Continuous epidural analgesia for postoperative pain: Technique and management

INTRODUCTION — Continuous epidural analgesia is commonly used for postoperative pain control after thoracic, abdominal, lower extremity, and rarely, upper extremity surgeries. This topic will discuss the technical aspects of epidural catheter placement that specifically relate to use for postoperative analgesia, choice of drugs, infusion techniques, and management of the infusion.

The technique for placement of an epidural catheter and other methods of postoperative pain control are discussed separately.

●(See "Epidural and combined spinal-epidural anesthesia: Techniques".)

●(See "Approach to the management of acute pain in adults".)

CHOOSING CATHETER INSERTION SITE — Epidural catheters are inserted in the thoracic or lumbar spine, at a level determined by the dermatome level of the planned surgery (figure 1). For catheter-incision congruency, thoracic epidural catheters are usually placed for thoracic or abdominal surgery, and lumbar epidural catheters are used for lower extremity surgery. The suggested catheter insertion sites for different types of surgery are shown in a table (table 1). Continuous epidural analgesia does not typically block sacral nerve roots, and therefore is not used for perineal analgesia [1-4]. Surface landmarks useful for determining the spinal level for epidural placement are shown in a photo (picture 1).

Catheter-incision congruency is relevant when epidural analgesia is maintained with local anesthetics (LAs). Opioids and other adjuvants can be administered at an incongruent site because their mechanisms of action are different from LAs. (See 'Opioids' below and 'Nonopioid adjuvants' below.)

In addition to the site at which the epidural catheter is inserted, the sensory level achieved can vary depending on patient, epidural procedure and medication-related factors (eg, the dose of LA injected, patient age, adjuvants, and patient position), and the final position of the epidural catheter tip. Thus, epidural drug concentrations and volumes may need to be modified to optimize analgesia. (See 'Inadequate analgesia' below.)

Thoracic versus lumbar catheter placement — Lumbar and thoracic epidural catheters require different considerations (ie, anatomy, technique, drug dose, potential side effects, and complications). For thoracic and upper abdominal surgery, lumbar catheters often require higher doses of local anesthetic (with associated motor block); thoracic epidural catheters may provide better analgesia, better coronary blood flow [5], and reduced postoperative ileus [3,6].

CONFIRMATION OF CATHETER PLACEMENT IN THE EPIDURAL SPACE — Effective postoperative epidural analgesia depends on siting the catheter correctly in the epidural space. The authors confirm correct placement of the epidural catheter by initiating the block immediately following a negative test dose. (See 'Initiation of epidural analgesia' below.)

Incorrect placement is common; in one study, 26 percent of thoracic epidural catheters placed using a loss-of-resistance to air or saline were not found to be in the epidural space by radiologic dye imaging [7]. Incorrect placement is particularly important in patients who have an epidural catheter placed preoperatively before general anesthesia, since lack of efficacy may not be apparent until the patient emerges from anesthesia with inadequate analgesia.

After siting the epidural catheter, the first step is administration of an epidural test dose (typically 1.5% lidocaine with epinephrine 1:200,000, 3 to 5 mL) to confirm that the catheter is not intravascular or intrathecal. Use of an epidural test dose is discussed separately. (See "Epidural and combined spinal-epidural anesthesia: Techniques", section on 'Epidural test dose'.)

A negative test dose confirms that the catheter is not sited in a blood vessel or intrathecal space, but may not confirm that the catheter is correctly sited in the epidural space.

The definitive method for confirming catheter placement in the epidural place is fluoroscopy with dye distribution in the epidural space, but this is rarely used clinically.

INITIATION OF EPIDURAL ANALGESIA — There are several options for initiating epidural analgesia with the goal of establishing an epidural block before or at the end of the surgical procedure. While some evidence suggests that, at least for painful surgeries such as thoracotomy, preoperative initiation of the block provides better postoperative analgesia than initiation after incision, a 2002 systematic review on the preoperative versus postincisional initiation of epidural analgesia did not show any significant benefits of one over the other in terms of pain relief [8-11].

Patients who have combined general and epidural anesthesia may require blood pressure support with vasopressors and intravenous (IV) fluid, due to the effects of the sympathectomy induced by the epidural block, and blunted compensatory vasoconstriction in unblocked spinal segments as a result of general anesthesia (figure 2). For this reason, some clinicians prefer to initiate analgesia shortly before emergence from general anesthesia. (See 'Hypotension' below.)

