ISMP articles on catheter misconnections
Epidural-IV route mix-ups: Reducing the risk of deadly errors
From the July 3, 2008 issue
Problem: A number of risks have been associated with epidural injections and infusions. One of the most significant risks involves erroneous infusions of epidural medications—particularly epidural infusions containing bupivacaine—by the IV route of administration. IV bupivacaine can quickly lead to cardiotoxicity. A Black Box warning for bupivacaine notes that the drug can cause profound disturbances in cardiac rhythm and contractility that are resistant to typical resuscitation efforts, making these mix-ups particularly deadly.(1-3) Likewise, medications intended for IV administration—particularly morphine and vincristine—have been administered via the epidural or intrathecal route, also leading to fatal outcomes.(4-5)
ISMP and the National Patient Safety Agency (NPSA) in the United Kingdom (UK) have published numerous alerts regarding mix-ups between epidural and IV medications.(1-5) Much like the error we reported1 in which a 16-year-old patient in labor died after receiving IV fentanyl and bupivacaine, the NPSA and British news media recently published information about another young woman who died after receiving IV bupivacaine.(2-6) In this case, the woman in labor should have received normal saline IV, but a nurse accidentally selected a virtually identical bag of bupivacaine located in the same unlocked drawer as the saline. (The bupivacaine infusion did not contain fentanyl, so the bag did not require locked storage.) Since the nurse thought she was hanging a bag of normal saline, she had no reason to consider asking another nurse to double-check the solution before administering it. The patient developed seizures and cardiac arrest that could not be treated successfully.
Unfortunately, these cases are not isolated events. Between 2000 and 2004, three additional deaths were reported in the United Kingdom following IV administration of epidural bupivacaine.(2) Between 2005 and 2006, another six events were reported in which epidural medications were given IV.(2) In the US, we are aware of three additional cases between 2004 and 2007, one fatal and two near fatalities. In one of the near-fatal cases, the patient received a small IV loading dose of bupivacaine and morphine intended for epidural patient-controlled analgesia. We have also published multiple cases in which IV medications, such as vincristine, were accidentally administered by the intrathecal route, again leading to fatalities.(5)
Safe Practice Recommendations: Unlike many threats to patient safety, those involving epidural-IV mix-ups are well understood and can be prevented by industry changes to make IV and epidural syringe and tubing connections incompatible with each other. ISMP and other safety agencies, both in the US and internationally, continue to work toward that goal. Until then, we encourage you to evaluate the risks in your organization and implement the recommendations in the checklist found on pages 2-3. (Sadly, many of these recommendations were previously published by ISMP and NPSA well before the most recent tragic events, but action had not been taken to prevent these errors.)
It’s important to note that individual recommendations in the checklist will not prevent all types of mix-ups between IV and epidural medications. A process that reflects careful consideration of all risk points in the medication system in your organization should be implemented to safeguard against both the accidental administration of an IV medication by the epidural route and administration of an epidural medication by the IV route.
Reduce the risk of epidural-IV route mix-ups(1-9)
Prescribing IV and epidural medications
- When appropriate, consider use of other agents for epidural administration that may be less cardiotoxic than bupivacaine (e.g., NAROPIN [ropivacaine]). (While there is evidence to suggest that newer agents have less cardiotoxicity than bupivacaine, controversy continues over whether bupivacaine should be replaced by the new agents.)
- Require the route to be clearly defined on all prescriptions/orders.
- To reduce the risk of tubing mix-ups, replace a peripheral IV infusion used only to keep a vein open with a saline lock to maintain IV access.
Dispensing IV and epidural medications
- Prepare infusions not available commercially in the pharmacy or outsource their preparation.
- Dispense epidural medications to clinical areas in the appropriate container needed for administration (e.g., properly labeled syringe, small volume bag).
- Use barcode technology when preparing and dispensing IV and epidural medications.
- Establish a process to ensure delivery of the correct epidural medication to the correct clinical unit. In low-volume use areas, dispense the epidural medication immediately before use and hand it to an authorized clinician. In high-volume areas (e.g., labor and delivery), place the epidural medication immediately in the appropriate storage location. Epidural medications should not be left in medication rooms for clinical staff to put away, or sent to units in pneumatic tubes.
Dispensing intrathecal medications
- Dispense intrathecal medications in overwraps that help differentiate these syringes and bags from other medications intended for IV administration.
Dispensing IV vincristine
- Dispense IV vincristine in a small volume bag to differentiate it from syringes used for intrathecal medications.
Labeling epidural medications
- Clearly label infusion bags and syringes that contain epidural medications with ‘For Epidural Use Only’ in a large font. Use color and design to differentiate these products from IV medications.
Storing IV and epidural medications
- Reduce the risk of mix-ups by separating the storage of epidural and IV infusions (including those stored among controlled substances).
Infusion pumps and administration sets
- For epidural infusions, use pumps that look different than pumps used for IV infusions.
- Clearly label pumps used to deliver epidural medications as “Epidural Only.”
- When possible, use smart pump technology when administering epidural and IV medications.
- Avoid the use of dual-channel pumps for simultaneous administration of IV and epidural infusions.
- Use yellow-lined tubing without injection ports for epidural infusions to set its appearance apart from typical IV tubing. Never use yellow-lined tubing for anything other than epidural administration.(8)
- Place a neon sticker on the epidural tubing stating “Epidural” (which is often included with the special, yellow-lined tubing).
