Content uploaded by Shuchisnigdha Deb
Author content
All content in this area was uploaded by Shuchisnigdha Deb on Feb 25, 2016
Content may be subject to copyright.
This article was downloaded by: [108.178.250.241]
On: 31 July 2015, At: 10:03
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick Place,
London, SW1P 1WG
Click for updates
IIE Transactions on Healthcare Systems Engineering
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/uhse20
Alarm fatigue and its influence on staff performance
Shuchisnigdha Deba & David Claudioa
a Department of Mechanical and Industrial Engineering, Montana State University, Bozeman,
MT 59717, , USA E-mail:
Published online: 29 Jul 2015.
To cite this article: Shuchisnigdha Deb & David Claudio (2015) Alarm fatigue and its influence on staff performance, IIE
Transactions on Healthcare Systems Engineering, 5:3, 183-196, DOI: 10.1080/19488300.2015.1062065
To link to this article: http://dx.doi.org/10.1080/19488300.2015.1062065
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
IIE Transactions on Healthcare Systems Engineering (2015) 5, 183–196
Copyright C
“IIE”
ISSN: 1948-8300 print / 1948-8319 online
DOI: 10.1080/19488300.2015.1062065
Alarm fatigue and its influence on staff performance
SHUCHISNIGDHA DEB and DAVID CLAUDIO∗
Department of Mechanical and Industrial Engineering, Montana State University, Bozeman, MT 59717, USA
E-mail: david.claudio@ie.montana.edu
Received December 2014 and accepted June 2015
An alarm is a warning of an approaching situation which requires a response. The Emergency Care Research Institute considered
alarm hazard as the number one health technology hazard for the years 2012 through 2014. In response, The Joint Commission set
a standard for all hospitals in the United States to assess alarm fatigue in their monitoring process and to develop a systematic,
coordinated approach to clinical alarm system management. In order to comply with this requirement, a working definition of
alarm fatigue is necessary. This observational study undertook the objective of defining alarm fatigue, measuring it and exploring its
role in performance deterioration. A conceptual model was developed considering the significance of working conditions and staff
individuality on alarm fatigue and, consequently, alarm fatigue on staff performance. The results show that in general, performance
deterioration is actually influenced by a combination of alarm fatigue, working conditions and staff individuality. In fact, in the case
of nurses and response time, alarm fatigue plays no role, only working conditions and staff individuality. These findings suggest that
the role of alarm fatigue as a health hazard in the clinical environment should be reevaluated.
Keywords: Mental workload, alarm fatigue, acute care, affects
1. Introduction
Alarm systems give audible, visual or other forms of sig-
nal to indicate a potential need or hazardous condition.
Over the last 30 years, the use of clinical alarms has be-
come more widespread and have been improved through
new technologies, bringing the number of different alarms
from six in 1983 to more than 40 in 2011 (Kerr and Hayes,
1983; Borowski et al., 2011). In spite of this progress in the
detection of critical situations, the alarms themselves create
several problems in the healthcare process and sometimes
even result in patient deaths. In 1974, the Emergency Care
Research Institute (ECRI) published its first sentinel event
report on an ignored alarm signal of a hypothermia ma-
chine. That unawareness resulted in serious patient burns
(ECRI, 1974; Clinical Alarms Task Force, 2007). By 1982,
researchers recognized the increasing number of monitor-
ing signals, with “no end in sight” (Stafford, 1982). A year
later, Kerr and Hayes (1983) identified that each patient
could have six or more alarms, a situation that could in-
crease the risk of caregiver confusion as to which alarm is
going off and would actually be contrary to the patient’s
best interest. According to The Joint Commission (TJC),
∗Corresponding author
Color versions of one or more of the figures in the article can
be found online at www.tandfonline.com/uhse.
each day hundreds of auditory alarm signals sound for ev-
ery patient; thousands of alarm signals ring in every unit,
and tens of thousands of alarm signals blare throughout ev-
ery hospital (as cited by Mitka, 2013). Interestingly, several
previous studies reported that 80–99% of these alarms are
false alarms requiring no action (Lawless, 1994; Tsien and
Fackler, 1997; Clinical Alarms Task Force, 2007; Cvach,
2012). This excessive number of false positive alarms results
in caregiver sensory overload and desensitization. As a con-
sequence, the staff gets disinclined to respond to real threats
due to what has been called alarm fatigue by TJC (Mitka,
2013). TJC reported 98 alarm-related events between
January 2009 and June 2012. Of the 98 reported events,
80 resulted in death, 13 in permanent loss of function, and
five in unexpected additional care or extended stay (The
Joint Commission, 2013). In more than 60% of the cases,
alarms were either inappropriately turned off or were not
audible in all areas (Crites, 2013). The Food and Drug Ad-
ministration (FDA) revealed 566 death reports associated
with alarms from 2005 to 2008 (Weil, 2009). The Emer-
gency Care Research Institute (ECRI) found 264 incidents
related to alarms between 2000 and 2006 (Clinical Alarms
Task Force, 2007). Alarm-related problems may include
technical failures as well as alarm fatigue. Nevertheless,
health organizations have found these numbers significant
enough to address and the ECRI considered alarm hazard
as the number one health technology hazard for years 2012
through 2015 (ECRI, 2011, 2012, 2013, 2014).
1948-8300 C
2015 “IIE”
Downloaded by [108.178.250.241] at 10:03 31 July 2015
184 Deb and Claudio
In April 2002, TJC developed the first set of national
patient safety goals (NPSG) where “improving the effec-
tiveness of clinical alarm systems” was one of six goals
(The Joint Commission, 2014). In 2004, this was no longer
an optional goal for hospital organizations and was incor-
porated into TJC standards for all hospitals in the United
States (Clinical Alarms Task Force, 2007). In October 2011,
a two-day summit was held by the ECRI, the Association
for the Advancement of Medical Instrumentation (AAMI),
and the American College of Clinical Engineering (ACCE),
along with the FDA and TJC. At that summit, multidisci-
plinary specialists gathered in a collaborative effort with a
mission of eliminating from healthcare the issue of patient
safety risk related to alarms by 2017 (Association for the
Advancement of Medical Instrumentation, 2011). In order
to comply with the mission, this research found it necessary
to develop a quantitative definition for alarm fatigue and
to identify the proper metrics to measure it.
Alarm fatigue has been defined in a qualitative man-
ner as the sensory overload and desensitization that make
users disinclined to respond to real threats (Cvach, 2012).
