Continuous Peripheral Nerve Blocks

DEFINITION OF THE PROBLEM

Complications related to continuous perineural local anesthetic infusion can take many forms, which in turn may occur during various stages of patient care. The major stages during which complications occur include: during catheter placement, during local anesthetic infusion, following infusion, and during ambulatory infusion.

SCOPE OF THE PROBLEM

Unfortunately, the relatively recent widespread use of perineural local anesthetic infusion and lack of large clinical studies make evaluating the incidence of related complications problematic. One of the largest prospective investigations to date involving 700 patients receiving an interscalene perineural infusion following shoulder surgery suggested that the incidence of related complications was very low—at least as low as single-injection techniques.² Additional studies involving other catheter sites suggest a similar incidence of complications.³^–⁶

CAUSATION AND MANAGEMENT

Many of the complications that occur during catheter placement result from the needle used to place the catheter and are therefore similar (or identical) to complications of single-injection peripheral nerve blocks. Issues related specifically to perineural catheters are highlighted below, and readers are referred to Chapters 14 to 18 for thorough discussions on complications common to single-injection peripheral nerve blocks.

Complications During Catheter Placement

Inaccurate Catheter Placement

Inaccurate catheter placement may occur in a substantial number of cases. There are multiple techniques and equipment available for catheter insertion. One common insertion technique involves giving a bolus of local anesthetic via the needle to provide the initial surgical block, followed by the introduction of a catheter.⁷^–⁹ Using this technique, it is possible to provide a successful surgical block, but inaccurate catheter placement.⁵^,⁸^–¹² The inadequate perineural infusion will not be detected until after surgical block resolution, hours after placement. The incidence of this complication is presumably dependent upon many factors, including the experience of the practitioner, equipment and technique, as well as patient factors such as body habitus.¹³^,¹⁴ While the use of ultrasound guidance might, at first, appear to decrease the risk of this complication, it is often difficult to visualize the catheter tip location relative to the target nerve⁸; and techniques such as air injection may give a false-positive or false-negative result.¹⁵^,¹⁶ The reported range of what has been called “secondary block failure” is 0% to 40% (Box 21-1).⁵^,⁸^,¹¹^,¹⁷ In an effort to decrease the chances of an unidentified misplacement, investigators have first inserted the catheter and then injected the initial local anesthetic via the catheter.²^,¹⁸^–²¹ If a surgical block does not develop, the catheter may be replaced. Unfortunately, this technique decreases the incidence of surgical block and is best reserved for providing solely postoperative analgesia.⁸

BOX 21-1 Secondary Block Failure

Initial bolus injection of local anesthetic via the needle or perineural catheter can produce an adequate primary block, but mask an improperly positioned catheter. This results in secondary block failure as the dense primary block dissipates, but the more dilute local anesthetic infusion fails to elicit an adequate analgesic block.

However, even using this bolus-via-catheter technique, practitioners must wait 5 to 15 minutes for surgical block onset to determine if the catheter must be replaced. In an attempt to further improve catheter-placement success rates and decrease insertion time, “stimulating” catheters have been developed that deliver current to the catheter tip.²²^–²⁵ This design provides feedback on the positional relationship of the catheter tip to the target nerve prior to local anesthetic dosing.¹⁸^–²⁰ There are data suggesting that stimulating catheters may be placed, on average, closer to the target nerve/plexus compared with nonstimulating devices.²⁶^–²⁸ However, it remains unclear if clinical benefits are provided compared with nonstimulating catheters.²⁹^–³² In addition, stimulating catheters may take far more time to insert, on average, than nonstimulating catheters, especially when ultrasound is utilized.⁹^,³³^–³⁶ The optimal placement techniques and equipment for perineural catheter placement have yet to be determined and require further investigation.⁸^,³⁷

