Intravascular placement of the epidural catheter can occur during epidural anaesthesia and is more frequent in the obstetric patient.1 Unrecognized intravascular placement can have potentially lethal complications. We report a case of late migration of the epidural catheter into the vascular compartment. The literature is reviewed and based on the present experience and the literature review suggestions are made to avoid serious complications of local anaesthetic systemic toxicity.
A 32-year-old American Society of Anesthesiology (ASA) class I, primiparous patient, without prior medical or surgical history, was transferred at 24 weeks' gestation to a tertiary care, university obstetric department, for confirmation of the diagnosis of partial agenesis of the corpus callosum and additional central nervous system structural abnormalities of her fetus. At 28 weeks' gestation, the patient was admitted to the labour and delivery ward experiencing significant painful contractions. She was in active labour. The diagnosis of intrauterine death was made following cardiotocography and ultrasound examination. No cause could be identified although placental abruption was suspected. Routine laboratory investigations were performed including a coagulation screen which were within normal limits.
At 11 p.m. the parturient requested epidural analgesia. Within 30 min the attending resident on call (5th year of training) initiated analgesia using a combined spinal epidural (CSE) technique as per protocol. The epidural space was identified using an 18G Tuohy needle (Perifix; B-Braun, Melsungen, Germany) with the loss of resistance to saline technique at the L3–L4 interspace. A 27G Pencil Point (Pencan; B-Braun) spinal needle was introduced through the epidural needle and through the dura mater into the cerebrospinal fluid. An injection of 2.5 ml ropivacaine 0.175% with sufentanil 0.75 μg ml−1 was administered intrathecally. A 20G multiorifice epidural catheter (Perifix) was advanced 7 cm into the epidural space. An aspiration test of the catheter was performed prior to connection of the filter. No blood could be aspirated. No epidural test dose was given. Within 10 min the patient experienced no further pain from her uterine contractions and was comfortable.
As is routine practice in our department, patient controlled epidural analgesia (PCEA) with background infusion was initiated 10 min after the spinal dose. The PCEA device was programmed to infuse 2 ml h−1 as a background infusion and 4 ml boli with a lock-out of 15 min of our standard epidural solution containing ropivacaine 0.175% with sufentanil 0.75 μg ml−1. The maximal total hourly dose a patient could receive was 18 ml of the analgesic solution. The patient remained without pain and comfortable and consumed 50 ml of the epidural solution by 8 a.m. the next morning. At this moment the epidural syringe was refilled with a further 50 ml of the standard anaesthetic solution. The patient remained without pain from her contractions. The block was checked by the attending resident anaesthetist and found to be at the T10 level bilaterally. At this moment no aspiration of the catheter was performed.
Around noon (13 h after initial epidural catheter placement), the contractions became painful and pain could not be managed using self-administered PCEA boli. The attending anaesthetist was called to treat breakthrough pain. The block was tested using cold sensation. No clear sensory block was noted at any dermatome. Again no aspiration test was performed. Therefore, the attending resident decided to administer an epidural bolus of 10 ml of the standard epidural solution, which also was used in the PCEA device. Within 1–2 min the patient reported visual disturbances, metal taste in her mouth, perioral tingling and difficulties to pronounce words. Immediately intravenous (i.v.) positioning of the catheter was suspected. The bacterial filter was disconnected from the epidural catheter and an aspiration test was performed. Blood was aspirated. The catheter was subsequently redrawn 3 cm. No blood could then be aspirated. The signs of local anaesthetic toxicity disappeared within 10 min, after which a formal test dose using 3 ml lidocaine 2% with adrenaline was administered. No tachycardia or hypertension was noted. A 10 ml bolus of our analgesic solution was given over a period of 10 min after which the patient became comfortable within 20 min. One hour later she delivered a dead fetus. The postpartum course was uneventful and she went home on the first postpartum day.
The epidural catheter in this patient had been providing satisfactory epidural analgesia for almost 12 h. Breakthrough pain required intervention by an anaesthetist. The injection of a local anaesthetic/opioid mixture, to alleviate labour pain, produced clear symptoms of an i.v. injection. Intravenous migration of the epidural catheter was confirmed following the aspiration of blood. Several other reports of epidural catheter migration into the vascular space have been previously reported, often but not exclusively in obstetric patients.2–9
The potential consequences of intravascular placement or migration include failed analgesia/anaesthesia, signs of local anaesthetic toxicity, seizures and cardiac arrest. The incidence of intravascular positioning of epidural catheters has been reported to be around 1% in nonpregnant patients.5 However, the incidence seems to be higher in pregnant patients with a reported incidence up to 15%.10–12 Pan et al.11 performed a large retrospective analysis of almost 15 000 regional anaesthetic techniques during labour or Caesarean section (C-section). Apart from a 6% initial vascular cannulation of epidural catheters, these authors also reported a 0.25% incidence of late migration of epidural catheters into the vascular space.11 These authors, interestingly, noted that intravascular placement of the epidural catheter occurred less frequently when the catheter was positioned as part of a CSE technique as opposed to a conventional epidural. But the present case, in which the catheter was positioned as part of a CSE, stresses the fact that it can occur with any technique.
In the present patient following 13 h of adequate analgesia, breakthrough pain occurred. The anaesthetist tested for the presence of a sensory block which was not present. It can be discussed as to whether an epidural top-up should have been given as initial treatment. We admit that an aspiration test or the administration of a formal test dose would have been preferential. The top-up was however a diluted local anaesthetic solution with virtually no risk of cardiac arrest. When a more concentrated solution would have been required, aspiration or a formal test dose would have been mandatory. However, probably it would be wise to perform an aspiration test before every manual top-up even with less concentrated solutions.