Various studies have reported on several influencing factors
that cause alarm fatigue, such as the number of false alarms
(Chambrin et al., 1999; Billinghurst, Morgan, and Arthur,
2003; Ashton, 2011; Pishori, 2012; Baillargeon, 2013);
workload (Weinger and Smith, 1993; Billinghurst, Morgan,
and Arthur, 2003; Pishori, 2012); noise level (Wallis, 2010;
Pishori, 2012); difficulty with alarm-creating equipment
and their settings (Welch, 2011); alarm type (Chambrin
et al., 1999; Pishori, 2012; Baillargeon, 2013); emotional,
psychological, and personality factors (Weinger and Smith,
1993; Cvach, 2012); time (Pishori, 2012); and shift periods
(day/night) (Chambrin et al., 1999; Sendelbach and Funk,
2013). Most of these studies identified the factors which
can cause alarm fatigue in staff. Several studies tracked
alarm monitoring processes and found high percentages of
false alarms which require no action. Based on this high
percentage of non-actionable alarms, the researchers con-
cluded there was a presence of alarm fatigue in staff during
alarm monitoring processes.
In this regard, Weinger and Smith (1993) conducted an
evocative study and stated that the caregivers are prone to
errors in vigilance and monitoring alarms. They revealed
many instigating factors behind those errors, such as num-
ber of false alarms, high noise level, equipment and sys-
tems, workload and task characteristics, and caregivers’
emotional factors, psychological and personality factors,
stress levels and training and experience. The researchers
also believed that alarms would be effective only if prop-
erly designed and implemented. Two years later, in 1995,
Bliss et al. conducted an observational study on alarm re-
sponses to assess the monitoring system. The influences of
increasing alarm reliability on alarm response frequencies
were assessed in terms of speed and accuracy. The results
indicated that a reduction in alarm reliability resulted in
a reduced response. In 1999, Chambrin et al. conducted
an innovative observational study to evaluate the existing
alarm monitoring system in an adult ICU. The researchers
tracked 1,971 hours in a critical care unit and found 3,188
alarms, of which only 5.7% required responses. The authors
considered false alarms a significant issue which can cause
alarm fatigue. Researchers at a nine-bed Coronary Respi-
ratory Care Unit (CRCU) took a new approach by doing
an observational study to examine the influences of remote
telemetry on nurses and patients (Billinghurst, Morgan,
and Arthur, 2003). The researchers found that 80.2% of
the total alarms were non-actionable and there was a wide
range of 60% to 100% in nurse detection and response to
valid alarms. The authors concluded that remote cardiac
telemetry placed unnecessary demand on CRCU nurses’
workload and negatively influenced patients (Billinghurst,
Morgan, and Arthur, 2003). In 2012, Pishori investigated
the routine processes of alarm monitoring in a telemetry
unit. The researcher simulated the existing system to quan-
tify the sources of noise contributing to nursing fatigue.
Real-time data was gathered to understand the state of care
provided to CRCU patients and existing response strategies
through an observational study. By altering several factors
and parameters, the results showed the positive influence
of reducing certain activities from the workflow on increas-
ing the idle time of the nurses at the central monitoring
area. According to a non-observational evocative study, in
order to reduce the alarm-related incidents due to alarm fa-
tigue, it is necessary to address the common causes behind
the fatigue (Association for the Advancement of Medical
Instrumentation, 2011). This study emphasized education,
customization and /or minimization of alarms, adequate
staffing, and preparation of skin and proper placement of
electrodes to reduce the number of false and non-actionable
alarms. The authors also recommended designing an ex-
periment using a randomized control trial and statistical
analysis focusing on potential outcomes, in order to find
the influence of different factors. However, they did not
conduct any observational study nor did any experimental
design.
The most common variables used to measure the influ-
ence of alarm fatigue on staff performance are the modes
of response to the alarms: whether it is responded to
or ignored, delayed, silenced or turned off (Graham and
Cvach, 2010; Ashton, 2011); response time (Bliss, Gilson,
and John, 1995; Pishori, 2012; Baillargeon, 2013); and the
number of ignored alarms (Bliss, Gilson, and John, 1995).
In 2013, a risk assessment study was performed by Bail-
largeon at an acute care hospital. Alarm frequency was
calculated and average response time to critical and leads-
off alarms were determined. The results showed that the
percentages of false, nuisance and technical alarms added
up to 52%, which the researcher considered a contribu-
tor to alarm fatigue. In fact, response times were found
at times to be over 10 minutes. The study concluded
that nurses were at risk for experiencing alarm fatigue
based on high alarm frequency, increased number of false
and nuisance alarms, and a delayed response to leads-off
conditions.
Downloaded by [108.178.250.241] at 10:03 31 July 2015
Alarm fatigue 185
Despite all the research being performed on this topic,
none of these previous studies attempted to find a
quantitative definition of alarm fatigue nor measure alarm
fatigue based on the influencing factors; rather they mea-
sured staff performance based on the influencing factors
and blamed alarm fatigue as the cause of performance de-
terioration. To the very best of our knowledge, there has not
been a study done which identifies all the possible signifi-
cant factors behind alarm fatigue and the influence of alarm
fatigue on staff performance using statistical analysis. Fur-
thermore, according to most of the qualitative definitions of
alarm fatigue, it has been found that staff get overwhelmed
and desensitized due to a high number of false alarms and
continuous beeping monitors, resulting in alarm fatigue;
however, no study has measured this overwhelming con-
dition or desensitization. This study undertook the task
of defining alarm fatigue in terms of mental workload to
measure the overwhelming situation and affect (boredom,
apathy, and distrust) to measure desensitization. At the
same time, this study intended to verify whether alarm fa-
tigue truly is the cause that deteriorates staff performance
and eventually causes adverse clinical incidents in terms
of patient deaths, damage to patient condition or extended
stays, among other incidents. The use of personality factors
as an influencing variable and mental workload and affect
as measures of alarm fatigue were unique considerations
for this kind of observational study.
2. Methodology
In this observational study, alarm fatigue was defined in
terms of mental workload and affect. Affect can be de-
fined as the experience of feeling an emotion (Hogg et al.,
2013). The study considered total number of alarms, staff
to patient ratio (workload), time elapsed since start of the
shift, alarm type, alarm criticality, noise level, task pri-
ority and staff personality as independent variables. Shift
(day/night) and Staff ID were considered blocking factors.
At the same time, this study intended to investigate the
influence of alarm fatigue on staff performance deteriora-
tion which eventually may cause adverse clinical incidents
in terms of patient deaths, damage to patient conditions or
extended stays among other accidental events. Therefore,
in the second step, alarm fatigue measures were considered
independent variables and the performance measures: re-
sponse to alarms (yes/no), response time and number of ig-
nored alarms were considered predictors. Figure 1 presents
a conceptual model for the method of study.
2.1. Approval for the study
Approval from the Institutional Review Board (IRB) was
obtained prior to the beginning of all surveys and data
collection. The quality department of the selected hospital
also endorsed this study. In addition, the manager of the
facility where the study took place along with the head
of nurses gave approval for the surveys that were to be
conducted.