Vascular Puncture

While puncturing a vessel is certainly a well-known complication of single-injection peripheral nerve blocks, this may be a more significant occurrence when placing a perineural catheter since the needle gauge is often larger to allow for endoluminal catheter insertion. The incidence of this complication is reportedly between 0% and 11% and most likely is influenced significantly by such variables as the anatomic block location, placement technique (e.g., ultrasound), and needle/catheter design.³^,⁶^,¹⁷^,¹⁹^,³⁸^,³⁹ Indirect evidence for the latter case was recently described by investigators who experienced no needle vascular punctures out of 76 infraclavicular catheters placed using the coracoid technique.⁴⁰^,⁴¹ However, these same investigators using the identical coracoid approach reported an 11% incidence of needle vascular puncture using a different needle/catheter set.¹⁹ Initial evidence suggests that vascular puncture can be reduced dramatically with the use of ultrasound, but definitive data have yet to be published.⁸^,⁹^,³³^–³⁶^,⁴²^–⁴⁴ Should vascular puncture occur, removal of the needle/catheter is indicated, direct pressure should be applied to the site, and distal vascular compromise ruled out. Prolonged Horner’s syndrome due to neck hematoma is a rare complication but has been reported.⁴⁵ While a hematoma may require weeks for resolution (months for a Horner’s syndrome), practitioners and patients should be reassured with the multiple case reports of complete neural recovery following hematoma resolution.³⁸^,⁴⁵^–⁴⁷ In some extraordinary cases, surgical incision and drainage may be necessary.⁴⁸

If vascular puncture does occur, it is still possible to successfully place a perineural catheter using a nerve stimulator and insulated needle following a period of direct pressure, although a resulting hematoma will conduct electrical current and may decrease the ability to stimulate the target nerve with subsequent attempts.¹⁹ Use of ultrasound guidance will render this issue mute.⁸ Of note, clinically significant hematoma formation has been reported in patients with a psoas compartment catheter who received low molecular weight heparin for anticoagulation.⁴⁶^,⁴⁷ These occurrences have led some practitioners to manage patients with a psoas compartment catheter in much the same way as those having neuraxial block when thrombo prophylaxis is ordered,⁴⁶ although others have questioned this practice.⁴⁹ The Third American Society of Regional Anesthesia consensus statement on neuraxial anesthesia and anticoagulation now explicitly recommends that the same precautions as neuraxial techniques be exercised for deep procedures such as posterior lumbar plexus blocks/catheters—specifically, that any catheter be removed prior to administration of various anticoagulants at certain doses (e.g., enoxaparin 30 mg twice daily).⁵⁰ While “deep procedures” is not defined, presumably this subset includes psoas compartment, high sciatic, and possibly infraclavicular catheter insertion.

Intravascular Local Anesthetic Injection

Even if the bolus of local anesthetic for initial surgical block placement is given via the catheter, practitioners should be cautioned that intravascular injection with subsequent toxicity is possible.⁵¹ As with all single-injection techniques, gentle aspiration should be performed prior to local anesthetic injection; repeated aspiration between divided doses of local anesthetic is warranted; and many, if not most, investigators have included epinephrine as an intravascular marker with all local anesthetic boluses.⁵² Use of ultrasound does not preclude this complication, although it may decrease its incidence.⁵³ When a bolus is given via the needle, subsequent unintentional intravascular catheter insertion is still possible.⁴⁰^,⁵¹ Therefore, investigators have recommended injecting a “test dose” containing epinephrine via the catheter prior to local anesthetic infusion initiation.⁵⁴ Practitioners should be aware of the signs, symptoms, and treatment for local anesthetic toxicity (Chapter 7).⁵²^,⁵⁵^,⁵⁶

