RESEARCH
ED SERVICES: THE IMPACT OF CARING BEHAVIORS
ON PATIENT LOYALTY
Authors: Sandra S. Liu, PhD, David Franz, MHA, FACHE, Monette Allen, RN, MSN, En-Chung Chang, PhD (c),
Dana Janowiak, MSHA, Patricia Mayne, RN, BSN, MHA, CEN, and Ruth White, RN, BSN,
Lafayette, IN, St Louis, MO, Kokomo, IN, Birmingham, AL, and Detroit, MI
Earn Up to 9.5 CE Hours. See page 515.
Introduction: This article describes an observational study of
caring behaviors in the emergency departments of 4 Ascension
Health hospitals and the impact of these behaviors on patient
loyalty to the associated hospital. These hospitals were diverse in
size and geography, representing 3 large urban community
hospitals in metropolitan areas and 1 in a midsized city.
Methods: Research assistants from Purdue University (West
Lafayette, IN) conducted observations at the first study site and
validated survey instruments. The Purdue research assistants trained
contracted observers at the subsequent study sites. The research
assistants conducted observational studies of caregivers in the
emergency departments at 4 study sites using convenience sampling
of patients. Caring behaviors were rated from 0 (did not occur) to 5
(high intensity). The observation included additional information, for
example, caregiver roles, timing, and type of visit. Observed and
unobserved patients completed exit surveys that recorded patient
responses to the likelihood-to-recommend (loyalty) questions, patient
perceptions of care, and demographic information.
Results: Common themes across all study sites emerged,
including (1) the area that patients considered most important
atient satisfaction is increasingly important in the
P health care sector. For example, the Consumer Assess-
ment of Healthcare Providers and Systems (CAHPS)
Hospital Survey, also known as Hospital CAHPS, or
HCAHPS, was developed by the Centers for Medicare &
Medicaid Services and is a standardized instrument for mea-
suring patients’ perspectives of hospital care.1 Voluntary data
Sandra S. Liu is Professor of Consumer Sciences and Retailing, Purdue
University, West Lafayette, IN.
David Franz is Manager, Ascension Health, St Louis, MO.
Monette Allen is Manager of Emergency Services, St Joseph Hospital, Kokomo, IN.
En-Chung Chang is PhD Candidate, Consumer Sciences and Retailing, Pur-
due University, West Lafayette, IN.
Dana Janowiak is Business Manager, Medical Services, St Vincent’s Birming-
ham, Birmingham, AL.
Patricia Mayne, Member, Huron Valley Chapter, is Administrative Director,
Emergency Services, St John Hospital and Medical Center, Detroit, MI.
404 JOURNAL OF EMERGENCY NURSING
to an ED experience (prompt attention to their needs upon
arrival to the emergency department); (2) the area that patients
rated as least positive in their actual ED experience (prompt
attention to their needs upon arrival to the emergency
department); (3) caring behaviors that significantly affected
patient loyalty (eg, making sure that the patient is aware of
care-related details, working with a caring touch, and making
the treatment procedure clearly understood by the patient); and
(4) the impact of wait time to see a caregiver on patient loyalty.
A number of correlations between caring behaviors and patient
loyalty were statistically significant (P < .05) at all sites.
Discussion: The study results raised considerations for ED
caregivers, particularly with regard to those caring behaviors that
are most closely linked to patient loyalty but that occurred least
frequently. The study showed through factor analysis that some
caring behaviors tended to occur together, suggesting an
underlying, unifying dimension to that factor.
Key words: Caring behaviors; Patient satisfaction; Patient
loyalty
collection using this instrument began in 2006, with the
goal of providing data for consumers to compare hospitals
in judging the patient experience of hospital care. Previous
studies have shown that physician and nurse interpersonal
interactions impact patients’ satisfaction and loyalty rat-
ings. As physicians or nurses exhibit more caring behaviors
including social connectedness,2 nonverbal expressions,3
Ruth White, Member, Metro Birmingham Chapter, is Emergency Department
Nurse Manager, St Vincent’s Birmingham, Birmingham, AL.
For correspondence, write: David Franz, MHA, FACHE, 4600 Edmundson
Rd, St. Louis, MO 63134; E-mail: dfranz@ascensionhealth.org.
J Emerg Nurs 2010;36:404-14.
Available online 23 February 2010.
0099-1767/$36.00
Copyright © 2010 by the Emergency Nurses Association. Published by
Elsevier Inc. All rights reserved.
doi: 10.1016/j.jen.2009.05.001
VOLUME 36 • ISSUE 5 September 2010

