insertion site).
Given the lack of evidence that the use of bacterial filters reduces infectious complications, we generally do not apply filters during short-term continuous epidural infusions [23-27].
SPINAL EPIDURAL ABSCESS
Incidence of spinal epidural abscess after neuraxial procedures — Spinal epidural abscess (SEA) is a rare complication of neuraxial blockade, with a reported incidence ranging from 1/1,930  to 1/205,000 , depending on the patient population. (See "Spinal epidural abscess", section on 'Epidemiology'.)
Pregnant patients appear to be relatively protected from this potentially life-threatening complication. Three cases were found among 1,331,171 epidural procedures for labor and delivery in a combined review of nine surveys spanning 1987 to 2009 . Another four cases were reported among over 251,000 obstetric neuraxial procedures (including combined spinal-epidural [CSE], spinal, continuous spinal, and epidural techniques) in the Serious Complication Repository (SCORE) project . Each of the four reported cases involved use of an epidural catheter. The combined data from these 10 surveys yields an incidence of approximately 1/208,770 after epidural procedures in obstetric patients; these data do not reflect several additional individual case reports in the literature.
Iatrogenic SEA formation after neuraxial techniques accounts for only a small portion (5 to 5.5 percent) of all cases of SEA. Most confirmed cases have been reported with the use of an epidural catheter, in either a CSE or standard epidural technique [10,12,30]. (See "Spinal epidural abscess", section on 'Pathogenesis'.)
Risk factors for spinal epidural abscess
●Patient risk factors – Risk factors for SEA that have been identified in other patient populations (eg, diabetes, tuberculosis, HIV, malignancy, intravenous drug abuse, alcohol dependence) may be present in obstetric patients. Obstetric patients may also have altered immunity related to pregnancy, or take immunosuppressive therapy for inflammatory conditions. (See "Spinal epidural abscess", section on 'Epidemiology'.)
●Procedural risk factors – While cases of iatrogenic abscess have been reported after uneventful single-shot spinal and caudal block, placement of an epidural catheter is the single greatest risk factor for SEA attributed to neuraxial procedures in obstetric patients. Epidural techniques have been implicated more commonly than CSE techniques in reported cases, although this may be a reflection of the higher number of epidural procedures performed. Placement of a catheter requires a large bore needle, with more tissue disruption for bacterial access, and ensures a continuous tract for bacterial spread. Iatrogenic abscess has also rarely been reported after uneventful single-shot spinal and caudal block in nonobstetric patients .
Several other procedure-related risk factors may predispose patients to SEA formation:
•Poor aseptic technique
•Multiple procedure attempts
•Prolonged catheterization, which is less likely in obstetric patients
•Placement in a contaminated environment (eg, soiled bed in the labor and delivery suite)
•Localized or systemic infection at the site/time of initiation of the blockade
•Administration of an opioid without local anesthetic; local anesthetics have antimicrobial effects
The presence of a small collection of blood after a traumatic epidural technique has been postulated to provide a nidus for infection, though SEA has not been reported after injection of autologous blood for a blood patch.
Clinical presentation of spinal epidural abscess — The presentation of epidural abscess can be highly variable, with respect to both clinical signs and symptoms and timing of diagnosis. Only a small percentage of patients present with the "classic triad," of back pain, fever, and neurologic changes . (See "Spinal epidural abscess", section on 'Clinical manifestations'.)
Back pain, often severe and progressive, is the most consistent finding , and typically starts 4 to 10 days after neuraxial instrumentation , although the interval may be much longer. The pain may extend to the neck and radiate to the sacral region and lower extremities. Fever is common and often precedes neurologic changes, when present . Progressive focal neurologic deficits may include sensory loss restricted to one or several dermatomes, lower extremity weakness and gait disturbance, radiculopathy, sacral numbness, bladder dysfunction, decreased reflexes, and paralysis. Local infection or irritation, tenderness to palpation at the catheter insertion site, and inflammation of the paraspinous muscles may also be present.
Neurologic compromise from epidural abscess may result from ischemia due to mechanical compression of the spinal cord and its vasculature by the space-occupying lesion; thrombophlebitis and associated venous thrombosis precipitated by infection may also play a role  by occluding blood flow to the spinal cord.
