Abstract

We report a female patient with osteoarthritis of both knees, who was scheduled for bilateral total knee arthroplasty under spinal and lumbar (L) epidural anesthe-sia.  An  epidural  catheter  was  inserted  at  the  L3−L4  intervertebral  space  and  threaded 9 cm cephalad beyond the tip of the Tuohy needle. Attempts to remove the catheter met with resistance. Because changing the patient’s position failed to relieve the resistance, we suspected that the epidural catheter had kinked. After explaining to the patient and her family the need for surgical removal of the cath-eter and its associated risks, the catheter was successfully removed by an orthope-dist by laminectomy. A butterfly-like knot was found about 5 cm away from the tip of the epidural catheter. She experienced no complications during surgery and no neurological sequelae were observed during her stay in hospital.

Keywords

anesthesia; epidural; catheters; indwelling: knotting;

1. Introduction

Knotting  of  a  lumbar  epidural  catheter  is  a  very  rare complication,^1,2 with an estimated incidence of 0.0015%.³ Single and double knots have been re-ported to date.^4,5 When threading the epidural cath-eter quickly,  greater  force  is  needed  and  the  catheter may split. An excessive length of the cath-eter in the epidural space may promote knot forma-tion. If it is suspected that a knot has formed, some authors  have  suggested:  using  a  small  and  steady  force for withdrawal; to stop pulling if the catheter begins to stretch too much; placing the patient in various positions (e.g. the same position as on in-sertion, the lateral decubitus position, and a flex-ion  or  extension  position);  and  injecting  normal  saline  through  the  catheter.⁵  Here  we  present  an  extremely rare case in which the lumbar epidural catheter entangled itself in a butterfly-like knot in the epidural space. The patient was admitted for bilateral total knee arthroplasty under spinal and lumbar epidural anesthesia.

2. Case Report

A  75-year-old  female  suffered  from  osteoarthritis  and  was  admitted  for  bilateral  total  knee  arthro-plasty under spinal and lumbar epidural anesthesia. The patient was 160 cm tall and weighed 88 kg, with a body mass index of 34.38 kg/m². Her past history revealed no back pain or back injury; hypertension and obesity were the only findings of note.

After  successful  spinal  anesthesia  was  per-formed  in  the  lateral  decubitus  position,  an  18-G  epidural  Tuohy  needle  (Portex®;  Smiths  Medical  International Ltd, Kent, UK) was inserted by a senior anesthetist resident via the paramedian approach in  the  L3−L4  intervertebral  space  to  support  an-esthesia and for postoperative pain control. Using the loss-of-resistance to air technique, the needle tip  was  advanced  in  the  cephalic  direction.  The  epidural space was identified at the first attempt at a depth of 8 cm. When the epidural catheter was inserted 4 cm into the epidural space, an unusual resistance was noted. The resident then pulled the catheter out smoothly, and the epidural needle was left in  situ.  On  applying  the  loss-of-resistance  to  air technique, the findings were still negative so the resident tried to thread the catheter in again. The resident  initially  inserted  the  catheter  approxi-mately 9 cm into the epidural space to determine whether there was blockade in the presumed epi-dural space, and intended to withdraw it and fix it at the mark of 5 cm, but this was unsuccessful. The final position of the catheter was noted as 17 cm. On withdrawal, a firm resistance was noted. This was not decreased by changing the position of the patient with varying degrees of lumbar flexion and extension  in  the  seated  position,  lying  in  the  lat-eral decubitus position, or in the standing position. Further attempts to withdraw the catheter caused it to stretch. The resident noted firm resistance to subsequent attempts to injecting 2 mL of 2% lido-caine  (20 mg/mL).  Accordingly,  the  resident  sus-pected  that  the  catheter  had  kinked,  and  called  for  assistance.  Despite  further  attempts  by  two  division directors of the Department of Anesthesi-ology, the catheter could not be removed. Because the  epidural  catheter  was  not  radio-opaque,  we  decided to remove the catheter surgically without imaging.  After  explaining  to  the  patient  and  her  family the need for surgical removal and its associ-ated  risks,  we  resolved  to  continue  the  proposed  operation  under  general  anesthesia  and  remove  the catheter by laminectomy.

After  completing  the  bilateral  total  knee  ar-throplasty, the patient was placed in the prone po-sition. The orthopedist performed the laminectomy with a small incision wound at the L3−L4 level. The catheter was found to be anchored in the epidural space and it was removed intact, without shearing. A tight knot of double-twisting loops in the shape of a butterfly was found about 5 cm away from the tip of the catheter (Figure 1). The patient was dis-charged home 4 days after the operation. The sur-gical wound healed well. No neurological sequelae were observed during the hospital stay.

Download full-size image

Figure 1 A butterfly-like knot was found about 5 cm away from the tip of the epidural catheter.

3. Discussion

Knotting  of  a  lumbar  epidural  catheter  is  a  rare  complication^1,2  with  a  very  low  incidence  of  just  0.0015%.³ Although single and double knots^4,5 have been reported, there were no previous reports in the literature about a butterfly-like knotting of a catheter requiring surgical removal.