2.2. Participants
The potential participants included all staff who were re-
sponsible for alarm monitoring in the selected acute care
area. Data for six unit clerks and 18 registered nurses was
collected for this study. The staff scheduling policy for this
particular facility for one 12-hour shift included one reg-
istered nurse for every two patients (exceptions were avail-
able depending on patient criticality) and one unit clerk.
Demographic information for both unit clerks and nurses
is summarized in Table 1. All the participants were assigned
a subject identity number (ID) for the confidentiality and
convenience of data analysis of the research. All personal
information of the participants was maintained highly con-
fidential and was not reported with the results.
2.3. Location
The facility was an eight-bed Intensive Care Unit (ICU).
Unit clerks were responsible for monitoring two teleme-
try monitors in the central monitoring area. One telemetry
monitor showed the alarms for the ICU patients while the
second monitor showed alarms for patients admitted to
other medical floors of the same hospital. Ventilator and
IV pump alarms for each patient were monitored from the
patients’ room by the nurses. Figure 2 displays a represen-
tation of the facility (not to scale).
2.4. Procedure
Several surveys and observations were conducted during
this research. The study started with a current state assess-
ment survey, proposed by the American College of Clini-
cal Engineering. Following this survey, Hierarchical Task
Analyses (HTA) were conducted for unit clerks and nurses
separately to determine whether all influencing factors were
considered for defining alarm fatigue. In order to gather
data for most of those influencing factors, some direct ob-
servations were taken through work sampling. Some ques-
tionnaires were applied to measure alarm fatigue in terms
of mental workload and affects. A personality type survey
was conducted via email. The procedures are described in
the following sections.
2.4.1. Clinical Alarm Survey
The Clinical Alarm Survey was proposed by the American
College of Clinical Engineering. It was conducted twice
with the collaboration of physicians, registered nurses, and
other professionals in the medical field across the United
States in 2004 and 2011 (Clinical Alarms Task Force,
2007; Clark and David, 2011; Association for the Advance-
Downloaded by [108.178.250.241] at 10:03 31 July 2015
186 Deb and Claudio
Fig. 1. Conceptual model for the method of study.
ment of Medical Instrumentation, 2011). This study started
with this current state’s assessment survey to evaluate the
attitudes of staff toward the existing alarm monitoring sys-
temintheICU.
2.4.2. Hierarchical Task Analysis (HTA)
The creation of Hierarchical Task Analysis (HTA) is gen-
erally ascribed to Annett and Duncan (1967). This method
produces a hierarchy of three levels of tasks: Goal, which
is the system state that a person wishes to achieve; Task,
which is a structured set of activities following some se-
quence which allows a person to achieve the goal; and
Operation or Action, which is what a person must actually
do to carry out each individual activity. In this study, tasks
were analyzed for nurse and unit clerks separately and the
analyses were reviewed by two nurses, one unit clerk and the
ICU manager at the hospital. The top-level goal of the sys-
tem is to respond to the alarms at ICU. This task analysis
helped to find an influencing factor(s) that can cause alarm
fatigue which was not found from the literature review.
2.4.3. Observational data collection
Twelve direct observational blocks were designed includ-
ing day and night shifts to collect data on all the influenc-
ing variables found from the literature review as well as
those discovered through the HTA. These 12 blocks (six
for days and six for nights) were randomly selected over
a two-month period. Six 15-minute-long observation slots
for each of those shifts were randomly generated using work
sampling. Work sampling is a valid and popular method
in the case of healthcare research (Wright, 1954; Abdellah
and Levine, 1954; Reid, 1975; Liptak et al., 1985). Dickson
(1978) looked at fixed-time intervals for work-sampling in
contrast to random observations and found no significant
differences.
Two researchers collected data from two telemetry moni-
tors during direct observations: one observer collected data
Table 1. Demographics summary of participants
Statistics/Percentage
Demographic category Unit clerk (N=6) Nurse (N=18)
Gender
Female 83.33% 84.20%
Male 16.67% 15.80%
Age (years)
Average (SD) 30.33 (8.04) 44.11(16.20)
Education
High school 16.67% —
Bachelor 83.33% 100%
Experience (years)
Average (SD) 7 (2.77) 10.33 (11.48)
Experience at ICU (years)
Average (SD) 5.5 (3.02) 8.08 (7.78)
Specialized training on alarm monitoring 100% 36.36%
Experience of participating in alarm related research None None
Downloaded by [108.178.250.241] at 10:03 31 July 2015
Alarm fatigue 187
Fig. 2. Representation of the ICU.
for ICU patients from one telemetry monitor, as well as
from phone calls and noise level, and the other observer
collected data for the patients from other floors through
the second telemetry monitor, as well as for the alarms
in the ICU rooms including the ventilators, IV pumps,
call-lights, and med-door-open alarms. The study collected
twelve replications from which eight observations were in-
cluded in the data analysis. During the observation period,
data were collected primarily on the number of audible
alarms for each alarm category along with their criticality
and nurses’ response to those alarms. Task priority was also
noted for each record. The frequency of phone calls, call
lights and med-door-open alarms were also recorded with
the response times. The noise level at the central monitoring
station was collected using a noise meter for each alarm.
The number of patients and nurses were recorded from the
floor and daily staff schedule. Performance measures: re-
sponse (yes/no) and response time were also recorded dur-
ing direct observations. The response times were collected
using stopwatches.
2.4.4. Mental workload measurement
Alarm fatigue is caused when staff get overwhelmed by a
large number of alarms along with the related informa-
tion processing, decision making and the high noise level.
Therefore, mental workload was determined to be a very
important variable to measure fatigue caused by this task.
This study used two mental workload measuring tools:
the Subjective Workload Assessment Technique (SWAT)
and National Aeronautics and Space Administration-Task
Load Index (NASA-TLX).
Initially SWAT was described by Nygren (1982). Af-
terwards, in 1988, Reid and Nygren described it more
completely. SWAT has been found to be a measure of
mental workload that is valid (Haworth, Bivens, and
Shively, 1987), sensitive (Reid, Eggemeier, and Shin-
gledecker, 1982), reliable (Griscomb, 1985) and relatively
unobtrusive (Eggemeier, 1988). This is a multidimensional
tool with three dimensions for time load (T), mental effort
load (E), and psychological stress load (S) at three discrete
levels: low, medium, and high. The SWAT rating can be
used in two ways: during the tasks and/or after the com-
pletion of the task (Reid, Potter, and Bressler, 1989). This
study used SWAT during the tasks to obtain an instant feel
for the mental workload for the staff. SWAT was carried
out in three steps. The first step consisted of a scale devel-
opment in order to rank the three different dimensions. The
scale development process can be performed in two ways:
card sorting and pair-wise comparison. This study decided
to conduct pair-wise comparison which is named Contin-
uous SWAT or C-SWAT. C-SWAT is less time consuming
and gives similar results to the SWAT with card sorting
(Luximon and Goonetilleke, 1998). Participants were re-
quested to do pair-wise comparison for three pairs of three
dimensions according to his or her perceptions of increas-
ing workload. Following scale development, in the second
step, participants were asked to rate (1, 2, and 3 as Low,
Medium, and High, respectively) the alarm monitoring
task with regard to time load, mental effort load and psy-
chological stress load dimensions. In the third step, each of
the three dimension scorings was converted into a numeric
score from 0 to 100 using the scale developed in the first
step.
