The opportunity for error increases as medicine becomes more complex. Despite safeguards to prevent errors, unintentional epidural administration of IV medications continues, occasionally with devastating consequences. Perioperative and critical care patients with multiple intravascular catheters, IV infusions, epidural or perineural catheters, and surgical drains are particularly vulnerable. The potential for numerous connections increases the risk of injecting medications into the wrong site, such as epidural or spinal medications accidentally given IV or a medication intended for IV infusion unintentionally connected to and infused into the epidural space.
When an IV medication is accidently injected into the epidural space, it is critical to anticipate sequelae and institute appropriate management. Although, institutions use safeguards to prevent accidental injections of IV medications into epidural or spinal catheters, multiple cases have been described, occasionally with devastating consequences ranging from temporary and permanent paralysis to death.1–8 The common link among these reports is the ubiquitous Luer lock connectors used in both IV and neuraxial catheter, permitting wrong-route drug administration.
Permission for publication of this case was given by our local IRB.
The patient was a 25-year-old male who sustained left-sided 2 to 11 rib fractures and a clavicle fracture in a pedestrian versus train accident. He was briefly hospitalized but was discharged home after observation. He subsequently developed empyema, requiring rehospitalization and a thoracotomy. After thoracotomy, decortication, and placement of chest tubes for the empyema, the comprehensive pain service was consulted in the postanesthesia care unit for severe refractory postoperative pain. On postoperative day (POD) 1, the pain service placed an epidural at the T6 level without complication. The protocol at our institution involves a member of the pain service ordering, connecting, and initiating the epidural infusion pump at the patient’s bedside. We ordered a 0.1% bupivacaine infusion at 10 mL/h via patient-controlled epidural analgesia, with a demand dose of 3 mL available every 15 minutes (GemStar Infusion Pump; Hospira Inc., Lake Forest, IL). A 250-mL infusion bag was prepared by the central pharmacy. A member of the pain service received the preprimed and programmed epidural pump from the operating room pharmacy, connected it to the epidural catheter, and initiated the epidural infusion. Our hospital’s policy is that subsequent epidural infusion bag changes are to be performed at the patient’s bedside by the primary nurse with a second nurse performing verification. All bag changes are to occur within the pump. No connections or disconnections are to occur at the epidural catheter itself.
The patient did well overnight with only mild breakthrough pain, well controlled with the epidural infusion in conjunction with patient-controlled analgesia (PCA) IV hydromorphone. The IV PCA was in use to provide analgesia for the clavicle fracture, which had been causing significant pain outside the dermatomal coverage of the epidural. On the morning of POD 2, the pain service evaluated the patient and after assessment of the epidural, determined it to be functioning well, but with a narrow dermatomal level. The epidural was manually bolused with 5 mL of 0.125% bupivacaine, resulting in improved pain relief. An order was written to increase the epidural infusion rate to 12 mL/h. The pain service also interrogated the epidural pump and evaluated the medication as per usual assessment on rounds and found it functioning normally.
Approximately 12 hours after the morning assessment and rate adjustment during rounds, the pain service was contacted by the bedside nurse caring for the patient. The nurse informed the pain service that soon after she started her evening shift, she found a bag of phenylephrine infusing through the epidural pump. She immediately discontinued the infusion and contacted the pain service. The pain service arrived at the patient’s bedside within a few minutes to find a partially empty bag of phenylephrine 120 μg/mL, which had been infusing at a rate of 12 mL/h through the epidural pump. We calculated that over approximately 12 hours, 135 mL of the phenylephrine solution for a total dose of approximately 16.2 mg would have infused through the epidural catheter.
On examination, the patient was found to be hypertensive with arterial blood pressures in the 160s/80s, whereas his systolic blood pressures were in the 100s/130s on the previous day. He was also complaining of significant pain at his thoracotomy incision, chest tube sites, and had no discernible dermatomal anesthetic level. A thorough neurologic examination demonstrated that the patient was alert and oriented with full strength and sensation in his upper and lower extremities. Given the urgency of the situation, there was concern that pooled phenylephrine in the epidural space could harm the patient. Having no specific guidelines for a neuroprotective treatment strategy, a plan of care was created based on clinical judgment to continue an epidural infusion with preservative-free normal saline at 12 mL/h for approximately 12 hours to dilute the phenylephrine in the epidural space, maintain continuous monitoring overnight, and perform hourly neurologic examinations. Pain was managed with multimodal therapy including IV PCA and oral nonsteroidal anti-inflammatory medications.
There were no changes in the patient’s neurologic examination overnight. The patient’s arterial blood pressure returned to baseline the following day. The patient continued to improve during the next few days with no changes in his neurologic examination. He was discharged home on POD 5 without further sequelae.