Perineuraxis Injection

When placing a catheter near the neuraxis as with the psoas compartment, interscalene, and paravertebral locations, it is possible to cannulate the epidural⁵⁷^–⁶¹ or intrathecal⁶² spaces. Injection of local anesthetic is potentially catastrophic and may result in unconsciousness and extreme hypotension requiring aggressive resuscitation.⁶³ In an effort to avoid intrathecal injection, gentle aspiration for cerebrospinal fluid should be conducted prior to local anesthetic injection. As with intravascular catheter placement, it is possible to accurately inject the initial bolus of local anesthetic via the needle, followed by cannulation of the epidural,⁵⁷ intrathecal,⁶² and even intrapleural spaces with the catheter.⁶⁴ Therefore, a “test dose” containing local anesthetic should be injected via the catheter prior to local anesthetic infusion.¹⁰^,⁴⁰^,⁶²^,⁶⁵^,⁶⁶ Obviously, a misplaced catheter should be removed and the patient observed for related complications.⁶² Practitioners should be aware of the signs, symptoms, and treatment for intrathecal, epidural, or intrapleural local anesthetic injection and have appropriate resuscitation equipment immediately available (Chapter 18).⁶³ Of note, when working close to the neuraxis, it is possible to get epidural local anesthetic spread even with an accurately placed perineural catheter, resulting in a sympathectomy and possible hypotension.³⁹^,⁵⁹^,⁶⁷^,⁶⁸ Additionally, some practitioners recommend placing a perineural catheter only a small distance past the needle tip in an effort to ensure that the catheter does not migrate further than the desired target.⁶⁰^,⁶¹^,⁶⁹ And while ultrasound guidance may render this complication a past relic, practitioners are urged to remain mindful of this possibility,at least until many years of further experience are gained with this relatively new modality.⁸

Nerve Injury

Nerve injury is a recognized complication following the placement of both single-injection and CPNBs, presumably related to needle trauma and/or subsequent local anesthetic/adjuvant neurotoxicity.⁷⁰ While one might assume that the larger gauge of most needles designed for catheter placement would increase the risk of nerve injury, the extremely limited evidence suggests otherwise.² In one prospective study of 521 patients with an interscalene block or catheter for shoulder surgery, 11% of patients with a continuous block reported either paresthesia, dysesthesia, or pain not related to surgery after 10 days, compared with 17% in the single-injection group (difference not statistically significant).³⁸ All but one of these cases resolved over the following 9 months. Another prospective study examined 1,398 CPNBs in 849 orthopedic patients.⁶ Thirteen of these patients (0.9%) presented with neurologic deficits following catheter discontinuation; symptoms were transient in twelve of these patients. Examination of the one permanent nerve injury revealed a retroperitoneal hematoma in a patient with a continuous femoral nerve catheter for a total knee replacement. The origin of the hematoma was believed to be from injury to the femoral artery, though direct diagnosis could not be made.⁶

While the true incidence of neural injury is most probably related to the needle/catheter design, anatomic block location, and subsequent infusate selection, practitioners may be reassured that the current available evidence suggests that placing a perineural catheter during a regional nerve block does not appear to increase the risk of neural injury (Box 21-2).²^,⁴^,³⁸ The use of ultrasound guidance may decrease the risk of nerve injury,⁴⁴^,⁷¹ but to date there are no data supporting this proposition, and nerve injury following ultrasound-guided peripheral nerve block does occur.⁷²^–⁷⁴ The identification and treatment of suspected neural injuries following catheter placement does not differ from single-injection techniques (other than possible catheter removal), and readers are referred to Chapter 14 for a thorough discussion regarding these important topics.⁷⁵

BOX 21-2 Clinical Caveat

Limited data suggest that the placement of perineural catheters does not increase the incidence of perioperative nerve injury beyond that seen with single-injection peripheral nerve block techniques.

Complications During Infusion

Dislodgement

One of the most common complications during perineural infusion is simply unintentional catheter dislodgement.⁵^,¹⁰^,²³^,⁴⁰^,⁶⁶^,⁷⁶^,⁷⁷ The reported incidence of dislodgement varies greatly between 0% and 30% and is most likely related to the anatomical location, equipment type, and technique used to secure the catheter.¹⁷^,⁷⁷^–⁷⁹ Every effort to optimally secure the catheter must be made to maximize patient benefits. Measures have included the use of sterile liquid adhesive (e.g., benzoin), sterile tape (e.g., “Steri-Strips”), securing of the catheter-hub connection with either tape or specifically designed devices (e.g., “Statlock”), subcutaneous tunneling of the catheter,²²^,⁸⁰ and the use of 2-octyl cyanoacrylate glue.⁸¹ Using a combination of these maneuvers,¹⁸^–²⁰^,⁸² investigators have reported a catheter retention rate of 95% to 100% for 6 to 27 days of infusion.⁸³^,⁸⁴

