This shows that a substantial portion of the claims against regional anesthesia complications are for temporary injury. Reproduced with permission of Wolters Kluwer Health, Inc. Lee LA, Posner KL, Domino KB, Caplan RA, Cheney FW. Injuries associated with Regional Anesthesia in the 1980s and 1990s—a closed claims analysis. Anesthesiology 2004; 101:143–52 
The primary damaging events for neuraxial anesthesia claims 1980–1999, obstetric versus nonobstetric cases
Obstetric (n = 368), no. (%)
Non-obstetric (n = 453), no. (%)
187 (51 %)*
186 (41 %)
62 (17 %)
84 (19 %)
Neuraxial cardiac arrest
20 (5 %)*
61 (13 %)
40 (11 %)*
7 (2 %)
21 (6 %)
19 (4 %)
27 (7 %)*
8 (2 %)
Unintentional intravenous injection
7 (2 %)
Other anesthetic event
58 (16 %)
51 (11 %)
55 (15 %)
75 (17 %)
25 (7 %)
33 (7 %)
14 (4 %)
19 (4 %)
9 (2 %)*
36 (8 %)
9 (2 %)*
30 (7 %)
Wrong drug or dose
8 (2 %)
15 (3 %)
3 (1 %)
5 (1 %)
0 (0 %)
3 (1 %)
Hypotension following central neural blockade in the obstetric patient occurs probably in more than 50 % of our patients [6, 7]. It may be due to several causes, chief of which is vasodilatation below the site of the block and hence pooling of blood in the lower limbs . When the venous return to the heart is decreased, the cardiac output will naturally fall. With the parturient placed in the supine position, venous pooling is made worse by the presence of the term fetus obstructing the venous return following the blockade.
The extent of the obstruction and the decreased venous return is rarely appreciated. I once had a provider who was advised to place the parturient in the left lateral tilt position following a supervised spinal done in the sitting position. The provider failed to follow the instruction. Help was requested when the blood pressure dropped to near 45 mmHg systolic after the spinal blockade. In spite of the administration of fluids and vasopressor agents to improve the situation, the blood pressure continued to be low. It only improved when the patient was properly positioned as initially advised.
Most patients who need central blockade for vaginal delivery or cesarean section are required to fast and not eat solids or liquids. While a short fast may be well tolerated, longer fasting periods in an active parturient or a parturient moving around in pain, often leads to dehydration that may not be recognized and this can have significant impact on the cardiovascular system.
There are strategies to overcome this problem. Coloading with crystalloids  during the insertion of the central neural blockade is now the preferred technique as opposed to preloading which was previously used. Crystalloids are frequently used although colloids are often thought to be a better option. McDonald et al. found in a randomized controlled trial that there was no significant difference in the maternal cardiac output when these fluid management strategies were compared . The cardiac output in the colloid group was, however, more sustained.
Vasopressors are often administered when coloading of fluids and positioning fail to sustain the blood pressure. There are providers who also administer vasopressors prophylactically in an attempt to prevent hypotension from occurring so as to reduce its impact on the parturient and the fetus [23, 24]. Previously ephedrine in small boluses of 5–6 mg per dose was favored as it has both direct (alpha and beta agonist) and indirect (release of norepinephrine from presynaptic terminals) effects and the beta agonist property was thought to better maintain uterine blood flow . Ngan Kee et al. showed that placental transfer of ephedrine causes a slightly lower pH value of the umbilical blood of the newborn at the time of birth, in those whose mothers received ephedrine compared to phenylephrine . A systematic analysis of the data of several studies confirmed the consistent finding of a slightly higher pH in those patients who were treated with phenylephrine for hypotension but there was definitely no fetal acidosis even in those who were given ephedrine . The jury may be out as to the significance of this but there is now a definite shift to the use of phenylephrine to manage hypotension as it appears to be more effective .