- Consider placing IV pumps and epidural pumps on opposite sides of the patient’s bed to better separate the two infusion systems.
- Always trace a tube or catheter from the patient to the point of origin before connecting any new device or infusion or adjusting the infusion rate.
- Always hang epidural and IV bags in pumps with the labels facing out, so they can be read. (Pharmacy labels should be applied to accommodate loading syringes or bags in a pump with the labels facing out.)
Administering IV and epidural medications
- Require an independent double-check at the bedside of all epidural medications and IV opioid medications so that verification of the patient, pump settings, and line attachment can be included along with verification of the drug and dose/concentration.
- Require the receiving nurse and transferring nurse to verify pump settings and line attachments during change of shifts or patient transfers.
- Use barcode technology when administering medications to verify patient and product selection.
- Establish a resuscitation protocol to treat the effects of bupivacaine toxicity wherever this drug is administered. Make the protocol and required medications readily accessible to staff on code carts or with other secured emergency supplies. (There is some evidence that the use of lipid emulsion is of benefit to treat bupivacaine toxicity.[2,9])
Staff education and competencies
- Develop a credentialing process to ensure that all practitioners expected to hang epidural infusions and program pumps are competent.
- Heighten awareness of the risk for mix-ups between epidural and IV infusions among clinical staff.
- Educate staff who prescribe, dispense, and care for patients receiving bupivacaine to recognize and manage toxicity using the established resuscitation protocol.
Recommendations for the medical device and pharmaceutical industry
- Design and manufacture unique epidural connectors that cannot be connected to IV ports, or vice versa.
- Design pumps used to administer epidural medications in a way that clearly differentiates them from pumps used to administer IV infusions.
- Provide more commercially available epidural medications in ready-to-use, bar-coded containers that look different than IV medication containers.
References: 1) Institute for Safe Medication Practices (ISMP). More to the story. Safety Briefs. ISMP Medication Safety Alert! August 24, 2006:1-2. 2) National Patient Safety Agency (NPSA). Safer practice with epidural injections and infusions. Patient Safety Alert. March 28, 2007:1-8. Accessed on June 30, 2008, at: www.npsa.nhs.uk/patientsafety/alerts-and-directives/alerts/epidural-injections-and-infusions/. 3) NPSA. Stakeholder consultation summary; safer use of epidural medications. June 2008; personal communication from David Cousins to Michael Cohen. 4) ISMP. IV potassium given epidurally: getting to the route of the problem. ISMP Medication Safety Alert! April 6, 2006:1-2. 5) ISMP. Fatal administration of IV vincristine. ISMP Medication Safety Alert! December 5, 2005:1-2. 6) Hill M. Epidural drug drip ‘killed’ new mother. BBC News. February 5, 2008. Accessed on June 30, 2008, at: http://news.bbc.co.uk/2/hi/uk_news/england/wiltshire/7219434.stm. 7) The Joint Commission. Tubing misconnections—a persistent and potentially deadly occurrence. Sentinel Event Alert. April 3, 2006; Issue 36. Accessed on June 30, 2008, at: www.jointcommission.org/SentinelEvents/SentinelEventAlert/sea_36.htm. 8) ISMP. Don’t use epidural tubing for an IV solution. ISMP Medication Safety Alert! January 17, 2006:1-2. 9) ISMP. IV lipid emulsion for bupivacaine toxicity. ISMP Medication Safety Alert! December 14, 2006:3.
IV tubing misconnected to tracheostomy collar
From the June 14, 2007 issue
While bathing a patient, a nursing staff member accidentally connected IV tubing from an insulin infusion to the port of a tracheostomy collar. (A tracheostomy collar secures the tube’s positioning once the surrounding balloon is inflated to minimize movement of the tracheostomy tube. Since a parenteral syringe is typically used for inflation, the port has a Luer connector.) Normally, patients wore gowns with sleeves that snapped closed. However, on the day of the error, this patient happened to be wearing a gown without snaps. In order to remove the gown, the staff member disconnected the patient’s IV insulin infusion and threaded it through the sleeve. When reconnecting the tubing, the nursing staff member connected the IV line to the Luer connection on the tracheostomy collar. Fluid began to further inflate the tracheostomy collar balloon, constricting the tube itself. The error was noticed when the patient began to decompensate and became cyanotic. The IV line was immediately disconnected from the tracheostomy collar, and approximately 30 mL of fluid was removed from the balloon. Fortunately, the patient suffered no permanent harm.
In our November 28, 2001 newsletter, we reported a similar error in which an IV solution was accidentally infused into the balloon of a tracheostomy collar, resulting in hyperinflation of the collar, constriction of the tube’s lumen, and an airway obstruction. High-volume, low-pressure collars are used to lower the long-term risk of tracheal injury, so they are compliant enough to accept large volumes of air—or fluid, in this case. The patient’s roommate alerted a nurse to the patient’s respiratory distress, and a code was called. The error was noticed when the code team was unable to inflate the lungs. The line was disconnected and the fluid was removed from the collar, but the patient died.
Several factors contributed to the 2001 error. In low, overhead lighting during the night, the nurse had mistaken the tracheostomy collar port as a triple lumen catheter port, especially since the unsecured tubing hung down at the same level as the tracheostomy collar tubing. Although the size of the tracheostomy collar line is distinctly thinner than regular IV tubing, triple lumen tubing is similarly thin like tracheostomy collar line. Furthermore, the connection ports on needleless tubing can appear very similar to the tracheostomy collar line–an interesting example of how even a powerful safety innovation—a needleless system—can lead to new, unexpected problems.