Infection

While catheter site bacterial colonization is relatively common,⁸⁵^,⁸⁶ clinically relevant infection is not.⁶^,³⁸^,⁸⁶^–⁸⁸ In one study involving 211 femoral catheters that were removed and cultured after 48 hours, 57% were positive for bacterial colonization.⁸⁵ Three patients (1.5%) had transitory signs of systemic bacteremia that were noted and resolved following catheter removal, and nine other patients (4%) had discomfort at the insertion site. However, ultrasound of the insertion site and the psoas muscle revealed no abscesses. In a retrospective study of 405 axillary catheters, the incidence of infection was 0.25%.⁸⁹ In prospective investigations of interscalene³⁸ and posterior popliteal⁴ catheters involving over 800 patients, no infections were identified. In a subsequent study of interscalene catheters by the same authors, six patients (0.8%) out of 700 developed signs and symptoms of catheter site infection after 3 to 4 days.²

Three large series illustrate the low incidence of infection.⁵^,⁶^,⁹⁰ These series consisted of 1,398, 3,491, and 2,285 catheters, respectively, and revealed inflammation in 0.6%, 4.2%, and 4.2%, self-resolving infection in 0.2%, 2.4%, and 3.2%, and infection requiring surgical intervention in 0%, 0.8%, and 0.9%, respectively.⁵^,⁶^,⁹⁰ In addition, there is one case report of a psoas abscess following 4 days of femoral perineural infusion.⁹¹ In these few cases, all infections completely resolved within 10 days.⁹¹ Overall, the relatively rare incidence of infection may be related to local anesthetics’ bactericidal and static properties.⁶¹ However, major infectious complications do occur. These range from a nonsurgical abscess to florid sepsis requiring a prolonged intensive care unit admission.⁴⁸^,⁹²^–⁹⁴ With adequate medical treatment, all of these cases resolved, and there has never been a published case of permanent injury due to a perineural catheter infection.

Risk factors for perineural catheter-related infection include an intensive care unit admission, duration of catheter use >48 hours, absence of antibiotic prophylaxis, axillary and femoral sites, and frequent dressing changes.⁸⁸^,⁹⁵ Signs and symptoms of catheter site infection are similar to those of infection of any foreign body: local erythema, induration, purulent exudate, localized discomfort, as well as signs and symptoms of bacteremia. Treatment should include catheter removal, with the tip of the catheter sent for bacteriologic examination to help guide antibiotic therapy.⁸⁵ Ultrasonography is useful to help rule out an abscess requiring surgical drainage.² The available evidence suggests that clinically relevant catheter site infection is a very rare occurrence, and that with adequate treatment does not result in lasting impairment.²^,⁴^,³⁸^,⁸⁵^,⁸⁹^,⁹¹ Limiting catheter use to 3 days will decrease the incidence of this complication, and practitioners should balance the need for analgesia with the risk of infection.⁸⁶^,⁹⁶ However, strict sterile catheter insertion technique is probably the most important factor in reducing the incidence of catheter site infection.⁹⁷

Catheter Migration

While there are case reports of initially misplaced catheters, spontaneous migration into adjacent anatomic structures following a documented correct placement has not been described,⁹⁸ but remains a theoretical risk.⁵¹^,⁵⁸^,⁶²^,⁶⁴^,⁹⁹ Possible complications include intravascular or interpleural migration resulting in local anesthetic toxicity and epidural/intrathecal migration when using an interscalene, intersternomastoid, paravertebral, or psoas compartment catheter. An early symptom may include a decrease in analgesia accompanied by signs of epidural/intrathecal anesthesia. Of note, it is possible to accidentally position the catheter tip in the epidural space (and presumably other structures) following partial catheter withdrawl.⁵⁷ Therefore, a test dose containing both local anesthetic and epinephrine should follow any catheter repositioning.⁵⁷^,¹⁰⁰