Downloaded by [108.178.250.241] at 10:03 31 July 2015
188 Deb and Claudio
NASA-TLX, developed by Hart and Staveland (1988) is
a trusted tool which assesses the mental workload imposed
by a task. NASA-TLX includes a two-part evaluation pro-
cedure making absolute and comparative ratings of a task
performed by the participants based on six separate factors
contributing to perceived workload (Steege and Nussbaum,
2013). These subscales can be measured on their own (“Raw
TLX”) or combined in a pair-wise comparison that mea-
sures the relative importance of each subscale to the par-
ticipants (Hart and Staveland, 1988). According to Hart
(2006), this Raw TLX is simpler to use and gives similar re-
sults for the total mental workload score using the weighing
method. In this study the Raw TLX was used to measure
mental workload. NASA-TLX forms were distributed at
the end of the observed shifts to all the staff involved to
collect their comments about the overall shift workload.
Participants were instructed to read a description of each
scale before making any ratings. All the participants rated
their experience with a given task on 20-step scales for each
of the six dimensions. A score from 0 to 100 (rounded to
the nearest 5 points) was obtained on each scale. An overall
mental workload score for NASA-TLX was obtained on a
scale of 0 through 100 for each participant by calculating
an average of those six ratings for six dimensions.
2.4.5. Affect measurement
Alarm fatigue can also be defined through measuring affect
representing factors that make the staff desensitized to re-
sponding to alarms. Three types of affects were considered
for this study: boredom, apathy, and distrust. Boredom is an
unpleasant, transient affective state caused by insufficient
workload in which the individual feels a lack of interest and
difficulty concentrating on his/her activity (Fisher, 1993).
During the alarm monitoring process, due to large numbers
of non-actionable alarms, staff become indifferent and dis-
play a lack of interest in responding to alarms. This state is
the result of the monotonous occurrence of non-actionable
alarms which causes desensitization. Therefore, boredom
was selected as a measure for alarm fatigue.
Apathy is defined as a lack of feeling, interest, or con-
cern (Godefroy, 2013). It is a state of indifference, or the
suppression of emotions such as concern, excitement, mo-
tivation, and/or passion. In an alarm monitoring system,
caregivers get overwhelmed by a high number of repeti-
tive non-actionable alarms and desensitized by high noise
level. As a consequence, they become demotivated and im-
mune to any alarm. This is the state of apathy which was
considered another measure of alarm.
Distrust is the feeling of not being able to rely on some-
thing and/or someone (Harrison and Chervany, 2001). In
the case of alarm monitoring, a large number of false or
nuisance alarms leads the staff to stop relying on all alarms.
This situation results in an undesirable response, such as a
delayed or no response, silencing, or turning off alarms or
changing threshold parameters outside the safe limits, with
the possibility of very dangerous consequences. Therefore,
distrust can be a very important variable to define and
measure alarm fatigue.
The study used some valid tools that have research ap-
plications in healthcare to measure these three types of af-
fect. Boredom was measured using the Boredom Proneness
Scale (BPS) and for apathy, the Apathy Evaluation Scale
(APS) was used. As several questionnaires were applied to
measure alarm fatigue, valid shorter versions of these tools
were used.
The original BPS was developed by Farmer and Sund-
berg in 1986 with 28 items. Several studies provided ev-
idence of the validity of BPS through comparisons with
other measures of boredom and similar constructs (Ahmed,
1990; Vodanovich, Wallace, and Kass, 2010). In 2010, a
short version of this tool, having 12 items, was proposed
and validated by Vodanovich, Wallace, and Kass. Ten items
were taken directly from that short version of the orig-
inal BPS for this study; two questions were rejected for
their irrelevance to this study. Similar to the NASA-TLX,
the BPS questionnaire was administered to each partic-
ipant at the completion of their shift as a retrospective
survey.
The original Apathy Evaluation Scale was developed and
validated by Marin et al. in 1991. It has 18 items requiring
subjective responses which are to measure apathy in an indi-
vidual. Much research has been conducted in healthcare in
order to provide evidence for the reliability and validity of
AES (Marin, Biedrzycki, and Firinciogullari, 1991; Hsieh
et al., 2012). In 2007, Lueken and colleagues proposed and
validated a short version of the Apathy Evaluation Scale.
The short version has ten items which were used for this
research to measure staff apathy. As with the NASA-TLX
and the BPS, the short version of the AES was admin-
istered as a retrospective survey after each observational
shift.
The literature review did not give any established and
valid tool to measure distrust for hospital staff during their
alarm monitoring task. Therefore, as part of this research, a
survey was created to measure staff distrust toward alarms.
The survey consisted of three questions; two of the three
questions were created on the basis of an Inter-Cultural
Scale to measure trust in automation (Chien et al., 2014).
A third question was added on expertise opinion. The ques-
tions were validated by two registered nurses and two Unit
Clerks along with the manager in the ICU at the facility
where the study took place. The survey was administered
during tasks along with the SWAT questions.
2.4.6. Personality type survey
Personality has been shown to influence various areas of
life, such as coping with stress, dealing with crises, and
job performance (Barry and McCarthy, 2001; Meeusen
et al., 2010; Franklin et al., 2013). It is reasonable, there-
fore, to consider the influences of personality as they ap-
ply to concerns with alarm fatigue, in areas such as cop-
ing with a constantly noisy environment, devoting long
Downloaded by [108.178.250.241] at 10:03 31 July 2015
Alarm fatigue 189
hours to alarm monitoring, making careful decisions un-
der stress, and adapting to new and changing technologies.
There are several personality type tests available which were
studied. Research has shown that the elements of the Big
Five Personality Test can be seen in nearly all personal-
ity measurement systems (Goldberg, 1981). Because of the
consistency he had seen in the results using the Big Five Fac-
tors, Goldberg (1981) suggested that any model for struc-
turing individual differences would have to encompass sim-
ilar dimensions. Pertinent to this study, Lakin et al. found
in 2007 that neuroticism and extraversion were the only
personality traits associated with fatigue. In light of the
overwhelmingly positive support for the Big Five model,
this study used the Big Five Personality Test on nurses and
unit clerks to determine whether personality affects how
quickly they get overwhelmed.
The Big Five Personality Test is available online and very
easy to take. All the participants received the test with the
link (http://www.outofservice.com/bigfive/) via e-mail.