ISMP has received reports of numerous other inadvertent misconnections to balloon ports of various catheters and tubing, including:
- Accidental injection of drugs into an endotracheal tube collar during resuscitation efforts
- Drugs inadvertently delivered into the balloon inflation ports of Foley catheters and gastrostomy tubes.
These adverse events can be grouped into a larger class of errors labeled as, “Wrong tube, wrong hole, wrong connector.” Ideally, inflation and infusion ports should be incompatible, and interconnectivity should be impossible through product redesign. Until this occurs, consider the following error-reduction strategies.
Identify error potential through failure mode and effects analysis on existing medical tubing and when introducing new tubes, catheters, and connectors into a healthcare system. If possible, include an assessment of near novices doing multiple connections on manikins or in other simulated environments to promote the identification of, and focus on, high-hazard conditions. When possible, do not purchase tubing and connectors for non-intravenous functions that are compatible with IV tubing connectors.
Provide training to nurses, pharmacists, physicians, and respiratory therapists before using new tubes,catheters, and connectors. Include discussion about possible sources of errors identified during failure modeand effects analysis and steps to avoid these errors. When possible, include tubing misconnections in simulation training during orientation and annual safety competencies.
Affix labels on lines near insertion sites if the patient has more than one potential connection to a port of entry into the body (e.g., IV, arterial, umbilical, enteral, bladder, tracheostomy, drainage tubes).
Promote a consistent process for tracing all lines from the source (and infusion pump if used) to the connection port to verify attachments before connecting or reconnecting tubing, and/or administering drugs, solutions, or other products. Remind staff that, for patients with multiple tubes, situational awareness of each tube’s location and insertion site can be lost, especially if tubing is obscured by bedclothes and sheets.
Staff who are allowed to connect, disconnect, or reconnect medical tubing should be limited to those with professional healthcare training who are more likely to know and follow safety measures (such as tracing the line from the source to the point of entry), and are knowledgeable about the serious ramifications of misconnections. During orientation, include prohibitions to connecting/disconnecting medical tubing so those who should not be involved in these activities are aware of the mandate and reasons that require it.
Improve the environmental conditions under which medications are administered. Not surprisingly, we have often received reports of errors that have taken place at the bedside under poor lighting, particularly at night. Practitioners should adjust lights as needed for critical tasks (using flashlights, if necessary).
Monitor patients appropriately (e.g., vital signs, frequent observation, pulse oximetry, capnography cardiac monitoring) to detect an error quickly, and minimize the consequences of an error.
NHS takes action on tubing misconnection issue
From the May 3, 2007 issue
The National Patient Safety Agency (NPSA) in the United Kingdom has done something that should serve as an example for regulatory authorities and manufacturers in the US: they have set deadlines for National Health Service (NHS) entities in England and Wales to adopt enteral feeding catheters that do not contain ports with female Luer connectors. Ports with female Luer connectors require the use of parenteral syringes to administer oral liquid medications.
Many enteral catheters manufactured in the US feature ports that accept parenteral syringes. Even if a liquid medication is prepared in an oral syringe, the medication must be transferred to a parenteral syringe for administration via an enteral catheter, risking the accidental administration of the drug via a parenteral line. ISMP has published reports of several such events.
The NPSA directive also requires enteral catheters to have end connectors that are unable to connect to intravenous or other parenteral lines. Ports and connectors on the enteral feeding catheter must feature a male Luer, catheter tip, or other non-female Luer design. For example, administration sets will have a female Luer connector to fit a male Luer on the nasogastric tube(reverse Luer connection).
The NPSA has held meetings with medical device industry representatives to inform them of their recommendations, and it is anticipated that the design recommendations will be incorporated into catheters an administration sets within 12 months. The design is similar to that mentioned in our June 15, 2006, newsletter issue in which we described an error where an infant accidentally received breast milk intravenously (IV) instead of through a nasogastric tube (www.ismp.org/Newsletters/acutecare/articles/20060615.asp). The March 28, 2007, NPSA Patient Safety Alert also announced that, when a spoon or graduated measure cannot be used, measurement and administration of oral liquid drug doses must utilize labeled oral syringes that cannot be connected to IV catheters or ports. The measures being undertaken also prohibit three-way taps or syringe tip adapters, as these connecting devices bypass safeguards. Oral syringes must be in use in clinical areas by September 30, 2007, with remaining recommendations in place by March 31, 2008. Special training of healthcare practitioners must occur and policies must be in place by the 2008 deadline.
The NPSA is a special health authority created within NHS to coordinate the safety efforts of all those involved in healthcare, and more importantly, to learn from adverse events occurring in the NHS. There is no equivalent organization in the US although The Joint Commission (TJC) has served a similar role as an accreditation organization. However, a potential TJC 2008 National Patient Safety Goal to prevent tubing misconnections was not approved after field review earlier this year. Representatives from the FDA, American Hospital Association, ISMP, and several other organizations have formed an ad hoc committee that is expected to publish a white paper in 2007 on the issue of catheter misconnections.