Phenylephrine is more expensive than ephedrine and is not so readily available . It has a rapid onset but is short acting and causes reflex bradycardia when the blood pressure improves. Ngan Kee et al. are now looking at the use of norepinephrine which with its weak beta agonist property may have less effect on the heart rate and cardiac output . We are unlikely to start using this vasopressor in obstetrics any time soon. There are still many unknowns in the causation of hypotension following spinal blockade in many of the parturients we see everyday .
Local Anesthetic Toxicity: Maternal Rescue with Lipid
Local anesthetic agents are administered in greater quantities during an epidural blockade compared to a spinal blockade. This is because significant quantities of the local anesthetic agents have to diffuse from the epidural space to the intrathecal space to block the nerve roots concerned. Large quantities of local anesthetic drugs are required to achieve an adequate level of epidural blockade in obstetric patients. These patients are at risk for systemic toxicity if the epidural catheter is inadvertently inserted into a vascular structure. Subsequent injections of local anesthetic through the catheter result in a rapid increase in blood concentration of the local anesthetic thereby causing systemic toxicity.
Systemic toxicity primarily targets the CNS and cardiovascular systems . The initial symptoms usually involve the CNS and include dizziness, tinnitus, or convulsions but with bupivacaine where the neurological/cardiovascular risk ratio is narrowed [32, 33], toxic manifestations in the form of life-threatening cardiac arrhythmias, cardiac collapse, or even cardiac arrest may occur in the early phase of toxicity. Local anesthetic toxicity is of particular concern for the obstetric patient as convulsions or arrhythmias can be life threatening not only to the parturient but the fetus too.
In order to avoid local anesthetic toxicity , there are many strategies that can be used. Slowly aspirating the contents of the catheter before injection is part of the process to exclude the possibility of intravascular placement after the initial insertion of the epidural. Test dosing is used before the administration of the larger epidural doses. These include the administration of at least 15 μg of adrenaline with the local anesthetic , but parturients in labor may manifest false positive values especially if the labor pain coincides with the administration . The best advice is to treat every epidural injection, whether it is the initial dose or the subsequent top-up doses, as a test dose. This essentially requires the provider to administer small aliquots of the local anesthetic agent and to refrain from the administration of a large bolus dose at any time.
While the use of adrenaline-containing local anesthesia agents may allow the safe administration of more local anesthetic agents in the nonpregnant situation, in the obstetric setting, there are concerns about the effects of systemic absorption of adrenaline on the fetus and the parturient, especially those in the high-risk category with either cardiac disease or those with severe preeclampsia .
When there are signs of a local anesthetic systemic toxicity (LAST) occurring, additional injection of local anesthetics should stop . Airway management and cardiovascular support are important considerations in the event of the occurrence of LAST. These measures to reduce the accompanying risk of hypoxia and acidosis complicating and prolonging the often difficult resuscitation of these patients . Seizures should also be terminated with judicious use of benzodiazepines.
Lipid rescue in local anesthetic toxicity has become the recommended strategy to provide a “lipid sink” for the rapid decrease of the blood levels of the local anesthetic. The ASRA practice advisory recommends the administration of 1.5 ml/kg of the 20 % lipid emulsion as a bolus, followed by 0.25 ml/kg per minute to be given until there is attainment of hemodynamic stability lasting for 10 min . An upper limit of 10 ml/kg of lipid emulsion for the first 30 min is advised. Those who have recovered from LAST should also be observed closely for the subsequent 12 h in order to prevent an occurrence especially with the redistribution of the local anesthetics back to the circulation.
While propofol is prepared in lipid, the lipid concentration in propofol is 10 % and hence it is not recommended to be used for the “lipid rescue” of local anesthetic toxicity . The dose required if propofol is used for the purpose of lipid rescue will also cause excessive hypotension.
For additional discussion on this topic, please refer to Chap. 3.
Local Anesthetic Toxicity: Fetal Bradycardia
Local anesthetic agents readily cross the placental barrier. It is possible to get direct fetal toxic effects from these agents in the fetal circulation exemplified by fetal bradycardia without clinically obvious maternal systemic toxicity .
