Content uploaded by Patricia Trbovich
Author content
All content in this area was uploaded by Patricia Trbovich on Feb 01, 2016
Content may be subject to copyright.
The Use of Multiple Methods to Explore the Impact of Interruptions on Intravenous (IV) Push
Delivery
Tara McCurdie
1
, Varuna Prakash
1,2
, Patricia Trbovich
2,3
1
Healthcare Human Factors, Techna Institute, Univeristy Health Network, Toronto, Canada
2
Faculty of Medicine, Institute for Biomaterials and Biomedical Engineering, University of Toronto, Toronto, Canada
3
HumanEra, Techna Institute, University Health Network, Toronto, Canada
Despite the safety-critical nature of healthcare, it is an interrupt-laden domain. Patient safety
organizations have long called for a reduction in interruptions to healthcare workers, in an effort to
reduce the likelihood of preventable medical error occurring during patient care. The goal of this
research was to examine the impact that interruptions have on nurses’ task performance during
medication administration, specifically intravenous (IV) push delivery, and the development of
appropriately designed interventions to mitigate the potential harmful effects of interruptions on
patient safety. IV push administration errors have been found to be common, and occur when doses
are administered faster or slower than recommended. Direct observations found that nurses were
interrupted every time they administered an IV push, sometimes more than once. Furthermore, the
percentage of nurses who made errors when performing IV pushes during simulated scenarios was
significantly higher when interrupted than in the uninterrupted condition. Data collected through
focus groups qualitatively described perceptions and preferences that led to the design of a
successful and appropriate intervention to solve this important patient safety problem.
INTRODUCTION
Following the Institute of Medicine’s “To Err is
Human” report (2000), the examination of interruptions
in healthcare has become increasingly prevalent. Various
patient safety organizations have also since
acknowledged that interruptions may be a contributing
factor in the occurrence of preventable medical errors
(Agency for Healthcare Research and Quality, 2003;
Wears, 2004; ISMP Canada, 2007). Since these reports
surfaced, a number of observational, laboratory and
simulator-based experiments have been conducted to
study the cognitive and clinical impact of interruptions
during clinical care. For instance, observations have
shown that many healthcare settings are interrupt-laden
(Westbrook et al., 2008, 2011; Trbovich et al., 2013;
McGillis Hall et al., 2010). Furthermore, simulation
research has shown that anaesthesiologists who engaged
with an interruption failed to notice a missed positive
patient identification check (Liu, Grundgeiger,
Sanderson, Jenkins, & Leane, 2009).
Various interventions have been proposed to reduce
interruptions, such as ‘do not interrupt’ zones (Kliger,
2010); however, Raban and Westbrook (2013) note there
is only weak evidence to support their effectiveness. In
their review, the authors note that a more comprehensive
understanding of the effect of interventions on
interruptions is needed; to enable this, researchers must
determine who or what the sources of interruptions are,
and the phases of the medication management process
during which they occur.
The medication administration stage is especially
important because it may be harder to recover from error
at this critical point in patient care. Bates et al. (1995)
found that errors resulting in preventable adverse drug
events occurred most often at the stages of ordering
(56%) and administration (34%). However, errors were
much more likely to be intercepted if the error occurred
earlier in the process: 48% at the ordering stage versus
0% at the administration stage.
Delivering an intravenous (IV) push is an example of
medication administration. An IV push is a small-volume,
syringe-based infusion that is administered manually by
nurses. Because of the entirely manual nature of the task,
IV pushes are inherently subject to human error and may
be more so when nurses are interrupted. An ethnographic
study out of the United Kingdom (Taxis & Barber, 2003)
revealed that errors occurred in 73% of 235 observed IV
push medication administrations; 95% of those errors
occurred because nurses administered doses faster than
recommended. The consequences of administering an IV
push too quickly include a wide range of adverse
reactions, such as shock, cardiac arrest, tinnitus and
vessel damage (Taxis & Barber, 2003; Paparella, 2004).
Thus, accurate delivery of IV push medications is
necessary to ensure patient safety.