Preventing accidental IV infusion of breast milk in neonates
From the June 15, 2006 issue
This week we heard from the mother of a hospitalized infant named Zoey who accidentally received breast milk intravenously (IV) instead of through a nasogastric (NG) tube. The baby was born with duodenal atresia—complete absence of the duodenal lumen—so surgery was necessary at birth. The procedure was successful, after which a NG tube was inserted in order to provide nutrition with regular feedings of 30 mL of fortified breast milk administered over 2 hours. At the time of the event, an IV syringe pump for medications was located on the left side of the baby's incubator and an identical pump used to deliver breast milk via the NG tube was on the right. The pumps used identical IV administration tubing. Although it's not clear how the tubing used for breast milk was connected to the NG tube, a nurse mistakenly connected a syringe containing breast milk to the wrong line. About 10 mL of milk was infused IV before the problem was recognized. The baby developed respiratory distress and also had seizures. She was treated supportively and, fortunately, she recovered and does not seem to have any lasting adverse effects. However, infusion of non-sterile, particulate fluid such as enteral feedings or breast milk can be fatal, as it carries the risk of sepsis, diffuse intravascular coagulation (DIC), or emboli to major organs, which can lead to organ damage and pulmonary embolism.
Review of the literature reveals cases of inadvertent IV administration of breast milk reported as early as 1972.1 As this case demonstrates, inadvertent IV administration of breast milk is still happening today despite recognition of the problem more than 3 decades ago. Ryan et al2 recently reported a similar case and noted that neonatal health professionals communicated eight previously unknown events to the authors after they posted a question about accidental milk infusion to an online, e-mail discussion group.
All hospital staff—particularly in neonatal units—need to take the risk of misconnectionsseriously and proactively eliminate all chances of IV infusion or direct injection of non-sterile, particulate fluids meant for enteral administration. An April 2006 Joint Commission Sentinel Event Alert on tubing misconnections (www.jointcommission.org/SentinelEvents/SentinelEventAlert/sea_36. htm) provided a number of excellent recommendations for preventing such tragedies, including: 1) tracing the tubing to the point of origin before making any connections or reconnections, 2) rechecking connections and tracing all patient tubes and catheters to their sources upon transfer to a new setting, and 3) labeling
tubes and administration sets. However, as emphasized in many of our prior newsletters as well as in the April Sentinel Event Alert, the best chance of eliminating the risk of injecting enteral solutions intravenously is to use an oral syringe which is incompatible with IV tubing.
Enteral pumps for adults cannot deliver feedings in the small amounts necessary for premature infants.3 Thus, as in the aforementioned event, staff in neonatal and pediatric intensive care units sometimes employ off-label use of parenteral syringe pumps to administer breast milk enterally, although infusion rates may vary slightly from the programmed rate.3,4 (Some pump manufacturers are in the process of adjusting IV syringe pumps to accommodate oral syringes.) In these cases, syringes with standard Luer connections should be avoided and NG tubes should connect only to oral syringes via syringe extension sets. Viasys Healthcare (www.viasyshealthcare.com) and Neo Devices (www.neodevices.com) offer such systems, with non-Luer feeding tubes and extension sets. For example, the CORFLO Enteral Feeding System from Viasys has an administration set, for use with a syringe pump, that can only be connected to a syringe tip on an oral syringe (see photo). Unlike typical IV sets with a male Luer at the distal end, the connector at the distal end of the set is female and will only connect to a proprietary male connector on the system's feeding tube, thus preventing possible connection with an IV line. There are no IV ports on the NG tube or the administration set. The feeding tube has a side port that only allows connection of an oral syringe. Both the administration set and feeding tube also have an orange stripe along their lengths to distinguish them from IV lines. Unless makeshift fittings are created, feedings and oral medications cannot be administered via an IV line when using this tubing and an oral syringe. Children's Medical Ventures (http://enteralextensionsets.respironics.com/) also offers non-Luer extension sets for enteral feedings, but the company does not provide feeding tubes.
Although IV administration of breast milk may not happen often, the risk of patient harm is high when it occurs. Its remedy is within reach of all providers: use an anti-IV NG tube and administration set, and an oral syringe. We also recommend labeling the pumps as "Medication" or "Breast Milk" as well as labeling the breast milk syringes. If your organization hasn't addressed this issue, put it on your safety agenda now! Zoey's mom wanted us to advocate for immediate action before another child is injured from this potentially fatal but preventable error.
References: 1) Wallace JR, Payne RW, Mack AJ. Inadvertent IV infusion of milk. Lancet 1972:1264-6. 2) Ryan CA, Mohammed I, Murphy B. Normal neurological and developmental outcome after an accidental IV infusion of expressed breast milk in a neonate. Pediatrics 2006;117(1):236-8. 3) Page L. Diligence, technology prevent IV and feeding tube mix-ups: finding the wrong fit. Materials Management in Health Care April 2006:24-28. 4) Copeland D, Appel J. Implementation of an enteral nutrition and medication administration system utilizing oral syringes in the NICU. Neonatal Network January/February 2006;25(1):21-24.
IV potassium given epidurally: getting to the “route” of the problem
From the April 6, 2006 issue
Problem: A nurse accidentally infused intravenous potassium chloride injection by the epidural route to a post-operative hypokalemic patient. She intended to connect the IV tubing from the potassium chloride minibag to the maintenance IV infusion line via a Y-site port. Instead she connected the potassium infusion to a Y-site located on the patient’s epidural line through which fentanyl and bupivacaine were infusing. The patient received a total of 20 mEq of potassium in 50 mL of solution over 2 hours, after which the nurse disconnected the minibag. A short while later, an anesthesia staff member arrived and discontinued the epidural line. After-wards, the nurse went into the patient’s room to hang another dose of potassium and realized that she must have connected the prior infusion to a Y-site on the epidural tubing. The anesthesiologist and admitting surgeon were immediately notified. Fortunately, the patient developed no symptoms during or after the potassium infusions, thus interventions other than frequent monitoring were deemed unnecessary.