●Preoperative initiation – It is common practice to inject incremental boluses of local anesthetic (LA; 3 to 5 mL) until adequate sensory blockade is confirmed. The initial dose selection depends on several factors, including patient age, surgical site, and a plan for epidural anesthesia rather than solely postoperative analgesia.

•A typical dosing regimen for catheters used for postoperative analgesia includes 0.125 to 0.25% bupivacaine injected in increments of 3 to 5 mL every five minutes while checking for sensory blockade in the corresponding dermatomes with a blunt-tip needle or cold stimulus. If rapid determination of onset is a priority, 1% lidocaine with or without epinephrine can be used for initiation. If there are no sensory changes after 20 to 30 minutes of LA bolus, the block is deemed a failure. (See 'Troubleshooting' below.)

•If the epidural catheter is used to provide surgical anesthesia as well as postoperative analgesia, a higher concentration of LA is initially injected for surgical anesthesia, followed by the use of dilute LA solutions postoperatively (table 2).

●Intraoperative initiation – For patients who have general anesthesia, two commonly used options for intraoperative epidural initiation for postoperative analgesia are as follows:

•Initiate an epidural infusion of a dilute LA/opioid solution several hours before the end of the surgery.

•Administer a bolus of dilute LA/opioid solution around the time of surgical wound closure, or approximately 30 minutes prior to planned emergence from general anesthesia. This timing reflects the expected latency of onset (ie, 15 to 20 minutes) of the long-acting LAs typically used for postoperative analgesia (table 3).

After initial loading dose with either technique, a continuous infusion is started, ideally using the pump for patient-controlled epidural analgesia (PCEA) for ease of transition after patient emergence from anesthesia. (See 'Mode of drug delivery' below.)

CHOICE OF EPIDURAL DRUGS FOR POSTOPERATIVE ANALGESIA — The most common type of drug solution used for postoperative epidural analgesia is a combination of dilute local anesthetic (LA) and an opioid. The combination of LA and opioid offers superior analgesia compared with LA or opioid-only modalities while decreasing the dose requirements and the subsequent side effects of both drugs [12]. The lowest effective concentration of LAs should be used to minimize motor block and hypotension, improve patient satisfaction, and facilitate monitoring for neurologic complications of neuraxial procedures [13].

Local anesthetic/opioid combination

●Rationale for combination – Epidural LAs and opioids are thought to act synergistically with respect to analgesia, allowing lower doses and a reduced incidence of side effects of each class of drugs [14-18]. The concentration/dose of LA must be a balance between improved analgesia (desired) versus the degree of motor and sympathetic blockade (undesired). The opioid concentration/dose must balance analgesia versus opioid-induced side effects (eg, ileus, pruritus, nausea and vomiting, urinary retention).

The literature on the analgesic benefits of combining LAs and opioids is conflicting, and results may depend on the specific drugs and patient populations studied. Most, though not all [19,20], studies have found that the combination of LAs and opioids provide more effective analgesia than infusion of either LA or opioid alone [21-25]. As an example, in one study of epidural analgesia after joint arthroplasty, the combination of levobupivacaine 0.125% with fentanyl 4 mcg/mL for patient-controlled epidural analgesia (PCEA) resulted in lower postoperative pain scores than either drug alone, without a difference in side effects [23]. In contrast, a small randomized trial found no difference in pain scores after abdominal or thoracic surgery in patients who received epidural bupivacaine 0.1% with fentanyl 10 mcg/mL at 6 mL/hour compared with epidural fentanyl alone [20].

In some studies, thoracic epidural analgesia using only LA has been shown to improved postoperative bowel motility compared with solutions that contain opioids [4,26]. Epidural opioids can delay the return of bowel function similarly to systemic opioids, and may increase the risk of urinary retention [27].

Epidural opioid infusion alone can offer advantages in some clinical circumstances. For example, opioid administration without LA avoids the sympathetic blockade and associated complications that occur with epidural LA administration, an approach that may be useful in patients with poor cardiopulmonary reserve [28]. Further research is needed on the safety and efficacy of pure opioid-based epidural solutions in the context of multimodal regimens for thoracoabdominal analgesia.