Diagnosis of spinal epidural abscess — SEA should always be considered after neuraxial anesthesia in febrile patients with back pain, and particularly when accompanied by an examination consistent with radiculopathy or focal neurologic findings . Spinal imaging (gadolinium-enhanced magnetic resonance imaging [MRI]) should be performed as soon as the diagnosis is considered, and the appropriate surgical team should be notified if imaging is consistent with SEA. (See "Spinal epidural abscess", section on 'Diagnosis'.)
Breastfeeding does not need to be interrupted for administration of gadolinium. (See "Diagnostic imaging in pregnant and nursing patients", section on 'Risks to breastfed infants'.)
Laboratory markers for SEA are non-specific. An elevated erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and peripheral white blood cell count should raise suspicion for abscess in the presence of persistent or worsening back pain after neuraxial procedures, particularly in the presence of any risk factors. (See "Spinal epidural abscess", section on 'Laboratory findings'.)
Once an SEA has been identified, it is important to try to isolate the etiologic agent from samples of blood, abscess contents, the epidural catheter tip if still in place, or from the catheter insertion site (if there is evidence of superficial infection) (see "Spinal epidural abscess", section on 'Diagnosis'). A diagnostic lumbar puncture is not recommended unless there is clinical suspicion for meningitis, as it may inadvertently introduce the causative organism into the cerebrospinal fluid (CSF) or precipitate neurologic deterioration.
Pathogens — The most common causative organism of epidural abscess is S. aureus. (See "Spinal epidural abscess", section on 'Microbiology'.)
Ports of entry of infection in spinal epidural abscess — The route of infection in cases of SEA associated with neuraxial techniques differs from the mechanism of infection in most other cases of SEA, in which hematogenous spread and contiguous spread from a nearby source of infection are the predominant mechanisms . In iatrogenic SEA in obstetric patients, bacteria can gain access to the epidural space via several mechanisms, including direct contamination from the catheter entry point (ie, the patient's skin or hair follicles, aerosolized droplets originating from the medical personnel's upper airway [36-38]), migration of organisms along the catheter, contaminated equipment or injectate, and, rarely, hematogenous spread. After traumatic neuraxial procedures, small blood collections (ie, hematomas) may provide suitable conditions for bacterial growth, after contiguous or hematogenous seeding [32,33,39,40].
Contaminated local anesthetics from multidose bottles, which are used for individual patients and immediately discarded in current practice, have been implicated in older cases .
Abscess formation in obstetric patients may also be unrelated to the neuraxial procedure, complicating the diagnosis. Spread can occur from a nearby infection, such as from adjacent muscle or bone (eg, osteomyelitis, discitis, tuberculosis with bone involvement). Of note, an epidural abscess that results from spread of a contiguous bone infection (eg, a vertebral body) is likely to originate in the anterior epidural space , as distinct from a catheter-related abscess, which is more likely to present in the posterior (dorsal) epidural space. Although less common in the obstetric population, epidural abscess can develop from hematogenous spread from a distant focus (eg, skin and soft tissue, and genitourinary, gastrointestinal, and respiratory tracts). As an example, at least two cases of SEA after epidural techniques in obstetric patients have been attributed to hematogenous spread of Group B streptococcal infection isolated from the vagina [39,41]. In many cases of SEA in both obstetric and nonobstetric patients, the source of infection cannot be identified [40,42].
Treatment of spinal epidural abscess — The mainstays of management of catheter-related epidural abscess include prompt diagnosis, immediate catheter removal (if still in place), early initiation of broad-coverage antibiotics (tailored later to the specific pathogen), and decompressive laminectomy as soon as possible after diagnosis for most patients. Delay in diagnosis and treatment may result in permanent neurologic deficits. Choice and duration of antibiotic treatment, options for treatment, and prognosis after SEA are discussed separately. (See "Spinal epidural abscess", section on 'Management'.)
Prevention of spinal epidural abscess — Epidural catheters should be removed as soon as they are no longer necessary, and at any sign of infection at the insertion site. Preventive measures for SEA are similar to the strategies for prevention of meningitis. (See 'Prevention of infectious complications' above.)
Spinal epidural hematoma — Spinal epidural hematoma (SEH) is a rare complication of neuraxial blockade in healthy obstetric patients. Proposed reasons for protection from SEH include the hypercoagulable state of pregnancy, a relatively low incidence of spinal pathology (eg, spinal stenosis), and a lower likelihood of therapeutic anticoagulation than other patient populations. . Risk factors for SEH include hematologic abnormalities (eg, overt coagulopathy, severe thrombocytopenia, disorders of platelet function), administration of antithrombotic medications, and preexisting spinal deformity (eg, spinal stenosis, intraspinal neoplasm).