At first glance, the underlying cause of this error seems clear–the improper use of IV tubing with access ports for an epidural infusion. However, examining why this error occurred in a facility that typically used special epidural tubing without access ports led to the discovery of other important causal factors.
Inadvertent continuation of a standing order. For patients with epidural infusions in place for analgesia, post-anesthesia care unit (PACU) staff typically attached special tubing without a port during the immediate post-operative period. In this case, the surgery occurred during the evening on a weekend, and recovery took place in ICU, not PACU. While the patient was there, an anesthesiologist noticed that the epidural catheter was leaking and he capped it, telling the nurse that he wasn’t sure if he’d be using it later or not. The patient was soon transferred to a medical-surgical floor. Standing orders for epidural analgesia, which had been placed on the patient’s chart in anticipation of use during the post-op period, remained in effect.
Faulty procedure and absence of proper tubing. Usually, when the floor nurses received a patient with an epidural infusion, the proper tubing was already attached. In fact, a patient had never come to the floor previously with an epidural access site capped. When the patient complained of pain, the nurses decided to start the epidural analgesic per the standing orders. In preparation, they read a recently written policy and procedure for epidural analgesia which, unfortunately, did not mention the need for special tubing without access ports. In fact, epidural tubing was not available in the unit’s supplies. Thus, regular IV tubing was used to connect the epidural analgesia. Later, the potassium infusion was accidentally piggybacked into the epidural infusion line.
New checking policy not known. The hospital had a rigorous policy regarding independent double checks on high-alert medications that included IV potassium infusions in concentrations greater than 60 mEq/L. If the policy had been followed as written, one of the checks would have included having the nurse show a colleague at the bedside exactly where she was going to attach the IV potassium infusion. However, the double-check policy had just been implemented a few weeks prior to the error and, at that time, most nurses were unfamiliar with its scope; most thought the policy required two nurses to double check the medication label and dose against a patient’s medication administration record, not to also go into the patient's room to track the tubing to the site of injection, among other bedside checks.
Safe Practice Recommendations: The use of yellow-lined tubing without injection ports for epidural infusions has become a standard in most hospitals to set its appearance apart from typical IV tubing and to prevent inadvertent administration of drugs intended for the IV route. This tubing with restrictive access is a key error-prevention strategy that should be clearly described in all policies, procedures, and standardized order sets related to epidural infusions. Some hospitals, like this one, also have a policy that high-alert drugs like potassium chloride, greater than 60 mEq/L (i.e., highly concentrated minibags), require an independent double check, particularly at the bedside so that verification of the patient, pump settings, and line attachment can be included along with verification of the drug and dose.
Unfortunately, you can readily see in this case how several deeper system failures served to thwart both of these safety practices, which were not employed fully in this hospital. For example, in addition to shoring up the above more obvious error-reduction strategies, the hospital needs to improve its processes for handoffs between caregivers and reconciliation of medication orders upon transfer, including by those who recover post-operative patients on weekends and evenings. Had the transferring nurse reviewed the patient’s orders, including prescribed medications, with the receiving nurse, the inadvertent continuation of the epidural analgesia order might have been detected. Furthermore, changes are necessary in how new policies and safety practices related to high-alert drugs are communicated to all staff, including evening and night staff. When changing an established procedure, it is advisable to document training by having each trainee demonstrate the concept being taught. Additionally, epidural infusions should only be started by practitioners with demonstrated competency, typically anesthesia and PACU staff.
There are several other recommendations to help prevent IV-epidural line mix-ups:
Place IV pumps and epidural pumps on opposite sides of the patient’s bed to better separate the two infusion systems
- Use a different make or model of pump for epidural infusions to differentiate it from pumps used for IV infusions
- Clearly label the pump as “Epidural Only”
- Place a neon sticker on the tubing stating “Epidural” (which is often included with the special epidural tubing)
- Replace a peripheral IV infusion being used only to keep a vein open with a saline lock to maintain IV access
- Avoid the use of dual channel pumps for simultaneous administration of IV and epidural infusions
- Heighten awareness of the risk for mix-ups between epidural and IV infusions among clinical staff.
This week, the Joint Commission published an excellent review article about tubing misconnections that provides various scenarios and recommendations for prevention. Also, in Paris this month, the International Standards Organization (ISO) and the Association for Advancement of Medical Instrumentation (AAMI) will begin efforts to establish design standards that address the all-too-frequent problem of misconnections between various catheters and tubings used in healthcare. We’ll keep you posted about any new developments.
Another risk of IV misconnections
From the October 6, 2005 issue
The V.A.C. INSTILL SYSTEMby KCI is a wound healing system that facilitates automated delivery of topical solutions to wound sites. While very different from an infusion pump, the device is programmed to deliver a desired rate of infusion; but it lacks protection from gravity free-flow if the tubing is removed. Unfortunately, the device is designed to accommodate IV tubing to deliver the topical solution (e.g., Dakin’s solution, silver nitrate solution, SULFAMYLON [mafenide acetate]). Thus, one could inadvertently attach the “IV” tubing intended for the V.A.C. to an IV port.
