The Case

A 65-year-old man presented to a Level III trauma center with bilateral lower extremity paralysis following a ground level fall. His past medical history was significant for type 2 diabetes mellitus, hypertension, and coronary artery disease status post coronary artery bypass grafting (CABG). The patient was then transferred to a Level I trauma center that could provide a higher level of care following an incidental finding of a 9-cm abdominal aortic aneurysm and cervical spinal cord injury. Post transfer, the patient was noted to have rapidly progressive ascending paralysis. Magnetic resonance imaging (MRI) revealed severe spinal stenosis involving C3-4 and post-traumatic cord edema/contusion involving C6-7. A continuous intravenous (IV) infusion of norepinephrine was initiated to maintain adequate spinal cord perfusion, with a target mean arterial pressure (MAP) goal of greater than 85 mmHg (also known as “MAP therapy”).

Norepinephrine was incorrectly programmed into the infusion pump for a weight-based dose of 0.5 mcg/kg/min rather than the ordered dose of 0.5 mcg/min, resulting in a dose that was 70 times greater than intended. The patient became hemodynamically unstable, alternating between hypertensive urgency and hypotension. On four separate occasions, MAP increased by more than 80 mmHg immediately after restarting the infusion. He then experienced sudden cardiac arrest and was emergently taken to the cardiac catheterization laboratory to rule out acute coronary occlusion. Following a percutaneous cardiac procedure, the patient experienced bradycardia and cardiac arrest within 9 hours of admission to the hospital. Cardiopulmonary resuscitation was initiated but failed to achieve return of spontaneous circulation.

The Commentary  

By Jeremiah Duby, PharmD, Kendra Schomer, PharmD, Victoria Oyewole, PharmD, Delia Christian, RN, BSN, CNRN, and Sierra Young, PharmD

Vasopressors are commonly used in the treatment of shock to support blood pressure, cardiac output, and end-organ perfusion.^1-3 These agents are associated with risk of serious adverse effects including ischemia (e.g. stroke, myocardial infarction, soft tissue necrosis), cardiac dysrhythmia, extravasation, acute kidney injury, increased duration of hospital stay, and higher mortality.^4-6 Their relatively short duration of action requires administration via intermittent bolus (primarily in the operating room [OR]) or continuous infusion (primarily in the emergency department [ED] and intensive care units [ICUs]). Conventionally, the OR and pediatric ICUs utilize weight-based dosing, whereas adult ICUs more typically rely upon non-weight-based dosing. These differences create potential discrepancies in transitions of care (i.e., OR to ICU) and inconsistencies in dosing strategies within the same units, patient populations, and institutions.

In this case, norepinephrine was used to achieve higher mean arterial pressure (MAP) goals in an attempt to augment spinal perfusion. The patient’s hemodynamic instability was attributed to sympathetic dysregulation, which is a complication of spinal cord injury that results in wide fluctuation of blood pressure. The combination of a supraphysiologic blood pressure target and the patient’s altered blood pressure regulation made it difficult for bedside clinicians to identify the medication error in real time. The patient’s response to norepinephrine appeared grossly disproportionate to the relatively low doses charted. The consequences of repeatedly receiving toxic doses of norepinephrine likely contributed to the sequelae of cardiac events leading to his ultimate deterioration. Within 9 hours of admission the patient’s hemodynamic lability progressed to cardiac arrest and death.

Approach to Improving Patient Safety

Recognize Cognitive Biases

There were several factors that contributed to this medication error, including the availability of two different dosing strategies (mcg/kg/min and mcg/min) and cognitive biases at multiple points of the medication administration process. The critical chain of events began when the ordering physician selected the non-weight-based dosing strategy (0.5 - 30 mcg/min) in the electronic health record (EHR). The EHR's clinical decision support logic ranked and pre-populated the provider’s order preference list based on recent patterns of use. This common application of artificial intelligence effectively automated a form of availability bias. This cognitive shortcut enables rapid decisions with negligible analysis but at the cost of critical thinking and potential performance errors.

The next pivotal event occurred when the nurse incorrectly selected the weight-based dosing strategy and maximum infusion rate (i.e. 0.5 mcg/kg/min) in the medication pump library. The vast majority of norepinephrine orders in the ICU were for non-weight-based dosing with a starting rate of 0.5 mcg/min. This prevailing practice pattern placed the nurse at risk for medication errors associated with confirmation bias. Familiarity and procedural memory—which function as basic safeguards and facilitate rapid execution of complex tasks—framed preconceived expectations of the starting dose and dosing units. In other words, the nurse then saw what she expected to see, with nominal conscious awareness, throughout the process of manually programming the pump.