Root-cause analysis identified that on the morning of POD 2 after the pain service was present at rounds, the bedside nurse accidentally replaced the epidural solution with a bag of phenylephrine, which had been at the patient’s bedside in preparation for IV infusion. The nurse was apparently unfamiliar with epidural management and had bypassed the policy that required a second nurse to confirm the correct medication, correct patient, and correct infusion route before any epidural infusion change.
Vigilance and effective systems are required when administering medications to ensure that the proper medication is being administered to the correct patient by the correct route. This is particularly critical when a neuraxial catheter is present. Despite safeguards to prevent wrong-route administration, ether,1 potassium,5 lipid emulsion,6 neuromuscular-blocking agent,7 and antibiotics,8 have been infused into neuraxial catheters. Sequelae have ranged from localized pain to brief paraplegia, permanent paraplegia, and death.1–8
Contributing factors to wrong-route errors were lack of staff familiarity with epidural pumps and medications, identical connecting system for IV and neuraxial catheters (i.e., Luer lock), and complex patients with multiple entry points for infusions (IV, arterial, peripheral venous, central venous, peripheral nerve, and neuraxial). In our case, an infusion bag of phenylephrine was available at the patient’s bedside. The bag of phenylephrine resembled the bag of medication prescribed for his epidural infusion. The nurse caring for the patient was not familiar with management of epidural infusions and failed to follow hospital policy for double-checking epidural medications with another nurse.
After a root-cause analysis for this case, our institution took several steps in an attempt to reduce future medication errors. The epidural infusion pumps are now designated with a sticker on the inside epidural pump cover (Fig. 1) to prompt the user to double-check patient information, infusion route, and drug dose before initiating an epidural infusion. These changes were in addition to the safeguards already in place, which included a bright yellow striped infusion tubing (Fig. 2), making it distinct from IV infusion tubing, a bright yellow sticker on the front of the epidural medication bag (Fig. 3), and the requirement for 2 nurses to complete a brief checklist when initiating or changing epidural medication bags.
Despite these changes, this case highlights the weak link that permits wrong-route drug administration: identical catheters, tubing, and even infusion pumps for IV and neuraxial medications. Differentiating the type of connectors used for neuraxial catheters and IV infusion tubing appears to be a potential solution for reducing medication errors.
In 2012, Birnbach and Vincent9 strongly advocated for the change of epidural catheter connectors to non-Luer connectors as a mechanism to prevent confusion with the Luer connectors used for IV tubing and reduce one type of preventable medication error. We find it astonishing that despite multiple reported events and years of calling for change, the process of progressing to a new connector system for neuraxial infusions continues to be tediously slow.10 Despite past hurdles such as clinician resistance, lack of quality replacements, and the increased cost to research and purchase new catheter connection points such as filters, tubing, and syringes, the change to new connector systems needs to be expedited for patient safety.
To provide continued attention to this problem, we also suggest that wrong-route drug delivery involving neuraxial administration be considered a sentinel event. It is consequential, and as avoidable as administering a mismatched unit of blood. If this is considered a sentinel event, with every case undergoing root-cause analysis, then health care providers will likely take the “double-check” requirements as seriously for neuraxial drug administration as they do for transfusion.
In conclusion, current safeguards to prevent wrong-route neuraxial and IV medication infusions are inadequate. As physicians called to “do no harm,” we must adopt and enforce policies to prevent wrong-route drug administration. Such policies can reduce the risk. However, there will always be a small risk until manufacturers develop noninterchangeable systems for neuraxial and IV infusions to completely eliminate the possibility of wrong-route drug administration.
1. Mappes A, Schaer HM. Accidental injection of ether into the epidural space. Anaesthesia. 1991;46:124–5
2. Hew CM, Cyna AM, Simmons SW. Avoiding inadvertent epidural injection of drugs intended for non-epidural use. Anaesth Intensive Care. 2003;31:44–9
3. Dyer C. Doctors suspended after injecting wrong drug into spine. BMJ. 2001;322:257
4. Beckers A, Verelst P, van Zundert A. Inadvertent epidural injection of drugs for intravenous use. A review. Acta Anaesthesiol Belg. 2012;63:75–9
5. Shanker KB, Palkar NV, Nishkala R. Paraplegia following epidural potassium chloride. Anaesthesia. 1985;40:45–7
6. Bickler P, Spears R, McKay W. Intralipid solution mistakenly infused into epidural space. Anesth Analg. 1990;71:712–3
7. McHaourab A, Ruiz F. Inadvertent caudal epidural injection of cefazolin. Anesthesiology. 2001;94:184
8. Sánchez-Bailén MJ, Roca P, Benlloch R. [Accidental administration of atracurium through an epidural catheter]. Rev Esp Anestesiol Reanim. 2006;53:455–6
9. Birnbach DJ, Vincent CA. A matter of conscience: a call to action for system improvements involving epidural and spinal catheters. Anesth Analg. 2012;114:494–6
10. . The Joint Commission. Managing risk during transition to new ISO tubing connector standards, Sentinel Event Alert #53. 2014