Repeated large boluses of bupivacaine have resulted in myonecrosis in both animal models¹⁰¹^,¹⁰² and patients,⁹⁹^,¹⁰³ suggesting that intramuscular migration may have pathologic consequences (Chapter 15).¹⁰⁴ In minipigs, a bolus of bupivacaine followed by 6 hours of infusion resulted in severe tissue damage, whereas ropivacaine induced fiber injury of a significantly smaller extent.¹⁰⁵ Furthermore, bupivacaine results in far more muscle fiber apoptosis than ropivacaine in an animal model.¹⁰⁵^–¹⁰⁷ Although muscle injury has not been reported following a continuous perineural local anesthetic basal infusion (as opposed to repeated large boluses) in humans, practitioners may want to consider avoiding bupivacaine when using a catheter that is specifically placed into a muscle belly (e.g., a posterior lumbar plexus catheter placed using a nerve stimulator and insulated needle inserted into the psoas muscle).⁵⁹^,⁶⁷^,¹⁰⁸

Signs and symptoms of myonecrosis include muscle tenderness, intensification of pain with stretch, pain relief with shortening, elevated serum levels of muscle-type creatine kinase, inflammatory or necrotic myopathy on electromyography, increased protein, blood flow, and edema in T1-weighted magnetic resonance imaging.¹⁰⁴ An abrupt decrease in surgical site analgesia accompanied by an increase in pain in the area of the catheter tip is suggestive of myonecrosis and warrants catheter removal.¹⁰⁴

Delayed Local Anesthetic Toxicity

All practitioners using continuous block techniques should consider systemic local anesthetic toxicity due to a postoperative perineural infusion. The maximum safe doses for the long-acting local anesthetics as well as the incidence of systemic toxicity are unknown. There have been cases of patients reporting early symptoms of toxicity, such as perioral numbness, that resolved with infusion termination.⁸⁹^,¹⁰⁹ However, two subjects participating in placebo-controlled studies have reported early symptoms of toxicity that resolved upon infusion discontinuation, and each of these patients was subsequently found to have been receiving normal saline (unpublished data, Ilfeld et al. 2001). Investigators have reported successful analgesia and blood concentrations well below toxic levels using the following schedule with dilute (e.g., 0.125%–0.2%) long-acting local anesthetics (e.g., bupivacaine or ropivacaine) in patients free of renal or hepatic disease: basal rate of 5 to 10 mL/h, bolus volume of 2 to 5 mL, and lockout duration of 20 to 60 minutes.⁷⁷^,⁸⁴^,⁸⁷^,¹⁰⁹^–¹¹⁸ Of particular importance is a prospective investigation of long-term (1–4 weeks) perineural ropivacaine 0.2% infusion that reported no evidence of local anesthetic toxicity, even though 2 of 15 subjects had serum plasma levels of ropivacaine in the toxic range.⁸⁴ Interestingly, the duration of the ropivacaine infusion was not correlated with the free concentration of local anesthetic.⁸⁴

As a preventative measure, practitioners should consider reviewing the signs and symptoms of local anesthetic toxicity with patients receiving continuous nerve blocks. In addition, providing patients with the ability to self-administer bolus doses decreases local anesthetic consumption.¹⁸^–²⁰^,¹¹⁹^–¹²⁶ Finally, some investigators have utilized elastomeric pumps that provide “bolus-only” dosing when the patient releases a clamp on the tubing connecting the pump and catheter.¹²⁷^–¹²⁹ The patient is instructed to re-clamp the tubing after a specified period of time.¹²⁷^,¹²⁸ If a patient forgets to re-clamp the tubing, it is possible for the entire contents of the local anesthetic reservoir to be administered in under an hour. This potentially devastating scenario has been reported, although no apparent morbidity has yet occurred.¹²⁹ While the safety of this method may be demonstrated in the future, practitioners should consider the relative risks and benefits now that multiple pumps are available providing controlled bolus dosing.¹²⁰^,¹³⁰^–¹³³