The specific objectives of the present study are: (1)
identify the current state of interruptions and the impact
Proceedings of the Human Factors and Ergonomics Society 58th Annual Meeting - 2014 738
Copyright 2014 Human Factors and Ergonomics Society. DOI 10.1177/1541931214581134
they have on medication administration, specifically IV
push administration, (2) develop interventions using
Human Factors methods aimed to reduce the negative
effects of the interruptions, (3) evaluate the effectiveness
of the interventions.
METHODOLOGY & RESULTS
A combination of clinical observations, laboratory-
based simulations and focus group sessions were
conducted to gain a comprehensive understanding of the
relationship between interruptions and errors, in order to
support the design of successful and appropriate
interventions. This multi-method research strategy, also
known as “triangulation”, allows for a more holistic and
contextual portrayal of the unit under study (Jick, 1979).
It was anticipated that triangulation would enrich our
understanding of interruptions in healthcare by allowing
for new or deeper dimensions to emerge that were not
previously considered.
Observations
Method. An initial direct observation study involving
oncology nurses at a large cancer hospital was conducted
to determine the frequency, nature and timing of
interruptions to the nurses during medication
administration. For the purposes of this research, an
interruption was defined as a break in the performance of
the medication administration task, due to a distraction
initiated by another person or by technology. Observers
shadowed 17 oncology nurses for 3 hours each, over a
period of two months. To minimize influencing nurses’
behaviour during shadowing (i.e., the Hawthorne effect),
the observers’ presence was as unobtrusive as possible. A
microtablet PC with Remote Analysis of Team
Environment (RATE) software, developed at the
University of Virginia (Osborne, Shin, Guo, & Guerlain,
2004), was used to timestamp all activities performed by
nurses over the shadowing period.
Descriptive statistics were performed on the data to
determine frequency of various types of interruptions and
the activities being interrupted. This information was then
used to inform the design of high-fidelity simulations.
Results. Observations showed that nurses cared for up to
five patients and 15 medications at once. Noise levels in
the unit were high, and nurses had to cope with frequent
interruptions from staff, patients, family members and
equipment. A quantitative analysis of the data collected
revealed that nurses were interrupted up to 14 times per
hour. Alarmingly, 16% of interruptions occurred during
tasks that had a high potential safety impact, such as
intravenous delivery of drugs through infusion pumps or
during manual IV pushes. To better understand the
frequency of interruptions during safety critical tasks, the
frequency of interruptions per task occurrence was
examined. The most frequently interrupted safety-critical
task was the IV push. IV push tasks were interrupted on
every occasion they were delivered, and in some
instances, they were interrupted twice during a single IV
push delivery. Consequently, the opportunity for error is
high.
Observations provided researchers with a rich,
detailed description of the complex phenomena of
interruptions in the local oncology context. However,
observations were limited, as errors in the live clinical
environment were difficult to discern.
Simulation
Method. Using the data from observations, interruption
scenarios were developed for use in a high-fidelity
simulation. The purpose of the simulation was to measure
the effect of interruptions on nurses’ ability to follow best
practice procedures when administering medications to
patients. The study was conducted in a laboratory that
allowed for a high fidelity simulation of the large cancer
hospital in which the earlier observations took place.
Specific interruptions observed were selected based on
the frequency with which they occurred, and the level of
impact they may have had on nurses’ task performance.
In total, 18 nurses were asked to perform a series of
twelve medication administration procedures, one of
which was delivery of an IV push (see Figure 1). For half
of these procedures participants were interrupted while
performing their tasks, and for the remaining procedures
no interruption occurred. An example of such an
interruption was a question posed by a patient actor to the
participant while they administered an intravenous push
volume over the pharmacy prescribed rate of 6-10
minutes.
Figure 1: Participant (left) administering a (simulated)
IV push to the patient actor.
Proceedings of the Human Factors and Ergonomics Society 58th Annual Meeting - 2014 739
The order of interruption and non-interruption
conditions was counter-balanced to avoid carry-over
effects. Nurses’ IV push administration rates were
measured under both the interrupted and the uninterrupted
conditions. The simulation was videotaped to aid in
subsequent analysis.