FIGURE 1
Ascension Health patient experience map. This patient experience map was developed through Ascension Health research, and the analysis shows the 6 most important
attributes that comprise the patient experience. These attributes fall into 3 realms.
and verbal communication skills,4,5 patient satisfaction
and/or likelihood-to-recommend scores increase. However,
there is limited research on which specific caring behaviors
are most important to a positive experience and patient
loyalty. Ascension Health (St Louis, MO) determined that
this relatively unstudied element of patient care was essen-
tial to providing a consistent, exceptional patient experi-
ence. Caring behaviors align with Ascension Health’s
values, support its mission of providing holistic care, and
contribute to improving patient loyalty, which is increas-
ingly important to health care organizations operating in
competitive markets. Caring behaviors have also been
shown to improve quality of care. Studies of emotional
care indicate that it fosters patients’ feelings of safety
and reduces anxiety6 and enables rapport building and
connection development between caregivers and patients,7
thereby benefiting long-term quality improvement and
patient-caregiver relationships.8 The purpose of this study
was to identify the impact of health care providers’ caring
behaviors that lead to patient loyalty. Emergency depart-
ments were studied because (1) the length of stay and
service structure of emergency departments are conducive
to observational study and (2) emergency departments are
a vital “front door” to most health care institutions, ser-
September 2010 VOLUME 36 • ISSUE 5
Liu et al/RESEARCH
ving a high volume of patients, many of whom are
admitted as inpatients.
Ascension Health completed extensive, unpublished
research in June 2006 that included a survey of approxi-
mately 2,000 patients from 6 Ascension Health hospitals
to define the experience desired by patients and their
families. The results were used to develop a patient
experience map that defined 6 attributes within 3 realms
(Figure 1).
In 2007 Ascension Health chose the Net Promoter
Score (Satmetrix Systems, Inc., Bain & Company, Fred
Reicheld) as the overarching metric to track Ascension
Health’s success in delivering a positive patient experience
(Figure 2). This loyalty metric, based on the familiar “like-
lihood-to-recommend” question, was chosen based on sig-
nificant research that showed a correlation of this metric
with market growth across other industries.9,10 Patient loy-
alty also is increasingly important to payers and a part of
most surveys, including HCAHPS.
On the Ascension Health experience map, realm 3
attributes are more highly correlated with the patient’s like-
lihood to recommend than realm 1 and 2 attributes. The
realm 3 attributes included (1) compassionate, respectful
care; (2) care responsiveness; and (3) communication and
WWW.JENONLINE.ORG 405
RESEARCH/Liu et al
FIGURE 2
Calculating Net Promoter Score (NPS). The method for determining the NPS,
based on the willingness-to-recommend question used in a number of
industries, is shown. NPS is calculated by taking the percentage of promoters
(individuals giving a rating of 9 or 10) and subtracting the percentage of
detractors (individuals giving a rating of ≤6). Reprinted with permission from
Bain & Co, Boston, MA.
empowerment. Thus this study of emergency departments
focused on behaviors related to these 3 attributes.
Methods
Study sites included a diverse mix of emergency depart-
ments in both size and geography and included the follow-
ing: St Joseph Hospital in Kokomo, IN, a 167-bed
community hospital with 23,000 annual visits to its emer-
gency department; St John Hospital and Medical Center in
Detroit, MI, a 658-bed urban teaching hospital with
98,000 annual visits to its level II trauma center; St Vin-
cent’s Birmingham in Birmingham, AL, a 372-bed com-
munity/urban hospital with 42,000 annual visits to an
emergency department that meets criteria for level III trau-
ma; and Providence Hospital in Washington, DC, an
urban hospital with 226 acute care beds and 47,000 annual
ED visits. Each hospital’s institutional review board
approved the study for its site.
Funded by Ascension Health, researchers from the
Department of Consumer Sciences and Retailing, Purdue
University (West Lafayette, IN), designed the study and
developed the study instruments based on literature review
and consultation with medical professionals. The research
team piloted the study at the emergency department of a
98-bed community hospital, validated the study instru-
ments, and traveled to study sites to extensively train con-
tracted observers using videos of role-playing on a range of
patient-caregiver interaction situations. Inter-rater reliability
was measured at 0.86 to 0.99 by use of correlation analysis,
with inter-rater consistencies found to be satisfactory with
reliabilities ranging from 0.86 to 0.99. The general process
used at all sites is depicted in Figure 3, using convenience
sampling of observation subjects. Patients at triage were
assessed to determine whether they met criteria to be
included in the study. Those patients who met the criteria
were asked to participate and sign a consent form, which
was cosigned by an emergency nurse to witness the consent.
406 JOURNAL OF EMERGENCY NURSING
The patients were observed until discharge or for a mini-
mum of 3 hours. The caregivers were not blinded to the
study; each caregiver interacting with the observed patient
was rated on a separate coding sheet, identified by caregiver
category. At discharge, each patient was asked to complete an
exit survey. The study also collected a targeted number of
exit surveys completed by unobserved patients from 3 of
the 4 sites to determine whether observation biased the find-
ings. Unobserved patients were given an exit survey to deter-
mine whether observers impacted the perception of caring
behaviors and overall loyalty ratings and to provide a more
robust data set for rating unobserved dimensions of care.
Two instruments were used to collect data for the study:
• The coding sheet was used by the research assistants to
record the caring behaviors for each encounter. The scal-
ing of behavior intensity ranged from 0 (did not occur)
to 5 (high intensity). The coding sheet also included
additional information such as caregiver roles, timing
information, and type of visit (Appendix Figure 1).
• The exit survey was completed by the patient at the
end of the visit and was used for both observed and
unobserved patients (Appendix Figure 2). It recorded
(1) the likelihood-to-recommend (loyalty) questions,
(2) the degree to which patients appreciated caring
behaviors with their top 3 important areas ranked,
and (3) demographic information.
The primary analysis of interest was a correlation analy-
sis of observed behaviors and patient loyalty measured by use
of 4 questions applied to the various caregiver categories.
Other analyses included correlating observed behaviors with
patient perceptions and time-stamp information to measure
wait time for care. Time-stamp information was collected for
observed patients including time from arrival to triage, from
triage to see a registered nurse, and from triage to see a phy-
sician. These data were also correlated with patient loyalty.
Patient and caregiver demographics were also studied. Factor
analysis was applied to data at all sites to determine under-
lying behavior dimensions that impact patient loyalty.
The following were key parameters of the study:
• The study was conducted at 4 emergency departments
and included a total of 728 observed patients and 619
unobserved patients. Of the observed patients, there
was a range of 150 to 200 observed at each site, with
a minimum target of 150 per site.
• The study was conducted from January to July 2007,
with each site conducting a 3- to 4-month study within
that time frame.
• Patient demographics varied significantly by site, pro-
viding a rich and diverse data set for the study.
VOLUME 36 • ISSUE 5 September 2010