The test has 45 basic questions to decide personality type
based on the five personality dimensions: Extraversion,
Openness, Agreeableness, Conscientiousness, and Neuroti-
cism. There are also four demographic questions that are
considered helpful in deciding the personality type prop-
erly. Each test taker sent the link of his/her result page to
the principal researcher which only gives the personality
type, not the test answers. The personality type dimensions
for each of the staff were recorded along with participant
ID.
2.4.7. Statistical analyses
This study intended to associate alarm fatigue measures
with factors that could cause alarm fatigue as well as study
the relationship between alarm fatigue and performance
measures. As multiple categorical factors, blocking factors,
and continuous variables were considered in this study, cor-
relation analyses were necessary for the data pertaining to
both unit clerks and nurses; if a correlation larger than
0.9 was found between any pair of factors, one of them
would be removed from further analyses. Before doing the
regression analyses, the normality, linearity and homogene-
ity assumptions were tested for all the five alarm fatigue
measures and three performance measures. Box-Cox and
Johnson transformations were applied when needed. In the
case of failure to normality, linearity, and homogeneity
assumptions, a variable reduction analysis Principal Com-
ponent Analysis (PCA) was conducted for the five alarm
fatigue measures with all the independent variables and
for the three performance measures with five alarm fatigue
measures. PCA is a powerful tool for reducing a number
of observed variables into a smaller number of artificial
variables combining all the similar factors together, which
account for most of the variance in the data set. PCA iden-
tified the cluster of artificial variables which were used in re-
gression analyses after the interpretation of the variance of
factors on each new component. If unable to produce suc-
cessful models with the components extracted from PCA,
non-parametric, non-linear regression analyses were per-
formed.
3. Results and discussion
3.1. Assessment of the existing alarm monitoring system
The study started with a clinical alarm survey to assess staff
attitudes toward the existing alarm monitoring system. This
survey showed that the existing alarm monitoring system
produces a large number of false alarms which is one of the
main causes initiating alarm fatigue in staff. A majority of
respondents confirmed the occurrence of frequent nuisance
alarms (84.6%) which disrupt patient care (84.6%), and re-
duce trust in alarms causing caregivers to disable alarms
(53.8%, neutral: 46.2%). Responses were divergent on the
statements about the complexity of hearing, recognizing
alarms, and responding to alarms. Participants (46.2%) dis-
agreed that staff were sensitive enough to respond to alarms
quickly. These responses make it clear that the chosen fa-
cility had the possibility of having alarm fatigue in staff.
Furthermore, some descriptive statistics found on the ex-
isting alarm monitoring conditions make the necessity for
alarm fatigue study stronger.
This study collected data on 1055 alarms for six unit
clerks and 64 alarms for 18 nurses over four day and four
night shifts. Descriptive statistics based on these observa-
tions show that 88% of these observed alarms were false,
and the average number of alarms per day per patient was
116. Both of these numbers comply with the conditions
considered hazardous by TJC (Mitka, 2013). In addition,
the noise level found in the observed facility was 50–70 dB
which is much higher than the recommended noise level in
an ICU (30–45 dB) (Konkani and Oakley, 2012). Therefore,
the chosen facility desperately required to assess alarm fa-
tigue in their staff in order to identify the factors affecting
their performance.
3.2. Hierarchical Task Analysis (HTA)
With the confirmation from the results of the Clinical
Alarm Survey that alarm-related problems existed in the
chosen facility further research was performed at the facil-
ity. As a next step, the Hierarchical Task Analysis (HTA)
was conducted. During task analyses, a significant issue ob-
served was the difference in the responsibilities of unit clerks
and nurses. This distinction was not confirmed by the pre-
vious studies in this field. These two types of subjects have
completely different assignments toward patient safety and
distributed responsibilities for responding to alarms. As
a consequence, the decision was made to perform all the
analyses separately, dividing the ICU staff into two types of
subjects: unit clerks and nurses. HTA was also conducted
separately for both unit clerks and nurses in order to en-
Downloaded by [108.178.250.241] at 10:03 31 July 2015
190 Deb and Claudio
sure the identification of all the possible factors that could
influence alarm fatigue.
Patient care involves several tasks other than alarm mon-
itoring. These tasks could have influence on the monitoring
task and on the alarm fatigue as well. Both HTA for unit
clerks and nurses found task distribution as an important
matter to study and task priority as an important factor to
include in the list of alarm fatigue causing factors. Prior-
itizing tasks requires unit clerks and nurses to sometimes
respond in delay or even ignore other alarms which creates
dilemma, increases mental workload and causes fatigue.
3.3. Definitions of alarm fatigue
This study was designed to define alarm fatigue in terms
of mental workload using SWAT and NASA-TLX mea-
sures and in terms of three types of affect: boredom, ap-
athy, and distrust. In this regard, data was collected for
both unit clerks and nurses to define and measure alarm
fatigue by quantifying alarm fatigue causing factors (work-
ing conditions and staff individuality) and alarm fatigue
measures. Several statistical analyses were performed to
associate alarm fatigue measures with working conditions
and staff individuality.
The correlation analyses with alarm fatigue causing fac-
tors gave evidence that none of the considered variables
should be removed from further analyses; they all played a
role and needed to be considered in defining alarm fatigue
for each of the participants. Linear regression analyses in-
vestigating the significant factors in defining alarm fatigue
show acceptable models for nurses but not for unit clerks,
even after transformations for all five alarm fatigue mea-
sures. The attempt to conduct linear regression using the
components extracted from the PCA was also unsuccess-
ful. Finally, non-parametric, non-linear analyses produced
good models pertaining to unit clerks for all the alarm
fatigue measures. Additionally, better models were gener-
ated for mental workload measures and distrust in terms of
lower root mean square error (a measure of the differences
between values predicted by a model or an estimator and
the values actually observed) and higher R2values than
the linear regression models for nurses. Figure 3 shows the
models for all five alarm fatigue measures for both groups
of participants with pertinent R2values. These successful
models confirm that alarm fatigue can be defined and mea-
sured using any of these variables. The figure also shows
that in the case of unit clerks, SWAT measures give a better
model (R2=0.997) than the model for NASA-TLX mea-
sures (R2=0.916). This difference is not very prominent in
the case of nurses, which implies that SWAT gave mental
workload measurements similar to NASA-TLX. In addi-
tion, three types of affect give similar models in terms of
R2values.
Table 2 summarizes the influencing variables for two
alarm fatigue definitions for nurses and unit clerks. In addi-
tion, further post-hoc analyses evaluated the independent
association types (positive/negative) of these influencing
variables to alarm fatigue measures (presented in Figs. 4
and 5). For example, an increase in number of alarms would
increase unit clerks’ mental workload and affect (in term
of distrust). Therefore, systems should be designed more
precisely to reduce number of nuisance alarms which will
consequently reduce total number of alarms. In the case
of nurses, extraverted people get easily fatigued in terms
of mental workload and affect. Extraversion represents
the characteristics of being more talkative and interactive.