Various tests to evaluate correct epidural catheter placement have been evaluated. A thorough review of all tests was recently published by Bell and Leslie.12 Ideally a test should have 100% specificity and 100% sensitivity. Unfortunately, such a test does not exist. All tests described produce false positive and false negative results. Aspiration is frequently used. Single orifice catheters should not be tested using aspiration as more than 50% false negative results have been reported.12 In contrast, multiorifice catheters can be tested using aspiration. Norris et al.13 noted very few false negative results, comparable with other methods of testing for intravascular position. Testing the epidural catheter using an adrenaline containing solution is the best known pharmacological strategy. However, especially in parturients this technique produces many false positive results.12 Furthermore, i.v. adrenaline may impair uteroplacental perfusion and may have tocolytic effects, whereas epidural adrenaline may increase motor block. Additionally, i.v. injection of adrenaline in preeclamptic patients is not without risks.14 However, the dose of adrenaline used in a test dose is small and very unlikely to produce significant problems. An important indirect method to test the catheter is failure to establish effective analgesia or anaesthesia. This was the first indication in the present case to suggest something was wrong.
Many anaesthetists would not consider giving a test dose to patients receiving labour epidural analgesia or adapt the test dose (excluding adrenaline or reduce the amount of local anaesthetic used).15 Because the risk of toxicity or high spinals is limited with the use of diluted solutions of local anaesthetic given in modern labour analgesia,16 many will now give no formal test dose or use the initial diluted loading dose as a test.15 This seems reasonable, especially if the catheter is tested whenever the catheter is subsequently used to administer large doses. The present case underlines that even a tested catheter can migrate and thus evaluation of the catheter remains essential whenever failure of the block occurs or whenever large doses of local anaesthetic are given.
Owen et al.17 reported that with ropivacaine and levobupivacaine fewer signs of neurotoxicity can be observed when i.v. injection is performed. However, the present case underlines that even with these new local anaesthetics, these symptoms might occur and it is therefore important to specifically ask for these symptoms if a top-up is administered.
Some might disagree with our choice to withdraw the catheter for 3 cm and using it again. Norris et al.13 demonstrated that catheters that are withdrawn often remain intravascularly despite a negative aspiration test. Many would state that the option should have been to replace this catheter. This most likely would have been our action as well if the risk of unplanned C-section was high. An unplanned C-section was however highly unlikely in the present scenario and therefore we opted for the above described choice. Furthermore, the patient was in well established labour and was expected to deliver within a reasonable time frame.
The present case describes the late migration of a well functioning labour epidural catheter into the intravascular space. The case underlines the importance of performing a sensory block evaluation when breakthrough pain occurs. The case also underlines that when sensory block is absent, intravascular positioning must be considered and epidural catheter aspiration should be performed prior to administering a top-up.
1 Camorcia M. Testing the epidural catheter
. Curr Opin Anaesthesiol 2009; 22:336–340.
2 Blouin RT, Ruby ST, Gross JB. Determination of intravascular migration of an epidural catheter
using the air technique. Anesthesiology 1993; 79:1427–1429.
3 Bush DJ, Kramer DP. Intravascular migration of an epidural catheter
during postoperative patient-controlled epidural analgesia. Anesth Analg 1993; 76:1150–1151.
4 McLean BY, Rottman RL, Kotelko DM. Failure of multiple test doses and techniques to detect intravascular migration of an epidural catheter
. Anesth Analg 1992; 74:454–456.
5 Ravindran R, Albrecht W, McKay M. Apparent intravascular migration of epidural catheter
. Anesth Analg 1979; 58:252–253.
6 Zebrowski ME, Gutsche BB. More on intravascular migration of an epidural catheter
. Anesth Analg 1979; 58:531.
7 Hodgkinson R, Husain FJ. Intravascular migration of epidural catheters. Anesth Analg 1979; 58:443–444.
8 Lee BB, Chen PP, Chiu A. Intravascular migration of epidural catheter
during cesarean section. Anesth Analg 1996; 83:1347–1353.
9 Dickson MAS, Doyle E. The intravascular migration of an epidural catheter
. Paediatr Anaesth 1999; 9:273–275.
10 Colonna-Romano P, Nagaraj L. Tests to evaluate intravenous placement of epidural catheters in laboring women: a prospective clinical study. Anesth Analg 1998; 86:985–988.
11 Pan PH, Bogard TD, Owen MD. Incidence and characteristics of failures in obstetric neuraxial analgesia and anesthesia: a retrospective analysis of 19259 deliveries. IJOA 2004; 13:227–233.
12 Bell DN, Leslie K. Detection of intravascular epidural catheter
placement: a review. Anaesth Intensiv Care 2007; 35:335–341.
13 Norris MC, Fogel ST, Dalman H, et al
. Labor epidural analgesia without an intravascular test dose
. Anesthesiology 1998; 88:1495–1501.
14 Mulroy M, Glosten B. The epinephrine test dose
in obstetrics: note the limitations. Anesth Analg 1998; 86:923–925.
15 Gardner IC, Kinsella SM. Obstetric epidural test dose
: a survey of UK practice. IJOA 2005; 14:96–103.
16 Van Zundert A, Vaes L, Soetens M, et al
. Every dose given in epidural analgesia for vaginal delivery can be a test dose
. Anesthesiology 1987; 67:436–440.
17 Owen MD, Gautier P, Hood DD. Can ropivacaine and levobupivacaine be used as test doses during regional anesthesia ? Anesthesiology 2004; 100:922–925.