Fetal bradycardia is most typically associated with the paracervical block . The localized effects of vasoconstriction and myometrial hypercontractility from the local anesthetic agent infiltration of the cervical milieu, can have an impact on placental perfusion, resulting in fetal bradycardia. Placental perfusion should be maximized by moving the parturient into the left lateral tilt position which will most effectively alleviate aortocaval compression . Maternal oxygen supplementation is probably better avoided if the intention is to alleviate fetal distress . Fetal bradycardia is usually transient but if persistent and judged to be due to excessive uterine tone, 0.25–0.5 mg subcutaneous bolus dose of terbutaline could be used to initiate tocolysis .
The Use of Paracervical Blocks
Paracervical blocks with infiltration of local anesthetic agents typically into the inferior cervical periphery between the 3 to 9 o’clock positions , have largely been superseded in routine obstetric practice due to its relative ineffectiveness for relief of labor pain  and major concerns about fetal safety . Correct infiltration can be challenging as the cervix undergoes dynamic changes of effacement and dilation in labor, altering landmarks and the cervical milieu. Paracervical blockade can cause local vasoconstriction and increase myometrial contractility resulting in diminished placental function culminating in fetal bradycardia, hypoxia, and even death [39, 44]. The incidence of fetal bradycardia following cervical block is about 15 % . Such effects can also occur with the inadvertent intrauterine injection of local anesthetics . Given the reliability and availability of neuraxial anesthesia, paracervical blockade is unlikely to make a comeback into routine obstetric practice for labor analgesia, though it retains a role in gynecological procedures restricted to the cervix , and to a more limited extent, transcervical intrauterine minor surgery .
Complications due to local anesthetic agent systemic toxicity to the mother caused by hypersensitivity, overdose, or inadvertent intravascular injection, should be treated accordingly as previously described, specific to the agent. Fetal bradycardia can be managed as outlined previously but if the response is insufficient, expedited delivery by emergency cesarean section is indicated, as paracervical block has been associated with fetal and neonatal mortality.
The Use of Pudendal Nerve Block
Pudendal nerve blockade is still in fairly common use but its use for obstetric analgesia is definitely declining . An effective bilateral pudendal block is the minimum analgesia required for operative delivery [50, 51], but this is insufficient for mid-cavity or rotational forceps for which central neuraxial anesthesia should be used . The pudendal block may also be used to augment inadequate local infiltration in postdelivery perineal tear or episiotomy repair [49, 53]; in this context, the total dose of local anesthetic agent infiltrated must be carefully accounted for to avoid toxicity from overdose. As the pudendal nerve supplies only part of the sensory innervation of the perineum, the pudendal block is less effective to relieve pain during the late second stage of labor compared to subarachnoid anesthesia . A successful pudendal block also inhibits maternal bearing down . Recent data suggest that practicing obstetricians can misidentify the exact injection site for transvaginal pudendal nerve blockade in a mannequin pelvis setting and they typically overestimate the block’s speed of action .
Potential complications in pudendal block are uncommon but include inadvertent maternal intravascular injections or overdose of local anesthetic agent. Hematomas can arise from injury to the pudendal artery. Large hematomas can extend to the ischiorectal fossa or into the retroperitoneal space [57, 58]. Ongoing pudendal artery bleeding may best be managed with embolization if interventional radiology expertise is available, as it is difficult to access surgically, it being close to important nerve bundles. Infection (retro-psoas and subgluteal abscess) can also occur , and suspicion should be aroused if there is severe hip or back pain associated with fever after delivery aided by a pudendal block. If not responsive to antibiotics, drainage of these abscesses may be required. Neonatal toxicity , from absorbed or inadvertent direct fetal injection of local anesthetic agent is rare but may be suspected if unexplained neonatal hypotonia, papillary mydriasis fixed to light, apnea, or seizures , occur in the newborn after delivery.
