| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Department of Anesthesia and Perioperative Care, University of California, San Francisco, California
Address correspondence and reprint requests to Kenneth Drasner, MD, Department of Anesthesia and Perioperative Care, San Francisco General Hospital, Room 3C-38, San Francisco, CA 94110. Address e-mail to kdrasner{at}anesthesia.ucsf.edu
Few, if any, issues in anesthetic practice have inspired such divergent opinions as placement of epidural catheters, particularly thoracic catheters, in anesthetized patients. Depending on whom one asks or where one is practicing, placing a thoracic epidural catheter while a patient is under general anesthesia might be considered compassionate care or outright malpractice. For example, after a 1994 report in the German literature of two cases of paraplegia (1), the German anesthesiology society, Deutsche Gesellschaft für Anästhesiologie und Intensivmedizin, developed guidelines that classified general anesthesia as an absolute contraindication to placement of a thoracic epidural in adult patients (2). By contrast, in a 1998 survey of British anesthetists (3), 60% of 324 practitioners considered it preferable to insert a thoracic catheter with the patient anesthetized, whereas an additional 20% reported varying their practice on the basis of patient assessment.
In 1999, two case reports—one from the United States and the other from Britain—rekindled this debate. In the former report, Bromage and Benumof (4) reported a case from the closed-claims database in which paraplegia was associated with difficult placement of a thoracic epidural in an anesthetized patient. Clear evidence of trauma was demonstrated on magnetic resonance imaging (MRI), as indicated by an air bubble in the spinal cord at T10. However, there was a region of increased T2-weighted signal in the anterior aspect of the cord between T4 and T5, which was indicative of infarction and was more consistent with the patient’s sensory examination. This latter finding, in combination with episodes of intraoperative and postoperative hypotension, shed doubt on the relationship between the air bubble and the injury, which the authors themselves termed "speculative." Although the report did present a complex argument to link the air bubble with the apparent "watershed" infarct, the conclusion that "the case reinforces the admonition against attempting epidural puncture above the termination of the cord in unconscious, areflexic patients" drew sharp criticism (5).
The second case, reported in the British literature by Mayall and Calder (6), described a patient who developed a hematoma anterior to the spinal cord after multiple failed attempts at placement of a thoracic epidural under general anesthesia. However, as the authors pointed out, insertion of the Tuohy needle into the cord did not, by itself, appear to result in injury because the patient remained asymptomatic for an appreciable time after surgery. Instead, similar to the cases reported in the German literature (1), paraplegia appeared to result from a hematoma that, in the present case, was compounded by a delay in diagnosis and decompression.
Whereas critics of these initial reports focused on the uncertain relationship between the timing of epidural placement and injury, thus deflecting attention to other elements of care, such as inexperience, technical difficulty, inappropriate persistence, and poor standards of perioperative management (5,7), a more recent case reported by Wilkinson et al. (8) cannot be so readily dismissed. After completion of surgery but while the patient was still under anesthesia, a thoracic epidural was inserted at T8, and 10 mL 0.1% bupivacaine was administered. On awakening, the patient complained of severe pain and weakness in both arms. Attempted injection of an additional 1 mL of 0.25% bupivacaine resulted in exacerbation of pain, and the catheter was consequently removed. Neurological evaluations noted persistent weakness of both upper limbs, with diminished sensation between C3 and T10. An MRI showed a clearly demarcated tubular lesion extending from C2 to T1 that was suggestive of central cavitation with peripheral edema; this was inconsistent with an infarct attributable to arterial occlusion or hypoperfusion. Power gradually improved over several weeks, but neuropathic pain persisted. A repeat MRI at 3 mo demonstrated resolution of edema but persistence of the cavitation.
In this issue of Anesthesia & Analgesia, Kao et al. (9) report a similar case of neural injury that was clearly related to spinal cord trauma from a thoracic catheter that had been inserted under general anesthesia. Although critics might focus on the multiple attempts and resistance to catheter passage, few clinicians are so skilled as to place catheters consistently with a single pass without encountering resistance to catheter advancement. The lack of patient complaint with postoperative administration of local anesthetic raises an important point—absence of pain might have been related to the slow rate of infusion or, alternatively, could have resulted from the prior administration of local anesthetic, an additional form of "masked mischief" as coined by Bromage (10).
It has been said that "hard cases make for poor guidelines as well as bad law" (7). However, there are no data, let alone large-scale controlled comparisons, evaluating outcomes for thoracic epidural anesthesia performed in awake versus anesthetized patients, nor are there likely to be. Although it is tempting to extrapolate from a recent large-scale retrospective review by Horlocker et al. (11) of lumbar epidural catheters inserted under general anesthesia, such extrapolation is inappropriate; as the authors appropriately caution, the study did not include thoracic catheters, and <2% were used during surgery or for postoperative infusion of local anesthetic. Additionally, the authors note that, even with a zero incidence in 4298 patients, the possibility of serious neurologic complications may still be as frequent as 0.08% (95% confidence interval), a point echoed in the accompanying editorial (12), which comments on the need for a larger study. Such analysis highlights the difficulty in establishing the safety of a technique when the feared complication is rare, yet catastrophic, because even a mammoth comparative study would have difficulty establishing a statistically significant difference in incidence. In such circumstances, one must dispense with evidence-based medicine and revert to logic-based practice. Clinical care is, by nature, imperfect, and it is inevitable that needles (or catheters) will inadvertently violate the cord. How often is anyone’s guess, but in some cases injury might be averted or minimized by a responsive patient, a point well illustrated in the case reported by Wilkinson et al. (8)
It should be obvious that this discussion does not apply to the performance of epidural anesthesia in anesthetized children, where a conscious patient would likely impart no benefit but would instead add substantial risk. Also, there are exceptional circumstances in which similar considerations would apply to adults. In such cases, informed consent should include a discussion of the potential, albeit undefined, risk.
In their recent editorial, Rosenquist and Birnbach (12) question the need to perform epidurals under general anesthesia, noting that more than a million lumbar epidurals are performed on laboring women each year without sedation or anesthesia. Not being a particularly stoic individual myself, and having had the experience of an epidural placed without sedation and with minimal analgesia, I would concur. And certainly more substantial justification is needed as one moves cephalad along the neuraxis, where the risk of injury is likely more frequent and the consequences likely more severe. In the absence of data, it comes down to a simple question often posed during my residency by one of the faculty: there are many things one can do standing up in a hammock, but why?
 |
References |
|---|
 Top References |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Ng and J. Swanevelder Pain relief after thoracotomy: is epidural analgesia the optimal technique? Br. J. Anaesth., February 1, 2007; 98(2): 159 - 162. [Full Text] [PDF]
|
|
|
|
 |
|
 S. A. Lang Asleep at the Wheel? Anesth. Analg., April 1, 2005; 100(4): 1214 - 1214. [Full Text] [PDF]
|
|
|
|
|
|
K. Drasner Regional Anesthesia Under General Anesthesia and Spinal Cord Injury Anesth. Analg., April 1, 2005; 100(4): 1215 - 1215. [Full Text] [PDF]
|
|
|
|
|
|
J. L. Benumof Regional Anesthesia Under General Anesthesia and Spinal Cord Injury Anesth. Analg., April 1, 2005; 100(4): 1214 - 1214. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