Nerve Injury

Continuous nerve blocks differ from single-injection blocks in that a catheter remains in situ and a larger total dose of local anesthetic may be delivered, albeit over a greater duration of time.¹³⁴ There is limited evidence that this increased exposure to local anesthetic may have negative consequences.¹⁰² In an animal model, repeated boluses of 0.5% bupivacaine over 3 days led to a marked degree of disruption and vacuolization of myelin sheaths.¹⁰² However, a 3-hour infusion of bupivacaine resulted in only minor injury to muscle tissue.¹⁰² Furthermore, the prospective clinical evidence from human subjects suggests that the incidence of neural injury from a perineural catheter and ropivacaine (0.2%) infusion is no higher than following single-injection regional blocks.³⁸^,⁸⁹ There are two case reports of interscalene perineural catheters possibly resulting in brachial plexus irritation.¹³⁵ In both of these cases, repeated boluses of 0.25% bupivacaine had been injected over a period of days, and patient discomfort ceased upon removal of the catheters.¹³⁵

In one case of prolonged sensory and motor deficit following a continuous femoral nerve block, postoperative electrophysiologic testing suggested subclinical polyneuropathy that was unknown preoperatively.¹³⁶ Long-term follow-up of this patient revealed complete sensorimotor recovery of the quadriceps muscles.¹³⁶ There is also evidence that in diabetes, the risk of local anesthetic-induced nerve injury is increased.¹³⁷^,¹³⁸ The identification and treatment of suspected neural injuries following catheter placement does not differ from single-injection techniques (apart from catheter removal),¹³⁹ and readers are referred to Chapter 14 for a thorough discussion regarding these important topics.⁷⁵

Falls

Although pain fiber inhibition is the primary aim of CPNBs in the perioperative setting, local anesthetics affect other afferent and efferent nerve fibers.¹⁴⁰ Consequently, perineural infusion induces muscular weakness,¹⁴¹ particularly concerning during perineural infusion involving the femoral nerve since ambulation is depended upon a functioning quadriceps femoris muscle. One meta-analysis of three randomized, double-masked, placebo-controlled studies found evidence that perineural infusion involving the femoral nerve does increase the risk of falling.¹⁴² The risk of falling must be minimized, given that a fall may be catastrophic in many patients, especially the elderly.¹⁴³^,¹⁴⁴ Unfortunately, decreasing the concentration of perineural local anesthetic often results in not only less muscle weakness, but analgesia as well.¹⁴⁵ Alternatively, retaining the total delivered dose by increasing the delivered volume while decreasing the local anesthetic concentration has little effect on either sensory or motor block.¹⁴⁶ Practitioners should consider interventions that may decrease the risk of falls, such as minimizing the dose/mass of local anesthetic¹⁴⁶; providing limited-volume patient-controlled bolus doses that allow for a decreased basal dose without compromising analgesia in some¹⁹^,¹⁴⁷—but not all—cases¹⁸; utilizing a knee immobilizer and walker/crutches during ambulation,¹⁴⁸ and educating physical therapists, nurses, and surgeons of possible CPNB-induced muscle weakness and necessary fall precautions.