Neurological Dysfunction in the Obstetric Patient
Dysfunction can be the result of failed care by the obstetrician or the anesthesiologist. It is often difficult to differentiate among the many causes but what is useful to understand is that the earlier the dysfunction is recognized , and attended to by the correct care provider—usually the involvement of a neurologist under these circumstances, the lesser the damage in the long term.
Postpartum sensory or motor dysfunction in obstetrics has an incidence approaching 1 % . This is often secondary to the mechanics of labor or fetal pressure on the nerves, which can be exacerbated by dense sensory blockade allowing persistent unawareness of the ongoing nerve damage as it occurs .
There are several sites where there is increased vulnerability to damage occurring:
The lumbosacral nerve trunks as they cross the posterior pelvic brim before descending anterior to the sacral ala. The trunks can be compressed by the fetal head resulting in unilateral foot drop on the opposite side to the fetal occiput and some sensory loss on the lateral lower leg and dorsum of the foot .
The lateral femoral cutaneous nerve (no motor component, damage resulting in meralgia paresthetica) , the femoral nerve or both can be compressed particularly by prolonged thigh flexion as they cross the anterior superior iliac spine or inguinal ligament, with the likelihood of neuropathy exacerbated by increased abdominal pressure (possibly even from external cardiotogram straps). Femoral neuropathy, which can be bilateral, may cause inability to climb stairs, decreased patellar reflex, and femoral distribution sensory loss [66, 67].
The common peroneal nerve can be injured from even inappropriate positioning of the patient’s hand against the distal posterior thigh under epidural anesthesia in labor , prolonged pushing in squatting , or lithotomy positions , causing foot drop and sensory loss, limited to a wedge-shaped area on the dorsal side and proximal to the big and second toe .
Fortunately, these nerve palsies are usually temporary with recovery expected in about 2 months . The occurrence of these palsies may be minimized by frequent changes of lower extremity positions, particularly if the second stage of labor is prolonged. One should avoid prolonged thigh flexion and extreme thigh abduction and external rotation. One should also avoid dense motor and inappropriately dense sensory blockade of labor regional anesthesia . Alpha-lipoic acid supplementation has shown some early promise if neuropathic pain is a prominent feature .
Preventing Spinal–Epidural Hematoma
Improvement in medical care has seen many high-risk patients, especially cardiac parturients, arrive at a stage of their lives where they can potentially become pregnant and deliver in spite of their medical conditions. In order to safely do that, many may be put on treatment or prophylactic doses of antithrombotic agents by their care providers to reduce the incidence of thrombosis. This is especially likely in those who have arrhythmias, those with prior correction of lesions in the heart, and those who need to have bed rest in order to reduce the strain on their heart. Equally important are those parturients on antithrombotic therapy due to thrombophilia, or patients with history of venous thromboembolism [73, 74].
These treatment or prophylactic doses of antithrombotic agents have an impact both on the outcome of the fetus as well as the way we handle our anesthetic technique, especially regional anesthesia for deliveries in the parturient . To reduce the impact of the antithrombotic agents on the fetus, most providers would opt for heparin or heparin products during the first trimester when organogenesis is taking place [73, 75]. This is to avoid the embryopathy associated with the use of warfarin. Heparin compared with warfarin, however, is a much less efficient antithrombotic agent to prevent thrombosis in the parturient . Most obstetric/cardiac care providers opt for heparin in the first trimester, cover the parturient in the second and third trimester with warfarin, and then move on back to heparin products by 36 weeks of gestation, in order to allow greater flexibility with antithrombotic management of the parturient during the delivery process involving either surgery or anesthesia .
When a parturient presents at term or near term for delivery and is in need of a central blockade for analgesia or anesthesia, it is important to determine the actual antithrombotic status of a patient. The primary physician would be able to advise if the parturient is on treatment or prophylactic doses of the antithrombotic agents. If the patient is on a treatment regime, central blockade is best avoided and other options like patient-controlled analgesia for labor pain and general anesthesia for cesarean section are safer choices. If the patient is on a prophylactic dose of the agent, it is possible to time or readjust the administration of the agent, in such a way as to reduce the risk of a spinal–epidural hematoma occurring [73, 74].