At the authors' institution we administer opioids by only one route (epidural or intravenous [IV]) as part of multimodal analgesia. Options include either PCEA with an opioid/LA solution, or use of a continuous epidural infusion of dilute LA plus IV patient-controlled analgesia opioids. We prefer one route of administration to minimize the risk of respiratory depression, and either of these techniques allows titration of the opioid.

●Drug combinations – There is no single optimal combination of epidural drugs for specific surgical procedures and patients, and a variety of combinations are used. Typically, institutions use several standard concentrations of LA/opioid combinations, tailored for different patient populations or surgical procedures (table 4).

Local anesthetics — Dilute solutions of long acting LAs are typically used for postoperative epidural analgesia.

●The primary reason for using long-acting (eg, bupivacaine, ropivacaine) rather than short-acting (eg, lidocaine) LAs is to prevent block regression.

●Dilute solutions are used primarily to minimize motor block [29]. The degree of sympathetic block and resulting cardiovascular effects depend on the extent and location of epidural spread; when using clinically relevant concentrations of LA, for the same extent and location of block, the incidence of hypotension is similar regardless of concentration of LA [30-32].

Ropivacaine is less potent than bupivacaine, and slightly higher concentrations of ropivacaine are typically used for epidural analgesia [33]. At concentrations used for surgical epidural anesthesia, ropivacaine may cause less motor block than bupivacaine. However, at equipotent doses, dilute ropivacaine and bupivacaine appear to produce clinically comparable motor block, and patients who receive ropivacaine must still be monitored for motor block [34]. (See "Epidural and combined spinal-epidural anesthesia: Techniques", section on 'Local anesthetics'.)

Much of the literature on motor block with dilute epidural ropivacaine involves neuraxial labor analgesia [35]. (See "Neuraxial analgesia for labor and delivery (including instrumental delivery)", section on 'Epidural analgesia drug choice'.)

Opioids

Mechanism and site of action — The mechanism of action of epidural opioids is due to effects on the dorsal horn of spinal cord (ie, spinal effect) and supra-spinal effects due to systemic absorption and rostral spread in the cerebrospinal fluid (CSF). Epidural opioids cross the dura and arachnoid membranes to reach the CSF and bind to opioid receptors in the dorsal horn of the spinal cord. Opioids differ with respect to lipid solubility which in turn influences the meningeal permeability of these drugs [36]. Lipophilic opioids (eg, fentanyl or sufentanil) have faster onset and shorter duration of action compared with hydrophilic opioids (eg, morphine or hydromorphone) (figure 3) [37,38].

●Lipophilic opioids readily distribute into epidural fat and the intravascular compartment, and therefore blood levels are higher and CSF concentrations are lower compared with hydrophilic opioids. Lipophilic opioids absorb into the spinal cord close to the dermatomal levels of infusion [39,40].

●In contrast, hydrophilic opioids are not readily absorbed into epidural fat, the vasculature, or the spinal cord, and therefore circulate in CSF for many hours. As CSF circulates in the neuraxial canal and cranium, hydrophilic opioids ascend and descend in the spinal column and are available to bind to opioid receptors over a much greater extent of the spinal cord and brainstem than lipophilic opioids. Thus, the risk of delayed respiratory depression (6 to 12 hours) after epidural administration is significant for hydrophilic opioids. In contrast, there is less risk of delayed respiratory depression with the epidural administration of lipophilic opioids.

The site of action of epidurally administered lipophilic opioids (eg, fentanyl, sufentanil) is uncertain [41]. There is some evidence that lipophilic opioids administered into the epidural space by infusion act primarily by absorption into blood, with subsequent effect on mu-opioid receptors throughout the central nervous system. However, addition of epinephrine to the lipophilic opioid-LA mixture may decrease the systemic uptake of the lipophilic opioid and increased penetration of the opioid through the dura-arachnoid, leading to a local spinal effect and improved analgesic efficacy [42]. Additionally, the mode of administration may affect the site of action. In a volunteer study, fentanyl delivered into the lumbar epidural space by infusion provided systemic analgesia, while the same dose administered as a bolus provided spinal analgesia [1,43].