FIGURE 3
General process for observed patients. The map shows the key steps involved in observing patients and the criteria used to determine whether to observe patients.
Results
The study findings provided meaningful information in a
number of important areas relating to an emergency
department’s ability to foster patient loyalty. Overall, the
study findings indicated that caring behaviors do have an
impact on patient loyalty, as evidenced by a number of sta-
tistically significant correlations (P < .05) (Table 1). The
data showed strong statistical relationships between
observed intensities or absence of caring behaviors and
patient loyalty across a diverse sample (Table 2). Making
sure the patient is aware of care-related details, working
with a caring touch, and making the treatment procedure
clearly understood by the patient showed the strongest cor-
relations with patient loyalty. Furthermore, differences in
patient loyalty scores and experience ratings between
observed and unobserved patients were insignificant. A
review of Press Ganey patient satisfaction surveys11 used
at 3 of the 4 sites showed consistency with study findings;
priority areas (priority index) based on correlation with
overall satisfaction on the Press Ganey survey were consis-
September 2010 VOLUME 36 • ISSUE 5
Liu et al/RESEARCH
tent with many of the nurse and physician behaviors
found to be significantly correlated in the study. A num-
ber of caring behaviors with significant correlations with
loyalty across sites showed some common patterns, but
there was significant variation among the 4 study sites in
the caring behaviors by caregiver role that impacted loy-
alty. Though not the primary focus of the study, wait time
to see caregivers had a significant impact on patient loy-
alty. It was reported as the most important area to patients
yet was consistently ranked as the least positive area in
their perceived experience. With regard to the caring beha-
viors that affected loyalty, there were differences based on
patient gender but not based on patient race or age. At 2
sites, loyalty was correlated with information-based caring
behaviors for male patients, whereas for female patients,
loyalty was correlated with relationship-based caring beha-
viors. Although the impact on loyalty varied among sites,
the pattern of the frequency of observed caring behaviors
by caregiver role was consistent among study sites, with
physicians exhibiting a greater number and intensity of
caring behaviors than emergency nurses.
WWW.JENONLINE.ORG 407
RESEARCH/Liu et al

