Correspondence  |   September 2004
Self-extraction of Intrathecal Pump Opioid
Author Affiliations & Notes
  • Keith P. Kittelberger, M.D.
  • * Critical Health Systems of North Carolina, and Carolina Pain Consultants, Raleigh, North Carolina.
Article Information
Correspondence   |   September 2004
Self-extraction of Intrathecal Pump Opioid
Anesthesiology 9 2004, Vol.101, 807. doi:
Anesthesiology 9 2004, Vol.101, 807. doi:
To the Editor:—
Neuraxial opioids have been used successfully to treat chronic malignant and nonmalignant pain since Wang et al.  1 demonstrated pain relief with intrathecal morphine in 1979.2 Intrathecal drug delivery systems are internalized devices capable of administering analgesic medications such as opioids, clonidine, and local anesthetics in precisely controlled doses. Noncoring port access device needles are used to gain access to the pump for refills and aspirations.
Candidates for an intrathecal drug delivery system have chronic intractable pain and, despite other methods of pain control, have not experienced sufficient pain relief or have developed intolerable side effects from systemic analgesics. Before implanting a permanent system it is important to document a reduction in pain intensity, improvement in function, and significant reduction in oral or systemic analgesics.3 We report the case of a patient who withdrew opioid from his intrathecal pump and injected it intravenously.
A 39-yr-old male was referred to Carolina Pain Consultants in March, 1999 with a diagnosis of lumbar postlaminectomy syndrome4 causing chronic low back and bilateral leg pain. Previously, he had undergone five spine surgeries, including a Steffee fusion. In addition, he had a spinal cord stimulator implanted before his referral. Despite improved symptomatology with the spinal cord stimulator, the patient had a persistently high opioid requirement. The patient had also suffered from a postoperative infection necessitating the insertion of a Portacath for antibiotic therapy.
In July, 2001 during an in-hospital trial with intrathecal hydromorphone, the patient's pain was reduced by 80% using the numeric rating scale.5 The next month, he underwent implantation of an intrathecal pump.
On four office visits for pump refills the patient was found to have residual volumes in his pump to be substantially less than expected. The first occurred in November, 2002 when a 13-ml residual volume was predicted but only 1 ml was actually aspirated. This was followed by no residual volumes being aspirated on his next three refills, with residual volume deficits being 2.4 ml, 4.7 ml, and 6.3 ml. During these office visits the discrepancies in pump volumes were discussed with the patient who expressed bewilderment as to the cause.
Because of the concern over persistent pump volume discrepancies a meticulous examination was performed at the time of refill, including the pump site. The patient was then asked to return in 1 week for reevaluation of his pump. The next week on physical examination he was found to have an additional distinct puncture site over his pump. He had no explanation for this finding. The residual volume of his pump was found to be 14 ml when there should have been 16.4 ml. When confronted, the patient admitted to having been given noncoring needles in October, 2002 by a physician not associated with our pain clinic to access his Portacath for the administration of promethazine for nausea. Our suspicion that he was withdrawing drug from his pump and injecting it directly into his Portacath was corroborated by a call from his mother who had witnessed him performing these injections.
The pump was emptied and turned off at this office visit. Oral and transdermal clonidine were prescribed to reduce the symptoms of opioid withdrawal.6 
He was sent to the emergency room for psychiatric evaluation and admission to a substance abuse program. We present this case to illustrate an important point. Accurate measurement of the residual volume in the pump reservoir at the time of refill and comparison to the expected residual volume is critical. Any deviation from the expected residual volume must be investigated. Our pain clinic has approximately 75 patients with intrathecal pumps. Our experience with hundreds of pump refills is that the actual residual volume aspirated from the pump consistently differs from the expected residual volume by less than 1 ml.
* Critical Health Systems of North Carolina, and Carolina Pain Consultants, Raleigh, North Carolina.
Wang JF, Nauss LA, Thomas JE: Pain relief by intrathecally applied morphine in man. Anesthesiology 1979; 50:149–51Wang, JF Nauss, LA Thomas, JE
Tollison CD, Satterthwaite JR, Tollison JW: Neuraxial opioid therapy, Practical Pain Management. Philadelphia, Lippincott Williams & Wilkins, 2002, pp 135–54Tollison, CD Satterthwaite, JR Tollison, JW Philadelphia Lippincott Williams & Wilkins
Waldman SD: Intraspinal analgesia for nonmalignant pain, Interventional Pain Management. Philadelphia, WB Saunders, 2001, pp 609–19Waldman, SD Philadelphia WB Saunders
Bonica, JJ: Low back pain, The Management of Pain. Philadelphia, Lea & Febiger, 1990, pp 1471–2Bonica, JJ Philadelphia Lea & Febiger
Bonica, JJ: Measurement of pain, The Management of Pain. Philadelphia, Lea & Febiger, 1990, p 581Bonica, JJ Philadelphia Lea & Febiger
Stoelting, RK. Antihypertensive drugs, Pharmacology and Physiology in Anesthetic Practice. Philadelphia, Lippincott, 1987, pp 297–8Stoelting, RK Philadelphia Lippincott