Bromage reported only one complication of epi-dural catheter knotting in more than 30,000 epidural catheter  placements  at  his  institution.⁶  In  other  reports, there were 16 instances of knotting of epi-dural catheters.^2−4,7−15 Of these, three were caudal placement, 12 were lumbar and one was a thoracic epidural catheter. Twelve cases involved obstetric procedures. In 12 cases, the knots arose near the tips of  the  catheters.  In  our  hospital,  there  was  only  one case with sustained knotting of a thoracic epi-dural catheter in 2001, with an estimated incidence of 0.0038% over 10 years.⁵

To  manage  a  knotted  catheter,  some  experts  have reported that the catheter may be retrieved without shearing by steady and firm traction.^3,6−8,10 In a literature review, the authors advocated using small  but  steady  forces  on  withdrawal,  to  stop  pulling if the catheter begins to stretch too much and  trying  again  several  hours  later,  placing  the  patient in a variety of positions (e.g. the same po-sition  as  on  insertion,  lateral  decubitus  position,  flexion or extension positions) and injecting normal saline  through  the  catheter  because  this  may  in-crease its turgidity.⁵ Meanwhile, Gozal et al stated that the catheter was better removed under gen-eral anesthesia with muscle paralysis.¹¹ In another report of two cases, the catheters were broken on withdrawal and required surgical removal.^2,9 Although pulling may induce shear of the catheter, it seems likely  that  the  catheter  can  be  successfully  re-moved  because  steady  and  gentle  stretching  may  decrease the size of the knot. It has been advocated that the catheter should be removed at the site of needle insertion, where the withdrawal forces are lowest;  because  the  insertion  point  and  removal  point are comparable.¹⁶ There is evidence indicating that the withdrawal force is reduced in the lateral decubitus position.¹⁷ It has also been reported that the length of a nylon catheter can be increased by about  30%  without  breaking  and  that  the  tensile  strength is several-fold greater than the force re-quired to remove the catheter.¹⁸

If knotting of an epidural catheter is suspected, imaging  may  help  to  visualize  the  structure  of  the knot and its position, assuming the catheter is radio-opaque. For an epidural catheter that is not radio-opaque, such as in the present case, the in-jection of contrast medium is needed for visualiza-tion. In one previously reported case, the catheter was found to be entrapped in a facet joint.¹⁹ Inser-tion of a guidewire to assist visualization has also been  proposed.¹³  Computed  tomography  or  ultra-sonography may also be used to identify the site of entrapment.^20,21

In  Beilin  et  al’s  study,  there  was  an  increased  incidence of vascular cannulation and inadequate analgesia if the catheter was inserted into the epi-dural space by 7 cm or more.²² The optimal length of insertion is 5−6 cm. The epidural catheter in our case probably curled back into the epidural space because  the  inserted  length  of  the  catheter  was  9 cm. If an excess length of the catheter is inserted, it may lead to kinking or twisting, and the catheter may double-back to pass outside an intervertebral foramen, or to wrap around a nerve. When thread-ing the catheter more rapidly, a greater force may be created, which may split the catheter. A study carried out by Muneyuki et al²³ concluded that if a catheter is inserted more than 4.5 cm into the lum-bar  epidural  space,  it  is  more  likely  to  curl  and  form a loop or a knot. They showed that inserting a longer length of the catheter into the thoracic epi-dural  space  is  safer  than  in  the  lumbar  epidural  space. The underlying mechanism may be due to the relative  absence  of  segmentation  in  the  posterior  extradural space at the thoracic level as compared with the lumbar level.²⁴

If a sheared catheter is retained in the body, it is debatable whether it should be removed surgi-cally. Some experts recommend that the catheter is sterile and unlikely to cause neurologic damage.²⁵ It is believed to be safe to leave a catheter in situ if no neurologic sequelae are noted. Surgery may increase the risk of neurologic damage because of the risk of infection, fibrosis or mechanical neural irritation.  However,  some  reports  have  there  are  adverse effects from leaving the sheared catheter in situ. For example, in one case report, a sheared catheter  was  left  in  situ  until  pain  and  weakness  were  noted  18  months  later.²⁶  No  catheter  splin-ters  were  found  after  surgical  decompression.  Another study described a patient who underwent cesarean  delivery  under  general  anesthesia  after epidural anesthesia had failed because of a broken catheter.²⁷ A radiculopathy at the L3 level emerged 7 months later due to a retained fragment of the epidural  catheter  and  was  relieved  by  decom-pressive  surgery.  In  addition,  other  case  reports  have  described  granuloma  formation  and  fibrosis  as  a  result  of  long-term  retention  of  epidural  catheters.²⁸⁻³⁰

In conclusion, the incidence of knotting of lumbar epidural  catheter  is  rare.  If  a  knotting  epidural  catheter is suspected and irretrievable by manipu-lation, it is better to remove it surgically. Clinicians should consider the associated risks and complica-tions of leaving the catheter in situ.