Results. A 2-way (status: non-interrupted vs. interrupted)
repeated measures analysis of variance (ANOVA) was
performed to determine whether differences exist
between the control (i.e., non-interrupted) and interrupted
conditions. The percentage of nurses who made errors
(i.e., administered the IV push at a rate outside of the
prescribed range) was significantly higher in the
interrupted condition (16/18; 89%) than in the
uninterrupted condition (8/18; 44%, p < .01). When
interrupted, nurses tended to administer the intravenous
pushes too quickly or too slowly, even though a watch
was always at their disposal.
Through use of simulation scenarios we controlled
the confounding influence of variables present in the live
clinical environment. Furthermore, the results provide us
with a basis for predicting that interruptions likely lead to
error. Thus, addressing interruptions during IV push
administration will be an important strategy for
improving patient safety.
Intervention Design/Focus Groups
Method. The third phase of this research consisted of
focus groups with healthcare providers to design
interventions to mitigate potential errors resulting from
interruptions. Specifically, the results of the simulation
study were presented to a total of nine oncology nurses
(three sessions with three nurses in each group). Possible
solutions to mitigate interruptions were presented to the
nurses in order to initiate discussion and brainstorming.
As the simulation revealed that nurses lost track of time
when they were interrupted during IV pushes, example
interventions included solutions that would eliminate or
reduce the number of interruptions experienced during
this medication administration task. Nurses were
encouraged to discuss the potential effectiveness of these
solutions and provide additional ideas. The nurses were
central to the development of mitigation strategies, since
the best-designed solutions often result from
understanding the needs of the people who will use them
(Raabe, 2010).
Data from the focus group discussions were analysed
using thematic analysis. In this approach, data emerging
from focus groups was iteratively coded into groupings,
which were further analysed to examine the
interconnections and relationships between themes (Pope
& Mays, 2006). At the conclusion of the focus groups,
Human Factors Specialists, Advanced Nurse
Practitioners, direct healthcare providers, and nursing
executives collaborated to determine which interventions
could be feasibly tested out in the subsequent simulation
study.
Results. Analysis of focus group discussions and
subsequent collaboration with other stakeholders revealed
that interruptions are unavoidable and often necessary
during IV pushes. It emerged that nurses use the IV push
time to ‘teach’ and interact with their patients, and that
questions and associated interruptions should not be
discouraged. The nurses indicated preference for
strategies that would instead allow them to safely
multitask without losing track of time. One such strategy
involves the use of a visual timer (see Figure 2) to
mitigate the potential for error.
Figure 2: Visual timer for IV push.
Prior to starting an IV push, the nurse is required to
set the visual timer to the prescribed delivery time (e.g.,
10 minutes). The timer counts down by reducing the
visible colour block until zero, without an audible alarm
or distraction. This low-cost solution was intended to
provide a quick and intuitive way of keeping track of
medication delivery time, even if the nurse is interrupted.
A follow-up simulation study was planned to evaluate the
effectiveness of the intervention.
Follow-Up Simulation
Method. A second simulation was conducted in which 19
nurse participants were trained in the use of the visual
timer for IV pushes. The nurses were then asked to use
the timer while performing those procedures (see Figure
3). To permit comparability, tasks and interruptions were
repeated from the original simulation study. Errors
committed in the follow-up simulation were compared to
data collected in the first simulation to determine the
Proceedings of the Human Factors and Ergonomics Society 58th Annual Meeting - 2014 740
efficacy of the interruption mitigation strategy.
Additionally, nurses’ use of various coping mechanisms
to respond to interruptions were also recorded and
compared to those in the first simulation (acquired
through retrospective video review). The study by Liu et
al. (2009) found that the mechanism of coping with
interruptions (i.e., engaging with the interruption,
multitasking, deferring, or blocking the interruption) was
significantly associated with the resulting error rate.
Participants who were more engaged with the interruption
were less likely to detect critical errors than those who
deferred or blocked the interruption, since the former
behaviour diverted attention from the critical task at hand.
In the present study, nurses’ coping mechanisms were
recorded according to source of interruption (e.g., nurse
request, pump beep, patient conversation).
It was hypothesized that nurses would commit fewer
interruption-related medication administration errors
when using the visual timer, than when no mitigation
strategy was employed. Specifically, it was hypothesized
that IV pushes would be more likely to be delivered at the
prescribed rate when the visual timer was used.