The incidence of epidural hematoma in obstetric patients, risk factors for SEH, and neuraxial anesthesia in obstetric patients with risk factors for bleeding (ie, thrombocytopenia, preeclampsia, HELLP syndrome, or patients receiving antithrombotic or anticoagulant medications) are discussed separately.
Intracranial hemorrhage — Intracranial hemorrhage (most commonly subdural hematoma) is a very rare but potentially lethal complication of dural puncture during neuraxial anesthesia. It should be considered in the differential diagnosis in patients with delayed onset of headache after dural puncture, (up to approximately 72 hours), and in patients whose headache becomes non-positional, who develop other neurologic symptoms, or whose symptoms persist after one or two epidural blood patch procedures. (See "Post dural puncture headache", section on 'Differential diagnosis'.)
Incidence of arachnoiditis — Arachnoiditis is an extremely rare and potentially devastating complication of neuraxial procedures in obstetric patients, confined to case reports and case series [43-47]. Arachnoiditis has not been reported in any large-scale surveys of serious complications of neuraxial blockade in obstetric patients, including the Serious Complication Repository (SCORE) project. One nonobstetric case of severe generalized arachnoiditis was reported in the Third National Audit Project (NAP3) survey from the United Kingdom .
Mechanism of injury — Arachnoiditis has been reported after injection of sulfite-containing preservatives in local anesthetic solutions [49-51], inadvertent injection of skin antiseptic solution into the epidural space [45,52], accidental intrathecal injection of a large volume of a potentially neurotoxic local anesthetic , and unintentional injection or spread of blood into the subarachnoid or subdural space during or after an epidural blood patch procedure [43,53,54]. Cases after single-shot spinal anesthesia have been attributed to neurotoxicity of local anesthetic or accidental introduction of contaminants (eg, residual detergents) into the subarachnoid space .
Precipitating agents are thought to provoke an inflammatory response that results in fibrosis and adhesions between nerve roots and the leptomeninges . Ultimately, collagen deposits encapsulate the nerve roots and interrupt their blood supply, resulting in nerve atrophy . In some cases, soft tissue replaces the tethered, atrophic nerve roots in the subarachnoid space and obstructs cerebrospinal fluid (CSF) flow . A link between routine, uncomplicated obstetric epidural blockade and arachnoiditis has not been established.
In the previously mentioned case from the NAP3, a young man had two attempts at spinal anesthesia, one of which resulted in persistent paresthesia, while the other attempt was uneventful but resulted in inadequate block requiring general anesthesia . The cause of this complication was not determined.
Clinical presentation of arachnoiditis — Arachnoiditis can present with a variety of symptoms [58,59]. Some patients who ultimately went on to develop arachnoiditis complained of headache or radicular pain during the neuraxial injection . Persistent or worsening low back pain is the most common initial complaint. The pain, often severe, may radiate to the perineum and lower extremities, and is typically described as a tingling or burning sensation . Headache, nausea, dizziness, diminished reflexes, muscle spasms in the back and legs, sensory changes (eg, numbness in the buttocks and lower extremities, allodynia within approximate dermatomes ), bowel and bladder dysfunction, and motor weakness or paralysis have also been described . Permanent quadriplegia may ultimately result.
In cases of excessive scarring leading to obstruction of CSF flow or in the presence of a syrinx, patients may present with signs and symptoms of elevated intracranial pressure.
Arachnoiditis may present weeks to decades after the precipitating event; some case reports describe severe pain during the neuraxial procedure, followed by a more rapid decline in neurologic status .
Because many of these signs and symptoms are associated with other devastating neurologic disorders, epidural abscess and other central nervous system infections (eg, meningitis), spinal cord tumors, and epidural hematoma must be considered in the differential diagnosis. In some cases, other severe neural compressive disorders, such as cauda equina syndrome, arachnoiditis ossificans, and syringomyelia , present as complications of arachnoiditis. Magnetic resonance imaging (MRI) should be performed if arachnoiditis is suspected.
Prevention of arachnoiditis — Routine precautions must be observed to avoid neuraxial injection of neurotoxic substances and minimize the remote risk of arachnoiditis associated with neuraxial blockade, as follows:
●Chlorhexidine with alcohol (ChloraPrep) solution applied for skin antisepsis should be allowed to dry for three minutes before inserting a needle. Appropriate use of single use ChloraPrep applicators has not been proven to cause neurologic complications. (See "Lumbar puncture: Technique, indications, contraindications, and complications in adults", section on 'Aseptic technique'.)