TABLE 1
Correlation coefficients between caring behaviors and loyalty
Absent percentage
Caring behavior Hospital site of caring behavior Intensity of caring behavior
Introduce himself or herself A, C −0.165 and −0.220
a b

Show concern for family member’s needs A −0.321a

and consideration
Address family member’s questions patiently A, C −0.385b and −0.451b
Positive facial expressions A, C 0.195a and 0.155a
Chat with patient to help tune out illness state B −0.193 b

Respect patient’s privacy B 0.153a

Make sure patient is aware of care-related detail B, C 0.157a and 0.273b
before exiting
Clearly communicate information to patient C 0.243b
Explain to patient what to expect next C 0.214b
Ask whether there is anything else that is needed C 0.183a
before leaving
Speak with a caring tone C 0.179a
Explain purpose of visit C 0.175a
Actively respond to patient’s need C 0.160a

A, St. Joseph Hospital; B, St. Vincent's Birmingham; C, Providence Hospital.

This table depicts correlations of caring behaviors and loyalty measured by the mean score of 2 measurements: “the likelihood of recommending the emergency department to
your family and friends” and “the likelihood of recommending the hospital to your family and friends.” Caring behaviors for 1 site are not significantly correlated with the
index of loyalty in the aggregate level.
a
P < .05.
b
P < .01.

Through factor analysis, 3 main factors emerged
through the combined data set at all 4 sites. Certain caring
behaviors within each of these factors tended to occur
together, although there was no causal relationship among
them. Nonetheless, there was an underlying dimension
that unified the behaviors that occurred within a factor.
The 3 factors and some of the associated behaviors were
as follows:
• Care concern and communication, which included
making sure the patient was aware of care-related
details, proactively sharing health care knowledge,
explaining what to expect next, making the treatment
procedure clearly understood by the patient, and show-
ing concern for non–treatment-related details, such as
the patient’s family and job
• Body language, which included facial expression, tone of
voice, eye contact, active listening, respect for privacy, lean-
ing toward the patient, and working with a caring touch
• Initial greetings, which included caregivers introducing
themselves and addressing the patient by name
COMMON THEMES ACROSS ALL STUDY SITES
Despite variations in the study results, the following signif-
icant common themes emerged through descriptive, corre-
lation, and factor analyses:
1. Areas rated as most important by patients, with regard to
their ED experience
• Patients at all 4 study sites rated prompt attention to
needs upon arrival to the emergency department as
most important.
• Explaining the patient’s condition was ranked second
in importance.
• Depending on the study site, either making the patient
feel comfortable or offering warm greetings ranked
third in importance.
2. Areas patients rated least positive in their actual ED experience
• There was consistency at all sites in what patients rated
as least positive on their surveys: prompt attention to
needs upon arrival to the emergency department.
• Two areas—inquiring about patient fears and concerns
and making the patient feel comfortable—accounted