Figure 3: Participant using the visual timer while
administering a simulated IV push.
Results. Fisher’s exact test was used to assess differences
in error rates between the pre-intervention and post-
intervention simulations. The percentage of nurses who
made errors when performing IV pushes was significantly
higher in the pre-intervention condition (16/18; 89%) than
in the post-intervention condition (6/19; 32%, p < .001).
As hypothesized, when visual timers were used to assist
in IV push time management the number of nurses who
committed IV push errors was significantly reduced.
Nurses’ mechanisms of coping with interruptions
during IV pushes were also compared in the pre- and
post-intervention conditions. In the post-intervention
condition, significantly fewer nurses engaged or
multitasked when interrupted by nurse requests during IV
push administration (44% to 5%). Rates of engagement
remained the same for pump and patient related
interruptions between the two simulation studies (see
Figure 4).
Figure 4: Percentage of nurses who engaged/multitasked
interruptions pre- and post-intervention.
DISCUSSION
Previous research has shown that rapid
administration of IV push doses occur frequently, and
may lead to patient harm. The present findings shed
additional light on this phenomenon as observations
demonstrated that IV push of medications by nurses is
also a procedure that is also often interrupted.
Furthermore, controlled simulations established that
despite wearing watches, nurses tend to lose track of IV
push administration time when interrupted during the
delivery process.
The focus groups were instrumental in uncovering
additional insights regarding the necessity of
interruptions, specifically from patients, during IV push
administration. These findings permitted a shift in the
focus of the intervention design that eventually led to
improved error rates in the intervention condition. The
visual timer was introduced to increase nurses’ resiliency
to expected interruptions, rather than to reduce those
interruptions. The success of this approach can be derived
from Figure 4; the percentage of nurses who engaged or
multitasked when interrupted by patients did not change
between pre- and post-intervention conditions. The nurses
were able to use the IV push time to provide patient
education (e.g., explaining the side effects of the
medications the patient was receiving), while at the same
time improving accuracy in delivery times. A potential
explanation for the decrease in the percentage of nurses
who engaged or multitasked interruptions from other
nurses may be that in the passage of time between
simulation studies, the awareness of interruptions and
Proceedings of the Human Factors and Ergonomics Society 58th Annual Meeting - 2014 741
their potential for error amongst the participant
population may have increased.
The various techniques used to study the impact of
interruptions on IV push delivery generated a
comprehensive understanding of the potential for error
and potential solutions. Observations quantitatively
relayed information about the interruptions that occur
during this safety critical task. Simulation research
allowed us to explore the consequences of such
interruptions, and data collected through focus groups
qualitatively described perceptions and preferences that
led to the design of a successful and appropriate
intervention to solve this patient safety problem.
Analogous to the results reported by Jick (1979),
“triangulation allowed for more confident interpretations,
for both testing and developing hypotheses” (p. 608), and
led to unpredicted but important context-related findings.
Limitations. Although every effort was taken to
ensure the simulation scenarios were as high fidelity as
possible, they were conducted in a simulated laboratory.
Some differences from the actual clinical environment
and workflow had to be made to ensure safety (e.g.,
mannequins were used instead of patients). This may
have caused some users to deviate from their regular
workflows. Additionally, the presence of researchers
may have altered participants’ behaviour in the
observation and simulation components of this research,
though post-session debriefs suggested that this was not
the case.
Future research. Testing out the visual timer
intervention in other care environments and over time is a
goal of future research. Additionally, a simulator-based
prospective randomized control trial to evaluate the use of
interventions would eliminate the potential for the
passage of time to influence nurses’ perceptions and
behaviours.
Conclusion. Earlier studies of interruptions and
patient safety have focused largely on defining the
negative impact of interruptions, without taking into
consideration the essential information sharing
opportunities interruptions can sometimes provide.
Additionally, traditional design and evaluation of
interventions has focused solely on the “problem” of
interruptions, with an aim to prevent or reduce them
altogether. The present research demonstrates, however,
that some interruptions cannot or should not be prevented
entirely. The use of multiple methods in our examination
contributed to this more thorough understanding of the
phenomenon.
