Professional Documents
Culture Documents
Inadvertent Overinfusion of Norepinephrine Using Infusion Pump Loading Dose 2015 Intensive and Critical Care Nursing
Inadvertent Overinfusion of Norepinephrine Using Infusion Pump Loading Dose 2015 Intensive and Critical Care Nursing
ScienceDirect
journal homepage: www.elsevier.com/iccn
CASE STUDY
KEYWORDS
Cardiac arrest;
Event log;
Infusion pump;
Loading dose;
NORepinephrine;
Over infusion;
Smart pump;
Volume to be infused
(VTBI)
Summary Programming infusion pumps has been recognised as a high-risk step and a source
of adverse events (Nuckols et al., 2008; Hyman, 2010). Literature describing infusion pump
loading dose errors and NORepinephrine complications is scarce (Girard et al., 2010). This case
study presents the rst ever report of an inadvertent overinfusion of NORepinephrine due to
the loading dose option on the infusion pump, and resulting cardiac arrest of the patient.
A patient was admitted to the emergency room and started on a NORepinephrine infusion
inadvertently as a loading dose rather than a primary infusion. Historical values for the loading
dose volume to be infused (VTBI) and primary rate were not adjusted during the setup. Eight
hours and 58 minutes later, the loading dose VTBI reached 0 mL and the pump reverted to the
historical primary rate of 999 mL/hour.
The event log showed that 37.1 mL of NORepinephrine was infused resulting in an equivalent
calculated bolus dose of 1.8 mg administered in two minutes. The patient suffered a cardiac
arrest and the infusion was stopped. No faults were found with the pump. Herein, we discuss
our analysis of the pump event logs and propose further safety strategies and interventions.
2015 Elsevier Ltd. All rights reserved.
Corresponding author at: Department of Biomedical Engineering, St. Pauls Hospital, 1081 Burrard Street, Vancouver, BC V6Z 1Y6, Canada.
Tel.: +1 604 682 2344; fax: +1 604 806 8172.
E-mail address: aibey@providencehealth.bc.ca (A.A.M. Ibey).
http://dx.doi.org/10.1016/j.iccn.2014.12.001
0964-3397/ 2015 Elsevier Ltd. All rights reserved.
376
Introduction
A previously healthy 40 year old male presented at
a quaternary care academic teaching hospital emergency department in December 2011 with a diagnosis of
community-acquired pneumonia and respiratory failure. The
patient was not taking any medications prior to admission. Systolic blood pressure was 80 mmHg on admission and
showed transient improvement with repeated uid boluses.
However, oxygen saturation could not be maintained on 100%
high ow oxygen via facemask and within four hours of
admission the patient required intubation and ventilation.
Following intubation, the blood pressure dropped once again
and a standard NORepinephrine infusion of 12 mg in 250 mL
(48 mcg/mL) was started at 10 mcg/minute. The patient was
transferred to the intensive care unit (ICU) for further management.
The ICU nurse took over the patients care at 16.15 hours
and continued to titrate the NORepinephrine between 3 and
10 mcg/minute to maintain a mean arterial pressure (MAP)
of 65 mmHg or above. At 17.39 hours the physiologic monitor alarms sounded. Upon entering the room, the patients
blood pressure was noted as 325/190 mmHg by arterial line
and the heart rate was 90 beats per minute with a wide complex rhythm. The NORepinephrine infusion pump was found
to be running at a continuous rate of 999 mL/hour and it
was discontinued immediately. The nursing notes read, the
Alaris infusion pump had malfunctioned and the Levophed
was running at 999 cc/hour.
At 17.46 hours, the patient was noted to have ST depression on the monitor strip and a 12-lead electrocardiogram
(ECG) was carried out. By 17.54 hours, the patients MAP had
once again fallen to 32 mmHg and the NORepinephrine was
restarted at 10 mcg/minute via a different infusion pump.
At 17.56 hours, the patient went into pulseless electrical
activity (PEA) and a Code Blue was called. The patient
was successfully resuscitated and transferred to the Cardiac Catheterisation Laboratory for coronary angiography,
which showed normal coronary arteries. A diagnosis of ST
segment elevation myocardial infarction (STEMI) secondary
to vasospasm was made.
Following the event, the pump was quarantined and sent
to Biomedical Engineering (BME) for analysis.