Choice of opioid — The choice of opioid for epidural analgesia depends on its pharmacokinetic properties (eg, onset time, duration of action), side effect profile (eg, risk of pruritus), and risk for adverse effects (eg, respiratory depression). The mostly commonly administered epidural opioids are fentanyl and hydromorphone. Alfentanil is rarely used or available as it offers no advantages over the other drugs.

If the epidural catheter requires removal (eg, need to administer an anticoagulant), a single bolus dose of morphine or hydromorphone before catheter removal is preferred, as they provide prolonged analgesia.

The characteristics of individual opioids used for epidural analgesia are discussed here.

●Fentanyl – The pharmacokinetics of epidurally administered fentanyl are complex, and the use of epidural rather than IV fentanyl for postoperative analgesia is controversial. Several studies have reported that the quality of analgesia, incidence of side effects, daily fentanyl usage, and plasma levels following 24 hours of infusion are similar whether fentanyl is administered epidurally or intravenously [44-46].

The best evidence suggests that the low opioid concentration epidural infusions that are commonly used for postoperative analgesia act primarily via systemic absorption [41]. Bolus administration with higher concentration fentanyl may act primarily segmentally through a spinal mechanism. (See 'Mechanism and site of action' above.)

●Hydromorphone – The lipid solubility of hydromorphone is between that of fentanyl and morphine, significantly closer to morphine; thus it behaves as a hydrophilic drug [47]. Compared with equipotent doses of epidural morphine, hydromorphone results in faster onset of analgesia, but with similar rostral migration and thus similar risk of delayed respiratory depression. Pruritus is less common than with morphine [37,48].

●Sufentanil – Sufentanil behaves similarly to fentanyl with respect to both spinal and systemic effects. The potency ratio of systemic sufentanil to fentanyl (10:1) is reduced (3:1) when the drugs are administered into the epidural compartment. This smaller difference in potency may explain why epidural sufentanil is associated with decreased risk of respiratory suppression [49] and sedation [50] compared with IV administration. Potent opioids such as sufentanil that provide analgesia while occupying fewer mu-opioid receptors than less potent opioids, provide better analgesia in the setting of opioid tolerance compared with other opioids, such as morphine, that require higher receptor occupancy to achieve the same level of analgesia [37].

Nonopioid adjuvants — Epidural adjuvants other than opioids have been proposed to augment analgesia, however, these drugs are not commonly used because of various associated side effects and a paucity of evidence justifying their use as a component of postoperative epidural analgesia [51,52].

Addition of epinephrine may decrease systemic absorption of LAs. However, given the dilute concentrations of LAs used for postoperative analgesia, the benefit is likely small, and addition of epinephrine makes the solution more difficult to compound.

There is no compelling evidence that adding clonidine to epidural solutions for postoperative analgesia is beneficial. The limited available studies have found mixed results regarding benefits and an increase in side effects. (See "Epidural and combined spinal-epidural anesthesia: Techniques", section on 'Alpha-adrenergic agonists'.)

MODE OF DRUG DELIVERY — Postoperative epidural analgesia may be maintained by several techniques. The most common are continuous epidural infusion (CEI) and patient controlled epidural analgesia (PCEA), with or without a continuous basal infusion. If infusion pumps are not available, epidural analgesia may be maintained with intermittent boluses of dilute local anesthetic (LA)/opioid by an anesthesia provider.

●Infusion based techniques – An infusion-based technique may be inherently safer than a bolus technique. If the LA solution is unintentionally administered into a blood vessel or into the intrathecal space, the slow rate of infusion is not likely to cause local anesthetic systemic toxicity (LAST) or high spinal anesthesia. Importantly, infusion pumps and tubing used for epidural infusions should be visually different than infusion pumps used for intravenous (IV) infusions. Drug errors in which solutions intended for the epidural space are administered intravascularly can result in LAST.

PCEA with a background infusion combines the advantages of both CEI and PCEA. A low-volume basal infusion is chosen by the clinician and the patient can treat any breakthrough pain with self-administered bolus doses.

•Continuous infusion alone provides continuous analgesia, but the chosen infusion rate may provide inadequate analgesia for individual patients, or an overdose for others.

•PCEA without a basal infusion allows the patient to titrate analgesia, but analgesia may be uneven; the patient must self-administer a bolus to re-establish analgesia as the effect of a prior bolus wanes. In addition, breakthrough pain can interfere with sleep.