These people try to communicate with others frequently.
As alarm monitoring requires a high level of concentra-
tion, they get easily fatigued. Therefore, hospitals should
test the personality type before the recruitment of staff in
this kind of job. Thus, the approach of considering the
influencing factors along with their association types (pos-
itive/negative) will help define and measure alarm fatigue
and confirm the necessity of considering those factors in the
improvement of alarm monitoring systems in hospitals.
3.4. Significance of alarm fatigue on staff performance
The first approach of this study was to define alarm fa-
tigue in terms of mental workload and affect. Alarm fatigue
definitions showed the associations of alarm fatigue with
working conditions and staff individuality measures. The
second approach of this study attempted to observe the sig-
nificance of alarm fatigue measures on staff performance.
Based on the conceptual model of this study, several types
of analysis were performed to determine if alarm fatigue
measures deteriorate alarm monitoring performance. Fig-
ures 4 and 5 exhibit the proposed models to define alarm
fatigue and identify the cause of performance deterioration
for unit clerks and nurses, respectively. The dark lines show
successful models for performance measures. These figures
also include the type of influences (positive/negative) for
all the individual associations.
Figure 4 shows that non-parametric, non-linear regres-
sion analyses on alarm fatigue measures generated poor
models for response (yes/no) and response time for unit
clerks. The same analyses yielded acceptable results when
associating working conditions and staff individuality with
performance, specifically response time and number of ig-
nored alarms. Finally, an option which included a combi-
nation of working conditions, staff individuality, and the
five measure of alarm fatigue resulted in a better model for
all three performance measures.
Similarly, as shown in Fig. 5, non-parametric, non-linear
regression analyses on alarm fatigue measures for nurses
also generated poor models for response (yes/no) and re-
sponse time. On the other hand, when only considering
working conditions and individuality measures, the model
yielded good results for all three performance measures.
The same analyses were conducted using the combined
alarm fatigue measures and working conditions and indi-
viduality measures. The study found that using a combi-
nation of alarm fatigue measures, working conditions, and
Downloaded by [108.178.250.241] at 10:03 31 July 2015
Alarm fatigue 191
Fig. 3. Association of working conditions and staff individuality with mental workload and affect showing R2values for the models
inside arrows for both groups of participants.
individuality measures produces a useful model for deter-
mining response (yes/no) and number of ignored alarms for
nurses. However, a model considering only working condi-
tions and individuality resulted in a better fit for response
time. This can be explained in that it is the nurses’ duty
to give care to patients, whether they are fatigued or not.
Since alarms coming from patients’ rooms can make the
patients tense and uncomfortable, nurses try to respond to
all the alarms in the patient rooms as quickly as they can.
Besides, having multi-level alerting systems, nurses, who are
directly connected to the patients, can take some preventive
actions. Meyer and Bitan (2002) found that the operators
Table 2. Summary of influencing variables from two definitions for both participants
Influencing variables for
mental workload
Influencing variables for
affect
Influencing variables for
mental workload
Influencing variables for
affect
Unit clerk Nurse
Shift∗∗
Staff ID
Wor k l oa d∗∗
Number of alarms∗∗
Alarm type
Noise level
Time elapsed∗∗
Task priority
Extraversion∗∗
Agreeableness∗∗
Shift∗∗
Number of alarms∗∗
Wor k l oa d∗∗
Time elapsed∗∗
Extraversion∗∗
Openness
Agreeableness∗∗
Conscientiousness
Shift∗∗
Staff ID∗∗
Number of alarms∗∗
Wor k l oa d∗∗
Alarm type∗∗
Alarm criticality∗∗
Noise level∗∗
Time elapsed∗∗
Task priority∗∗
Extraversion∗∗
Openness∗∗
Agreeableness∗∗
Conscientiousness∗∗
Neuroticism∗∗
Shift∗∗
Staff ID∗∗
Number of alarms∗∗
Wor k l oa d∗∗
Alarm type∗∗
Alarm criticality∗∗
Noise level∗∗
Time elapsed∗∗
Task priority∗∗
Extraversion∗∗
Openness∗∗
Agreeableness∗∗
Conscientiousness∗∗
Neuroticism∗∗
∗∗ Indicates the factors common to mental workload and affect.
Downloaded by [108.178.250.241] at 10:03 31 July 2015
192 Deb and Claudio
Fig. 4. Proposed model for unit clerks.
Downloaded by [108.178.250.241] at 10:03 31 July 2015
Alarm fatigue 193
Fig. 5. Proposed model for nurses.
Downloaded by [108.178.250.241] at 10:03 31 July 2015
194 Deb and Claudio
who take some kind of preventive actions respond less to
warnings because they know that those alarms are false. On
the other hand, the operators, who take fewer preventive ac-
tions, respond more quickly to warnings, which in their case
are more likely to be an actual hazard. Preventive actions
increase the trust in system outcomes and operators delay
their responses as a strategy of avoiding fatigue (Bolton,
G¨
oknur, and Bass, 2013).
Before this study, most studies measured alarm fatigue
using response time. However, response time is not a mea-
sure of alarm fatigue but rather it is an outcome influenced
by alarm fatigue, not alarm fatigue itself. In addition, pre-
vious studies did not differentiate between nurses and unit
clerks. One of the most important findings of this is the
fact that delayed response time for nurses does not seem
to be caused by alarm fatigue, but rather by the working
conditions and staff individuality. This is important for two
reasons: (i) until now researchers blamed delayed response
time on alarm fatigue, which as shown in this study cannot
be claimed, and (ii) because of this false assumption, what
has been measured is, in fact, not alarm fatigue. In order
to find the best association of alarm fatigue measures with
staff performance deterioration, this study proposes two
successful models, one for unit clerks and one for nurses.
4. Conclusions
This study began with the intent of defining and measur-
ing alarm fatigue in terms of all the possible influencing
factors. These factors were gleaned from the literature re-
view and Hierarchical Task Analyses and included working
conditions and staff individuality. Results from the HTA
showed the need to separate nurses and unit clerks which
other reported studies do not differentiate.
Data were then collected in an Intensive Care Unit, us-
ing surveys, questionnaires and observations to quantify
the alarm fatigue factors, alarm fatigue measures (mental
workload and emotional affects), and staff performance.
Several failed attempts to associate alarm fatigue and per-
formance through various regression analyses revealed that
the original conceptual model is invalid and that alarm fa-
tigue alone is not the cause of deterioration in staff per-
formance, as has been strongly purported until now. The
role of alarm fatigue in the deterioration of staff perfor-
mance must, therefore, be reevaluated. Until now, the gen-
erally held view has been that alarm fatigue is the cause
of performance deterioration (i.e., delayed response time).