408 JOURNAL OF EMERGENCY NURSING VOLUME 36 • ISSUE 5 September 2010

 Liu et al/RESEARCH

4. Impact of wait time on patient loyalty

TABLE 2
• The area with the most significant gap in meeting
Patient demographics
patient expectations was wait time. This was based
Observed cases Unobserved cases on patient-ranked areas of importance, the perceived
(n = 728) (%) (n = 619) (%)
experience, and the high correlation of wait time and
Gender patient loyalty. Although most emergency departments
Male 274 (37.6) 242 (39.1) are aware of research that indicates the positive impact
Female 454 (62.4) 377 (60.9) of reducing wait times on patient satisfaction,12,13 this
Age study suggested that it is most important to reduce the
<25 y 124 (17.0) 109 (17.6) time patients wait to see a physician or a primary emer-
25-35 y 167 (22.9) 123 (19.9) gency nurse after being triaged.
36-49 y 208 (28.6) 198 (32.0) • Time-stamp data also showed that wait time signifi-
cantly impacted patient loyalty at all sites.
50-65 y 150 (20.6) 129 (20.8)
>65 y 79 (10.9) 60 (9.7)
Discussion
Primary health insurance
Medicare 128 (17.6) 104 (16.8) The findings of this study have implications for emergency
Medicaid 123 (16.9) 123 (19.9) departments:
Private insurance 369 (50.7) 320 (51.7) • Supported patient experience research: These findings
Self-pay 108 (14.8) 72 (11.6) were consistent with other data linked to patient satis-
Ethnicity faction. For example, the Press Ganey Emergency
White 289 (39.7) 147 (23.7) Department Pulse Report 2008 cited survey items
African American 387 (53.2) 439 (70.9) “How well were you kept informed about delays”
Hispanic 30 (4.1) 21 (3.4) and “Degree to which staff cared about you as a per-
Asian 4 (0.5) 2 (0.3) son” as the number 1– and number 2–ranked priorities
of patients that correlated to patient ratings of “likeli-
Other 18 (2.5) 10 (1.6)
hood of your recommending our emergency depart-
ment to others.”14
• Created a framework of caring behaviors: Although the
degree to which specific behaviors correlated with loy-
alty varied by site, factor analysis produced a distinct
for patients’ second and third least positive–rated areas, framework of the 3 factors referenced previously: care
although the order varied site to site. concern and communication, body language, and initial
3. Caring behaviors with significant impact on patient loyalty greetings. This framework can be used to develop and
• On the basis of the number of significant correlations at prioritize interventions, because certain caring beha-
all sites between caring behaviors and patient loyalty, sta- viors tended to occur in the presence of other caring
tistically significant behaviors emerged for all caregiver behaviors. Focusing on a specific factor and exploring
roles. Most of these behaviors were grouped in the same ways to promote that factor may reveal additional caring
factor category of care concern and communication: behaviors that foster patient loyalty but that were not
• Make the treatment procedure clearly understood identified in the study. For example, a focus on care
by the patient. Patients expected this behavior concern and communication could point to an under-
and responded with loyalty only when it was done lying dimension of caregiver-patient dialogue. Explor-
well. If the behavior was absent, patient loyalty ing ways to enhance this dimension in the ED setting
decreased. could lead to new caring behaviors that relate to the
• Show concern for non–treatment-related details. physical, emotional, and/or spiritual care of the patient.
This was an expected behavior, and if not present, • Identified caring behaviors focus areas: Although the
loyalty decreased. correlations between patient loyalty and specific caring
• Make sure the patient is aware of care-related behavior varied by site, a summary table of correlations
details, and explain what to expect next. These were for all sites showed several behaviors that had a high
2 behaviors that significantly increased patient loy- number of correlations with patient loyalty. Using
alty when they were performed well. these data in conjunction with an aggregate frequency