Programmed intermittent epidural bolus (PIEB) is a new method of administration that is increasingly used to maintain epidural labor analgesia [53]. The literature on the use of PIEB for postoperative analgesia is limited, with uncertain benefits and side effects, and without optimal settings having been established [54-57]. Use of PIEB for thoracic epidural has not been well studied, and further study is required before recommending routine use. A concern is that PIEB may be associated with increased hypotension, particularly for high thoracic epidural [54,58].

●Pump settings – Optimal PCEA and infusion settings vary depending on the spinal level of the epidural catheter, type of surgical procedure, and patient factors. More dilute LA concentration solutions (eg, 0.0625% bupivacaine) are usually administered as a higher rate/bolus volume than more concentrated solutions (eg, 0.1% bupivacaine). Typical infusion pump settings for maintaining epidural analgesia are shown in a table (table 4).

MONITORING DURING EPIDURAL ANALGESIA — Monitoring epidural analgesia requires a multidisciplinary approach consisting of system wide protocols and algorithms for assessment and management [59].

●Anesthesia clinician monitoring – In addition to nursing protocols, all patients with postoperative epidural catheters in place should be examined by a clinician at least daily to assess the following:

•Vital signs

•Adequacy of pain relief and level of activity tolerated

•Degree of motor blockade

•Nausea and pruritus

•Signs suggestive of localized infection (ie, erythema, tenderness, swelling, discharge) at epidural catheter site

•Neurological changes suggestive of spinal hematoma (See "Neuraxial anesthesia/analgesia techniques in the patient receiving anticoagulant or antiplatelet medication", section on 'Spinal epidural hematoma (SEH)'.)

●Nursing protocols – Nursing protocols for patients who have received intrathecal opioids and epidural infusions vary by institution but should include:

•Scheduled monitoring of vital signs, level of sedation, motor function, and pain, with guidelines for notification of the pain service clinician

•Maintenance of patent intravenous (IV) access

•Immediate availability of naloxone and a vasopressor (eg, ephedrine)

•Avoidance of parenteral or oral opioids or sedatives without pain service clinician approval

●Monitoring for respiratory depression – Respiratory depression is a risk for any patient who receives opioids by any route. We suggest the following protocol for respiratory monitoring for patients who have received neuraxial opioids [60]:

•All patients receiving neuraxial opioid infusions should be monitored for adequacy of ventilation, including rate and depth of respiration, oxygenation (pulse oximetry if indicated), and level of sedation [60]:

-Monitor once per hour for the first 12 hours

-Monitor once every two hours from 12 to 24 hours

-Monitor at least every four hours after 24 hours if stable for the entire time the infusion is used

•Patients with conditions which place them at increased risk for respiratory depression (eg, obesity, obstructive sleep apnea, extremes of age, concomitant administration of systemic sedatives or opioids) may require more intensive monitoring (eg, continuous pulse oximetry, capnography).

TROUBLESHOOTING — The epidural infusion may require adjustment because of inadequate analgesia, excessive motor block, hypotension, or oversedation.

Inadequate analgesia — The reported incidence of failure of postoperative epidural analgesia to provide satisfactory analgesia is between 13 and 41 percent, depending on the definition of failure. In a single center review of over 25,000 postoperative epidurals, the failure rate of thoracic epidurals was slightly higher than lumbar epidurals (32 versus 27 percent), with failure defined as the requirement for epidural catheter replacement or addition of another major treatment modality (eg, intravenous [IV] patient-controlled analgesia) [61]. However, surgical site congruency may be more important for successful analgesia than whether the catheter is placed in the thoracic versus lumbar epidural space.

Inadequate epidural analgesia may be the result of technical issues related to the epidural catheter or infusion pump, patient factors, or surgical issues. The technical factors that may cause epidural failures include the following:

●Insertion site incongruent with the required dermatome

●Inadequate local anesthetic (LA) delivery (inadequate dose, pump failure)

●Catheter dislodgement [62]

Intravascular catheter migration should always be considered when inadequate analgesia occurs. We use a systematic approach for the evaluation and management of breakthrough pain during epidural analgesia, as follows (algorithm 1):

●Assess the pain – Evaluate the intensity, location, and laterality of the pain. Consider surgical causes for increased pain.

