Patient Safety Issues
©2020 American Association of Critical-Care Nurses
doi:https:
doi.org/10.4037/ajcc2020120
Background Nurses in intensive care units are exposed
to hundreds of alarms during a shift, and research shows
that most alarms are not clinically relevant. Alarm fatigue
can occur when a nurse becomes desensitized to alarms.
Alarm fatigue can jeopardize patient safety, and adverse
alarm events can lead to patients dying.
Objective To evaluate how a process intervention affects
the number of alarms during an 8-hour shift in an inten-
sive care unit.
Methods A total of 62 patients from an intensive care unit
were included in the study; 32 of these patients received
the intervention, which included washing the patient’s
chest with soap and water and applying new electrocar-
diography electrodes at the start of a shift. The number
of alarms, clinical diagnoses, and demographic variables
were collected for each patient. A Poisson regression
model was used to evaluate the impact of the interven-
tion on the overall number of clinical alarms during the
shift, with no adjustments to the alarm settings or other
interventions.
Results After relevant covariates are controlled for, the
esults suggest that patients in the intervention group
presented significantly fewer alarms than did patients in
the control group.
Conclusions Managing clinical alarms is a main issue in
terms of both patient safety and staff workload manage-
ment. The results of this study demonstrate that a rela-
tively simple process-oriented strategy can decrease the
number of alarms. (American Journal of Critical Care.
2020;29:390-395)
SKIN PREPARATION AND
ELECTRODE REPLACEMENT
TO REDUCE ALARM FATIGUE
IN A COMMUNITY HOSPITAL
INTENSIVE CARE UNIT
By De
ie Leigher, BSN, RN, CNML, Paula Kemppainen, BSN, RN, and
David M. Neyens, PhD, MPH
This article is followed by an AJCC Patient Care Page on
page 396.
XXXXXXXXXXAJCC AMERICAN JOURNAL OF CRITICAL CARE, September 2020, Volume 29, No XXXXXXXXXXwww.ajcconline.org
www.ajcconline.org AJCC AMERICAN JOURNAL OF CRITICAL CARE, September 2020, Volume 29, No XXXXXXXXXX
Alarm events can lead to
potential harm for patients.
W
ithin intensive care units (ICUs), most equipment has safety alarms embed-
ded that alert staff of changes in various patient parameters and situations.1
This technology can increase the already large number of alarms—sometimes
hundreds—nurses encounter during shifts1,2; many of those alarms are not
clinically relevant.3 Frequent auditory alarms can result in unintended con-
sequences that have implications for patient safety (eg, patient injuries, fatalities) and quality
of care.4 Managing these clinical alarms has been identified as a “top 10” safety concern.5
Although all alarms must be acknowledged or
dismissed, clinically relevant alarms require nursing
intervention, whereas false alarms do not. Various
factors can create false or nonactionable alarms: the
patient’s motion, inco
ect alarm parameter settings,
the patient’s condition, care being provided to the
patient (eg, bathing or turning), improper skin prepa-
ation or electrocardiography (ECG) electrode place-
ment, or faulty connections of leads or electrodes.
Here we use the term false alarms to describe both
false alarms and nonactionable alarms.
In its 2013 Sentinel Alert, the Joint Commission
eported that between January 2009 and June 2012,
alarm-related events led to the death of 80 patients
and to a permanent loss of function in 13 patients.6
Because of the critical importance of patient safety
and the rising number of alarm-related events, the
Joint Commission issued a national patient safety
goal related to alarm management, mandating that
hospitals make establishing an alarm safety system a
hospital priority and identify the most important
alarms to manage.6 In addition, the Joint Commis-
sion required accredited hospitals to implement
policies and procedures to manage alarms and
appropriately educate staff.7 When clinical alarms are
more likely to be false than clinically relevant, a
work culture can emerge wherein nurses may delay
esponding to alarms, especially when the setting
has a large patient census or a high patient to nurse
atio, and thus may miss critical alarms.8-10
The ever-increasing number of alarms can lead
to a phenomenon known as alarm fatigue.7,11 Alarm
fatigue can occur when a nurse is exposed to frequent
alarms and becomes desensitized to them.8 Alarm
fatigue has been described as the most common
factor contributing to alarm-related events,3,7,12,13
and it is well known that alarm fatigue can jeopar-
dize patient safety and that adverse alarm-related
events can lead to patient fatalities3,7,12 and staff
workload management issues.13 A recent study
showed that alarm management and nuisance
alarms remain problems.14
Much of the previous research on alarm fatigue
has examined the issue from the perspective of
technology and alarm parameters.8,15,16 A technology-
only intervention will not, however, completely
alleviate alarm fatigue because factors related to
organizational best practices and nursing best prac-
tices influence alarm management and subsequent
alarm fatigue. The litera-
ture focuses on special-
ized ICUs and ICUs in
large academic hospi-
tals,2,3,7,15,17 but the
patient population in a
community hospital’s ICU is typically more diverse
than that in specialty ICUs at larger facilities. Com-
mon diagnoses and conditions among patients in a
community ICU—like the one included in this
study—include congestive heart failure, pneumonia,
gastrointestinal bleeding, sepsis, cardiac a
hythmia,
alcohol withdrawal, suicide attempt, postoperative
complications, diabetic ketoacidosis, and chronic
obstructive pulmonary disease. This diversity makes
managing alarm fatigue through technology-centric
strategies (eg, by adjusting alarm parameters) a chal-
lenge for nurses.8,16,18,19 Therefore, alternative strate-
gies are needed to reduce the number of false
alarms and to address and reduce the effects of
alarm fatigue and increase patient safety.