Hence, until now, researchers have measured those per-
formance metrics and then blamed alarm fatigue for their
deterioration. The problem with this approach is that the
performance metrics such as delay in response are not
measures of alarm fatigue but rather consequences. This
research study measured alarm fatigue through mental
workload and affect and then studied its influence on staff
performance.
The results show that performance deterioration is ac-
tually the effect of a combination of alarm fatigue with
working conditions and staff individuality and in the case
of nurses and response time, alarm fatigue plays no role,
only working conditions and staff individuality. As care-
givers’ performance cannot be deteriorated only by alarm
fatigue, the performance measures should not be used for
defining fatigue induced by alarms. For example, for nurses,
response to alarms or number of ignored alarms is highly
dependent on some work environment and staff personal-
ity factors along with alarm fatigue. Their responses are
primarily based on workload, time elapsed, noise level,
personality build from experience along with their mental
workload and affect. Hence, performance measure degra-
dation is mainly the result of a combined influence of alarm
fatigue, working conditions, and staff individuality. How-
ever, response time would be applied to the alarms for which
the staff had taken some actions. Usually nurses take most
of the actions on the alarms which they find risky to the pa-
tients, on patient calls, and also on nuisance alarms to stop
those repeated alarms. These variations in actions (from
risky to nuisance) cannot be connected directly to only
alarm fatigue or even to the combination of fatigue along
with the working conditions and staff individuality. There-
fore, the approach of previous studies in defining alarm
fatigue as response time to alarms, in the case of healthcare
arena should be reconsidered.
The study was conducted in an acute care area. There-
fore, several limitations were experienced during research
work to ensure patient safety. The presence of researchers
(two at a time) may have interrupted the usual alarm moni-
toring tasks. In addition, this study was focused on one In-
tensiveCareUnit(ICU)withthepresenceofunitclerksand
a second monitor for patients from other medical floors.
This scenario may not be representative of other ICUs. For
this reason, the findings of this study may not be general-
ized outside of this particular setting.
Nevertheless, this study opens up possibilities for future
research into defining and measuring alarm fatigue and its
influence on staff performance. The definitions of alarm fa-
tigue proposed by this study can be validated by conducting
further research on the acute care areas of the same hos-
pital, as well as on several acute care departments of other
hospitals. In addition, physiological measures can be used
along with mental workload and affects to produce a com-
plementary definition of alarm fatigue.
References
Abdellah, F., and Levine, E. (1954) Work-sampling applied to the study
of nursing personnel. Nursing Research,3(1), 11–16.
Ahmed, S. M. (1990) Psychometric properties of the Boredom Proneness
Scale. Perceptual and Motor Skills,71(3), 963–966.
Ashton, J. (2011) “Alarm fatigue” blamed in hospital’s deaths.
http://www.cbsnews.com/news/alarm-fatigue-blamed-in-
hospital-deaths/ (accessed October 24, 2014).
Downloaded by [108.178.250.241] at 10:03 31 July 2015
Alarm fatigue 195
Association for the Advancement of Medical Instrumentation. (2011) A
siren call for action. Summit Publications: Clinical Alarms, Arling-
ton, VA
Baillargeon, E. (2013) Alarm fatigue: A risk assessment. Rhode Is-
land College: Master of Science in Nursing. http://digitalcommons.
ric.edu/school of nursing/23 (accessed October 24, 2014).
Billinghurst, F., Morgan, B., and Arthur, H. M. (2003) Patient and nurse-
related implications of remote cardiac telemetry. Clinical Nursing
Research,12(4), 356–370.
Bliss, J. P., Gilson, R. D., and John, E. D. (1995) Human probability
matching behaviour in response to alarms of varying reliability.
Ergonomics,38(11), 2300–2312.
Bolton, M. L., Goknur, S., and Bass, E. J. (2013) Framework to sup-
port scenario development for human-centered alerting system
evaluation. IEEE Transactions on Human-Machine Systems,43(6),
595–607.
Borowski, M., Gorges, M., Fried, R., Such, O., Wrede, C., and Imhoff,
M. (2011) Medical device alarms. Biomed Tech (Berl),56(2),
73–83.
Chambrin, M. C., Ravaux, P., Calvelo-Aros, D., Jaborska, A., Chopin, C.,
and Boniface, B. (1999) Multicentric study of monitoring alarms in
the adult intensive care unit (ICU): A descriptive analysis. Intensive
Care Medicine,25(12), 1360–1366.
Chien, S.-Y., Semnani-Azad, Z., Lewis, M., and Sycara, K. (2014) To-
wards the development of an inter-cultural scale to measure trust in
automation. 16th International Conference on Human-Computer
Interaction, pp. 35–46. Heraklion, Crete, Greece.
Clark, T., and David, Y. (2011) Advancing the Safety of Healthcare
Technology. http://www.who.int/medical devices/global forum/
D05.pdf?ua=1
Clinical Alarms Task Force. (2007) Impact of clinical alarms on patient
safety: A report from the American College of Clinical Engineering
Healthcare Technology Foundation.Journal of Clinical Engineering,
32(1), 22–33. http://thehtf.org/white%20paper.pdf
Crites, A. (2013) Too much noise from hospital alarms poses risk for
patients. Health & Science. http://www.washingtonpost.com/sf/
feature/wp/2013/07/07/too-much-noise-from-hospital-alarms-
poses-risk-for-patients/
Cvach, M. (2012) Monitor Alarm Fatigue: An Integrative Review. AAMI,
Baltimore.
Dickson, W. M. (1978) Measuring pharmacist time use: a note on the
use of fixed-interval work sampling. American Journal of Hospital
Pharmacy,35(10), 1241–1243.
ECRI. (1974) Hazard: Gaymar hypothermia machine. Health Devices,
3(9), 229–230.
ECRI. (2011) Top ten health technology hazards for 2012. Health
Devices. https://www.ecri.org/Documents/secure/Health Devices
Top 10 Hazards 2012.pdf
ECRI. (2012) Top Ten health technology hazards for 2013. Health
Devices. https://www.ecri.org/Documents/Secure/Health Devices
Top 10 Hazards 2013.pdf
ECRI. (2013) Top ten health technology hazards for 2014. Health
Devices. https://www.ecri.org/Documents/Secure/Health Devices
Top 10 Hazards 2014.pdf
ECRI. (2014) Top Ten health technology hazards for 2015. Health
Devices,43(11). www.ecri.org/2015hazards
Eggemeier, F. T. (1988) 1988, Properties of workload assessment tech-
niques. In P. A. Hancock, and N. Meshkati, Human Mental Work-
load (pp. 41–62). North-Holland, Amsterdam.