Previously reported misconnections of seemingly dissimilar tubing (e.g., blood pressure, oxygen, and air supply tubing) to IV ports should serve as a reminder that such an event is possible. In fact, there are a few additional conditions that heighten the risk of connecting the V.A.C. tubing to an IV line. First, most V.A.C. topical solutions are prepared in “IV” bags, so bags and tubing look similar. Next, the device provides a detachable “IV” pole (as described in product literature) on which to hang the “IV” bag of solution. Finally, the device is indicated for recalcitrant wounds, so it’s used infrequently. Thus, staff may be unfamiliar with the device and unaware of the risk of misconnections. Furthermore, there will be no protection from gravity free-flow if the tubing is erroneously attached to an IV site and is removed from the device.
The reporting nurse has asked KCI to develop specific tubing that connects only to their equipment. Meanwhile, prepare topical solutions in a container dissimilar to typical IV solutions, such as a 500 mL bottle. Special adapter caps for the irrigation solutions will accommodate IV tubing for use with the V.A.C. Instill System. Misconnections are also less likely if you label all lines; affix bold cautionary labels to topical solutions; physically trace all lines from the source solution to the port of insertion; and require an independent double check before initial use of the device, or when replacing bags of solutions.
Problems persist with life-threatening tubing misconnections
From the June 17, 2004 issue
In our June 12, 2003 newsletter, we reported several cases in which the tubing from a portable blood pressure monitoring device was inadvertently connected to the patient’s IV line – in one case, leading to a fatal air embolism. In our September 4, 2003 issue, we wrote about an inadvertent connection of an air supply hose from a sequential compression device (SCD), also referred to as intermittent pneumatic compression, to a needleless IV tubing port. In that case, the SCD was turned off and the misconnection was found before any harm occurred.
Figure 1. Tube delivering oxygen fell off nebulizer.
Figure 2. The oxygen tubing was connected to a Baxter Clearlink needleless port.
Sadly, we recently learned about another tragic error involving the connection of oxygen tubing to a pediatric patient’s IV line. The child had been receiving medication via a nebulizer to treat asthma. While still attached to a wall outlet, the oxygen tubing (AIRLIFE, from Allegiance Healthcare Corporation) became disconnected from the nebulizer fluid chamber (see Figure 1 in the PDF version of the newsletter). Later, a staff member accidentally reconnected the oxygen tubing to the injection port on a Baxter CLEARLINK Needleless Access System IV tubing Y-site.
While oxygen tubing does not have a Luer connector, Baxter’s investigation into this event confirmed that such a connection is possible with their Clearlink valve (with oxygen tubing encompassing the Y site), albeit with excessive force (see Figure 2 in the PDF version of the newsletter). According to the FDA report on this event, the oxygen tubing disconnected from the IV tubing in seconds, but not before the pressure of the compressed oxygen supply forced the needleless valve open and allowed air into the tubing. The child died instantly. Baxter has since issued a safety alert advising of this hazard. However, the May 13, 2004, alert was sent only to directors of nursing, so others who need to know, including respiratory therapists, may have missed it.
Although these types of misconnections might be uncommon, the risk of a similar tragic event persists and is likely to be present in most hospitals – including yours! Baxter is exploring modifications to the Clearlink valve to prevent the potential for a recurrence elsewhere. Allegiance has also redesigned the specific oxygen tube to nebulizer connection with hard plastic to reduce separation. Still, we tested a variety of additional needleless devices that also accommodated oxygen tube connectors. Right now, oxygen tubing could very well connect to your needleless or standard IV tubing, including systems from Alaris, Baxter, B. Braun, BMP, Inc., Catheter Innovations, Clave, and Hospira (formerly Abbott). Clearlink is a clear needleless system, similar to some oxygen tubing, but the color on other needleless systems may not be a deterrent. Additionally, all medical gases and most fluids are clear, making it virtually impossible to distinguish between them when observed through transparent medical tubing. As in this case, many medical tubing “attachment” sites look quite similar, further increasing the risk of an error. With typical oxygen flow rates in liters per minute, a fatal pneumatic misconnection would require only seconds.
The best solution lies with eliminating interconnectivity between various medical tubings. The Association for the Advancement of Medical Instrumentation (AAMI) offers a publication (HE-74, Human factors design process for medical devices) that provides human factors engineering guidance to aid in the proactive testing of tubing interconnectivity in realistic settings. However, we are unaware of any specific domestic or international standards organization looking at mandatory standards for connector designs.
Take the time to review the medical equipment used in your facility to identify the potential for misconnections to IV tubing. Educate all staff, including non-clinical employees who work in patient care units, about this hazard, perhaps using storyboards to communicate the point effectively. Before tubing is connected or reconnected to a patient, require staff to completely trace it from the patient to the point of origin for verification. Appropriately labeled IV lines could help alert staff if they are about to access that line accidentally.
Also identify all types of staff in your organization who might connect, disconnect, or reconnect various forms of tubing (including IV tubing) attached to patients (e.g., technicians who work in diagnostic units, transport staff, nursing assistants, students, physicians), and consider whether this task falls within a safe and acceptable scope of practice. Certainly, trained licensed practitioners can inadvertently connect the wrong tubing to an IV line. However, untrained staff (e.g., ancillary personnel, medical and nursing students) are less likely to know and follow safety measures (such as tracing the line from the patient to the point of origin), or to be knowledgeable about the serious ramifications of misconnections. In fact, many unlicensed, untrained staff may disconnect or reconnect various tubing, or be asked inappropriately to perform specific tasks such as turning off pumps or pressing the PCA button before patient transport. Include prohibitions for these tasks during orientation, and when possible, offer new ancillary staff practice in turning down requests to connect or disconnect medical tubing.