●Assess the extent of the sensory block to assure that it is appropriate for the surgical procedure (table 5) – Using a cold or sharp stimulus, begin at the dermatome level of catheter insertion and assess the sensory block in both the caudad and cephalad directions. If indicated, assess bilateral block.

●Assess the equipment – Check the infusion pump settings and function, and pump tubing-epidural catheter connection. Assess whether the patient has been appropriately requesting patient-controlled bolus doses.

●Assess the epidural catheter position – Aspirate the epidural catheter, particularly looking for blood as a sign of intravascular catheter migration. Confirm the position of the epidural catheter at the skin insertion site. If the catheter has migrated out, consider replacing it. If the catheter was initially placed more than 3 cm into the epidural space, consider withdrawing the catheter 1 to 2 cm, particularly if the sensory block is unilateral (catheter should be sited 3 to 5 cm in the epidural space).

●Administer an epidural bolus of LA to confirm the catheter is in the epidural space – Assess vital signs and hemodynamic status. If stable, assure negative catheter aspiration and administer an epidural test dose to rule out intravascular catheter migration (see "Epidural and combined spinal-epidural anesthesia: Techniques", section on 'Epidural test dose'). After a negative test dose, administer a bolus of LA (eg, 5 mL of 3% 2-chloroprocaine, 3 to 5 mL of 1.0 to 2% lidocaine or 5 mL of 0.125 to 0.25% bupivacaine), which should produce a sensory and motor block in a properly sited epidural catheter.

•If no sensory block occurs after the bolus of LA solution, replace the epidural catheter or switch to alternative methods for analgesia.

•If a sensory block occurs and the patient's pain is resolved, it is likely that the epidural catheter is correctly sited in the epidural space. Modify the infusion pump settings or the concentration of the LA-opioid solution.

●Modify the epidural drugs or pump settings – We modify the epidural infusion based on the sensory examination found on initial pain assessment, after verification that the epidural catheter is sited correctly in the epidural space.

•If the extent of sensory block was inadequate, increase the basal infusion rate by 1 to 3 mL/hour and/or increase the patient-controlled bolus volume by 1 to 3 mL. For patients with normal or low blood pressure, consider decreasing the increment in infusion rate or bolus volume to avoid hypotension.

•If the extent of sensory block was adequate, increase the density of block by increasing the concentration of the LA or opioid in the epidural solution.

If these measures fail, consider replacing the epidural catheter or discontinuing the epidural catheter and transitioning to an alternative analgesic regimen.

Motor block — Patients should be assessed daily for motor block, and whenever a new weakness occurs. New-onset lower extremity weakness during a stable infusion of dilute LA raises concern for extensive epidural blockade or the possibility of serious adverse events (eg, spinal epidural hematoma or intrathecal migration of the catheter). A motor block is a common complication of a lumbar epidural catheter, but is rare for a mid- or high-thoracic catheter [61]. Therefore, a motor block in the setting of a thoracic catheter should be evaluated immediately. We use a systematic approach to evaluate motor block during epidural analgesia, as follows:

●Decrease the rate or pause epidural infusion during evaluation or in case the patient cannot be evaluated (eg, postoperative mechanical ventilation)

●Assess the degree of motor block, time course of onset, laterality, and the sensory level. For new onset motor block unexplained by recent epidural LA bolus:

•Aspirate epidural catheter to assess for CSF or blood to rule out intrathecal migration or spinal epidural hematoma.

•Monitor motor block with frequent examinations over two hours. Motor block related to epidural LA should wane over this time period.

-If no improvement, arrange for spine magnetic resonance imaging (MRI) and neurosurgical consultation (see "Neuraxial anesthesia/analgesia techniques in the patient receiving anticoagulant or antiplatelet medication", section on 'Evaluation, management, and prognosis'). Possible epidural catheter artifact should be discussed with the radiology team prior to scanning. Note that many wire-reinforced epidural catheters are not MRI compatible and must be removed prior to MRI [63]. Catheter-specific MRI safety information is available at MRIsafety.com. Prior to removal of the catheter, any underlying coagulation disturbance may need to be corrected. (See "Overview of neuraxial anesthesia", section on 'Spinal-epidural hematoma (SEH)'.)

-For excessive motor block that improves after discontinuing the infusion (ie, thought to be a side effect of epidural LA), decrease the concentration LA in the epidural infusion or decrease the basal infusion rate.