ACKNOWLEDGEMENTS
This work was funded by the Canadian Patient
Safety Institute (CPSI). We thank the oncology nurses
who participated in all phases of this work, as well as the
following students for their assistance in conducting the
simulation experiments: Karin Ayanian, Michelle
Dowling, Archana Gopal, Melissa Griffin, and Diane
Kostka.
REFERENCES
Agency for Healthcare Research and Quality. (2003). The effect of
health care working conditions on patient safety. Rockville, MD:
Hickam, D.H., Severance, S., Feldstein, A., Ray, L., Gorman, P.,
Schuldheis, S., … Helfand, M.
Bates, D. W., Cullen, D. J., Laird, N., Petersen, L. A., Small, S. D.,
Servi, D., … Leape, L. L. (1995). Incidence of adverse drug events
and potential adverse drug events: Implications for prevention.
JAMA, 274(1), 29-34.
Institute for Safe Medication Practices Canada. (2007). Fluorouracil
incident root cause analysis report. Toronto, ON: U, D., Hyland,
S., Greenall, J., Gosbee, J., & Lane, S.
Institute of Medicine. (2000). To err is human: Building a safer health
system. Washington, DC: National Academies Press.
Jick, T. D. (1979). Mixing qualitative and quantitative methods:
Triangulation in action. Administrative Science Quarterly, 24(4),
602-611.
Kliger, J. (2010). Giving medication administration the respect it is
due. Arch Intern Med, 170(8), 690-692.
Liu, D., Grundgeiger, T., Sanderson, P. M., Jenkins, S. A., & Leane, T.
A. (2009). Interruptions and blood transfusion checks: Lessons
from the simulated operating room. Anesthesia and
Analgesia, 108(1), 219-222.
Osborne, N., Shin, T., Guo, H., & Guerlain, S. (2004). Remote
Analysis of Team Environments (Version 8/11/04) [Software].
Available from
http://www.sys.virginia.edu/hci/Rate/RATE%20Manual.pdf.
McGillis Hall, L., Pendersen, C., Hubley, P., Ptack, E., Hemingway,
A., Watson, C., & Keatings, M. (2010). Interruptions and pediatric
patient safety. Journal of pediatric nursing, 25(3), 167-175.
Paparella, S. (2004). Avoiding disastrous outcomes with rapid
intravenous push medications. J Emerg Nurs, 30(5), 478-480.
Pope, C.M., & Mays N. (Eds). (2006). Qualitative Research in Health
Care (3
rd
ed.). Malden, MA: Blackwell Publishing Ltd.
Raabe, P. (2010). User Centred Design [Information graphic].
Retrieved from http://www.paznow.com/ucd/.
Raban, M. Z. & Westbrook, J. I. (2013). Are interventions to reduce
interruptions and errors during medication administration
effective?: a systematic review. BMJ Qual Saf, doi:10.1136/bmjqs-
2013-002118.
Taxis, K., & Barber, N. (2003). Ethnographic study of incidence and
severity of intravenous drug errors. BMJ, 326: 684.
Trbovich, P., L., Griffin, M.C., White, R. E., Bourrier, V., Dhaliwal,
D., Easty, A. C. (2013). The effects of interruptions on
oncologists’ patient assessment and medication ordering practices.
Journal of Healthcare Engineering, 4(1), 127-144.
Wears, R. L. (2004). Caution, interrupted [online]. Retrieved from
http://www.webmm.ahrq.gov/case.aspx?caseID=%2073#case.
Westbrook, J. I., Ampt, A., Kearney, L., & Rob, M. I. (2008). All in a
day’s work: an observational study to quantify how and with
whom doctors on hospital wards spend their time. MJA, 188(9),
506-509.
Westbrook, J. I., Duffield, C., Li, L., & Creswick, N. J. (2011). How
much time do nurses have for patients? A longitudinal study
quantifying hospital nurses’ patterns of task time distribution and
interactions with health professionals. BMC Health Services
Research, 11, 319.
Proceedings of the Human Factors and Ergonomics Society 58th Annual Meeting - 2014 742