Improvised drug delivery: A cause for concern
From the April 22, 2004 issue
Problem: Have you ever used IV tubing and/or an IV pump to administer an oral solution or liquid nutrition to patients via a gastric or nasogastric tube? Before you say "no," don't overlook the potential for purposefully using this method of delivering enteral solutions. For example, GoLYTELY bowel prep has been administered via nasogastric tube to older children and adults due to vomiting or intolerance to the large volume necessary for effectiveness. For some patients, an enteral infusion pump is not capable of delivering the solution at the desired infusion rate (e.g., 600-1,000 mL over an hour). Thus, we have heard about many instances in which an IV pump has been used to administer GoLYTELY.
This is often accomplished by placing GoLYTELY in a plastic enteral container and jury-rigging the solution administration sets. For example, an IV pump set is cut just below its drip chamber, and the end of the enteral solution apparatus (with attached feeding tube connector) is then jammed into the cut IV tubing and secured with tape to prevent leakage. The solution is then administered via a nasogastric tube using an IV pump. Of course, this form of improvised drug delivery could result in accidentally connecting the IV tubing to an IV access site.
In fact, cases of accidental IV administration of GoLYTELY, or a similar high molecular weight polyethylene glycol solution, have been reported in the literature.(1, 2) Many years ago, nine patients received polyethylene glycol 300 intravenously; seven developed renal tubular necrosis but recovered, and two patients died as a result of polyethylene glycol toxicity. More recently, a 4-year-old child received GoLYTELY intravenously. 1 The child had presented to the ED after ingesting a large number of 6-mercaptopurine tablets. After treatment with activated charcoal, the child was started on GoLYTELY, which was to be administered using IV tubing attached to a nasogastric tube. After 1 hour, a nurse discovered that the solution was actually being administered through an IV access line; 391 mL had already infused. Luckily, the child showed no evidence of acidosis or renal failure, and glycol levels were undetectable. He was discharged several days later without further complication.
With opaque IV medications in use today, healthcare professionals can no longer rely on visual appearance to determine the suitability of administering solutions IV. Thus, it's not surprising that enteral feedings have also been administered via IV infusion pumps.
Ready-to-hang, closed enteral nutrition containers are easily spiked with an IV infusion set, allowing the formula to flow freely or to be delivered via an IV infusion pump. In our March 20, 2003 newsletter, we reported several instances in which this occurred. In one case, the nurse couldn't find an enteral feeding set, so she improvised and used IV tubing and an IV pump until another nurse recognized that this was an error waiting to happen. In another case, the patient received the enteral feeding IV for 2 hours, but luckily suffered no harm.
Safe Practice Recommendation: Of course, the most obvious solution is to prohibit the use of IV tubing and IV pumps to administer enteral solutions. However, simply having such a policy in place is not sufficient; nor can violation of such a policy be considered the root cause if an error occurs. As implied above, healthcare providers use these devices to overcome obstacles to enteral administration. Thus, it might be helpful to hold focus groups with nurses to dig deeper into both the obvious and subtle incentives for using IV tubing and IV pumps for enteral administration. The reasons are often rooted in system-based problems for which safer solutions can be found.
For example, if enteral solutions like GoLYTELY must be administered quickly in large volumes, you might be able to use an adapter to connect two enteral feeding pumps, each delivering half the desired volume simultaneously. Also, some enteral pumps are capable of delivering higher volumes per hour (e.g., 500 mL per hour with the Ross Embrace pump), and some nasogastric tubes have a dual port to facilitate connection to two enteral pumps simultaneously. Bold labels that state "WARNING! For enteral use only" should also appear on the containers of all enteral products that could possibly be connected to IV tubing. It's equally important to investigate other potential uses of IV tubing and IV infusion pumps for enteral administration. For example, neonates have sometimes been fed breast milk or formula via a nasogastric tube using an IV pump to permit very slow delivery.
(1) Guzman DD, Teoh D, & Velez LI: Accidental intravenous infusion of Golytely(R) in a 4-year-old female (abstract). J Toxicol Clin Toxicol 2002; 40(2):361-362.
(2) Tuckler V, Cramm K, & Martinez J: Accidental large intravenous infusion of Golytely (abstract). J Toxicol Clin Toxicol 2002; 40(5):687.
Misconnection of IV tubing.
From the September 4, 2003 issue
We recently learned about an inadvertent connection of an air supply hose from an Albahealth sequential compression device(SCD, also referred to as intermittent pneumatic compression) to B. Braun’s needleless IV tubing. In our June 12, 2003 issue, we reported a similar situation in which blood pressure monitor tubing was connected to the IV port of needleless systems. The SCD is used to prevent deep vein thrombosis and resulting pulmonary embolism. A hospitalized patient was easily able to connect the air supply tubing from the SCD (see http://albahealth.alba1.com/pulstar.htm for photo) to his needleless IV tubing, risking an air embolism. Fortunately, the device was turned off at the time and the connection was found before any harm occurred. Please investigate whether a similar potential problem exists with SCDs at your hospital. The manufacturer has apparently used this connector for many years, so other Albahealth pumps (as well as other brands of pumps) could also be affected. If you have a compatible connection, keep SCD tubing away from needleless IV tubing, and label the tubing to identify that it supplies air. Albahealth is sending labels to their identified customers, and is working on a new connector design, but the process may take several months.