Hypotension — Hypotension is a common side effect of epidural analgesia due to the associated sympathectomy. The reported rates of hypotension following LA-based epidural infusions range from 8 to 14 percent [12] (8 percent for lumbar and 14 percent for thoracic epidurals) but lower rates (0.7 to 3 percent) [12,64] have been reported in large-scale observational studies. The risk of hypotension following thoracic or lumbar epidural analgesia varies with patient factors (eg, comorbidities, type of surgery, intravascular fluid volume) and the degree of sympathetic blockade.

The degree of sympathectomy depends on the location and extent of dermatomal spread of block, even with low concentrations of LA; the resulting cardiovascular changes can vary depending on the balance of these effects and patient factors. (See "Overview of neuraxial anesthesia", section on 'Cardiovascular'.)

●Epidural block from T10 to L2 blocks vasoconstrictor fibers to the lower limbs, which causes venous pooling, reduced venous return, and reflex vasoconstriction in the upper limbs and increases in cardiac output and heart rate.

●Epidural block from T6 to L1 blocks splanchnic nerves, causing blood pooling in the gut, and adrenal medullary catecholamine secretion, which may cause hypotension and bradycardia.

●Epidural analgesia from T1 to T4 blocks cardioaccelerator fibers and sympathetic vasoconstriction in the head, neck and arms, leaving compensatory vasoconstriction intact in the gut and lower extremities. Reductions in cardiac output and heart rate may result.

If unacceptable hypotension occurs, follow these steps:

●Pause the epidural infusion and treat with IV fluid and/or vasopressor as necessary.

●Evaluate for and treat other causes of hypotension (eg, hypovolemia from blood loss or fluid shifts) or acute cardiopulmonary events (eg, myocardial ischemia, pulmonary embolism).

●Assess the sensory and motor levels of block. Extensive or dense block may indicate migration of the catheter into the subdural or subarachnoid space; remove the catheter if migration is suspected.

For most patients, the epidural infusion can be restarted at a lower infusion rate, or using a lower concentration LA solution, after resolution of the hypotension. Sympathetic nerve fibers are sensitive to even low concentrations of LA; reducing the concentration in this setting may be effective by reducing the mass of LA, thereby reducing the dermatomal spread and resultant sympathectomy.

Sedation — If sedation occurs, the epidural infusion should be stopped while the patient is evaluated. Sedation may be related to the opioid in the infusion. Sedation is not likely to be due to the LA in the epidural solution, except as a sign of local anesthetic systemic toxicity (LAST). (See "Local anesthetic systemic toxicity".)

●Review the patient's medications for concurrent administration of other opioids, sedatives (especially benzodiazepines), antihistamines, or other perioperative analgesic medications (gabapentin).

●Monitor vital signs, level of sedation, and respiratory status with pulse oximetry for several hours or until sedation wanes, as sedation may be a sign of neuraxial opioid-induced respiratory depression.

After sedation resolves, consider removing the opioid from the epidural infusion solution, or decreasing the opioid concentration.

DISCONTINUATION OF EPIDURAL ANALGESIA

Duration of infusion — Epidural analgesia is typically used for up to approximately a week after major surgery during the period of maximal postoperative pain [65], although longer duration may be used. Epidural analgesia may be stopped earlier for technical reasons (eg, unwitnessed catheter disconnect, leakage, dislodgement), or for side effects that persist despite adjustments to the infusion (eg, hypotension, urinary retention, pruritus, or interference with physical therapy or mobilization). More prolonged epidural analgesia may be beneficial for patients with difficult pain control (eg, high preoperative pain scores, history of opioid tolerance), and for patients who have difficulty transitioning to oral analgesics.

For patients who receive perioperative anticoagulation or antiplatelet medication, timing of epidural catheter removal must be coordinated with medication administration due to risk of spinal epidural hematoma. (See "Neuraxial anesthesia/analgesia techniques in the patient receiving anticoagulant or antiplatelet medication".)

Transition to alternate analgesia — Transitioning from primarily epidural analgesia to an alternate analgesic regimen may be complex and the timing and method are patient dependent. For patients who are able to use oral analgesics, a natural transition may occur when they regain gastrointestinal function.