Several events describing inadvertent high dose administration of epinephrine intravenously have been documented
in the medical literature (Fyfe et al., 1991; Hall et al.,
1987; Kanwar et al., 2010; Karch, 1989; Kurachek and
Rockoff, 1985; Novey and Meleyco, 1969; Taneli Vyrynen
et al., 2005). Generally, these cases describe an overdose due to an inadvertent intravenous bolus injection of
epinephrine mistaken for another drug, or, the correct drug
Data extraction
The pump used to infuse the NORepinephrine is a 7200
Alaris SE dual channel large volume pump software version 2.79 (Carefusion Corporation, San Diego, CA, USA).
BME completed performance verication on the pump in
question. Functionally, the pump operated within the manufacturers specications and no faults were found. The only
anomaly noted was that the pumps event log date/time
stamp wasnt calibrated to real time. This had no bearing
on the event but complicated the investigation as all times
recorded in the event log (referred to as log throughout) had to be converted to match the real time of nursing
documentation.
A logged event is dened as any keystroke on the pump
or action item recorded by the device (example, power off).
Interrogation of the pump logs yielded 2000 historical events
dating back 35 calendar days. Two hundred and eighty-nine
logged events were attributed to this patients care.
When the pump was initially turned on for this patient
for the NORepinephrine infusion, the log showed that the
nurse navigated through the pump options and arrived at
the loading dose rate. This rate was set to 13 mL/hour
(10 mcg/minute). There was a previously set volume to be
infused (VTBI) of 89.7 mL that was not adjusted. These
settings were accepted by the nurse and the infusion was
started. The log and the documentation show that over the
next several hours the pump rate was titrated between 4
and 13 mL/hour (3 and 10 mcg/minute) according to the
patients physiological condition. At no point in time was
the VTBI adjusted.
Eight hours and 58 minutes after the start of the NORepinephrine infusion, the VTBI reached 0 mL and the pump
defaulted to a previously set primary rate of 999 mL/hour.
377
KVO
OPT
HLD
PRI
SEC
A
B
Options
VTBI
Limit
This was determined to be a historical setting from a previous patient and was not manually conrmed by the current
users.
The pump infused NORepinephrine at 999 mL/hour for
two minutes, following which the patient suffered a cardiac arrest. The log showed that 37.1 mL of the drug was
infused during the elapsed time, resulting in an equivalent
calculated bolus dose of 1.8 mg of NORepinephrine in two
minutes.
Discussion
The pump was tested by BME and performed precisely as it
was designed to do. The loading dose and the continuous primary rate were drawn from the same polyvinylchloride bag.
The loading dose mode allows the clinician to set up an infusion rate different to the primary rate until a specic volume
has infused. It operates such that it maintains priority in
the pumping sequence over the primary rate and, once the
VTBI has been delivered, the pump will default back to the
programmed primary rate (Alaris Medical Systems, 2003).
A retrospective review of pump logs can help decipher
incidents, but provides only one perspective of the incident.
The logs can only show the keystrokes on the pump and do
not provide contextual information, such as the nature of
the environment, stafng levels, clinical choices presented
to the user at the time, conversations or decisions made
about care, or even who keyed the entries. In this way, the
logs are much less informative than black boxes used in
aviation (Bitan and Nunnally, 2007).
During the interview with the nurse we discovered that
he was not aware that the pump had been programmed as a
loading dose and not a primary rate. Not having recognised
that the infusion was running as a loading dose (and the
implications of the pre-existing VTBI of 89.7 mL), he was ultimately unaware of the latent consequences of those actions.
Further, it was determined that the pump settings were not
reviewed at the time of transfer to ICU.
A cognitive mechanism contributing to medical device
use errors is dened by Zhang et al. as conrmation bias
which is the the tendency to interpret the outcome as
a piece of conrming, positive, or consistent evidence for
ones hypothesis or anticipation (pg. 199, Zhang et al.,
2004). Photo 1 shows what the user would have seen when
Photo 1
programming the pump (Table 1 denes the pump nomenclature). While the current status of the pump is shown in
block letters as LOADING DOSE, the authors believe it could
be interpreted that the Primary Settings label directly
under the Rate and VTBI settings conrmed the users belief
that they were indeed in the correct mode of infusion.
The old adage looks can be deceiving could not be
more tting as a summary of this event. It was discovered
that two software versions (v2.79 and v4.54) of the pump
were mixed in the hospital eet. To the user, these pumps
appear identical but in fact they behave quite differently.