We use single-use, color-tinged chlorhexidine with alcohol applicators (ChloraPrep), which can be distinguished from other solutions in and around the sterile field. We ensure that the antiseptic solution does not contaminate the local anesthetics and sterile equipment by applying it before opening the spinal and epidural kits, or by activating the single-use applicator away from the sterile field. We permit the antiseptic solution to dry prior to needle insertion, and keep the stylet in place while piercing the skin.
●Epidural catheters should be aspirated frequently for CSF, to detect misplaced catheters and avoid accidental injection of large volumes of potentially neurotoxic local anesthetic (and inadvertent high spinal). When dosing an in situ epidural for cesarean delivery anesthesia, we aspirate for CSF prior to administration of each small aliquot of lidocaine, or, in emergency cesarean delivery, prior to administration of large volumes of chloroprocaine. Unrecognized spinal catheters are one of the most frequently encountered complications in obstetric anesthesia and often occur despite an uneventful test dose; frequent aspiration of an epidural catheter for CSF is highly sensitive and specific for detecting misplaced catheters . The use of flexible wire-reinforced catheters rather than stiffer nylon catheters may reduce the risk of catheter migration [62-65].
●During an epidural blood patch procedure, we do not inject blood until we ensure that the needle tip is in the epidural space. We reconfirm loss of resistance to saline between syringes of blood (if more than one syringe is used), and reidentify the epidural space if there is any doubt. (See "Post dural puncture headache", section on 'EBP technique'.)
●Spinal microcatheters are no longer used for continuous spinal anesthesia, to avoid possible neurotoxicity and cauda equina syndrome. (See "Spinal anesthesia: Technique", section on 'Continuous spinal'.)
●In current clinical practice, preservative-free local anesthetic solutions are routinely used for neuraxial administration. Adjuvants used for neuraxial anesthesia and analgesia, such as morphine, hydromorphone and epinephrine are preservative free.
Treatment of arachnoiditis — Treatment for arachnoiditis is largely supportive, and includes medications for pain relief, glucocorticoids, and, occasionally, spinal cord stimulation. Patients with severe or deteriorating symptoms may elect to undergo surgical removal of scar tissue (ie, microlysis). However, outcomes are mixed, ranging from no neurologic improvement, to slight improvement, to a decline in functional status and ultimately paraplegia . In the setting of elevated intracranial pressure from obstruction of CSF flow (due to excessive scarring), ventriculoperitoneal shunting (various shunting procedures) and surgical decompression of the foramen magnum may be necessary [43,67]. A shunt may also be necessary to drain a syrinx and prevent progression of the symptoms .
SPINAL CORD AND NERVE ROOT INJURY — Injury to the spinal cord or nerve roots can occur as a result of neuraxial needle or catheter trauma, or injection of anesthetic solution into the spinal cord or nerve. Injury can also occur after inadvertent injection of a drug intended for another route of administration (eg, neuraxial administration of medication intended for the intravenous route).
Incidence of spinal cord or nerve root injury — The incidence of direct spinal cord or nerve root injury related to neuraxial procedures in obstetric patients is unknown, but likely very low. In large reviews of serious complications of neuraxial procedures, very few cases of direct spinal cord or nerve root trauma have been reported in obstetric patients [2,3,5,10,69]. In a prospective study of complications associated with 122,989 obstetric neuraxial procedures in the United Kingdom, no cases of spinal cord trauma occurred . There were 46 cases of postpartum neuropathies involving a single spinal or a peripheral nerve (ie, 3.4 per 10,000 cases). Most reported cases of traumatic injury involve single nerve roots and resolve over time. However, permanent injury, including damage to the conus medullaris and paraplegia, can occur [70,71].
An analysis of obstetric anesthesia data from the Anesthesia Closed Claims Project database reported that 35 claims (14 percent of all claims) were for nerve injury felt to be associated with neuraxial anesthesia techniques . Fifteen claims involved spinal cord injury (including one epidural abscess and two cases of arachnoiditis), and severe permanent injury occurred in one-third of the claims for nerve injury.
Risk factors for spinal cord or nerve root injury — Permanent, severe neurologic deficits resulting from spinal cord or nerve root injury are more likely to occur during a spinal technique than an epidural procedure.