Epidural analgesia is typically employed as part of multimodal postoperative analgesia, which may include acetaminophen, nonsteroidal anti-inflammatory drugs, and gabapentinoids. Oral medications are encouraged unless gut function is impaired, and patients may be transitioned to oral analgesic therapy when they have evidence of return of bowel function. In practice, we usually wait until patients are tolerating a full liquid or general diet before initiating an oral analgesic regimen.

Patients who have delayed return of gut function (eg, postoperative ileus) or impaired absorption (eg, short gut syndrome, malabsorption syndromes), inability to swallow, chronic pain syndromes or opioid tolerance, or other sources of pain (eg, chest tubes) may benefit from longer duration epidural analgesia.

If the patient had early postoperative initiation of feeding (eg, non-gastrointestinal surgery, as a component of an enhanced recovery after surgery [ERAS] protocol), the decision to discontinue epidural analgesia depends on the extent of pain coverage with the multimodal oral analgesics and the anticipated severity of pain. If these are difficult to assess, one strategy is to discontinue the epidural infusion for two to four hours without removing the epidural catheter. If analgesia is satisfactory, the epidural catheter is then discontinued.

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: Local and regional anesthesia".)

SUMMARY AND RECOMMENDATIONS

●Epidural catheter insertion

•For catheter-incision congruency, thoracic epidural catheters are usually placed for thoracic or abdominal surgery, and lumbar epidural catheters are used for lower extremity surgery (table 1). (See 'Choosing catheter insertion site' above.)

•The authors confirm correct catheter placement within the epidural space by initiating the block immediately after a negative test dose. (See 'Confirmation of catheter placement in the epidural space' above.)

●Initiation of epidural analgesia – Epidural block used primarily for postoperative analgesia can be initiated with boluses of dilute local anesthetic (LA; eg, 0.25% bupivacaine, 1% lidocaine) either preoperatively or during surgery, with no clear advantage to either approach. (See 'Initiation of epidural analgesia' above.)

●Choice of epidural drugs

•A combination of dilute LA and opioid is typically used for epidural analgesia, to allow lower doses and a reduced incidence of side effects of each class of drug (table 4). (See 'Local anesthetic/opioid combination' above.)

•Hydrophilic (eg, morphine or hydromorphone) or lipophilic (eg, fentanyl or sufentanil) opioids can be used for epidural analgesia. Hydrophilic opioids may be used when the epidural catheter insertion site is not congruent with the surgical site, since they travel cephalad and caudad within cerebrospinal fluid (CSF). (See 'Choice of opioid' above.)

●Mode of drug delivery

•Epidural analgesia is typically maintained with continuous epidural infusion (CEI) or with patient controlled epidural analgesia (PCEA) with or without background infusion.

•PCEA with a background infusion provides continuous stable analgesia and allows the patient to self-administer boluses for breakthrough pain (table 4). (See 'Mode of drug delivery' above.)

●Monitoring and troubleshooting during epidural analgesia

•Protocols should be in place to monitor for adequacy of analgesia and occurrence of side effects and complications. (See 'Monitoring during epidural analgesia' above.)

•Inadequate analgesia requires a systematic assessment of the pain characteristics, the epidural catheter placement, and equipment function. If after this assessment an epidural bolus or pump change does not provide adequate analgesia, the catheter may need to be replaced or discontinued (algorithm 1). (See 'Inadequate analgesia' above.)

•For patients who develop motor block, the epidural infusion should be stopped while the patient is assessed, and the catheter should be aspirated looking for blood or CSF. (See 'Motor block' above.)

-If motor block resolves over two hours, the dose of epidural LA should be reduced with either a reduced concentration or volume.

-If motor block does not resolve, arrange for an immediate spine MRI and evaluation for spinal epidural hematoma.

•If hypotension occurs, the infusion should be stopped and intravenous (IV) fluid and/or vasopressors should be administered. Patients should be assessed for possible intrathecal catheter migration. For most patients the infusion can be restarted with a lower rate or LA concentration. (See 'Hypotension' above.)

●Discontinuation

•Epidural analgesia is typically used for up to approximately a week after surgery, though longer administration may be indicated. (See 'Duration of infusion' above.)

•The transition from epidural to alternate forms of analgesia is patient and procedure dependent, and may be initiated once the patient is able to take oral medication as part of multimodal analgesia. (See 'Transition to alternate analgesia' above.)