The risk of a hematoma is smaller in an intrathecal injection compared to that of an epidural injection—estimated to be around 1 in 150,000 for epidurals and 1 in 220,000 for spinals , although there are concerns that these are underestimated . Catheter utilization increases the risk of spinal hematoma in these patients. They must also be inserted and removed at a time when the effect of the antithrombotic agent is at a minimum. Guidelines do exist on this subject [74, 76]. An excellent guide as to how to adjust these drugs for the various procedures appears in Table 18.2.
Guide to adjustment and administration of antithrombotic agents for regional anesthesia/analgesia
Timea before puncture/catheter manipulation or removal
Time after puncture/catheter manipulation or removal
Unfractionated heparins (for prophylaxis, ≤15,000 IU per day)
Platelets during treatment for more than 5 days
Unfractionated heparins (for treatment)
i.v. 4–6 h
aPTT, ACT, platelets
s.c. 8–12 h
Low-molecular-weight heparins (for prophylaxis)
Platelets during treatment for more than 5 days
Low-molecular-weight heparins (for treatment)
Platelets during treatment for more than 5 days
Fondaparinux (for prophylaxis, 2.5 mg per day)
(anti-Xa, standardized for specific agent)
Rivaroxaban (for prophylaxis, 10 mg q.d.)
(PT, standardized for specific agent)
Apixaban (for prophylaxis, 2.5 mg b.i.d.)
Dabigatran (for prophylaxis, 150–220 mg)
Contraindicated according to the manufacturer
INR ≤ 1.4
After catheter removal
Hirudins (lepirudin, desirudin)
aPTT, ECT, ACT
After catheter removal
After catheter removal
6 h after catheter removal
6 h after catheter removal
5 h after catheter removal
While it is important to be guided by them, it is as important to weigh the options and consequences of each individual patient and determine what is best for the patient concerned. A patient on antithrombotic agents may have had a previous cesarean section done under general anesthesia with a Grade 1 laryngeal view—this should be a license in an elective situation, to proceed with another general anesthetic, instead of proceeding with the potentially hazardous risk of a spinal hematoma no matter how remote the possibility.
All parturients who have delivered with a central blockade should have monitoring regular enough to pick up a spinal–epidural hematoma. This is especially so for those who are on antithrombotic agents, even those on prophylactic doses of the agent. Monitoring should include sensory/motor functions of the lower limb and the ability to pass urine or pass flatus, plus the state of recovery of the anal sphincter tone. In order to facilitate monitoring for spinal hematoma, it is imperative that continued analgesia should have reduced concentrations of local anesthetic agents in them and preferably only analgesic agents should be used in the catheter infusion to facilitate recognizing the signs of a hematoma. A spinal hematoma picked up and definitively evacuated within 8 h of occurrence of the event has a better prognosis compared to one which is evacuated much later .
Newer drugs are increasingly brought into the market and some of these drugs may be used by the parturients . The collective experience of these drugs in the global arena is limited—one should preferably err on the side of caution and opt for the safer alternative of general anesthesia in a parturient already proven to have an easily controllable airway or for patient-controlled intravenous analgesia for labor requirements especially if in doubt about the antithrombotic status.
Infection (Meningitis/Epidural Abscess)
Anesthesiology providers globally have been able to provide a very high level of cleanliness/sterility during the process of delivering central blockade to parturients. Infectious complications in the form of epidural abscess and meningitis are rare events considering that over 90 % of obstetric anesthesia conducted worldwide is in the form of regional anesthesia [78, 79]. As it is a rare event, it is difficult to prove that our stringent standard of wearing masks, caps, and sterile caps/gowns during the process is responsible for this clean or almost clean record . Baer et al., however, noted that nearly 50 % of the meningitis in his historical series of 179 cases was related to viridans streptococcus, a mouth commensal, so presumably the risk of infection may be related to the way the care provider wears his face mask . Whether related or otherwise, it still behooves us to be vigilant and to keep the standard of cleanliness high as every case of abscess or meningitis has the potential of causing permanent harm .