September 2010 VOLUME 36 • ISSUE 5 WWW.JENONLINE.ORG 409

RESEARCH/Liu et al
FIGURE 4
Correlations to loyalty and frequency of occurrence. Depicted are 8 behaviors that highly correlate to loyalty, but they were not frequently exhibited during
patient observations.
of behavior occurrence for all sites showed potential
areas of focus for all emergency departments (Figure
4). The study showed that the following behaviors had
the highest number of correlations to patient loyalty com-
bined with the lowest frequency of occurrence. The first 5
of these 8 behaviors fell under the factor category of care
concern and communication. Because these 5 were among
those behaviors most linked to loyalty, increasing their fre-
quency could significantly improve patient loyalty scores.
• Make sure the patient is aware of care-related details.
• Explain to the patient what to expect next.
• Make the treatment procedure clearly understood by
the patient.
• Chat with the patient to help tune out the illness state.
• Show concern for non–treatment-related details.
• Respect the patient’s privacy.
• Work with a caring touch.
• Make sure the family is aware of care-related details.
Limitations
As with many studies, a number of issues were identified
that could strengthen future investigations of caring beha-
viors in the emergency department and their impact on
patient loyalty:
1. Patients with a triage acuity level of 1 or 2 by use of the
5-level Emergency Severity Index were excluded from the
study because of the difficulties in observing such
410 JOURNAL OF EMERGENCY NURSING
patients, who typically require more intensive medical
attention at a more accelerated pace. Ideally, a future
study would include such patients to determine whether
the same caring behaviors lead to patient loyalty among a
population that is more acutely ill.
2. The observation periods were limited to the schedules of
the research assistants, roughly 7 AM to 11 PM or midnight.
At times, an observation was aborted because the patient
visit outlasted the research assistant’s schedule. Thus a
future study would benefit from 24-hour observation, both
to ensure an accurate reflection of the entire 24-hour cycle
of ED admissions and to ensure that observations that were
begun could be brought to completion.
3. Although the study sites were diverse, not all populations/
regions were captured. A larger, more inclusive study would
capture demographic subsets and determine whether the same
caring behaviors lead to patient loyalty among these subsets.
4. The exit surveys of observed patients and unobserved patients
were conducted at different times, largely because of logistics.
Ideally, a future study would conduct both simultaneously.
Implications for Emergency Nurses
Emergency departments planning to develop goals and
strategies for improving the patient experience in the emer-
gency department should consider these points:
1. Focus initially on those caring behaviors that were identi-
fied as correlating strongly with loyalty yet were not con-
VOLUME 36 • ISSUE 5 September 2010

