To the Editor:-I read with interest and consternation Drs. Jaeger and Madsen's [1]correspondence on what they presumed to be delayed subarachnoid migration of an epidural Arrow FlexTip Plus catheter. There are several areas in this case report that concern me.

First and foremost is in regard to patient monitoring and appropriate documentation. It is worrisome that the only references to the patient's level of consciousness or motor function on the evening preceding the event are the following “… Sometime that evening the patient's husband recalled that she was more somnolent and seemed weaker …” and “… At 2:00 AM the next day, the nurse found the patient barely responsive …”. Incomplete documentation at the time of the critical incident is also apparent, when no arterial blood gas is drawn, when fluid is aspirated from the epidural catheter but is not tested to determine its nature, and when the catheter is removed before any confirmation of its location. Thankfully, the patient was successfully treated.

Vigilance in the operating room should not be left at the door as we tread into the unfamiliar territory of intensive care units, “step-downs,” and hospital wards. It is therefore mandatory that any group of clinicians that sets itself up as an Acute Pain Service (APS) ensures that patients in its care are appropriately monitored and that evidence of monitoring be documented in the nursing record. Standing orders should be clear as to when to alert the APS, particularly with respect to level of consciousness and motor function, and the APS must be available to respond 24 h per day.

Second, the epidural infusion of bupivacaine 0.125% with hydromorphone 6 mg/ml, I trust, is a printing error. At 14 ml/h, this would result in 84 mg hydromorphone delivered! For our routine thoracic cases we use T4-T8 hydromorphone 0.05 mg/ml without local anesthetic, at rates of 2–5 ml/h (0.1–0.25 mg/h) to achieve dynamic pain control. When we do use thoracic local anesthetic/opioid admixtures, we use bupivacaine 0.1% with 0.005 mg/ml fentanyl at rates of 6–12 ml/h. We have considered bupivacaine/hydromorphone admixtures, but there is no current literature to guide us in this regard. Based on other admixtures however, bupivacaine 0.1–0.125% with hydromorphone 0.010 mg/ml would seem to be appropriate.

Finally, with respect to their conclusion that the catheter migrated into the subarachnoid space, it is at least as plausible that the catheter remained in the epidural space but that the infusate found its way intrathecally via at least one rent in the dura. There was obvious technical difficulty in attempting to secure epidural catheter placement, before surgery, at T7-T8; successful placement did not occur, they state, despite “multiple” attempts. The fact that “doses of local anesthetic” failed to confirm placement of a catheter or that blood, cerebrospinal fluid, or paresthesia were not noted during these attempts does not rule out dural puncture at this level. Second, there was documentation that a dural puncture did occur at T11-T12 during attempts to secure epidural catheter placement after surgery. Eventually, a catheter appears to have been successfully passed at T10-T11. The fact that the Tuohy needle was “oriented cephalad,” however, does not rule out catheter tip placement adjacent to the documented T11-T12 dural puncture and may have resulted in catheter placement close to an unrecognized dural puncture at T7-T8, particularly when it appears that an excessive amount of catheter was threaded into the epidural space.

Gordon O. Launcelott, M.D., F.R.C.P.C.

Director Acute Pain Service; Department of Anaesthesia; QE II Health Sciences Centre; Halifax, Nova Scotia, Canada

(Accepted for publication January 20, 1998.)


Jaeger JM, Madsen ML: Delayed subarachnoid migration of an epidural Arrow FlexTip Plus catheter. Anesthesiology 1997; 87:718-9.