The subarachnoid space is more vulnerable to permanent injury as a result of chemical injury because its nerve roots (particularly the sacral nerve roots) are poorly myelinated and have little protection from chemical irritants and large volumes of potentially neurotoxic local anesthetics that may pool in the area. Nerve roots in the epidural space, in contrast, are surrounded by three meningeal protective layers; injury in the epidural space is more likely to resolve over the course of days to months.
Accidental administration of medications intended for intravenous use (eg, tranexamic acid, chemotherapy drugs, potassium chloride, ephedrine, etc) into the subarachnoid space also has greater potential for harm, including hemodynamic and neurologic changes and death, than wrong route errors into the epidural space. A number of reports of maternal death after accidental subarachnoid administration of tranexamic acid were reported between 2005 and 2017 .
Direct spinal cord injury resulting from needle trauma or injection of drug solution is more likely to occur if the needle is inserted at a level higher than the L3-L4 interspace. Among obstetric anesthesia closed malpractice claims for spinal cord injury from 2000 to 2011, eight patients had magnetic resonance imaging (MRI) evidence suggestive of initiation of neuraxial blockade above the intended L3-4 interspace (most commonly between T11 and L1) . Patient obesity contributing to difficulty performing the neuraxial procedure and patient movement during the procedure were identified as complicating factors.
Certain vertebral anomalies and preexisting spinal disease also can predispose patients to cord damage after neuraxial blockade, including a low-lying conus medullaris, severe spinal stenosis, and tethered cord syndrome. Coexisting disease (eg, diabetes) may also be implicated in neurologic injury after neuraxial procedures.
Clinical presentation and outcome — Paresthesias often herald needle or catheter contact with nerve roots or the spinal cord. Thus, when a patient experiences a paresthesia, the needle should be redirected, and the spinal level reassessed. It is generally considered acceptable to proceed if the paresthesia resolves completely. Although paresthesias are neither sensitive nor specific for intraneural injection, all seven of the patients with permanent disability after spinal and combined spinal-epidural (CSE) blockade in a series of cases experienced pain during spinal needle insertion on the side that developed deficits . Similarly, seven of eight of the cases with MRI evidence of spinal cord damage following neuraxial blockade in the Anesthesia Closed Claims Project reported severe paresthesia or pain on injection during the procedure .
Outcomes after spinal cord injury related to neuraxial anesthesia tend to be poor. In a series of seven cases of spinal cord injury after spinal or CSE anesthetics, all patients suffered persistent unilateral sensory loss, with or without urinary symptoms and foot drop, and most had MRI evidence of lesions (eg, an infarct, intramedullary hemorrhage, or fluid collection) in the conus medullaris . Direct injection into the parenchyma of the spinal cord is most likely to result in paraplegia.
Prevention of spinal cord and nerve root trauma — Direct trauma to the spinal cord can be minimized by taking precautions to identify the correct intervertebral space; redirecting the needle in the presence of persistent paresthesia; avoiding injection of neuraxial medication if the patient experiences severe pain; and avoiding neuraxial blockade in the presence of a tethered cord, severe spinal stenosis, and some other spinal pathologies (see "Obstetric and nonobstetric anesthesia for patients with neurologic disorders", section on 'Spine and spinal cord abnormalities'). Trauma to nerve roots can be minimized by taking the same precautions with respect to paresthesias, and possibly by using soft tip wire reinforced epidural catheters.
●Paresthesias – Because a large percentage of patients may not spontaneously report a paresthesia during needle or catheter insertion, we routinely ask the parturient if she is experiencing a burning, tingling, or shock-like sensation that radiates to a lower extremity. We redirect the spinal or epidural needle in the setting of paresthesias, and we do not inject intrathecal or epidural medications in the presence of severe pain. If the paresthesia or pain resolves completely, we proceed with needle or catheter advancement.
●Spinal needle placement – Spinal needles for spinal or CSE techniques should always be placed at or below the level of the L3-L4 interspace to avoid injury to the conus medullaris. The termination of the conus medullaris varies between the middle one-third of the T12 vertebra, and the upper one-third of L3 , but may be even lower in patients with spinal cord abnormalities (eg, tethered cord). (See "Spinal anesthesia: Technique", section on 'Anatomy'.)