From the March 20, 2003 issue
Recently, when a nurse couldn’t find an enteral feeding set, she improvised and spiked the bottle’s cap with IV tubing. Since the enteral pump would not accept the IV tubing, she used an IV pump to deliver the feeding. The patient was being weaned off a three-in-one total parenteral nutrition (TPN) solution and both pumps were on the same pole. The solutions were similar in appearance (white with a yellow tint for TPN, tan for enteral product), and the tubings hard to distinguish because they wound around one another. The enteral feeding continued infusing in this manner until the next morning when a nurse noticed that this was an accident waiting to happen. Fortunately, an error did not occur, but just yesterday we received another report where similar conditions led to an actual error.
An enteral feeding bag was inadvertently spiked with a standard IV set and infused through an IV line (50 mL/hr) using an IV pump. Luckily, another nurse noticed the error within two hours and the patient was not harmed. In 2001, we published another error that resulted in a patient’s death (ISMP Medication Safety Alert! IV administration set spiked into enteral nutrition container. March 21, 2001). In that case, a nurse administered 200 mL of enteral feeding IV to a patient who was supposed to receive three-in-one TPN solution.The patient expired four days later.
This potential hazard and related risk reduction efforts are WorthRepeating. With the ever-increasing use of opaque IV fat emulsions, three-in-one TPN, and lipid-based drug products, healthcare professionals no longer can rely on visual appearance to determine the suitability of administering solutions intravenously. A warning appears on the label of enteral containers, but it
is not prominent enough to eliminate the risk of IV administration. Use large, bold auxiliary labels that state “WARNING! For enteral use only - NOT for IV use.” It’s been suggested that FDA and standards setting organizations take action to prevent an IV set from attaching to an enteral feeding container. In the meantime, use a rubber band to attach an appropriate enteral administration set to all enteral feedings before distribution to (or storage in) patient care units.
On a broad level, hold focus groups with nurses to discuss the reasons that underlie the need to “improvise” during drug administration (e.g., lack of ready access to enteral administration sets) and other “at risk” behaviors that are often driven by common system problems. Incidentally, some providers may consider this to be a rare, isolated incident that does not require immediate action. We disagree. Although this type of an error may not happen often, the risk of patient harm is high and its remedy is clearly within the reach of all providers.
IV administration set spiked into enteral nutrition container
From the March 21, 2001
PROBLEM: If you use oral syringes that cannot be connected to IV tubing, and enteral feeding tubes with connectors that accept oral syringes, you can significantly lessen the likelihood of inadvertently administering an enteral product by the IV route. But there’s still another trap that could lead to such a disastrous event. Current efforts to avoid harmful connections between enteral and parenteral drug delivery devices have focused on the distal ends of administration sets. What if an enteral feeding container could be spiked at its proximal end with an IV administration set and the container looked like a 3-in-one parenteral nutrition container? Several dietitians, nurses, pharmacists, and physicians have alerted us to just that.
Many ready-to-hang, closed enteral nutrition containers (e.g., Mead Johnson Kangaroo, Nestle UltraPak, Ross Ready-to-Hang, etc.) can be spiked with a standard IV infusion set and will allow the formula to flow freely. Even if a filter is in-line, it may not prevent administration IV. If administration sets with a 0.22-micron filter are used, it will likely occlude right away. But with the larger filters (1-5 micron), it may be possible for some volume of the enteral feeding suspension to enter the vascular system before an infusion pump alarms. Although the enteral containers have labeling to warn against IV administration, this is not sufficient to eliminate that risk. With the ever-increasing use of opaque IV fat emulsions, 3-in-1 parenteral nutrition formulas, and lipid-based drug products, healthcare professionals can no longer rely on visual appearance to determine the suitability of administering solutions by the IV route. Recent introduction of the modified Nestle UltraPak Enteral Closed System bag, which is filled with creamy white enteral feeding (Nutren 1.0), heightens the risk of confusion because it closely resembles a 3-in-1 parenteral nutrition bag. We recently heard about a fatality with this product. An agency nurse administered 200 mL of enteral feeding via an IV infusion over 1-2 hours in a patient who was to receive TPN (total parenteral nutrition). The enteral product had been discontinued one week prior but not removed from the nursing unit refrigerator. The nurse had never seen enteral products dispensed in anything but a can and mistakenly thought the enteral bag was TPN. The patient expired 4-5 days after the event.
SAFE PRACTICE RECOMMENDATIONS: If you are using a closed enteral nutrition system, alert practitioners to this potential for error. Affix large, bold auxiliary warning labels in prominent places on the container stating, “WARNING! For enteral use only - not for IV use,” and attach the appropriate enteral administration set with a rubber band before the product is stocked in or dispensed to patient care units. As with any medications dispensed for specific patients, discontinued enteral products should be removed from patient care areas immediately and returned to their original source. An independent double-check system for TPN, which compares the original order with the container’s contents before administration, may have prevented the fatality. ISMP and several individual practitioners have alerted the FDA and standards-setting organizations and suggested the development of standards for IV and enteral administration sets that would prevent connections at their proximal ends with containers intended for the opposite route of administration.