The patient immediately began experiencing hypertensive urgency. However, the wide fluctuations in blood pressure observed were consistently attributed to the patient’s spinal cord injury. This was an example of premature closure associated with anchoring bias. The bedside clinicians persisted with their preformed impression (sympathetic dysregulation) based on the initial presentation (spinal cord injury), thereby overlooking other possible diagnoses (medication error).⁷

Eliminate Multiple Dosing Strategies

In reviewing the case, it became clear that maintaining two different dosing strategies for adults within one healthcare system created an unnecessary and untenable risk to patient care. There are persuasive considerations to support the use of either strategy; however, the benefits of non-weight-based dosing may outweigh the risks. In patients experiencing shock, weight is an insignificant factor in dosing norepinephrine, as catecholamine concentrations in this disease state are orders of magnitude above the normal physiologic levels. Furthermore, the pharmacokinetic properties of norepinephrine – relatively short half-life and low volume of distribution (Vd) – allow for rapid titration to the clinical effect, i.e. the target MAP or systolic blood pressure (SBP).^4,5 In practical terms, non-weight-based dosing (mcg/min) offers the operational advantage of not requiring a weight for initiation of treatment, which may expedite care for critically ill patients in the emergency department. Finally, dosing in mcg/min (e.g. 0.5 - 30 mcg/min) mitigates the risk of decimal point errors associated with weight-based dosing (e.g. 0.03 - 0.5 mcg/kg/min). The use of weight-based dosing may be an advantage in patient populations where physiologic doses of vasopressor are used to counteract the effects of anesthesia in the OR and where the volume of distribution is a critical factor as in pediatrics.⁸  The disadvantage of non-weight-based dosing in this population is that omission of the size of the patient, i.e. not accounting for their weight, could lead to under- or overdosing.

Unfortunately, there are no guideline-based recommendations on which dosing strategy (weight-based versus non-weight-based) of norepinephrine is more efficacious or safe. As a result, it is common to see both dosing strategies utilized within a single institution. However, it is imperative that institutions choose one dosing strategy for the adult population and one dosing strategy for the pediatric population to minimize the risk of error.⁹  Education and training for physicians, nurses, and pharmacists must be provided prior to any major practice change to ensure smooth implementation and minimize risk of errors.

Implement Smart Pump Technology

Additionally, infusion pump entry was carried out in this case via manual selection (selecting mcg/kg/min instead of mcg/min) which created a significant opportunity for error. Barcode scanning was used to confirm the correct patient and medication. However, integration between the infusion pump and the EHR was not available at the time of this event. Infusion pump interoperability with the EHR improves pump programing safety by guiding accurate administration and documentation of continuous infusions. Interoperability  translates the ordered medication and dosing strategy into the infusion pump, thereby reducing instances of manual programming. This relatively new technology likely would have prevented the nurse’s manual entry of the incorrect dosing strategy in this case. However, implementing new technologies can result in unforeseen workflow inefficiencies; therefore on-going training efforts are necessary to mitigate potentially unsafe adaptive practices or workarounds. Training can include simulations and education related to the availability of onsite experts. 

Review and Update Infusion Pump Safety Guardrails

The numeric value of the intended starting rate of 0.5 (mcg/min) was also coincidentally the same as the suggested maximum for a weight-based dose, i.e. 0.5 (mcg/kg/min). The nurse incorrectly programmed the infusion as 0.5 mcg/kg/min instead of 0.5 mcg/min, and the program logic of the existing pump did not trigger a high-dose guardrail alert at the time of entry. As a result, norepinephrine guardrail soft limits were reduced to 0.49 mcg/kg/min to avoid accidentally starting at a maximum rate of 0.5 mcg/kg/min when an initial rate of 0.05 mcg/kg/min is intended. Furthermore, requiring nurses to perform independent verification with another nurse prior to exceeding any guardrail limit when starting a critical medication is another potential strategy for minimizing the possibility of error in a high-intensity environment.

Reduce Distractions and Utilize Comprehensive Closed-loop Communication

Finally, there was a breakdown of bedside communication in this case caused by the chaos of managing a critically ill patient. This high-intensity situation can result in the separation of tasks among multiple clinicians, disrupting the process of programming, administering, and charting that would normally have been completed by a single nurse during routine non-urgent ICU care. Closed-loop communication, in which the person speaking addresses a person by name and the receiving person repeats back the message to the sender, should always be utilized when there is a separation of tasks among multiple clinicians caring for one patient. In all communications among healthcare providers, complete dosing units should always be stated.

Take-Home Points

• Dosing strategies for norepinephrine can be weight-based (mcg/kg/min) or non-weight-based (mcg/min) – institutions should choose one dosing strategy for the adult population and one dosing strategy for the pediatric population to minimize risk for error.
• Bedside clinicians commonly rely on previous experience to facilitate quick decision-making and efficiency. However, cognitive biases associated with this practice can contribute to increased risk for error, especially in critically ill, complex patients.
• Infusion pump interoperability with EHRs is a relatively new technology that can be employed to mitigate manual programming errors and improve documentation.
• Active surveillance of medication pump entry data and guardrail overrides can assist in identifying new errors that may emerge with system changes. 
• Reduce distractions and utilize closed-loop communications to reduce risk of errors.
• Once mitigation strategies are deployed, education and training efforts with nurses, physicians, and pharmacists should be performed.

Jeremiah Duby, PharmD, BCCCP, BCPS FCCM
Clinical Pharmacist, Adult Critical Care
UC Davis Health
jjduby@ucdavis.edu

Kendra Schomer, PharmD, BCCCP
Clinical Pharmacist, Adult Critical Care
UC Davis Health
kjschomer@ucdavis.edu

Victoria Oyewole, PharmD
Clinical Pharmacist, Adult Critical Care
UC Davis Health
vaoyewole@ucdavis.edu

Delia Christian, RN, BSN, CNRN
Clinical Nurse III, NSICU
UC Davis Health
dchristian@ucdavis.edu

Sierra Young, PharmD, BCCCP
Clinical Pharmacist, Adult Critical Care
UC Davis Health
sryoung@ucdavis.edu