One possible way to reduce the number of false
alarms is to improve skin preparation before placing
ECG electrodes. Cvach et al20 reported that daily
electrode changes reduced by 46% the number of
alarms per bed day in 2 acute care units. Hermens
et al21 recommended changes in sensor placement
procedures in an effort to reduce the number of false
About the Authors
De
ie Leigher is a nurse manager and Paula Kemppainen
is an assistant nurse manager, Greer Memorial Hospital,
Prisma Health System, Greer, South Carolina. David M.
Neyens is an associate professor, Department of Indus-
trial Engineering, Clemson University, Clemson, South
Carolina.
Co
esponding author: David M. Neyens, PhD, MPH, Depart-
ment of Industrial Engineering, Clemson University,100
Freeman Hall, Clemson, SC XXXXXXXXXXemail: dneyens@
clemson.edu).
XXXXXXXXXXAJCC AMERICAN JOURNAL OF CRITICAL CARE, September 2020, Volume 29, No XXXXXXXXXXwww.ajcconline.org
Red and yellow alarms
were counted for an
8-hour shift in an
intensive care unit at
a community hospital.
alarms related to surface electromyography, and they
proposed that preparing patients’ skin could improve
electrode–skin contact, thereby resulting in fewer non-
elevant alarms. In addition, an American Association
of Critical-Care Nurses Practice Alert outlined 7 nurs-
ing actions related to false alarms that may reduce the
number of such alarms22: properly preparing the skin
for ECG electrodes, changing ECG electrodes daily,
customizing alarm parameters and levels on ECG mon-
itors, customizing delay and threshold settings for
oxygen saturation via pulse oximetry, providing initial
and ongoing nursing education about devices with
alarms, establishing interprofessional teams to address
issues related to alarms (eg, developing policies and
procedures), and monitoring only those patients who
present clinical indications for
monitoring.1,22 Several of these
clinical decision–related or
technology-mediated interven-
tions can affect the number of
alarms that occur in an ICU,
and they are well documented
in the literature.5,8,15,16,18 We
must, however, further evalu-
ate how preparing the skin
for ECG electrodes and changing the electrodes
affect the number of alarms while accounting for spe-
cific patient types and characteristics. Therefore, the
objective of this study was to evaluate how a process
intervention of preparing the skin (ie, washing a
patient’s chest with soap and water) and changing
electrodes at the start of each shift affected the number
of alarms throughout an 8-hour day shift in an ICU.
Methods
Study Design
This study included 2 groups. For patients in
the intervention group, a nurse prepared their skin
for electrode placement (by washing the patient’s
chest with soap and water) and changed the elec-
trodes daily (before 8:00 AM). The same clinical staff
member prepared the skin and changed electrodes
throughout the entire study. We collected data only
on weekdays to ensure that the intervention was
consistent and done by the same provider. Patients
in the control group received standard care that
included changing electrodes only as needed, per
standard hospital procedure. We used 3M Red Dot
monitoring electrodes with foam tape and sticky gel
and Philips IntelliVue MP70 Patient Monitors for all
patients. Throughout the study, we did not modify
or adjust any parameters (eg, alarm thresholds) for
the equipment and monitors for any patient.
Study Setting and Sample
This study was conducted after we obtained
approval through the Prisma Health institutional
eview board (no. Pro XXXXXXXXXXThe study took
place in the ICU at a community hospital. The ICU
is not specialized and is similar to a medical-surgical
ICU in that the patient population varies daily and
can include patients with diagnoses of cardiac, respi-
atory, or gastrointestinal diseases, sepsis, alcohol
withdrawal, postsurgical complications, suicide
attempt, and others.
Data Collection
The study included 100 patients, with 50 patients
in each group. Each patient was included in the
study for a single 8-hour shift, and no individual
patient was included in both groups. If a patient
experienced no alarms during the 8-hour period,
they were excluded from the analysis. After exclu-
sions, the study included 62 patients. We counted
alarms hourly during the 8-hour period to calculate
the total number of alarms during the work shift.
Several demographics were collected for each patient:
age, primary and secondary diagnoses, body mass
index, activity level