There has been debate on the choice of the best skin disinfectant to decontaminate the skin prior to the procedure. Chlorhexidine and povidone iodine are available as aqueous or alcohol-based solutions and may have comparable antiseptic properties . However, chlorhexidine-based solutions have been shown to be superior in antiseptic properties to aqueous or alcohol-based povidone solution in other studies [83, 84]. The better efficacy of one agent over another may have to be explored in the context of whether alcohol [85, 86] was part of the test solution or the concentration of the solution involved . Chlorhexidine is cheaper, faster in onset, and provides sterility of longer duration, and so is often preferred [82, 88].
To further enhance the safety record, most of us use occlusive dressings after the epidural catheters are inserted and bacteria filters for epidural infusions. We may on most occasions avoid regional analgesia/anesthesia techniques in those who are febrile as there has always been an element of doubt whether it would be safe to conduct intrathecal and epidural insertions in patients who possibly may have septic foci elsewhere in the body. Goodman EJ et al. have shown that it is safe to perform spinal and epidural anesthesia in parturients with chorioamnionitis . Their experience with 517 parturients with epidurals and 14 with spinals, whose placentas were found to be subsequently positive for chorioamnionitis, had no adverse events in the form of meningitis or epidural abscess. As many as 18 % of his patients were febrile and there were 14 % having leukocytosis. From this collection, it is possibly safe, even without prior antibiotic therapy, to conduct regional anesthesia in these parturients.
Infectious complications of central neuraxial blockade although rare can be life threatening when they occur. It is therefore important to be aware of the possible symptoms and signs these patients may reveal. Any parturient who presents with fever, headache, backache , following a regional blockade and especially associated with photophobia, neck stiffness with any neurological deficits following a central neuraxial blockade, should be viewed as likely to have infection in/around the spinal cord, unless proven otherwise.
Cultures from the blood and cerebrospinal fluid should be requested for and if the catheter is still in place, the catheter tip should be sent for culture. Imaging studies may provide further details about collections of abscesses . Involvement of the neurologists and other appropriate experts in the management of the patient early is essential . Correct antibiotic therapy , guided by cultures and with the hospital infection team involved, are urgent considerations. It might be useful to know that the most common isolate for meningitis related to central blockade, is alpha hemolytic streptococci due to the proceduralist [78, 79]. Abscesses as may need drainage and these can be done either by surgeons or through percutaneous drainage with the help of a radiologist .
For additional discussion on this topic, please refer to Chap. 9.
Chronic Adhesive Arachnoiditis
This condition although rare has generated a fair amount of interest in light of two parturients experiencing quadriparesis , following the administration of regional techniques during their deliveries. The first patient has been awarded compensation for chronic adhesive arachnoiditis following the spinal she received for her cesarean section [17, 91–93]. The condition has been judged as due to a minute amount of chlorhexidine contaminating the bupivacaine which was injected into the epidural space. In the second patient’s situation, it was due to a mix up of the chlorhexidine meant for cleaning and the normal saline meant for determining the loss of resistance and the subsequent administration of 8 ml of the former into the epidural space [18–20].
In both these cases, the condition is characterized by an initial complaint of pain at the time of administration, the subsequent complaint of headache and backache with the MRI demonstration of clumping of the nerve roots and demyelination in the cord with the formation of syringomyelia and hydrocephalus which needed drainage [17–20, 91–93]. They both ended up with paralysis/paresis in all four limbs with impairment of micturition and bowel control. While there have been doubts whether minute amounts of chlorhexidine are neurotoxic, in the case of the second patient where chlorhexidine has been accidentally injected, the cleaning agent was implicated as being responsible for the signs and symptoms she had.