When the v4.54 pump is powered on, the user sees a display asking New Patient?, and the user is required to
conrm yes or no (see Photo 2). If the user presses Yes,
the historical infusion rate and VTBI for both the loading
dose and the primary rate are set to 0.1 mL/hour and 0.1 mL
Photo 2
378
respectively, forcing the user to manually key in a value
for each parameter. Conversely, if the user presses No,
all historical settings on the pump are preserved. Unfortunately, the v2.79 pump involved in this incident does not
include the New Patient? prompt and, therefore, lacks
this important safeguard during pump programming. In v2.79
historical settings are preserved unless the user manually
resets them. The presence of this simple prompt could well
have aided in the prevention of this adverse event.
Disclosures
The views and opinions expressed in this article are those of
the author alone and do not necessarily represent the views
or opinions of CareFusion or Becton, Dickinson and Company.
Funding
The authors have no sources of funding to declare.
Conict of interest
The authors have no conict of interest to declare.
Conclusion
References
A patient was started on a NORepinephrine infusion inadvertently as a loading dose rather than a primary infusion.
Historical values were not adjusted during the setup. Almost
nine hours later, the loading dose VTBI reached 0 mL and the
pump reverted to the historical primary rate of 999 mL/hour.
The log showed that 37.1 mL of NORepinephrine was infused
before it was shut off, resulting in an equivalent calculated
bolus dose of 1.8 mg. The patient suffered a cardiac arrest.
Pump testing revealed the pump to be functioning according
to manufacturers specications and operation as designed.
A eet analysis of the hospital found two separate software
versions of the pump, creating a situation where pumps look
the same but behave differently.
By incorporating information obtained from the pump log
into our usual incident analysis framework (Incident Analysis
Collaborating Parties, 2012), we were able to reconstruct a
timeline and understanding of the sequence of events that
led to the critical incident. A complete understanding would
not have been possible without the information provided by
the log.
Alaris Medical Systems. Signature Edition GOLD infusion pump models 7130/7131 and 7230/7231 Directions for Use, Inc. San Diego,
CA, USA; 2003. p. 1478598103.
Bitan Y, Nunnally ME. Can a log of infusion device events be used to
understand infusion accidents? J Patient Saf 2007;3:20813.
Brown SL, Bogner MS, Parmentier CM, Taylor JB. Human error
and patient-controlled analgesia pumps. J Intraven Nurs
1997;20:311.
Cafazzo JA, St-Cyr O. From discovery to design: the evolution of
human factors in healthcare. Healthc Q 2012;15(April (Special
Issue)):249.
Fyfe AI, Daly PA, Dorian P, Tough J. Reversible cardiomyopathy after
accidental adrenaline overdose. Am J Cardiol 1991;67.4:3189.
Girard C, Payen C, Tchenio X, Holzapfel L, Descotes J. Severe reaction to inadvertent intravenous administration of a large dose of
norepinephrine. Am J Emerg Med 2010;28:113.e57.
Hall AH, Kulig KW, Rumack BH. Intravenous epinephrine abuse. Am
J Emerg Med 1987;5.1:645.
Harding AD. Intravenous smart pumps. J Infus Nurs 2013;36:1914.
Hicks RW, Sikirica V, Nelson W, Schein JR, Cousins DD. Medication
errors involving patient-controlled analgesia. Am J Health Syst
Pharm 2008;65:42940.
379
analysis of corresponding adverse drug events. J Gen Intern Med
2008;23:415.
Notcutt WG, Knowles P, Kaldas R. Overdose of opiod from patientcontrolled analgesia pumps. Br J Anaesth 1992;69:957.
Pinkney S, Trbovich P, Fan M, Rothwell S, Cafazzo JA, Easty A. Do
smart pumps actually reduce medication errors? Biomed Instrum
Technol 2010;44:649.
Taneli Vyrynen MJ, Luurila HH, Mtt TK, Kuisma MJ. Accidental intravenous administration of racemic adrenaline: two
cases associated with adverse cardiac effects. Eur J Emerg Med
2005;12.5:2259.
Syed S, Paul JE, Hueftlein M, Kampf M, McLean RF. Morphine overdose from error propagation on an acute pain service. Can J
Anaesth 2006;53:58690.
Trbovich PL, Jeon J, Easty A. Smart medication delivery systems: infusion pumps; 2009, http://ehealthinnovation.org/
wp-content/uploads/SmartMedicationDeliverySystems
FullReport1.pdf (accessed 21.04.14).
Vicente KJ, Kada-Bekhaled K, Hillel G, Cassano A, Orser BA. Programming errors contribute to death from patient-controlled
analgesia: case report and estimate of probability. Can J Anaesth
2003;50:32832.
Zhang J, Patel VL, Johnson TR, Shortliffe EH. A cognitive taxonomy
of medical errors. J Biomed Inform 2004;37(3):193204.