In our practice, we increasingly rely on ultrasound guidance to identify the correct interspace. Palpation and visual inspection of surface anatomic landmarks are often unreliable for identification of the correct interspace, particularly in parturients, and clinicians often enter a higher interspace than intended during neuraxial anesthesia [74-77]. The intercristal line (ie, the line between the superior aspects of the iliac crests) is typically used as a rough guide for spinal needle placement. In many patients, this line crosses the L4 spinous process or L4-L5 interspace, though it may cross the spine anywhere from L1 to L2 to L4 to L5, and tends to be higher in obese and female patients [73,75]. The intercristal line may be higher in parturients as a result of exaggerated lumbar lordosis, or fat deposition over the iliac crests. In a study of 50 parturients who were to undergo neuraxial anesthesia, ultrasound was used to determine the interspace identified by clinicians using the intercristal line . The intercristal line was at least one vertebral level higher than the ultrasound-determined level 40 percent of the time.
●Flexible epidural catheters – We use flexible reinforced epidural catheters and thread the catheter no more than six cm into the epidural space to minimize the risk of damage to nerve roots.
Flexible reinforced catheters reduce the incidence of paresthesia compared with stiffer catheters, and presumably reduce the incidence of nerve injury (see "Epidural and combined spinal-epidural anesthesia: Techniques", section on 'Epidural anesthesia equipment'). Flexible catheters may also curl or change course as they brush against nerve roots or other obstacles in the epidural space .
●Prevention of wrong route administration of medication – Every effort should be made to reduce or eliminate the risk of administration of an incorrect drug or administration via an incorrect route. Strategies for risk reduction are discussed separately. (See "Prevention of perioperative medication errors", section on 'Solutions for wrong route errors'.)
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Bacterial meningitis in adults".)
SUMMARY AND RECOMMENDATIONS
●Incidence – Serious neurologic complications of obstetric neuraxial analgesia and anesthesia techniques are very rare. Most postpartum nerve injuries have obstetric etiologies, but neuraxial procedures can cause both transient, mild neurologic complications, as well as serious, life-threatening neurologic injuries. (See 'Incidence of neurologic injuries' above.)
●Neurology consultation – A neurology consultant should be involved in the evaluation of all postpartum neurologic complaints. (See 'Evaluation of postpartum neurologic complications' above.)
●Meningitis – Iatrogenic meningitis can occur after dural puncture during a spinal or combined spinal-epidural (CSE) technique, or after an unintentional dural puncture during an epidural technique (see 'Meningitis' above). The most important risk factor for iatrogenic meningitis is poor aseptic technique during neuraxial procedures. (See 'Risk factors and ports of entry' above.)
●Spinal epidural abscess – Placement of an epidural catheter is a risk factor for spinal epidural abscess (SEA). SEA should always be considered after neuraxial anesthesia in febrile patients with back pain, and particularly when accompanied by an examination consistent with radiculopathy or focal neurologic findings. Treatment includes prompt diagnosis, appropriate antibiotic coverage, and decompressive laminectomy as soon as possible after diagnosis for most patients. (See 'Spinal epidural abscess' above.)
●Spinal epidural hematoma – Spinal epidural hematoma (SEH) following neuraxial blockade is rare in obstetric patients. Risk factors include hematologic abnormalities (eg, overt coagulopathy, disorders of platelet function, severe thrombocytopenia), administration of antithrombotic medications, and preexisting spinal deformity (eg, spinal stenosis, intraspinal neoplasm). (See 'Spinal epidural hematoma' above.)
●Arachnoiditis – Arachnoiditis is an extremely rare complication of neuraxial blockade that has been reported after inadvertent neuraxial injection of chemical substances, local anesthetic preservatives, neurotoxic local anesthetics, or subarachnoid blood. Paraplegia can result. (See 'Arachnoiditis' above.)
●Spinal cord and nerve root injury – Injury to the spinal cord or nerve roots can result from neuraxial needle or catheter trauma, or injection of local anesthetic solution into the spinal cord or nerve. Trauma to the spinal cord can be minimized by avoiding spinal needle placement (single-shot spinal or CSE techniques) above the L3-L4 interspace and avoiding neuraxial procedures in some patients with spinal abnormalities. (See 'Prevention of spinal cord and nerve root trauma' above.)
For patients who report paresthesias during neuraxial procedures, it is reasonable to proceed with needle or catheter advancement if they resolve. For persistent paresthesias or pain, the needle should be redirected or the catheter should be withdrawn, and medication should not be injected.