sistently displayed (Figure 4). The loyalty-linked behaviors
that ED caregivers neglected offer the greatest potential for
increasing patient loyalty. Designing efforts to concur-
rently address specific behaviors that are significantly cor-
related with loyalty and that fall within 1 of the 3 primary
factors (care concern and communication, body language,
and initial greetings) could lead to better understanding of
the factor’s underlying dimension.
2. Develop and implement tactics to increase caregiver
knowledge of caring behaviors and mechanisms to ensure
their consistent delivery.
• Use the ED study to build learning experiences. The
measures used for this study could lead to an effective
shadowing checklist when focusing on specific beha-
viors. For new recruits and in-service training, for
example, caregivers could be paired during various
patient encounters; a modified version of the observa-
tion tool used in this study could be developed for the
shadowing experience.
• Consider ways to more effectively communicate medical
information in a proactive, timely, clear manner, espe-
cially during key patient-caregiver interactions. For
example, St Vincent’s Birmingham established 30-min-
ute targets for caregiver rounds for its ED patients.
• Use the results of the ED study to better define staffing
roles. Some examples are as follows:
• Drawing on data from the ED study on the impor-
tance of patient communication and addressing family
needs, St Joseph Hospital assessed staffing patterns to
ensure that a staff member is present in the waiting
area to communicate and interact with families in the
triage area.
• As a result of the ED study, St John Hospital and
Medical Center tested a “care and comfort assistant”
position for the purpose of attending to such needs.
The first month of data after the implementation of
the position showed a 12.2% increase in the “likely-
to-recommend” question on the patient exit survey.
3. Implement strategies to decrease wait time to see a physi-
cian or primary emergency nurse.
• Recognize the importance of patients seeing a physi-
cian or primary emergency nurse promptly. Consider
adopting a time target for this initial introduction. As
a result of this study finding, St Vincent’s Birmingham
and St John Hospital and Medical Center each
initiated a “door-to-doctor” time target of 30 and 32
minutes, respectively.
4. Set specific targets within patient satisfaction categories
and monitor progress to measure effectiveness.
• Develop an ED survey to assess the impact of inter-
ventions to increase the frequency of caring behaviors
September 2010 VOLUME 36 • ISSUE 5
Liu et al/RESEARCH
and to reduce wait time. St Vincent’s Birmingham cre-
ated standards of behavior based on the ED study,
conducted staff education about the behaviors, and
developed and displayed posters in the emergency
department that reinforced the behaviors. To measure
the effectiveness of the program, a 5-question survey
of ED patients was developed that addressed the stan-
dards of behavior.
• Set goals to increase the mean score for specific areas
and Net Promoter Score as an overall measure of
patient loyalty.
Conclusion
This study revealed a correlation between caring behaviors
shown by ED caregivers and subsequent patient loyalty,
with new information emerging about this relationship.
Some behaviors that were particularly important to
patient loyalty were among those that occurred least fre-
quently in the study, for example, chatting with patients
to help tune out the illness state and showing concern for
non–treatment-related details. Behaviors that were most
closely linked to loyalty but occurred least frequently
are logical targets for improvement. The factor category
of care concern and communication showed the most
potential because a number of its associated behaviors
strongly correlated with patient loyalty but were not fre-
quently exhibited during patient observations.
In the study prompt attention to patients’ needs
upon arrival to the emergency department not only
emerged as most important to the patients’ ED experi-
ence but also proved to be the area patients rated as
least positive in their actual ED experience. Time-stamp
data also showed the link between wait time and patient
loyalty. The study showed that the most important ele-
ment of the wait time was between arrival at the emer-
gency department and seeing a physician or primary
emergency nurse.
Providing personnel dedicated to care and comfort,
creating standards of caring behavior and a correspond-
ing patient survey, establishing time targets for care-
giver rounds, and establishing wait time targets are
examples of how to use the ED study results to
improve patient loyalty.
REFERENCES
1. HCAHPS Fact sheet. http://www.cms.hhs.gov/HospitalQualityInits/.
September 2007. Accessed June 4, 2009.
2. Hausman A. Modeling the patient-physician service encounter: Improv-
ing patient outcomes. J Acad Market Sci. 2004;32(4):403-17.
WWW.JENONLINE.ORG 411
RESEARCH/Liu et al
3. Godkin J, Godkin L. Caring behaviors among nurses: Fostering a con-
versation of gestures. Health Care Manage Rev. 2004;29(3):258-67.
4. Detmar SB, Muller MJ, Wever LD, et al. Patient-physician communi-
cation during outpatient palliative treatment visits: An observational
study. JAMA. 2001;285(10):1351-7.
5. Otani K, Kurz RS. The impact of nursing care and other healthcare
attributes on hospitalized patient satisfaction and behavior intentions.
J Healthc Manag. 2004;49(3):181-96.
6. Pålsson M-BE, Norberg A. Breast cancer patients’ experiences of nursing
care with the focus on emotional support: The implementation of a nur-
sing intervention. J Adv Nurs. 1995;21(2):277-85.
7. Davies B, Oberle K. Dimensions of the supportive role of the nurse in
palliative care. Oncol Nurs Forum. 1990;17(1):87-94.
8. Skilbeck J, Payne S. Emotional support and the role of Clinical Nurse
Specialists in palliative care. J Adv Nurs. 2003;43(5):521-30.
412 JOURNAL OF EMERGENCY NURSING
9. Reichheld F. The Ultimate Question: Driving Good Profits and True
Growth. Watertown, MA: Harvard Business School Press; 2006.
10. Reichheld F. The one number you need to grow. Harvard Business
Review. December 1, 2003;81(12), 46-54.
11. Press Ganey Partners in Improvement. http://www.pressganey.com/.
Accessed July 1, 2008.
12. Kumar P, Kalwani M, Dada M. The impact of waiting time guarantees
on customers’ waiting experiences. Mark Sci. 1997;16(4):295-314.
13. Indiana ranks in study of emergency room wait times. Inside Indiana
business with Gerry Dick Web site. http://www.insideindianabusiness.
com/newsitem.asp?ID=18352. Updated June 2, 2008. Accessed June
11, 2008.
14. Emergency department pulse report: Patient perspectives on American
health care. Press Ganey. http://www.pressganey/com/galleries/default-
file/2008_ED_Pulse_Report.pdf. Accessed June 9, 2009.
VOLUME 36 • ISSUE 5 September 2010
 Liu et al/RESEARCH

APPENDIX FIGURE 1
Observer coding sheet.

September 2010 VOLUME 36 • ISSUE 5 WWW.JENONLINE.ORG 413

RESEARCH/Liu et al

APPENDIX FIGURE 2
Patient exit survey.

414 JOURNAL OF EMERGENCY NURSING VOLUME 36 • ISSUE 5 September 2010