Complications Following Infusion

Catheter Knotting and Retention

Multiple cases of catheter retention have been published,¹⁹^,¹⁴⁹^–¹⁵² and one retrospective study of nearly 6,000 catheters reported an incidence of 0.13%.¹⁵³ The most common etiology is knot formation below the skin or fascia and has been reported in fascia iliaca,¹⁴⁹ femoral,¹⁵⁰ psoas compartment,¹⁵¹ and sciatic catheters.¹⁵² Occasionally, a case requires surgical exploration for catheter removal.¹⁵⁰^,¹⁵¹ However, removal of a knotted fascia iliaca catheter has been achieved using simple hip flexion,¹⁴⁹ or a relatively noninvasive technique involving fluoroscopy.¹⁵³ In all reported cases of knot formation, the catheter had been advanced more than 5 cm past the needle tip. Advancing the catheter more than 3 to 5 cm is often attempted in an effort to decrease the risk of dislodgement, or to “thread” the catheter tip toward the lumbar plexus when using the femoral or fascia iliaca insertion points.⁶^,¹²^,⁹⁵^,¹²¹^,¹²³^,¹⁵³^–¹⁵⁸ Unfortunately, this practice probably increases the risk of a retained catheter due to a knot formed under the skin or fascia iliaca.¹⁴⁹^,¹⁵³ Retention rates of 95% to 100% have been reported using a maximum distance of 5 cm,¹⁰^,¹⁸^–²⁰^,⁴⁰^,⁴¹^,⁶⁶^,¹⁵⁹ and—in the absence of using a catheter-over-wire Seldinger technique¹²¹^,¹²³^,¹⁵⁷^,¹⁵⁸—the catheter tip rarely reaches the lumbar plexus following a femoral insertion even when the catheter is advanced 15 to 20 cm.¹⁵⁴ Therefore, although there is no consensus regarding the optimal distance of catheter insertion, the available data suggest that insertion >5 cm is unnecessary and most probably increases the risk of catheter knotting.¹⁴⁹

In an unusual case, a posterior lumbar plexus catheter tip was adhered to surgical tissue after intraoperative inadvertent coagulation forceps contact by the surgeon.¹⁶⁰ Catheter retention has also occurred when the metallic tip of a stimulating catheter became “caught” on underlying tissue.¹⁹ Infraclavicular catheter insertion and postoperative infusion were uneventful, and the complication was discovered only after attempts at removal resulted in severe pain. Fluoroscopy did not reveal a knot, and the catheter was extracted surgically under general anesthesia. The overall incidence of this unusual complication is unknown, but of more than 10,000 stimulating catheters placed by one investigator, there was only one incidence of a retained catheter (Andre Boezaart, MD, personal communication, May 2003). Because the etiology of this complication remains unknown, steps for prevention are, at best, speculative. However, it should be noted that increased forward force placed on the catheter during insertion might increase the risk of the metallic end getting caught on surrounding tissue, and a paresthesia during catheter removal should be viewed as a warning sign.

Catheter Shearing or Breaking

It is possible to “shear off” a segment of catheter if, following insertion past the needle tip, the catheter itself is withdrawn back into the needle.⁶ Therefore, this maneuver should only be attempted when using needle/catheter combinations that have been specifically designed for catheter withdrawal. And when using specifically designed needle/catheter combinations—such as with some stimulating catheters—catheter withdrawal should cease with any resistance, and the needle itself retracted until the catheter resistance resolves.⁶^,¹⁸^–²⁰^,¹⁵⁹^,¹⁶¹ In one reported case, a 6-cm femoral catheter fragment was sheared off and remained in situ for 1 week, causing persistent pain of the ipsilateral groin, thigh, and knee.¹⁶² Despite an embedded radio-opaque strip, the catheter fragment could not be visualized with plain radiographs. However, a computerized tomographic scan did localize the fragment and the femoral nerve neuralgia resolved in the week following surgical extraction of the fragment.¹⁶² In an additional case, an axillary catheter fragment was diagnosed with ultrasonography and surgically extracted.⁸⁹ Practitioners should be reassured that in all of the case reports of retained catheters/fragments, no patient has experienced persistent symptoms following removal.¹⁹^,⁸⁹^,¹⁶² Finally, practitioners should document successful catheter tip extraction following catheter removal.⁸²

Complications of Ambulatory Infusion

Outpatients may theoretically experience the same level of analgesia previously afforded only to those remaining hospitalized by combining the perineural catheter with a portable infusion pump.⁸² However, complications that could be managed routinely within the hospital may take longer to identify or be more difficult to manage in medically unsupervised patients at home.¹⁶³ Because not all patients desire, or are capable of accepting, the extra responsibility that comes with the catheter and pump system, appropriate patient selection is crucial for safe ambulatory local anesthetic infusion (Box 21-3).⁸²

BOX 21-3 Key Concepts for Successful Ambulatory Infusion Strategies

images Appropriate patient selection—medically, psychologically, and with appropriate support systems—is crucial