Farmer, R., and Sundberg, N. D. (1986) Boredom proneness—The de-
velopment and correlates of a new scale. Journal of Personality As-
sessment,50(1), 4–17.
Godefroy, O. (2013) The Behavioral and Cognitive Neurology of Stroke.
Cambridge University Press.
Goldberg, L. R. (1981) Language and individual differences: The search
for universals in personality lexicons. Review of Personality and So-
cial Psychology,2(1), 141–165.
Graham, K., and Cvach, M. (2010) Monitor alarm fatigue: Standardiz-
ing use of physiological monitors and decreasing nuisance alarms.
American Journal of Critical Care,19(1), 28–34.
Griscomb, C. (1985) Survey of SWAT Use in Flight Test.BDMCorp.,
Albuquerque.
Harrison, D. M., and Chervany, N. L. (2001) Trust and distrust def-
initions: One bite at a time. In R. Falcone, M. Singh, and T.
Y-H (Eds. ) , Trust in Cyber-Societies (pp. 27–54). Springer, Berlin
Heidelberg.
Haworth, L. A., Bivens, C. C., and Shively,R. J. (1987) An investigation of
single-piloted advanced cockpit and control configuration for nap-
of-the-earth helicopter mission tasks. Proceedings of the 42nd of the
Annual Forum of the American Helicopter Society, (pp. 657–671).
California.
Hogg, M. A., Abrams, D., and Martin, G. N. (2013). Social condition
and attitudes. In G. N. Martin, N. R. Carlson, and W. Buskist,
Psychology. Pearson Education Limited, Harlow.
Hsieh, C. J., Chu, H., Cheng, J. J., Shen, W. W., and Lin, C. C. (2012)
Validation of apathy evaluation scale and assessment of severity of
apathy in Alzheimer’s disease. Psychiatry and Clinical Neurosciences,
66(3), 227–234.
Kerr, J. H., and Hayes, B. (1983) An “alarming” situation in the intensive
therapy unit. Intensive Care Medicine,9(3), 103–104.
Konkani, A., and Oakley, B. (2012) Noise in hospital intensive care
units—a critical review of a critical topic. Journal of Critical Care,
27(5), 522e1–522e9.
Lakin, K. P. (2007) HCBS recipients are increasingly likely to live with
parents or other relatives. Intellectual and Developmental Disabili-
ties,45(5), 659–661.
Lawless, S. (1994) Crying wolf: False alarms in a pediatric intensive care
unit. Critical Care Medicine,22, 981–985.
Liptak, G. S., Super, D. M., Baker, N., and Roghmann, K. J. (1985).
An analysis of waiting times in a pediatric emergency department.
Clinical Pediatrics (Phila),24(4), 202–209.
Lueken, U., Seidl, U., V¨
olker, L., Schweiger, E., Kruse, A., and
Schr¨
oder, J. (2007) Development of a short version of the Ap-
athy Evaluation Scale specifically adapted for demented nursing
home residents. The American Journal of Geriatric Psychiatry,15(5),
376–385.
Luximon, A., and Goonetilleke, R. S.(1998) Continuous subjective work-
load assessment technique. Ergonomics for Global Quality and Pro-
ductivity.
Marin, R. S., Biedrzycki, R. C., and Firinciogullari, S. (1991) Reliability
and validity of the Apathy Evaluation Scale. Psychiatry Research,
38(2), 143–162.
Meyer, J., and Bitan, Y. (2002) Why better operators receive worse warn-
ings. Human Factors,44(3), 343–353.
Mitka, M. (2013) Joint Commission warns of alarm fatigue: Multitude
of alarms from monitoring devices problematic. JAMA,309(22),
2315–2316.
Nygren, T. E. (1982) Conjoint Measurement and Conjoint Scaling: A
User’s Guide. Air Force Aerospace Medical Research Lab, Wright-
Patterson AFB, OH.
Pishori, T. (2012) A stochastic simulation model of alarm response
strategies on a telemetry floor. (Masters of Science). Paper 256.
University of Connecticut. http://digitalcommons.uconn.edu/gs
theses/256/
Reid, G. B., and Nygren, T. E. (1988) The subjective workload assessment
technique: A scaling procedure for measuring mental workload.
Advances in Psychology,52, 185–218.
Reid, G. B., Eggemeier, F. T., and Shingledecker, C. A. (1982) Pro-
ceedings of the workshop on flight testing to identify pilot work-
load and pilot dynamics. In M. L. Frazier, and R. B. Crom-
bie. Subjective Workload Assessment Technique Edwards AFB,
CA.
Reid, G. B., Potter, S. S., and Bressler, J. (1989) Subjective Workload
Assessment Technique (SWAT): A user’s guide. WrightPatterson Air
Downloaded by [108.178.250.241] at 10:03 31 July 2015
196 Deb and Claudio
Force Base, OH: Harry G. Armstrong Aerospace Medical Research
Laboratory.
Reid, R. A. (1975) A work sampling study of midlevel health professionals
in a rural medical clinic. Medical Care,13(3), 241–249.
Sendelbach, S., and Funk, M. (2013) Alarm fatigue: a patient safety
concern. AACN Advanced Critical Care,24(4), 378–386.
Stafford, T. (1982). Whether monitoring? Critical Care Medicine,10(11),
792–795.
The Joint Commission. (2013) Sentinel Event Alert. (50). http://
www.jointcommission.org/assets/1/18/sea 50 alarms 4513
final1.pdf
The Joint Commission. (2014) The Joint Commission. https://
www.premierinc.com/safety/topics/patient safety/index 3.jsp
(accessed March 8, 2014).
Tsien, C. L., and Fackler, J. C. (1997) Poor prognosis of existing monitors
in the intensive care unit. Critical Care Medicine,25(4), 614–619.
Vodanovich, S. J., Wallace, J. C., and Kass, S. J. (2010) A confirmatory
approach to the factor structure of the Boredom Proneness Scale:
Evidence for a two-factor short form. Journal of Personality Assess-
ment,85(3), 295–303.
Wallis, L. (2010, April 3) Alarm fatigue linked to patient death. American
Journal of Nursing,110(7), 16.
Weil, K. M. (2009) Alarming monitor problems. Nursing,39(9), 58.
Weinger, M. B., and Smith, N. T. (1993) Vigilance, alarms and integrated
monitoring systems. In Ehrenwerth, J. and Eisenkraft, J. (eds): Anes-
thesia Equipment: Principles and Applications. Mosby, St. Louis,
MO.
Welch, J. (2011) An evidence-based approach to reduce nuisance alarms
and alarm fatigue. Biomedical Instrumentation & Technology: Alarm
Systems,45(1), 46–52.
Wright, M. J. (1954) Improvement of Patient Care. G. P. Putnam’s Sons,
New York.
Downloaded by [108.178.250.241] at 10:03 31 July 2015
